Spacetime Physics - Taylor, Wheeler

V 'Mr
SPACETIME PHYSICS
introduction to special relativity
Second Edition
Edwin F. Taylor
M assachusetts Institute o f Technology
John Archibald Wheeler

Princeton U niversity and
U niversity o f Texas a t A u stin
W . H. Freeman and Com pany

N ew York
Library o f Congress Cataloging-in-Publication Data
Taylor, Edwin F.
Spacetime physics introduction to special relativity / Edwin F.
Taylor, John Archibald Wheeler. — 2nd ed.
p. cm.
Includes bibliographical references and index.
ISBN-13: 978-0-7167-2327-1
ISBN-10: 0-7167-2327-1
1. Special relativity (Physics) I. Wheeler, John Archibald,
1911- .II. Title.
QC173.65T37 1991 92-722
530.1T— dc20 CIP
© 1992 by Edwin F. Taylor and John Archibald Wheeler
No part o f this book may be reproduced by any mechanical, photographic, or electronic

process, or in the form o f a phonographic recording, nor may it be stored in a retrieval
system, transmitted, or otherwise copied for public or private use, without written
permission from the publisher.
Printed in the United States o f America

Eleventh printing
Both males andfemales make competent observers. We ordinarily treat the laboratory
observer as male and the rocket observer as female. Beyond this, to avoid alternating
"his ” and "her” in a single chapter, we use female pronouns fo r an otherwise
undesignated observer in odd-numbered chapters and male pronouns in even-numbered
chapters.
Epigram, facing page: Einstein remark to his assistant Ernst Straus, quoted in Mainsprings of Scientific
Discovery by Gerald Holton in The Nature of Scientific Discovery, Owen Gingerich, Editor (Smithsonian
Institution Press, Washington, 1975).
)Nhat I’m really interested in is whether God
could have made the world in a different way;
that is, whether the necessity of logical simplicity
leaves any freedom a t all.
— Albert Einstein
Edwin F. Taylor and John Archibald W heeler have w ritten a general
relativity sequel to Spacetime Physics, namely: Exploring Black Holes:
Introduction to General R elativity Addison Wesley Longman, San
Francisco, 2000
ISBN 0-201-38423-X
CONTENTS
Chapter 1 SPACETIME: OVERVIEW 1
The great u n ity is spacetime; its measure, the spacetime interval, is the same
fo r a ll observers.
1.1 Parable of the Surveyors 1 1 .2 Surveying Spacetim e 5 1 .3 Events

and Intervals Alone! 9 1 .4 Sam e Unit for Sp ace and Time: M eter, Second,
Minute, or Y e a r 11 1 .5 Unity of Spacetim e 15 References 18
Acknowledgments 19 Introduction to the Exercises 19 Exercises 20
Chapter 2 FLOATING FREE 25

Jum p off the roof: On the w a y down — in free flo a t— we have an (almost!)
perfect setting fo r conducting experiments.
2.1 Floating to the Moon 25 2 .2 The Inertial (Free-Float) Frame 26
2 .3 Local C h aracte r of Free-Float Frame 30 2 .4 Regions of Spacetim e 34
2 .5 Test Particle 36 2 .6 Locating Events With a Latticework of Clocks 37
2 .7 O b server 39 2 .8 M easuring Particle Speed 4 0 2 .9 Rocket
Frame 41 2 .1 0 Summary 43 References 4 4 Exercises 45
Chapter 3 SAME LAWS FOR ALL 53

W ithout looking out o f the w indow , we cannot tell which free-float fram e we
are in.
3.1 The Principle of Relativity 53 3 .2 W hat Is NOT the Sam e in Different

Fram es 56 3 .3 W hat IS the Sam e in Different Frames 6 0 3 .4 Relativity of
Simultaneity 62 3 .5 Lorentz Contraction of Length 6 3 3 .6 Invariance of
Transverse Dimension 65 3 .7 Invariance of the Interval Proved 6 7
3 .8 Invariance of the Interval for ALL Free-Float Fram es 71 3 .9 Summary 73
References 76 Acknowledgments 7 7 Exercises 78
Special Topic
LORENTZ TRANSFORMATION 95
Observe an event in the laboratory; predict its space an d time readings in the
rocket.
L .l Lorentz Transformation; Useful or Not? 95 L.2 Faster Than Light? 96
L .3 First Steps 9 9 L.4 Form of the Lorentz Transformation 100
L.5 Completing the Derivation 101 L .6 Inverse Lorentz
Transformation 102 L .7 Addition of Velocities 103 L.8 Summary 111
Reference 11 1 Exercises 1 12
Chapter 4 TRIP TO CANOPUS 121
T ravel quickly to a d ista n t sta r an d return, to fin d we have traveled into the
future.
4.1 Invitation to C anopus 121 4 .2 Stripped-Down Free-Floot Frame 121

4 .3 Faster Than Light? 122 4 .4 All of Sp ace is Ours! 123 4 .5 Flight
Plan 124 4 .6 Twin Paradox 125 4 .7 Lorentz Contraction 126
4 .8 Time Traveler 127 4 .9 Relativity of Simultaneity 128
4 .1 0 Experimental Evidence 131 References 134 Exercises 135
Chapter 5 TREKKING THROUGH SPACETIME 137

Move or stan d still; in either case we soar through spacetime.
5.1 Time? No. Spacetim e M ap ? Y es. 137 5 .2 Sam e Events; Different

Free-Float Frames 139 5 .3 Invariant H yperbola 143
5 .4 W orldline 143 5 .5 Length Along a Path 147 5 .6 W ristwatch Time
Along a W orldline 148 5 .7 Kinked W orldline 152 5 .8 Stretch
Factor 155 5 .9 Touring Spacetim e Without a Reference Frame 160
5 .1 0 Summary 162 References 163 Exercises 163
Chapter 6 REGIONS OF SPACETIME 171

The speed o f light is a m ighty barrier th a t preserves cause an d effect.
6.1 Light Speed : Limit on C ausality 171 6 .2 Relation Between Events:

Timelike, Spacelike, or Lightlike 172 6 .3 Light C on e: Partition in
Spacetim e 177 Exercises 183
Chapter 7 MOMENERGY 189

A second great unity is momentum-energy (momenergy); its measure, mass, is
the same fo r a ll observers.
7.1 M om energy: Total C onserved in a Collision 189 7 .2 M om energy
Arrow 191 7 .3 M om energy Com ponents and M agnitude 195
7 .4 Momentum: “ Sp ace Part” of M om energy 199 7 .5 Energy: “ Time Part" of
M om energy 201 7 .6 Conservation of M om energy and its
C onsequences 2 0 7 7 .7 Summary 211 Acknowledgment 213
Exercises 214
C h apters COLLIDE. CREATE. ANNIHILATE. 221

Convert mass to energy and energy to mass.
8.1 The System 221 8 .2 Three M odest Experiments 2 2 2 8 .3 M ass of a

System of Particles 22 4 8 .4 Energy Without M ass: Photon 228 8 .5 Photon
Used to C rea te M ass 23 3 8 .6 M aterial Particle Used to C re a te M ass 234
8 .7 Converting M ass to U sable Energy: Fission, Fusion, Annihilation 2 3 7
8 .8 Summary 244 D ia lo g : Use and Abuse of the Concept of M ass 246
References 251 Acknowledgments 2 5 2 Exercises 253
Vi
Cliapter 9 GRAVITY: CURVED SPACETIME IN
ACTION 275
G ra vity is not a force reaching across space but a distortion — curvature ! —
o f spacetime experienced right where you are.
9.1 G ravity in Brief 275 9 .2 G alileo , Newton, and Einstein 275 9 .3 Local
Moving O rd ers for M ass 2 7 7 9 .4 Spacetim e Curvature 2 8 0 9 .5 Parable
of the Two Travelers 281 9 .6 Gravitation a s Curvature of Spacetim e 284
9 .7 G ravity W av es 28 8 9 .8 Black Hole 2 9 2 9 .9 The Cosm os 2 9 6
References 29 6
ANSWERS TO ODD-NUMBERED EXERCISES 299

INDEX 303
v ii
. 'i
.. ,-i^‘
■■~-ut'
■■.'■:A.,■■ «,■
f c . - s s ;
■ '■^■■■^.
' •% i'->'
'r | '
■'y;t;- ’tA
> # V ,; *
I . %'4"- .
• . •
' F '•>■
‘ -1 ^ ■■■. ■
. '/;:^:I4;
■ a f /_' ■■.
'-. ■
■ •;
■ ■ ;:.'
,■ ■■■
' . . ■
.' i '
\ \ •
• •’ ' • /■ '. y
• i ■'',' .. j*' ,' - 5 '- is
, . ■' f t ' -
', - - s
' * ' ■'
- ■■ "V' . i
' - ■- i S
' ' '■ ••
-ft-
■■ . '- . ■ >
% 5 « --^ r* ¥ v
. , : 5 f ' 4 ' ' n.“ t-
■ - j ^ y - '- .y - = -v .:
^ -IJ’ -#• ' •>•
•' ■•vV’j’- S ' ; '

■■ ---
"j "
vi fK iJv;
-^•- .
■' . > '

,*'; ‘^ ■ • U t _
'* ‘ ■Vt'*'
.^.-1 '
' I'>,-K'. - i '- '- i
' ■• '■'’
f% .' i f t '-
’•'“ •r'^,
^■frLva-JP. ' ,
-.j( . , :
-;J i ’ in ;.
•X‘^:=;;a‘ > ' ‘-- y ' ^ ' M m
■-■'- . ~ ™ - - T ' ; r _ . ■■.;■■;
tr: —
: ' y’ ' ;.!.r' ' ' *.^^^?^V ■
■'V
'
--'*'--'•’ î ' ^ ' "V ’'
' >• '
■■‘V- ; •< ■
SPAariME:
O ur imagination is stretched to the utmost, not, as
in fiction, to imagine things which are not really there,
hut ju st to comprehend those things which are there.
Richard P. Feynman
1.1 PARABLE OF THE SURVEYORS

disagree on northward and eastward
separations; agree on distance
Once upon a time there was a Daytime surveyor who measured off the king’s lands.
Daytime surveyor uses
He took his directions of north and east from a magnetic compass needle. Eastward
magnetic north
separations from the center of the town square he measured in meters. The northward
direction was sacred. He measured northward separations from the town square in a
different unit, in miles. His records were complete and accurate and were often
consulted by other Daytimers.
A second group, the Nighttimers, used the services of another surveyor. Her north
and east directions were based on a different standard of north: the direction of the
Nighttime surveyor uses
North Star. She too measured separations eastward from the center of the town square
North-Star north
in meters and sacred separations northward in miles. The records of the Nighttime
surveyor were complete and accurate. Marked by a steel stake, every corner of a plot
appeared in her book, along with its eastward and northward separations from the
town square.
Daytimers and Nighttimers did not mix but lived mostly in peace with one another.
However, the two groups often disputed the location of property boundaries. Why?
Because a given corner of the typical plot of land showed up with different numbers in
the two record books for its eastward separation from the town center, measured in
meters (Figure 1-1). Northward measurements in miles also did not agree between the
two record books. The differences were small, but the most careful surveying did not
succeed in eliminating them. No one knew what to do about this single source of
friction between Daytimers and Nighttimers.
One fall a student of surveying turned up with novel open-mindedness. Unlike all
previous students at the rival schools, he attended both. At Day School he learned
CHAPTER 1 SPACETIME: OVERVIEW
magnetic North-Star
north north
magnetic
east
North-Star
east
NIGHTTIME: NORTH-STAR NORTH

FIGURE 1-1. The town as plotted by Daytime and Nighttime surveyors. Notice that the line of
Daytime magnetic north just grazes the left side of the north gate, while the line of Nighttime North-Star
north just grazes the right side of the same gate. Steel stakes A, B, C, D driven into the ground mark the
comers of a disputed plot of land. Ar shown, the eastward separation of stake A from the north-south line
measured by the Daytime surveyor is different from that measured by the Nighttime surveyor.
from one expert his method of recording locations of gates of the town and corners of
plots of land based on magnetic north. At Night School he learned the other method,
based on North-Star north.
As days and nights passed, the student puzzled more and more in an attempt to find
some harmonious relationship between rival ways of recording location. His attention
was attracted to a particular plot of land, the subject of dispute between Daytimers and
Nighttimers, and to the steel stakes driven into the ground to mark corners of this
disputed plot. He carefully compared records of the two surveyors (Figure 1-1, Table
1- 1).
In defiance of tradition, the student took the daring and heretical step of converting
Student converts miles to meters
northward measurements, previously expressed always in miles, into meters by multi
plying with a constant conversion factor k. He found the value of this conversion factor
to be ^ = 1609.344 meters/mile. So, for example, a northward separation of 3 miles
could be converted to ^ X 3 miles = 1609.344 meters/mile X 3 miles = 4828.032
meters. "At last we are treating both directions the same!” he exclaimed.
Next the student compared Daytime and Nighttime measurements by trying
various combinations of eastward and northward separation between a given stake
and the center of the town square. Somewhere rhe student heard of the Pythagorean
Theorem, that the sum of squares of the lengths of two perpendicular legs of a right
triangle equals the square of the length of the hypotenuse. Applying this theorem, he
discovered that the expression
D aytim e D aytim e
/ northward \ 2 eastward
k X 1 separation I + separation ( 1- 1)
\ (miles) / _ (meters)
1.1 PARABLE OF THE SURVEYORS
------------------------------ iC^^ ^ B L E 1 - 1 ^ -------------------------------
TWO DIFFERENT SETS OF RECORDS; SAME PLOT OF LAND

Daytime surveyor’s axes Nighttime surveyor's axes
oriented to magnetic north oriented to North-Star north
Eastward Northward Eastward Northward
(meters) (miles) (meters) (miles)
Town square 0 0 0 0
Corner stakes:
Stake A 4010.1 1.8330 3950.0 1.8827
Stake B 5010.0 1.8268 4950.0 1,8890
Stake C 4000.0 1.2117 3960.0 1.2614
Stake D 5000.0 1.2054 4960.0 1.2676
based on Dayrime measurements of the position of steel stake C had exaaly the
same numerical value as the quantity
N ig h ttim e N ig h ttim e
/ northward \ 2 eastward
k X 1 separation 1 + separation ( 1- 2)
V (miles) / _ (meters)
computed from the readings of the Nighttime surveyor for stake C (Table 1-2). He
-C j^ B L E
“ INVARIANT DISTANCE” FROM CENTER OF TOWN SQUARE TO STAKE C

(Data from Table 1-1)
Daytime measurements Nighttime measurements
Northward separation Northward separation

1.2117 miles 1.2614 miles
Multiply by Multiply by
k = 1609.344 meters/mile k = 1609.344 meters/mile
to convert to meters: to convert to meters:
1950.0 meters 2030.0 meters
Square the value 3,802,500 (meters)^ Square the value 4,120,900 (meters)^
Eastward separation Eastward separation
Square the value and add -b 16,000,000 (meters)^ Square the value and add + 15,681,600 (meters)^
Sum of squares = 19,802,500 (meters)^ Sum of squares = 19,802,500 (meters)^
Expressed as a Expressed as a
number squared = (4450 meters)^ number squared = (4450 meters)^
This is the square This is the square
of what measurement? 4450 meters of what measurement? 4450 meters
i i
SAME
DISTANCE
from center of Town Square
CHAPTER 1 SPACETIME: OVERVIEW
magnetic north
town square
DAYTIME: MAGNETIC NORTH
North-Star north
FIGURE 1-2. The distance between stake A

and the center of the town square has the same
value for Daytime and Nighttime surveyors,
even though the northward and eastward sepa
rations, respectively, are not the samefor the two
surveyors.
NIGHHIME; NORTH-STAR NORTH
cried the same comparison on recorded positions of stakes A, B, and D and found
agreement here too. The student’s excitement grew as he checked his scheme of
comparison for all stakes at the corners of disputed plots — and found everywhere
agreement.
Flushed with success, the student methodically converted all northward measure
ments to units of meters. Then the student realized that the quantity he had calculated,
the numerical value of the above expressions, was not only the same for Daytime and
Discovery: Invariance of distance
Nighttime measurements. It was also the square of a length: (meters)^. He decided to
give this length a name. He called it the d istance from the center of town.
northward 2 eastward
(distance)^ — separarion + separarion ( 1- 3)
(meters) (meters)
He said he had discovered the p rin cip le o f invariance o f distance; he reckoned

exactly the same value for distance from Daytime measurements as from Nighttime
measurements, despite the fact that the two sets of surveyors’ numbers differed
significantly (Figure 1-2).
After some initial confusion and resistance, Day timers and Nighttimers welcomed
rhe srudent’s new idea. The invariance of distance, along with further results, made it
possible to harmonize Daytime and Nighttime surveys, so everyone could agree on the
location of each plot of land. In this way the last source of friction between Day timers
and Nightrimers was removed.
1.2 SURVEYING SPACETIME

disagree on separations in space and time;
agree on spacetime interval
The Parable of the Surveyors illustrates the naive state of physics before the discovery
of special relativ ity by Einstein of Bern, Lorentz of Leiden, and Poincare of Paris.
Naive in what way? Three central points compare physics at the turn of the twentieth
century with surveying before the student arrived to help Daytimers and Nighttimers.
First, surveyors in the mythical kingdom measured northward separations in a
The second: A sacred unit
sacred unit, the mile, different from the unit used in measuring eastward separations.
Similarly, people smdying physics measured time in a sacred unit, called the second,
different from the unit used to measure space. No one suspected the powerful results
of using the same unit for both, or of squaring and combining space and time
separations when both were measured in meters. Time in meters is just the time it takes
a light flash to go that number of meters. The conversion factor between seconds and
meters is the speed of light, c = 299,792,458 meters/second. The velocity of light c
(in meters/second) multiplied by time t (in seconds) yields ct (in meters).
The speed of light is the only natural constant that has the necessary units to convert
Speed of light converts seconds
a time to a length. Historically the value of the speed of light was regarded as a sacred
to meters
number. It was not recognized as a mere conversion factor, like the factor of conversion
between miles and meters — a factor that arose out of historical accident in human
kind’s choice of units for space and time, with no deeper physical significance.
Second, in the parable northward readings as recorded by two surveyors did not
differ much because the two directions of north were inclined to one another by only
the small angle of 1.15 degrees. At first our mythical student thought that small
differences between Daytime and Nighttime northward measurements were due to
surveying error alone. Analogously, we used to think of the separation in time between
two electric sparks as the same, regardless of the motion of the observer. Only with the
publication of Einstein’s relativity paper in 1905 did we learn that the separation in
Time between events: Different
time between two sparks really has different values for observers in different states of
for different frames
motion — in different fram es.
Think of John standing quietly in the front doorway of his laboratory building.
Suddenly a rocket carrying Mary flashes through rhe front door past John, zooms
down the middle of the long corridot, and shoots out the back door. An antenna
projects from the side of Mary’s rocket. As the rocket passes John, a spark jumps across
rhe 1-millimeter gap between the antenna and a pen in John’s shirt pocket. The rocket
continues down the corridor. A second spark jumps 1 millimeter between the antenna
and the fire extinguisher mounted on the wall 2 meters farther down the corridor. Still
latet other metal objects nearer the rear receive additional sparks from the passing
rocket before it finally exits through the rear door.
John and Mary each measure the lapse of time between “pen spark” and “fire-
O n e o bserver uses laboratory
extinguisher spark.” They use accurate and fast electronic clocks. John measures frame
this time lapse as 33.6900 thousand-millionths of a second (0.0000000336900
second = 33.6900 X 10“^ second). This equals 33.6900 nanoseconds in the
terminology of high-speed electronic circuitry. (One nanosecond = 10~^ second.)
Another observer uses rocket
Mary measures a slightly different value for the time lapse between the two sparks, frame
33.0228 nanoseconds. For John the fire-extinguisher spark is separated in space by
2.0000 meters from the pen spark. For Mary in the rocket the pen spark and
fire-extinguisher spark occur at the same place, namely at the end of her antenna. Thus
fot her their space separation equals zero.
Later, laboratory and rocket observers compare their space and time measurements
between the various sparks (Table 1-3). Space locations and time lapses in both frames
are measured from the pen spark.
6 CHAPTER 1 SPACETIME: OVERVIEW
-------------------- C [ ^ B L E l - 3 ^ > --------------------
SPACE AND TIME LOCATIONS OF THE SAME

SPARKS AS SEEN BY TWO OBSERVERS
Distance and time between sparks as measured by observer who is
standing in laboratory (John) moving by in rocket (Mary)
Distance Time Distance Time
(meters) (nanoseconds) (meters) (nanoseconds)
Reference spark 0 0 0 0
(pen spark)
Spark A 2.0000 33.6900 0 33.0228
(fire-extinguisher
spark)
Spark B 3.0000 50.5350 0 49.5343
Spark C 5.0000 84.2250 0 82.5572
Spark D 8.0000 134.7600 0 132.0915
The third point of comparison between the Parable of the Surveyors and the state of
physics before special relativity is this: The mythical student’s discovery of the concept
of distance is matched by the Einstein - Poincare discovery in 1905 of the in v arian t
spacetim e in terv al (formal name L orentz in terv al, but we often say just in te r
Discovery: Invariance of
val), a central theme of this book. Let each time measurement in seconds be converted
spacetime interval
to meters by multiplying it by the “conversion factor c " the speed of light:
c = 299,792,458 meters/second = 2.99792458 X 10* meters/second

= 0.299792458 X 10^ meters/second = 0.299792458 meters/nanosecond
Then the square of the spacetime interval is calculated from the laboratory observer’s
measurements by subtracting the square of the space separation from the square of the
time separation. Note the minus sign in equation (1-4).
L aboratory L aboratory
/ time \ 2 space
(interval)^ = c X 1 separation 1 — separation ( 1-41
V (seconds) / _ (meters)
The rocket calculation gives exactly the same value of the interval as the laboratory
calculation.
R ocket R ocket
/ time \ 2 space
(interval)^ = c X 1 separation 1 separation (1-5)
V(seconds) / _ (meters)
even though the respective space and time separations are not the same. Two observers
find different space and time separations, respectively, between pen spark and fire-
extinguisher spark, but when they calculate the spacetime interval between these
sparks their results agree (Table 1-4).
The student surveyor found that invariance of distance was most simply written
with both northward and eastward separations expressed in the same unit, the meter.
Likewise, invariance of the spacetime interval is most simply written with space and
-C ^ A B L E
“ INVARIANT SPACETIME INTERVAL” FROM REFERENCE SPARK TO SPARK A

(Data from Table 1-3]
Laboratory measurements Rocket measurements
Time lapse Time lapse

33.6900 X 10-« seconds 33.0228 X 10-9 seconds
= 33.6900 nanoseconds = 33.0228 nanoseconds
Multiply by Multiply by
r = 0.299792458 f = 0.299792458
meters per nanosecond meters per nanosecond
to convert to meters: to convert to meters:
Square the value 102.010 (meters)^ Square the value 98.010 (meters)^
Spatial separation Spatial separation
2.000 meters zero
Square the value and subtract — 4.000 (meters)^ Square the value and subtraa - 0
Result of subtaction = 98.010 (meters)^ Result of subtaction = 98.010 (meters)^
expressed as a expressed as a
number squared = (9.900 meters)^ number squared = (9.900 meters)^
This is the square This is the square
of what measurement? 9.900 meters of what measurement? 9.900 meters
i i
SAME SPACETIME
INTERVAL
from the reference event
time separations expressed in the same unit. Time is converted to meters: t (meters) =
£• X t (seconds). Then the interval appears in simplified form:
time 2 space
(interval)^ — separation — separation ( 1- 6 )
(meters) (meters)
The in v arian ce o f th e spacetim e in terv al — its independence of the state of

motion of the observer — forces us to recognize that time cannot be separated from
Sp a ce and time are
space. Space and time are part of a single entity, spacetim e. Space has three
part of spacetime
dimensions: northward, eastward, and upward. Time has one dimension: onward!
The interval combines all four dimensions in a single expression. The geometry of
spacetime is truly four-dimensional.
To recognize the unity of spacetime we follow the procedure that makes a landscape
take on depth— we look at it from several angles. That is why we compare space and
rime separations between events A and B as recorded by two different observers in
relative motion.
Why the minus sign in the equation for the interval? Pythagoras tells us to A D D the
squares of northward and eastward separations to get the square of the distance. Who
tells us to SUBTRACT the square of the space separation between eventsfrom the square
of their time separation in order to get the square of the spacetime interval?
Shocked? Then you’re well on the way to understanding the new world of very fast
motion! This world goes beyond the three-dimensional textbook geometry of Euclid,
in which distance is reckoned from a sum of squares. In this book we use another
kind of geometry, called Lorentz geometry, more real, more powerful than Euclid
for the world of the very fast. In Lorentz geometry the squared space separation is
combined with the squared time separation in a new way— by subtraction. The
result is the square of a new unity called the spacetime intervalhtvf/ttn events. The
numerical value of this interval is invariant, the same for all observers, no matter
how fast they are moving past one another. Proof? Every minute of every day an
experiment somewhere in the world demonstrates it. In Chapter 3 we derive the
invariance of the spacetime interval— with its minus sign— from experiments.
They show the finding that no experiment conducted in a closed room will reveal
whether that room is “at rest’’ or “in motion” (Einstein’s Principle of Relativity).
We won’t wait until then to cash in on the idea of interval. We can begin to enjoy the
payoff right now.
SAMPLE PROBLEM l-i;

S P A R K I N G AT A FASTER RATE
Another, even faster rocket follows the first, enter John, the laboratory observer. The second flash
ing the ftont door, zipping down the long corridor, jumps when the rocket antenna reaches a door
and exiting through the back doorway. Each time knob 4.00000000 meters farther along the hall as
the rocket clock ticks it emits a spark. As before, measured by the laboratory observer, who records
the first spark jumps the 1 millimeter from the the time between these two sparks as 16.6782048
passing rocket antenna to the pen in the pocket of nanoseconds.
a. W hat is the time between sparks, measured in meters by John, the laboratory
observer?
b. W hat is the value of the spacetime interval between the two events, calculated
from John’s laboratory measurements?
c. Predict: W hat is the value of the interval calculated from measurements in the
new racket frame?
d. W hat is the distance between sparks as measured in this rocket frame?
e. W hat is the time (in meters) between sparks as measured in this rocket frame?
C om pare w ith the tim e between the sam e sparks as m easured by Jo h n in the
laboratory frame.
f. W hat is the speed of this rocket as measured by John in the laboratory?
SOLUTION
a. Time in meters equals time in nanoseconds multiplied by the conversion factor,
the speed of light in meters per nanosecond. For John, the laboratory observer,
16.6782048 nanoseconds X 0.299792458 meters/nanosecond

== 5.00000000 meters
b. The square of the interval between two flashes is reckoned by subtracting the
square of the space separation from the square of the time separation. Using
laboratory figures:
(interval)^ = (laboratory time)^ — (laboratory distance)^

= (5 meters)^ — (4 meters)^ = 25 (meters)^ - 16 (meters)^
= 9 (meters)^ ~ (3 meters)^
1.3 EVENTS AND INTERVALS ALONE! 9
Therefore the interval between the two sparks has the value 3 meters (to nine
significant figures).
c. W e strongly assert in this chapter that the spacetim e in terv al is in v aria n t —
has the same value by whomever calculated. Accordingly, the interval between
the two sparks calculated from rocket observations has the same value as the
interval (3 meters) calculated from laboratory measurements.
d. From the rocket rider’s viewpoint, both sparks jump from the same place, namely
the end of her antenna, and so distance between the sparks equals zero for the
rocket rider.
e. W e know the value of the spacetime interval between two sparks as computed in
the rocket frame (c). And we know that the interval is computed by subtracting
the square of the space separation from the square of the time separation in the
rocket frame. Finally we know that the space separation in the rocket frame
equals zero (d ). Therefore the rocket time lapse between the two sparks equals the
interval between them;
(interval)^ = (rocket time)^ — (rocket distance)^

(3 meters)^ = (rocket time)^ — (zero)^
from which 3 meters equals the rocket time between sparks. Compare this with 5
meters of light-travel time between sparks as measured in the laboratory frame.
f. Measured in the laboratory frame, the rocket moves 4 meters of distance (state
ment of the problem) in 5 merers of light-travel time (a). Therefore its speed in
the laboratory is 4 /5 light speed. Why? Well, light moves 4 meters of distance in
4 meters of time. The rocket takes longer to cover this distance: 5 meters of time.
Suppose that instead of 5 meters of time, the rocket had taken 8 meters of time,
twice as long as light, to cover rhe 4 meters of disrance. In that case it would be
moving at 4 /8 — or half— the speed of light. In the present case the rocket
travels the 4 meters of distance in 5 meters of time, so it moves at 4 /5 light speed.
Therefore its speed equals
(4 /5 ) X 2.99792458 X 10® meters/second

2.3983397 X 10® meters/second
1.3 EVENTS AND INTERVALS ALONE!

tools enough to chart matter and motion
without any reference frame
In surveying, rhe fundamental concept is place. The surveyor drives a steel stake to
Surveying locates a place
mark the corner of a plot of land — to mark a place. A second stake marks another
corner of the same plot — another place. Every surveyor — no matter what his or her
standard of north — can agree on the value of the distance between the two stakes,
between the two places.
Every stake has its own reality. Likewise the distance between every pair of srakes
also has its own teality, which we can experience direcrly by pacing off the straight line
from one stake to the other stake. The reading on our pedometer— the distance
lO CHAPTER] SPACETIME: OVERVIEW
between stakes— is independent of all surveyors’ systems, with their arbitrary choice
of north.
More: Suppose we have a table of distances between every pair of stakes. That is all
we need! From this table and the laws of Euclidean geometry, we can constmct the
map of every surveyor (see the exercises for this chapter). Distances between stakes:
That is all we need to locate every stake, every place on the map.
In physics, the fundamental concept is event. The collision between one particle
Physics locates an event
and another is an event, with its own location in spacetime. Another event is the
emission of a flash of light from an atom. A third is the impact of the pebble that chips
the windshield of a speeding car. A fourth event, likewise fixing in and by itself a
location in spacetime, is the strike of a lightning bolt on the rudder of an airplane. An
event matks a location in spacetime; it is like a steel stake driven into spacetime.
Every laboratory and rocket observer— no matter what his or her relative velocity
— can agree on the spacetime interval between any pair of events.
Every event has its own reality. Likewise the interval between every pair of events
W ristwatch m easures
also has its own reality, which we can experience directly. W e carry our wristwatch at
interval directly
constant velocity from one event to the other one. It is not enough just to pass through
the two physical locations— we must pass through the actual events', we must be at
each event precisely when it occurs. Then the space separation between the two events
is zero for us — they both occur at our location. As a result, our wristwatch reads
directly the spacetime interval between the pait of events:
time space
(interval)^ — separation separation
(meters) (meters)
time time
separation — [zero]^ separation [wristwatch time}
(meters) (meters)
The time read on a wristwatch carried between two events — the interval between
those events — is independent of all laboratory and rocket reference frames.
More: To chart all happenings, we need no more than a table of spacetime intervals
between every pair of events. That is all we need! From this table and the laws of
Lorentz geometry, it turns out, we can construct the space and time locations of events
as observed by every laboratory and rocket observet. Intervals between events: That is
all we need to specify the location of every event in spacetime.
In brief, we can completely describe and locate events entirely without a reference
‘Do science” with intervals alone
frame. W e can analyze the physical world— we can “do science” — simply by
cataloging every event and listing the interval between it and every other event. The
unity of spacetime is reflected in the simplicity of entries in our table: intervals only.
O f course, if we want to use a reference frame, we can do so. We then list in our table
the individual northward, eastward, upward, and time separations between pairs of
events. However, these laboratory-frame listings for a given pair of events will be
different from the corresponding listings that our rocket-frame colleague puts in her
table. Nevertheless, we can come to agreement if we use the individual separations to
reckon the interval between each pair of events:
(interval)^ — (time separation)^ — (space separation)^
That returns us to a universal, frame-independent description of the physical world.
When two events both occur at the position of a certain clock, that special clock
measures directly the interval between these two events. The interval is called the
p r o p e r tim e (or sometimes the local tim e). The special clock that records the
proper time directly has the name p r o p e r clo ck for this pair of events. In this book
1.4 SAME UNIT FOR SPACE AND TIME: METER, SECOND, MINUTE, OR YEAR 11
we often call the proper time the w ristw atc h tim e and the proper clock the
w ristw atc h to emphasize that the proper clock is carried so that it is “present” at
each of the two events as the events occur.
In Einstein’s German, the word for proper time is Eigenzeil, or “own-time,”
implying “one’s very own time.” The German word provides a more accurate
description than the English. In English, the word “proper” has come to mean
“ following conventional mles.” Proper time certainly does not do that!
Hey! I just thought of something: Suppose two events occur at the same time in myframe
but very fa r apart, for example two handclaps, one in New York City and one in San
Francisco. Since they are simultaneous in my frame, the time separation between
handclaps is zero. But the space separation is not zero— they are separated by the width
of a continent. Therefore the square of the interval is a negative number:
{interval^ = (time separation^ — (space separation)^

= (zero)^ — (space separation)^ = — (space separation)^
How can the square of the spacetime interval be negative?
In most of the situations described in the present chapter, there exists a reference
frame in which two events occur at the same place. In these cases time separation
predominates in all frames, and the interval squared will always be positive. W e call
these intervals tim e lik e in tervals.
Euclidean geometry adds squares in reckoning distance. Hence the result of the
calculation, distance squared, is always positive, regardless of the relative magni
tudes of north and east separations. Lorentz geometry, however, is richer. For your
simultaneous handclaps in New York City and San Francisco, space separation
between handclaps predominates. In such cases, the interval is called a sp acelik e
in te rv al and its form is altered to
(interval)^ = (space separation)^ — (time separation)^ [when spacelike]
This way, the squared interval is never negative.

The timelike interval is measured directly using a wristwatch carried from one
event to the other in a special frame in which they occur at the same place. In contrast,
a spacelike interval is measured directly using a rod laid between the events in a
special frame in which they occur at the same time. This is the frame you describe in
your example.
Spacelike interval or timelike interval: In either case rhe interval is invariant— has
the same value when reckoned using rocket measurements as when reckoned using
laboratory measurements. You may want to skim through Chapter 6 where timelike
and spacelike intervals are described more fully.
1.4 SAME UNIT FOR SPACE AND TIME:

METER, SECOND, MINUTE, OR YEAR
meter for particle accelerators; minute for
planets; year for the cosmos
The parable of the surveyors cautions us to use the same unit to measure both space
M easure time in meters
and time. So we use meter for both. Time can be measured in meters. Let a flash of
light bounce back and forth between parallel mirrors separated by 0.5 meter of
distance (Figure 1-3). Such a device is a “clock” that “ticks” each time the light flash
0.5 meter
arrives back at a given mirror. Between ticks the light flash has traveled a round-trip
distance of 1 meter. Therefore we call the stretch of time between ticks 1 m e te r o f
lig h t-trav el tim e or more simply 1 m e te r o f tim e.
One meter of light-travel time is quite small compared to typical time lapses in
FIGURE 1-3. This two-mirror “clock" sends to
our everyday experience. Light travels nearly 300 million meters per second
the eyeflash afterflash, each separatedfrom the
next by 1 meter of light-travel time. A light
(300,000,000 meters/second = 3 X 10® meters/second, four fifths of the way to
flash (represented by an asterisk) bounces back Moon in one second). Therefore one second equals 300 million meters of lighr-travel
and forth between parallel mirrors separated time. So 1 meter of light-travel time has the small value of one three-hundred-mil-
from one another by 0.5 meter of distance. The lionth of a second. [How come? Because (1) light goes 300 million meters in one
silver coating of the right-hand mirror does not second, and (2) one three-hundred-millionth of that distance (one meter!) is covered in
reflect perfectly: It lets 1 percent of the light pass
through to the eye each time the light pulse hits
one three-hundred-millionrh of that time.] Nevertheless this unit of time is very useful
it. Hence the eye receives a pulse of light every when dealing with light and with high-speed particles. A proton early in its travel
meter of light-travel time. through a particle accelerator may be jogging along at “only” one half the speed of
light. Then it travels 0.5 meter of distance in 1 meter of light-travel time.
We, our cars, even our jet planes, creep along at the pace of a snail compared with
light. W e call a deed quick when we’ve done it in a second. But a second for light
means a distance covered of 300 million meters, seven trips around Earth. As we dance
around the room to the fastest music, oh, how slow we look to light! Not zooming.
Not dancing. N ot creeping. Oozing! That long slow ooze racks up an enormous
number of meters of light-travel time. That number is so huge that, by the end of one
step of our frantic dance, the light that carries the image of the step’s beginning is well
on its way to Moon.
In 1983 the General Conference on Weights and Measures officially redefined the
M eter officially defined
meter in terms of the speed of light. T h e m e te r is now d efined as th e d istan ce
using light speed
th a t lig h t travels in a v acu u m in th e fractio n 1/299,792,458 o f a second.
(For the definition of the second, see Box 3-2.) Since 1983 the speed of light is, by
definition, equal to c = 299,792,458 meters/second. This makes official the central
position of the speed of light as a conversion factor between time and space.
This official action defines distance (meter) in terms of time (second). Every day we
use time to measure distance. “My home is only ten minutes (by car) from work.”
“The business district is a five-minute walk.” Each statement implies a speed — the
speed of driving or walking— that converts distance to time. But these speeds can
vary— for example, when we get caught in traffic or walk on cmtches. In contrast, the
speed of light in a vacuum does not vary. It always has the same value when measured
over time and the same value as measured by every observer.
We often describe distances to stars and galaxies using a unit of time. These
M easure distance in light-years
distances we measure in light-years. One light-year equals the distance that light
travels in one year. Along with the light-year of space goes the year of time. Here again,
space and time are measured in the same units— years. Here again the speed of light is
the conversion factor between measures of time and space. From our everyday per
spective one light-year of space is quite large, almost 10,000 million million meters: 1
light-year = 9,460,000,000,000,000 meters = 0.946 X 10*® meters. Nevertheless
it is a convenient unit for measuring distance between stars. For example, the nearest
star to our Sun, Proxima Centauri, lies 4.28 light-years away.
Any common unit of space or time may be used as the same unit for both space and
time. For example. Table 1-5 gives us another convenient measure of time, seconds,
compared with time in meters. We can also measure space in the same units,
light-seconds. Our Sun is 499 light-seconds — or, more simply, 499 seconds — of
distance from Earth. Seconds are convenient for describing distances and times among
events that span the solar system. Alternatively we could use minutes of time and
light-minutes of distance: Our Sun is 8.32 light-minutes from Earth. W e can also use
hours of time and light-hours of distance. In all cases, the speed of light is the
conversion factor between units of space and time.
1.4 SAME UNIT FOR SPACE AND TIME; METER, SECOND, MINUTE, OR YEAR 13
-------- d ^ B L E --------
SOME LIGHT-TRAVEL TIMES

Time in seconds
of light-travel time Time in meters
Telephone call one way:

New York City to San Francisco 0.0138 4.139,000
via surface microwave link
New York City to San Ftancisco 0.197 59,000,000
via Earth satellite
New York City to San Francisco 2.51 752,000,000
bounced off Moon
Flash of light:
Emitted by Sun, 499.0 149,600,000.000
received on Earth
Expressing time and space in the same unit m e te r is convenient for describing
motion of high-speed particles in the confines of the laboratory. Time and space in the
Use convenient units,
same unit second (or m in u te or h o u r) is convenient for describing relations among the sam e for sp ace and time
events in our solar system. Time and space in the same unit year is convenient for
describing relations among stars and among galaxies. In all three arenas spacetime is
the stage and special relativity is the spotlight that illuminates the inner workings of
Nature.
W e are not accustomed to measuring time in meters. So as a reminder to ourselves

we add a descriptor: meters of light-travel time. But the unit of time is still the meter.
Similarly, the added words “seconds of distance" and ‘‘light-years’’ help to remind
us that distance is measured in seconds or years, units we usually associate with time.
But this unit of distance is really just second or year. The modifying descriptors are
for our convenience only. In Nature, space and time form a unity: spacetime!
The words sound OK. The mathematics appears straightforward. The Sample Problems
seem logical. But the ideas are so strange! Why should I believe them? How can
invariance of the interval be proved?
N o wonder these ideas seem strange. Particles zooming by at nearly the speed of
light — how far this is from our everyday experience! Even the soaring jet plane
crawls along at less than one-millionth light speed. Is it so surprising that the world
appears different at speeds a million times faster than those at which we ordinarily
move with respect to Earth?
The notion of spacetime interval distills a wealth of real experience. W e begin with
interval because it endures: It illuminates observations that range from rhe core of a
nucleus to the center of a black hole. Understand the spacetime interval and you
vault, in a single bound, to the heart of spacetime.
Chapter 3 presents a logical proof of the invariance of the interval. Chapter 4
reports a knock-down argument about it. Chapters that follow describe many
experiments whose outcomes are rorally incomprehensible unless the interval is
invariant. Real verification comes daily and hourly in the on-going enterprise of
experimental physics.
SAMPLE PROBLEM 1- 2
P R O T O N , R O C K , AND S T A R S H I P
a. A proton moving at 3 /4 light speed (with respect to the laboratory) passes
through two detectors 2 meters apart. Events 1 and 2 are the transits through the
two detectors. W hat are the laboratory space and time separations between the
two events, in meters? W hat are the space and time separations between the
events in the proton frame?
b. A speeding rock from space streaks through Earth’s outer atmosphere, creating a
short fiery rrail (Event 1) and continues on its way to crash into Sun (Event 2) 10
minutes later as observed in the Earth frame. Take Sun to be 1.4960 X 10“
meters from Earth. In the Earth frame, what are space and time separations
between Event 1 and Event 2 in minutes? W hat are space and time separations
between the events in the frame of the rock?
c. In the twenty-third century a starship leaves Earth (Event 1) and travels at 95
percent light speed, later arriving at Proxima Centauri (Event 2), which lies 4.3
light-years from Earth. W hat are space and time separations between Event 1 and
Event 2 as measured in the Earth frame, in years? W hat are space and time
separations between these events in the frame of the starship?
SOLUTION
a. The space separation measured in the laboratory equals 2 meters, as given in the
problem. A flash of light would take 2 meters of light-travel time to travel
between the two detectors. Something moving at 1/ 4 light speed would take four
times as long: 2 meters/( 1/4 ) = 8 meters of light-travel time to travel from one
detector to the other. The proton, moving at 3 /4 light speed, takes 2 meters/
(3/4) = 8 /3 meters = 2.66667 meters of light-travel time between events as
measured in the laboratory.
Event 1 and Event 2 both occur at the position of the proton. Therefore the
space separation between the two events equals zero in the proton frame. This
means that the spacetime interval — the proper time— equals the time between
events in the proton frame.
(proton time)^ — (proton distance)^ = (interval)^ = (lab time)^ — (lab distance)^

(proton time)^ — (zero)^ = (2.66667 meters)^ ~ (2 meters)^
= ( 7. 1111— 4) (meters)^
(proton time)^ = 3.1111 (meters)^
So time between events in the proton frame equals the square root of this, or
1.764 meters of time.
b. Light travels 60 times as far in one minute as it does in one second. Its speed in
meters per minute is therefore:
2.99792458 X 10® meters/second X 60 seconds/minute

= 1.798754748 X 10^® meters/minute
So the distance from Earth to Sun is
1.4960 X 10“ meters

— 8.3169 light-minutes
1.798754748 X 10^® meters/minute
1.5 UNITY OF SPACETIME 15
This is the distance between the two events in the Earth frame, measured in
light-minutes. The Earth-frame time between the two events is 10 minutes, as
stated in the problem.
In the frame traveling with the rock, the two events occur at the same place; the
time between the two events in this frame equals the spacetime interval — the
proper time— between these events:
(interval)^ = (10 minutes)^ — (8.3169 minutes)^

= (100 - 69.1708) (minutes)^
= 30.8292 (minutes)^
The time between events in the rest frame of the rock equals the square root of
this, or 5.5524 minutes.
c. The distance between departure from Earth and arrival at Proxima Centauri is
4.3 light-years, as given in the problem. The starship moves at 95 percent light
speed, or 0.95 light-years/year. Therefore it takes a time 4.3 light-years/(0.95
light-years/year) = 4.53 years to arrive at Proxima Centauri, as measured in the
Earth frame.
Starship time between departure from Earth and arrival at Proxima Centauri
equals the interval:
(interval)^ = (4.53 years)^ ~ (4.3 years)^

= (20.52 - 18.49) (years)2
= 2.03 (years)^
The time between events in the rest frame of the starship equals the square root of
this, or 1.42 years. Compare with the value 4.53 years as measured in the Earth
frame. This example illustrates the famous idea that astronaut wristwatch time
— proper time — between two events is less than the time between these events
measured by any other observer in relative motion. Travel to stay young! This
result comes simply and naturally from the invariance of the interval.
1.5 UNITY OF SPACETIME

time and space: equal footing but distinct nature
When time and space are measured in the same unit— whether meter or second or
year— the expression for the square of the spacetime interval between two events
takes on a particularly simple form:
(interval)^ = (time separation)^ — (space separation)^

= ,2 _ „2 [same units for time and space]
This formula shows forth the unity of space and time. Impressed by this unity,
Spacetim e is a unity
Einstein’s teacher Hermann Minkowski (1864-1909) wrote his famous words,
“ Henceforth space by itself, and time by itself, are doomed to fade away into mere
shadows, and only a union of the two will preserve an independent reality. ’’Today this
union of space and time is called spacetime. Spacetime provides the tme theater for
PAYOFF OF THE PARABLE

from distance in space to interval in spacetime
DISCUSSION SURVEYING TOWNSHIP ANALYZING NATURE

Location marker Steel stake driven in ground Collision between two particles
Emission of flash from atom
Spark jumping from antenna to pen
General name for such a location Point or place Event
marker
Can its location be staked out for all Yes Yes
to see, independent of any scheme
of measurement, and independent
of all numbers?
Simple descriptor of separation Distance Spacetime interval
between two location markers
Are there ways directly to measure Yes Yes
this separation?
With enough markers already Specify distances from other Specify spacetime intervals from
staked out, how can we tell some points. other events.
one where we want the next one?
Instead of boldly staking out the By locating point relative to a refer- By locating event relative to a ref
new marker, or instead of position ence frame erence frame
ing it relative to existing markers,
how else can we place the new
marker?
Nature of this reference frame? Surveyor’s grid yields northward Lattice frame of rods and clocks
and eastward readings of point yields space and time readings of
(Chapter 1). event (Chapter 2).
Is such a reference frame unique? No No
How do two such reference frames Tilt of one surveyor’s grid relative Uniform velocity of one frame rela
d iffer from o n e a n o t h e r ? to the other tive to the other
What are names of two such possi Daytime grid: oriented to magnetic Laboratory frame
ble reference frames? north Rocket frame
Nighttime grid: oriented to North-
Star north
What common unit simplifies analy The unit meter for both northward The unit meter for both space and
sis of the results? and eastward readings time readings
What is the conversion factor from Converting miles to meters: Converting seconds to meters using
conventional units to meters? k= 1609.344 meters/mile the speed of light:
c = 299,792,458 meters/second
1.5 UNITY OF SPACETIME 17
DISCUSSION SURVEYING TOWNSHIP ANALYZING NATURE

For convenience, all measurements A common origin (center of town) A common event (reference spark)
are referred to what location?
How do readings for a single Individual northward and eastward Individual space and time readings
marker differ between two refer readings for one point — for one for one event — for one spark —
ence frames? steel stake — do not have the same do not have the same values re
values respectively for two survey spectively for two frames that are in
ors’ grids that are tilted relative to motion relative to one another.
one another.
When we change from one marker Subtract: Figure the difference be Subtract: Figure the difference be
to two, how do we specify the offset tween eastward readings of the tween space readings of the two
between them in reference-frame two points; also the difference in events; also the difference in time
language? northward readings. readings.
How to figure from offset readings Figure the distance between the Figure the spacetime interval be
a measure of separation that has two points. tween the two events.
the same value whatever the choice
of reference frame?
Figure how? (distance)^ = (interval)^ =
/ difference in V / difference in V
\northw ard readin g s/ \time read in g s/
_l_ / difference in V _ / difference in
\e a stw a rd reading s/ \sp a c e readin g s/
Result of this reckoning? Distance between points as figured Interval between events as figured
from readings using one surveyor’s from readings using one lattice-
grid is the same as figured from work frame is the same as figured
readings using a second surveyor’s from readings using a second
grid tilted with respect to first grid. frame in steady straight-line motion
relative to first frame.
Phrase to summarize this identity of Invariance of the distance between Invariance of the spacetime inter
separation as figured in two refer points val between events.
ence frames?
Conclusions from this analysis? (1) Northward and eastward di (1) Space and time dimensions are
mensions are part of a single entity: part of a single entity: spacetime.
space.
(2) Distance is the simple measure (2) Spacetime interval is the simple
of separation between two points, measure of separation between
natural because invariant: the same two events, natural because invar
for different surveyor grids. iant: the same for different refer
ence frames.
every event in the lives of stars, atoms, and people. Space is different for different
observers. Time is different for different observers. Spacetime is the same for everyone.
Minkowski’s insight is central to the understanding of the physical world. It focuses
attention on those quantities, such as spacetime interval, electrical charge, and particle
mass, that are the same for all observers in relative motion. It brings out the merely
relative character of quantities such as velocity, momentum, energy, separation in
time, and separation in space that depend on relative motion of observers.
Today we have learned not to overstate Minkowski’s argument. It is right to say
that time and space are inseparable parts of a larger unity. It is wrong to say that time is
Difference between
identical in quality with space.
time and space
Why is it wrong? Is not time measured in meters, just as space is? In relating the
positions of two steel stakes driven into the ground, does not the surveyor measure
northward and eastward separations, quantities of identical physical character? By
analogy, in locating two events is not the observer measuring quantities of the same
nature: space and time separations? How else could it be legitimate to treat these
quantities on an equal footing, as in the formula for the interval?
Equal footing, yes; same nature, no. There is a minus sign in the formula for the
interval squared = (time separation)^ — (space separation)^ that no sleight of hand
can ever conjure away. This minus sign distinguishes between space and time. No
twisting or turning can ever give the same sign to real space and time separations in
the expression for the interval.
The invariarxe of the spacetime interval evidences the unity of space and time while
also preserving — in the formula’s minus sign — the distinction between the two.
The principles of special relativity are remarkably simple— simpler than the
axioms of Euclidean geometry or the principles of operating an automobile. Yet both
Euclid and the automobile have been mastered— perhaps with insufficient surprise
— by generations of ordinary people. Some of the best minds of the twentieth century
stmggled with the concepts of relativity, not because nature is obscure, but because (1)
people find it difficult to outgrow established ways of looking at namre, and (2) the
world of the very fast described by relativity is so far from common experience that
everyday happenings are of limited help in developing an intuition for its descriptions.
By now we have won the battle to put relativity in understandable form. The
concepts of relativity can now be expressed simply enough to make it easy to think
correctly — “to make the bad difficult and the good easy. ’’ This leaves only the second
difficulty, that of developing intuition — a practiced way of seeing. W e understand
distance intuitively from everyday experience. Box 1.1 applies our intuition for
d istance in space to help our intuition for in terv a l in spacetim e.
To put so much into so little, to subsume all of Einstein’s teaching on light and
motion in the single word spacetime, is to cram a wealth of ideas into a small picnic
basket that we shall be unpacking throughout the remainder of this book.
REFERENCES
Introductory quote: Richard P. Feynman, The Character of Physical Law (MIT
Press, Cambridge, Mass., 1967), page 127.
Quote from Minkowski in Section 1.5: H. A. Minkowski, “Space and Time,” in
H. A. Lorentz et al., The Principle of Relativity (Dover Publications, New York,
1952), page 75.
ACKNOWLEDGMENTS 19
Quote at end of Section 1.5: “ to make the bad difficult and the good easy, ’’ ‘‘rend
le maldifficile et le bien facile.” Einstein, in a similar connection, in a letter to the
architect Le Corbusier. Private communication from Le Corbusier.
For an appreciation of Albert Einstein, see John Archibald Wheeler, “Albert
Einstein,” in The World Treasury of Physics, Astronomy, and Mathematics, Timo
thy Ferris, ed. (Little, Brown, New York, 1991), pages 5 6 3 -5 7 6 .
ACKNOWLEDGMENTS
Many students in many classes have read through sequential versions of this text,
shared with us their detailed difficulties, and given us advice. We asked students
to write down comments, perplexities, and questions as they read and turn in
these reading memos for personal response by the teacher. Italicized objections in
the text come, in part, from these commentators. Both we who write and you who
read are in their debt. Some readers not in classes have also been immensely
helpful; among these we especially acknowledge Steven Bartlett. No one could
have read the chapters more meticulously than Eric Sheldon, whose wide knowl
edge has enriched and clarified the presentation. William A. Shurcliff has been
immensely inventive in devising new ways of viewing the consequences of relativ
ity, a few of these are specifically acknowledged in later chapters. Electronic-mail
courses using this text brought a flood of comments and reading memos from
teachers and students around the world. Richard C. Smith originated, organized,
and administered these courses, for which we are very grateful. The clarity and
simplicity of both the English and the physics were improved by Penny Hull.
Some passages in this text, both brief and extended, have been adapted from
the book A Journey into Gravity and Spacetime by John Archibald Wheeler
(W .H. Freeman, New York, 1990). In turn, certain passages in that book were
adapted from earlier drafts of the present text. W e have also used passages, logical
arguments, and figures from the book Gravitation by Charles W . Misner, Kip S.
Thorne, and John Archibald Wheeler (W. H. Freeman, New York, 1973).
INTRODUCTION TO THE EXERCISES

Important areas of current research can be analyzed exercises and problems in this text evoke a wide range
very simply using the theory of relativity. This analy of physical consequences of the properties of space-
sis depends heavily on a physical intuition, which time. These properties of spacetime recur here over
develops with experience. Wide experience is not easy and over again in different contexts:
to obtain in the laboratory— simple experiments in
• paradoxes
relativity are difficult and expensive because the speed
of light is so great. As alternatives to experiments, the • puzzles
20 EXERCISE 1-1 COMPARING SPEEDS
• derivations than five lines to write down. On the other hand, the
exercises require some “ruminarion time.”
• technical applications
In some chapters, exercises are divided inro rwo
• experimental results categories. Practice and Problems. The Practice exer
cises help you to get used to ideas in the text. The
• estimates Problems apply these ideas to physical systems,
• precise calculations thought experiments, and paradoxes.
w heeler ’s first moral principle : Never make
• philosophical difficulties
a calculation until you know the answer. Make an
The text presents all formal tools necessary to solve estimate before every calculation, try a simple physical
these exercises and problems, but intuition — a prac argument (symmetry! invariance! conservation!) be
ticed way of seeing — is best developed without fore every derivation, guess the answer to every para
hurry. For this reason we suggest continuing to do dox and puzzle. Courage: No one else needs to know
more and more of these exercises in relativity after you what the guess is. Therefore make ir quickly, by
have moved on to material outside this book. The instinct. A right guess reinforces this instinct. A wrong
mathematical manipulations in the exercises and guess brings the refreshment of surprise. In eirher case
problems are very brief: only a few answers take more life as a spacetime expert, however long, is more fun!
CHAPTE3R 1 EXERCISES
1-2 images from Neptune

PRACTICE
At 9:00 P.M. Pacific Daylight Time on August 24,
1-1 comparing speeds 1989, the planetary probe Voyager 11 passed by the
planet Neptune. Images of the planet were coded and
Compare the speeds of an automobile, a jet plane, an
rransmitted to Earth by microwave relay.
Earth satellite. Earth in its orbit around Sun, and a
It took 4 hours and 6 minutes for this microwave
pulse of light. Do this by comparing the relative
signal to travel from Neptune to Earth. Microwaves
distance each travels in a fixed tim e. Arbitrarily
(electromagneric radiation, like light, but of fre
choose the fixed time to give convenient distances. A
quency lower than that of visible light), when propa
car driving at the USA speed limit of 65 miles/hour
gating through interplanetary space, move at the
(105 kilometers/hour) covers 1 meter of distance in
‘‘standard ’’ light speed of one meter of distance in one
about 35 milliseconds = 35 X 10“ ^ second.
meter of light-travel time, or 299,792,458 meters/
a How far does a commercial jetliner go in 35
second. In the following, neglect any relative motion
milliseconds? (speed: 650 miles/hour = 1046
among Earth, Neptune, and Voyager 11.
kilometers/hour)
a Calculate the distance between Earth and
b How far does an Earth satellite go in 3 5 milli
Neptune at fly-by in units of minutes, seconds, years,
seconds? (speed: 17,000 miles/hour ~ 27,350
meters, and kilometers.
kilometers/hour)
b Calculate the time the microwave signal takes
C How far does Earth travel in its orbit around
to reach Earth. Use the same units as in part a.
Sun in 35 milliseconds? (speed: 30 kilometers/se-
cond)
d How far does a light pulse go in a vacuum in
1-3 units of spacetime
35 milliseconds? (speed: 3 X 10® meters /second). Light moves at a speed of 3.0 X 10® meters/second.
This distance is roughly how many times the distance One mile is approximately equal to 1600 meters.
from Boston to San Erancisco (5000 kilometers)? One furlong is approximately equal to 200 meters.
EXERCISE 1-6 MAPMAKING IN SPACE 21
a How many meters of time in one day? second line of the table. And so on. Complete the
b How many seconds of distance in one mile? table.
c How many hours of distance in one furlong?
d How many weeks of distance in one light-year?
e How many furlongs of time in one hour?
1-5 where and when?
Two firecrackers explode at the same place in the
1>4 time stretching and the laboratory and are separated by a time of 3 years as
spacetime interval measured on a laboratory clock.
a W hat is the spatial distance between these two
A rocket clock emits two flashes of light and the
events in a rocket in which the events are separated in
rocket observer records the time lapse (in seconds)
time by 5 years as measured on rocket clocks?
between these two flashes. The laboratory observer
b W hat is the relative speed of the rocket and
records the time separation (in seconds) and space
laboratory frames?
separation (in light-seconds) between the same pair of
flashes. The results for both laboratory and rocket
observers are recorded in the first line of the table. 1-6 mapmaking in space
Now a clock in a different rocket, moving at a
The table shows distances between cities. The units
different speed with respect to the laboratory, emits a
are kilometers. Assume all cities lie on the same flat
different pair of flashes. The set of laboratory and
plane.
rocket space and time separations are recorded on the
a Use a ruler and a compass (the kind of compass
that makes circles) to construct a map of these cities.
■<^[^[e X E R C IS E Choose a convenient scale, such as one centimeter on
the map corresponds to ten kilometers on Earth.
SPACE AND TIME SEPARATIONS D iscussion: How to start? W ith three arbitrary
decisions! (1) Choose any city to be at the center of the
Rocket Laboratory Laboratory
map. (2) Choose any second city to be “due north”
time lapse time lapse distance
(seconds) (seconds) (light-seconds) — that is, along any arbitrary direction you select. (3)
Even with these choices, there are two places you can
Example 20 29 21 locate the third city; choose either of these two places
arbitrarily.
a > 10.72 5.95 b If you rotate the completed map in its own
b 20 99 plane — for example, turning it while keeping it flat
c 66.8 72.9 p on the table— does the resulting map also satisfy the
d ? 8.34 6.58 distance entries above?
e 21 22 ? C Hold up your map between you and a light,
with the marks on the side of the paper facing rhe
C ^ ^ ^ X E R C IS E -----
DISTANCES BETWEEN CITIES

Distance
to city A B C D E F G H
from city
A 0 20.0 28.3 28.3 28.3 20.0 28.3 44.7
B 0 20.0 20.0 44.7 40.0 44.7 40.0
C 0 40.0 40.0 44.7 56.6 60.0
D 0 56.6 44.7 40.0 20.0
E 0 20.0 40.0 72.1
F 0 20.0 56.6
G 0 44.7
H 0
22 EXERCISE 1-7 SPACETIME MAP
light. Does the map you see from the back also satisfy instruction requires transmission of data from the
the table entries? memory (where data is stored) to the processor (where
D iscussion: In this exercise you use a table con the computation is carried out) and transmission of
sisting only of distances between pairs of cities to the result back to the memory for storage.
construct a map of these cities from the point of view a W hat is the maximum average distance be
of a surveyor using a given direction for north. In tween memory and processor in a “one-megaflop”
Exercise 5-3 you use a table consisting only of space- computer? Is this maximum distance increased or
time intervals between pairs of events to draw a decreased if the signal travels through conductors at
“spacetime m ap” of these events from the point of one half the speed of light in a vacuum?
view of one free-float observer. Exercise 1-7 previews b Computers are now becoming available that
this kind of spacetime map. operate at “one gigaflop,” that is, they carry out 10^
sequential instructions per second. W hat is the maxi
mum average distance between memory and proces
1-7 spacetime map sor in a “one-gigaflop” machine?
The laboratory space and time measurements of c Estimate the overall maximum size of a “one-
events 1 through 5 are plotted in the figure. Compute teraflop” machine, that is, a computer that can carry
the value of the spacetime interval out 10*^ sequential instructions per second.
a between event 1 and event 2. d D iscussion question: In contrast with most
b between event 1 and event 3. current personal computers, a “parallel processing”
c between event 1 and event 4. computer contains several or many processors that
d between event 1 and event 5. work together on a computing task. One might think
e A rocket moves with constant velocity from that a machine with 10,000 processors would com
event 1 to event 2. That is, events 1 and 2 occur at the plete a given computation task in 1/10,000 the time.
same place in this rocket frame. W hat time lapse is However, many computational problems cannot be
recotded on the rocket clock between these two divided up in this way, and in any case some fraction
events? of the computing capacity must be devoted to coordi
nating the team of processors. W hat limits on physi
cal size does the speed of light impose on a parallel
event processing computer?
2
event
4 1-9 trips to Andromeda by
event '3
rocket
t
time
event
The Andromeda galaxy is approximately two million
light-years distant from Earth as measured in the
(meters)
event Earth-linked frame. Is it possible for you to travel
1, from Earth to Andromeda in your lifetime? Sneak up
on the answer to this question by considering a series
of trips from Earth to Andromeda, each one faster
than the one before. For simplicity, assume the Earth-
0 1 2 3 4 5 6
Andromeda distance to be exactly two million light-
-----space (meters) — ►
years in the Earth frame, treat Earth and Andromeda
EXERCISE 1-7. Spacetime map of some events. as points, and neglect any relative motion between
Earth and Andromeda.
a TRIP 1. Your one-way trip takes a time 2 .0 1 X
10^ years (measured in the Earth-linked frame) to
PROBLEMS cover the distance of 2.00 X 10^ light-years. How
long does the trip last as measured in your rocket
frame?
1-8 size off a computer b W hat is your rocket speed on Trip 1 as mea
In one second some desktop computers can carry out sured in the Earth-linked frame? Express this speed as
one million instructions in sequence: One instruction a decimal fraction of the speed of light. Call this
might be, for instance, multiplying two numbers to fraction, p = where is speed in conven
gether. In technical jargon, such a computer operates tional units, such as meters/second. D iscussion: If
at “one megaflop.” Assume that carrying out one your rocket moves at half the speed of light, it takes
EXERCISE 1-11 TIME STRETCHING WITH /i-MESONS 23
4 X 10^ years to cover the distance 2 X 10^ light- in the common rest frame of Transporter and Re
years. In this case ceiver.
a How much does Samantha age during her
2 X 10^ light-years 1 outward trip to Zircon?
4 X 10^ years b Samantha collects samples and makes obser
vations of the Zirconian civilization for one Earth-
Therefore . . . year, then beams back to Earth. How much has Sa
c TRIP 2. Your one-way Earth-Andromeda trip mantha aged during her entire trip?
takes 2.001 X 10® years as measured in the Earth- C How much older is Earth and its civilization
linked frame. How long does the trip last as measured when Samantha returns?
in your rocket frame? W hat is your rocket speed for d Earth has been taken over by a tyrant, who
Trip 2, expressed as a decimal fraction of the speed of wishes to invade Zircon. He sends one warrior and has
light? him duplicated into attack battalions at the Receiver
d TRIP 3. Now set the rocket time for the one end. How long will the Earth tyrant have to wait to
way trip to 20 years, which is all the time you want to discover whether his ambition has been satisfied?
spend getting to Andromeda. In this case, what is e A second transporternaut is beamed to a much
your speed as a decimal fraction of the speed of light? more remote galaxy that is moving away from Earth
D iscussion: Solutions to many exercises in this text at 87 percent of the speed of light. This time, too, the
are simplified by using the following approximation, traveler stays in the remote galaxy for one year as
which is the first two terms in the binomial expansion measured by clocks moving with the galaxy before re
turning to Earth by Transporter. How much has the
(1 -b z)” ~ 1 + nz l « 1 transporternaut aged when she arrives back at Earth?
(Careful!)
Here n can be positive or negative, a fraction or an
integer; z can be positive or negative, as long as its
magnitude is very much smaller than unity. This 1-11 time stretching with
approximation can be used twice in the solution to muons
part d.
At heights of 10 to 60 kilometers above Earth, cosmic
rays continually strike nuclei of oxygen and nitrogen
atoms and produce muons (muons: elementary parti
1-10 trip to Andromeda by cles of mass equal to 207 electron masses produced in
Transporter some nuclear reactions). Some of the muons move
In the Star Trek series a so-called Transporter is used vertically downward with a speed nearly that of light.
to “beam” people and their equipment from a star- Follow one of the muons on its way down. In a given
ship to the surface of nearby planets and back. The sample of muons, half of them decay to other ele
Transporter mechanism is not explained, but it ap mentary particles in 1.5 microseconds (1.5 X 10~®
pears to work only locally. (If it could transport to seconds), measured with respect to a reference frame
remote locations, why bother with the starship at all?) in which they are at rest. Half of the remainder decay
Assume that one thousand years from now a Trans in the next 1.5 microseconds, and so on. Analyze the
porter exists that reduces people and things to data results of this decay as observed in two different
(elementary bits of information) and transmits the frames. Idealize the rather complicated acmal experi
data by light or radio signal to remote locations. There ment to the following roughly equivalent situation:
a Receiver uses the data to reassemble travelers and All the muons are produced at the same height (60
their equipment out of local raw materials. kilometers); all have the same speed; all travel straight
One of your descendants, named Samantha, is the down; none are lost to collisions with air molecules on
first “transporternaut” to be beamed from Earth to the way down.
the planet Zircon orbiting a star in the Andromeda a Approximately how long a time will it take
Nebula, two million light-years from Earth. Neglect these muons to reach the surface of Earth, as mea
any relative motion between Earth and Zircon, and sured in the Earth frame?
assume: (1) transmission produces a Samantha iden b If the decay time were the same for Earth
tical to the original in every respect (except that she is observers as for an observer traveling with the muons,
2 million light-years from home!), and (2) the time approximately how many half-lives would have
required for disassembling Samantha on Earth and passed? Therefore what fraction of those created at a
reassembling her on Zircon is negligible as measured height of 60 kilometers would remain when they
24 EXERCISE 1-12 TIME STRETCHING WITH 7T+-MES0NS
reached sea level on Earth? You may express your

answer as a power of the fraction 1/2.
c An experiment determines that the fraction TIME STRETCHING WITH 7T+-MES0NS
1/ 8 of the muons reaches sea level. Call the rest frame
of the muons the rocket frame. In this rocket frame, "Characteristic distance”
how many half-lives have passed between creation of Time for half to (speed of light
decay (measured multiplied by
a given muon and its arrival as a survivor at sea level? Particle in rest frame) foregoing time)
d In the rocketframe, what is the space separation
between birth of a survivor muon and its arrival at the muon 1.5 X 10"^ second 450 meters
surface of Earth? (Careful!) (207 times
e From the rocket space and time separations, electron mass)
find the value of the spacetime interval between the TT^-meson 18 X 10“* second 5.4 meters
birth event and the arrival event for a single surviving (273 times
electron mass)
muon.
Reference: Nalini Easwar and Douglas A. Macintire, American Jour
nal of Physics, Volume 59, pages 5 8 9 - 5 9 2 (July 1991).
target inside the accelerator. Mesons leave this target

1-12 time stretching with with nearly the speed of light. If there were no time
TT^-mesons stretching and if no mesons were removed from the
Laboratory experiments on particle decay are much resulting beam by collisions, what would be the
more conveniently done with TT^-mesons (pi-plus greatest distance from the target at which half of
mesons) than with //-mesons, as is seen in the table. the mesons would remain undecayed?
In a given sample of TT^-mesons half will decay to b The TT^-mesons of interest in a particular ex
other elementary particles in 18 nanoseconds (18 X periment have a speed 0.9978 that of light. By what
10“^ seconds) measured in a reference frame in which factor is the predicted distance from the target for
the TT^-mesons are at rest. Half of the remainder will half-decay increased by time dilation over the
decay in the next 18 nanoseconds, and so on. previous prediction — that is, by what factor does this
a In a particle accelerator TT^-mesons are pro dilation effect allow one to increase the separation
duced when a proton beam strikes an aluminum between the detecting equipment and target?
CHAPTER 3
<• __ ,- _■_. »-
FLOATING FREE
A t that moment there came to me the happiest
thought of my life . . . for an observer falling freely
from the roof of a house no gravitational field exists
during his fa ll . . .
A lb ert Einstein
2.1 FLOATING TO MOON

will the astronaut stand on the floor— or float?
Less than a month after the surrender at Appomattox ended the American Civil War
(1 8 6 1 -1 8 6 5 ), theFrenchauthorJulesV erne began writing A Trip From the Earth to
the Moon and A Trip Around the Moon. Eminent American cannon designers, so the
story goes, cast a great cannon in a pit, with cannon muzzle pointing skyward. From
this cannon they fire a ten-ton projectile containing three men and several animals
(Figure 2-1).
As the projectile coasts outward in unpowered flight toward Moon, Verne says, irs
Jules Verne:
passengers walk normally inside the projectile on the end nearer Earth (Figure 2-2). As
Passenger stands on floor
the trip continues, passengers find themselves pressed less and less against the floor of
the spaceship until finally, at the point where Earth and Moon exert equal but opposite
gravitational attraction, passengers float free of the floor. Later, as the ship nears
Moon, they walk around once again — according to Verne — but now against the end
of the spaceship nearer Moon.
Early in the coasting portion of the trip a dog on the ship dies from injuries susrained
at takeoff. Passengers dispose of its remains through a door in the spaceship, only to
find the body floating outside the window during the entire trip (Figure 2-1).
This story leads to a paradox whose resolution is of crucial importance to relativity.
Parad ox of p a ssen g er and dog
Verne thought it reasonable that Earth’s gravitational attraction would keep a passen
ger pressed against the Earth end of the spaceship during the early part of the trip. He
also thought it reasonable rhat the dog should remain next to the ship, since both ship
and dog independently follow the same path through space. But since the dog floats
outside the spaceship during the entire trip, why doesn’t the passenger float around
inside the spaceship? If the ship were sawed in half would the passenger, now
"outside,” float free of the floor?
25
26 CHAPTER 2 FLOATING FREE
AN INCORRECT PREDICTION THE CORRECT PREDICTION
Point of equal
gravitational ^
attrartion A
> TIIK m>l)Y OK SATKI.I.ITE.
FIGURE 2 -1 . I l l u s t r a t i o n f r o m an FIGURE 2 -2 . I n c o r r e c t p r e d ic tio n : J u le s Verne believed th a t a passenger insid e a fre e projectile w o u ld

e a r ly e d i t i o n o f K T r i p A r o u n d the s ta n d a g a in st the end o f the projectile nearest E a rth or M oon, w hichever h a d greater g r a v ita tio n a l
M o o n . S a te llite is the nam e o f the un fo r a ttra c tio n — h u t th a t the dog w o u ld jio a t along beside the projectile f o r the entire trip . C o r r e c t p r e d ic tio n :
tu n a te dog. Verne w a s rig h t a bout the dog, b u t a passenger also flo a ts w ith respect to the fre e projectile d u rin g the entire
trip .
Our experience with actual space flights enables us to resolve this paradox (Figure
2-2). Jules Verne was wrong about the passenger’s motion inside the unpowered
Reality;
spaceship. Like the dog outside, the passenger inside independently follows the same
P a s s e n g e r f lo a t s in s p a c e s h ip
path through space as the spaceship itself. Therefore he floats freely relative to the ship
during the entire trip (after the initial boost inside the cannon barrel). True: Earth’s
gravity acts on the passenger. But it also acts on the spaceship. In fact, with respect to
Earth, gravitational acceleration of the spaceship just equals gravitational acceleration
of the passenger. Because of this equality, there is no relative acceleration between
passenger and spaceship. Thus the spaceship serves as a referen ce fram e relative to
which the passenger does not experience any acceleration.
To say that acceleration of the passenger relative to the unpowered spaceship equals
zero is not to say that his velocity relative to it necessarily also equals zero. He may jump
from the floor or spring from the side— in which case he hurtles across the spaceship
and strikes the opposite wall. However, when he floats with zero initial velocity
relative to the ship the situation is particularly interesting, for he will also float with
zero velocity relative to it at all later times. He and the ship follow identical paths
through space. How remarkable that the passenger, who cannot see outside, never
theless moves on this deterministic orbit! W ithout a way to control his motion and
even with his eyes closed he will not touch the wall. How could one do better at
eliminating detectable gravitational influences?
2.2 THE INERTIAL (FREE-FLOAT) FRAME

goodbye to the "force of gravity"
It is easy to talk about the simplicity of motion in a spaceship. It is hard to think of
conditions being equally simple on the surface of Earth (Figure 2-3). The reason for
2.2 THE INERTIAL (FREE-FLOAT) FRAME 27
recovery release mechanism

apparatus thruster nozzle
release
apparatus
■thruster gas tanks
inner cylinder
release mechanism
capsule
■inner cylinder
for experiments
mechanical
braking shock absorbers
shock-
data transmission
and batteries
FIGURE 2 -3 . T h e J a p a n M i c r o g r a v ity C e n te r ( J A M I C ) i n s t a l l e d i n a n a b a n d o n e d c o a l m i n e
7 1 0 m e te r s d e e p i n t h e s m a l l t o w n o f K a m i s u n a g a w a o n t h e n o r th e r n i s l a n d o f H o k k a id o ,
J a p a n . The capsule carrying the exp erim en ta l a p p a ra tu s provides a fre e-flo a t fr a m e fo r 10 seconds as i t fa lls
4 9 0 meters through a vertic a l tube, a ch ievin g a m a x im u m velocity o f nearly 1 0 0 m etersj second. I t is g u id e d
by tw o contact-free m a gnetic suspensions along the tube. The vertic a l tube is not evacuated; d o w n w a rd -
th ru s tin g g a s je ts on the capsule compensate f o r a i r d ra g a s the capsule drops. The capsule is slow ed dow n in
a n a d d itio n a l d ista n ce o f 2 0 0 meters near the bottom o f the tube by a ir resistance a fte r thrusters are tu rn e d
off, fo llo w ed by m echanical bra kin g . T w e n ty meters o f cushioning m a te r ia l a t the very bottom o f the tube
provide emergency stopping. The fa llin g capsule is nearly 8 meters long a n d nearly 2 meters in d ia m e te r w ith
a m ass o f 5 0 0 0 kilogram s, in c lu d in g 1 0 0 0 kilogram s o f exp erim en ta l equipm ent contained in a n inn er
cylin d er 1 .3 meters in d ia m eter a n d 1 .8 m eters long. The space between capsule a n d experim en ta l cylinder is
evacuated. T he in n e r exp erim en ta l cylin d er is released j u s t before the outer capsule itself. O ptica l m onitoring
o f the ve rtic a l p o sition o f the in n er cylin d er triggers d o w n w a rd -p u sh in g thrusters a s needed to overcome a ir
resistance. T h u s the exp erim en ta l cylin d er i ts e lf a cts a s a n in te rn a l “conscience," ensuring th a t the capsule
takes the sam e course th a t i t w o u ld have ta k e n h a d both resistance a n d th ru st been absent. The result? A
nearly fre e-flo a t fra m e , w ith a m a x im u m acceleration o f 1 .0 54 1 0 ~ ^ g in the experim en ta l capsule, where
g is the acceleration o f g ra v ity a t E a rth ’s surface. E xperim ents carried out in th is f a c ility benefit fro m
conditions o f “no a ir pressure, no h ea t convection, no flo a tin g or s in k in g buoyancy, no resistance to m o tio n ,"
a s w e ll a s m uch low er cost a n d less enviro n m en ta l dam age th a n those involved in la u n c h in g a n d m onitoring
a n E a rth sa tellite. The f a c ility is designed to carry o ut 4 0 0 drops p er year, w ith experim ents such a s fo rm in g
large superconducting crystals, crea tin g alloys o f m a teria ls th a t do not norm ally m ix , stu d y in g transitio n s
between g a s a n d liq u id phases, a n d b u rn in g u n d er zero-g. (See also Figure 9 -2 .)
FIGURE 2 -4 . I l lu s i o n a n d R e a l it y , The sam e b a ll throw n fro m the sam e com er o f the sam e room in the
sam e direction w ith the sam e speed is seen to undergo very d ifferent m otions depending on w hether i t is
recorded by a n observer w ith a floor p u sh in g up a g a in st h is fe e t or by a n observer in “fre e f a l l " ( “fre e f l o a t ”)
in a house saw ed fre e fro m the cliff. In both descriptions the b a ll arrives a t the sam e p la c e — rela tive to
M o th er E a r th — a t the sam e in sta n t. L et each b a ll s q u ir t a j e t o f in k on the w a ll w e are looking a t. The
resulting record is as crisp fo r the arc as fo r the stra ig h t line. Is the arc real a n d the str a ig h t lin e illu sio n ? Or
is the stra ig h t line real a n d the arc illusion? E in stein tells us th a t the tw o in k tra ils are equally v a lid . W e
ha ve only to be honest a n d say w hether the house, the w a ll, a n d the describes o f the m otion are in fre e flo a t or
w hether the describer is co n tin u a lly being d riv en a w a y fro m a condition o f free flo a t by a p u sh a g a in st his
fe e t. E in stein also tells us th a t physics a lw a y s looks sim plest in a free-flo a t fra m e . F inally, he tells us th a t
every tru ly local m a n ifesta tio n o f “g r a v ity " can be elim in a te d by observing m otion fro m a fr a m e o f reference
th a t is in fre e flo a t.
concern is not far to seek. We experience it every day, every minute, every second. We
call it gravity. It shows in the arc of a ball tossed across the room (Figure 2-4, left).
How can anyone confront a mathematical curve like that arc and not be trapped again
in that tortuous trail of thought that led from ancient Greeks to Galileo to Newton?
They thought of gravity as a force acting through space, as something mysterious, as
something that had to be “explained.”
Einstein put forward a revolutionary new idea. Eliminate gravity!
Where lies the cause of the curved path of the ball? Is it the ball? Is it some
mysterious “force of gravity”? Neither, Einstein tells us. It is the fault of the viewers
— and the fault of the floor that forces us away from the natural state of motion: the
Concept of free-float frame
state of free fall, or better put, free float. Remove the floor and our motion
immediately becomes natural, effortless, free from gravitational effects.
Let the room be cut loose at the moment we throw the ball slantwise upward from
rhe west side at floor level (Figure 2-4, right). The ball has the same motion as it did
before. However, the motion looks different. It looks different because we who look at
it are in a different frame of reference. We are in a free-float fram e. In this free-float
frame the ball has straight-line motion. W hat could be simpler?
Even when the room was not cut away from the cliff, the floor did not affect the
midair flight of the ball. But the floor did affect us who watched the flight. The floor
forced us away from our natural motion, the motion of free fall (free float). We
blamed the curved path of the ball on the “force of gravity” acting on the ball. Instead
we should have blamed the floor for its force acting on us. Better yet, get rid of the floor
by cutting the house away from the cliff. Then our point of view becomes the natural
one: We enter a free-float frame. In our free-float frame the ball flies straight.
2.2 THE INERTIAL (FREE-FLOAT) FRAME 29
W hat’s the fault of the force on my feet?

What pushes my feet down on the floor?
Says Newton, the fa u lt’s at Earth’s core.
Einstein says, the fa u lt’s with the floor;
Remove that and gravity’s beat!
— Frances Towne Ruml
How could humankind have lived so many centuries without realizing that the
“arc” is an unnecessary distraction, that the idea of local “gravity” is superfluous —
the fault of the observer for not arranging to look at matters from a condition of free
float?
Even today we recoil instinctively from the experience of free float. W e and a
companion ride in the falling room, which does not crash on the ground but drops into
a long vertical tunnel dug for that purpose along the north-south axis of Earth. Our
companion is so filled with consternation that he takes no interest in our experimental
findings about free float. He grips the door jamb in terror. “W e’re falling!” he cries
out. His fear turns to astonishment when we tell him not to worry.
“A shaft has been sunk through Earth,” we tell him. “ It’s not the fall that hurts
anyone but what stops the fall. All obstacles have been removed from our way,
including air. Free fall,” we assure him, “is the safest condition there is. That’s why we
call it free float.”
“You may call it float,” he says, “but I still call it fall.”
“Right now that way of speaking may seem reasonable,” we reply, “but after we
pass the center of Earth and start approaching the opposite surface, won’t the word
Free-float through Earth
‘fair seem rather out of place? Might you not then prefer the word ‘float’?” And with
“ float” our companion at last is happy.
W hat do we both see? Weightlessness. Free float. Motion in a straight line and at
uniform speed for marbles, pennies, keys, and balls in free motion in any direction
within our traveling home. No jolts. No shudders. No shakes at any point in all the
long journey from one side of Earth to the other.
For our ancestors, travel into space was a dream beyond realization. Equally beyond
our reach today is the dream of a house floating along a tunnel through Earth, but this
dream nonetheless illuminates the simplicity of motion in a free-float frame. Given the
necessary conditions, nothing that we observe inside our traveling room gives us the
slightest possibility of discriminating among different free-float frames: one just above
Earth’s surface, a second passing through Earth’s interior, a third in the uttermost
reaches of space. Floating inside any of them we find no evidence whatever for the
presence of “gravity.”
W a i t a m i n u t e ! I f th e id e a o f lo c a l “g r a v i t y ” is u n n e c e s sa r y , w h y does m y p e n c i l b e g in
to f a l l w h e n I h o ld i t in th e a i r a n d le t g o ? I f th e re is no g r a v i ty , m y p e n c i l s h o u ld r e m a in
a t rest.
And so it does remain at test — as observed from a free-float frame! The natural
motion of your pencil is to remain at rest or to move with constant velocity in a
free-float frame. So it is not helpful to ask: “Why does the pencil begin to fall when I
let go? ”A more helpful question: “Before I let go, why must I apply an upward force
to keep the pencil at rest?” Answer: Because you are making observations from an
unnatural frame: one held fixed at the surface of Earth. Remove that fixed hold by
dropping your room off a cliff. Then for you “gravity” disappears. For you, no force
is required to keep the pencil at rest in your free-float frame.
2.3 LOCAL CHARACTER OF FREE-FLOAT

FRAME
tidal effects intrude in larger domains
First to strike us about the concept free float has been its paradoxical character. As a
first step to explaining gravity Einstein got rid of gravity. There is no evidence of
gravity in the freely falling house.
Well, almost no evidence. The second feature of free float is its local character.
Riding in a very small spaceship (Figure 2-5, left) we find no evidence of gravity. But
the enclosure in which we ride— falling near Earth or plunging through Earth —
cannot be too large or fall for too long a time without some unavoidable relative
changes in motion being detected between particles in the enclosure. Why? Because
widely separated particles within a large enclosed space are differently affected by the
Earth’s pull nonuniform:
nonuniform gravitational field of Earth, to use the Newtonian way of speaking. For
Large spaceship
not a free-float frame example, two particles released side by side are both attracted toward the center of
Earth, so they move closer together as measured inside a falling long narrow horizontal
railway coach (Figure 2-5, center). This has nothing to do with “gravitational attrac
tion’’ between the particles, which is entirely negligible.
As another example, think of two particles released far apart vertically but directly
above one another in a long narrow vertical falling railway coach (Figure 2-5, right).
This time their gravitational accelerations toward Earth are in the same direction.
FIGURE 2 -5 . T h r e e v e h ic le s i n f r e e f a l l n e a r E a r th : s m a l l sp a c e c a p s u le , E i n s t e i n ’s o l d - fa
s h io n e d r a i l w a y c o a ch i n f r e e f a l l i n a h o riz o n ta l o r ie n ta tio n , a n d a n o th e r r a i l w a y c o a c h i n
v e r tic a l o r ie n ta tio n .
2.3 LOCAL CHARACTER OF FREE-FLOAT FRAME 31
according to the Newtonian analysis. However, the particle nearer Earth is more
strongly attracted to Earth and slowly leaves the other behind: the two particles move
farther apart as the coach falls. Conclusion: the latge enclosure is not a free-float frame.
Free-float frame is local
Even a small room fails to qualify as free-float when we sample it over a long
enough time. In the 42 minutes it takes our small room to fall through the tunnel from
North Pole to South Pole, we notice relative motion between test particles teleased
initially from rest at opposite sides of the room.
Now, we want the laws of motion to look simple in our floating room. Therefore we
want to eliminate all relative accelerations produced by external causes. “Eliminate”
means to reduce these accelerations below the limit of detection so that they do not
interfere with more important accelerations we wish to study, such as those produced
when two particles collide. We eliminate the problem by choosing a room that is
sufficiently small. Smaller room? Smaller relative accelerations of objects at different
points in the room!
Let someone have instruments for detection of relative accelerations with any given
degree of sensitivity. No matter how fine that sensitivity, the room can always be made
so small that these perturbing relative accelerations ate too small to be detectable.
Within these limits of sensitivity our room is a free-float frame. “Official” names for
such a frame are the in ertial reference fram e and the L orentz referen ce fram e.
Here, however, we often use the name free-float fram e, which we find more
descriptive. These are all names for the same thing.
Free-float (inertial) frame
A reference frame is said to be an “inertial” or “free-float” or “Lorentz”
formally defined
reference frame in a certain region of space and time when, throughout that
region of spacetime — and within some specified accuracy — every free test
particle initially at rest with respect to that frame remains at rest, and every
free test particle initially in motion with respect to that frame continues its
motion without change in speed or in direction.
Wonder of wonders! This test can be carried out entirely within the free-float frame.
The observer need not look out of the room or refer to any measurements made
external to the room. A free-float frame is “local” in the sense that it is limited in space
and time — and also ‘‘local ’’ in the sense that its free-float character can be determined
from within, locally.
Sir Isaac Newton stated his First Law of Motion this way: “ Every body perseveres in
its state of rest, or of uniform motion in a right [straight} line, unless it is compelled to
change that state by forces impressed upon it.” For Newron, in ertia was a property of
objects that described their tendency to maintain their state of motion, whether of rest
or constant velocity. For him, objects obeyed the “Law of Inertia.” Here we have
turned the “ Law of Inertia” around: Before we certify a reference frame to be inertial,
we require observers in that frame to demonstrate that every free particle maintains its
initial state of motion or rest. Then Newton’s First Law of Motion defines a reference
frame— an arena or playing field — in which one can study the motion of objects and
draft the laws of their motion.
W h e n is th e room , th e sp a c e sh ip , o r a n y o th e r v e h ic le s m a ll e n o u g h to be c a lle d a lo c a l
fr e e - flo a t f r a m e ? O r w h e n is th e r e la tiv e a c c e le r a tio n o f tw o f r e e p a r tic le s p l a c e d a t
o p p o site e n d s o f th e v e h ic le too s lig h t to he d e te c te d ?
“Local” is a tricky word. For example, drop the old-fashioned 20-meter-long

railway coach in a horizontal orientation from rest at a height of 315 meters onto the
surface of Earth (Figure 2-5, center). Time from release to impact equals 8 seconds,
or 2400 million meters of light-travel time. At the same instant you drop the coach,
release tiny ball bearings from rest — and in midair — at opposite ends of the coach.
THE TIDE-DRIVING POWER
OF MOON AND SUN
N o t e ; N e it h e r a s t r o n o m e r s n o r n e w s p a p e r s s a y “ th e V e n u s ” o r “ th e M a r s . " A ll
s a y sim p ly “ V e n u s " o r “ M a r s . " Astronomers fo llo w th e s a m e s n a p p y p r a c tic e
f o r E a rth , Moon, a n d Sun. More a n d m o re o f th e r e s t o f th e w o r ld n o w fo llo w s —
os do w e in this b o o k — th e re c o m m e n d a tio n s o f th e In te rn a tio n a l A s tro n o m ic a l
U n io n .
The ocean’s rise and fall in a never-ending rhythmic cycle bears witness to
the tide-driving power of Moon and Sun. In principle those influences are no
different from those that cause relative motion of free particles in the vicinity
of Earth. In a free-float frame near Earth, particles separated vertically in
c r e a s e their separation with time; particles separated horizontally decrease
their separation with time (Figure 2-5). More generally, a thin spattering of
free-float test masses, spherical in pattern, gradually becomes egg-shaped,
with the long axis vertical. Test masses nearer Earth, more strongly attracted
than the average, move downward to form the lower bulge. Similarly, test
masses farther from Earth, less strongly attracted than the average, lag be
hind to form the upper bulge.
By like action Moon, acting on the waters of Earth — floating free in space
— would draw them out into an egg-shaped pattern if there were water
everywhere, water of uniform depth. There isn’t. The narrow Straits of G i
braltar almost cut off the Mediterranean from the open ocean, and almost kill
all tides in it. Therefore it is no wonder that Galileo Galilei, although a great
pioneer in the study of gravity, did not take the tides as seriously as the more
widely traveled Johannes Kepler, an expert on the motion of Moon and the
planets. Of Kepler, Galileo even said, “ More than other people he was a
person of independent genius . . . [but he] later pricked up his ears and
became interested in the action of the moon on the water, and in other occult
phenomena, and similar childishness.’’
Foolishness indeed, it must have seemed, to assign to the tiny tides of the
Mediterranean an explanation so cosmic as Moon. But mariners in northern
waters face destruction unless they track the tides. For good reason they
remember that Moon reaches its summit overhead an average 50.47 minutes
later each day. Their own bitter experience tells them that, of the two high
tides a day — tw o because there are two projections on an egg — each also
comes about 50 minutes later than it did the day before.
Geography makes Mediterranean tides minuscule. Geography also
makes tides in the Gulf of Maine and Bay of Fundy the highest in the world.
How come? Resonance! The Bay of Fundy and the Gulf of Maine make
together a g r e a t bathtub in which water sloshes back and forth with a natural
period of 13 hours, near to the 12.4-hour timing of Moon’s tide-driving
power — and to the 12-hour timing of Sun’s influence. Build a big power-
producing dam in the upper reaches of the Bay of Fundy? Shorten the length
of the bathtub? Decrease the slosh time from 13 hours to exact resonance
with Moon? Then get one-foot higher tides along the Maine coast!
Want to see the highest tides in the Bay of Fundy? Then choose your visit
according to these rules: (1) Come in summer, when this northern body of
water tilts most strongly toward Moon. (2) Come when Moon, in its elliptic
orbit, is closest to Earth — roughly 10 percent closer than its most distant
point, yielding roughly 35 percent greater tide-producing power. (3) Take
into account the tide-producing power of Sun, aksout 45 percent as great as
that of Moon. Sun’s effect reinforces Moon’s influence when Moon is dark,
dark because interposed, or almost interposed, between Earth and Sun, so
Sun and Moon pull from the same side. But an egg has two projections, so Sun
and Moon also assist each other in producing tides when they are on oppo
site sides of Earth; in this case we see a full Moon.
The result? Burncoat Head in the Minas Basin, Nova Scotia, has the great
est mean range of 14.5 meters (47.5 feet) between low and high tide when
Sun and Moon line up. At nearby Leaf Basin, a unique value of 16.6 meters
(54.5 feet) was recorded in 1953.
High and low tides witness to the relative accelerations of portions of the
ocean separated by the diameter of Earth. High tides show the “ stretching”
relative acceleration at different radial distances from Moon or Sun. Low
tides witness to the “ squeezing” relative accelerations at the same radial
distance from Moon or Sun but at opposite sides of Earth.
During the time of fall, they move toward each other a distance of 1 millimeter— a
thousandth of a meter, the thickness of 16 pages of this book. Why do they move
toward one another? Not because of the gravitational attraction between the ball
bearings; this is far too minute to bring about any “coming together.” Rather,
according to Newton’s nonlocal view, they are both attracted toward the centet of
Earth. Their relative motion results from the difference in direction of Earth’s
gravitational pull on them, says Newton.
As another example, drop the same antique railway coach from rest in a vertical
orientation, with the lower end of the coach initially 315 meters from the surface of
Earth (Figure 2-5, right). Again release tiny ball bearings from rest at opposite ends
of the coach. In this case, during the time of fall, the ball bearings move apart by a
distance of 2 millimetets because of the greater gravitational acceleration of the one
neater Earth, as Newton would put it. This is twice the change that occurs for
horizontal separation.
In either of these examples let the measuring equipment in use in the coach be just
shott of the sensitivity required ro derect this relative motion of the ball bearings.
Then, with a limited time of observation of 8 seconds, the railway coach— or, to use
the earlier example, the freely falling room— serves as a free-float frame.
When the sensitivity of measuring equipment is increased, the railway coach may
no longer serve as a local free-floar frame unless we make additional changes. Eithet
shorten the 20-meter domain in which observations are made, or decrease the time
given to the observations. Or better, cut down some appropriate combination of
space and time dimensions of the region under observation. Or as a final alrernative,
shoot the whole apparatus by rocket up to a region of space where one cannor detect
locally the “differential gravitational acceleration” between one side of the coach and
another— ro use Newton’s way of speaking. In another way of speaking, relative
accelerations of particles in different parts of the coach must be too small to perceive.
Only when these relative accelerations are too small to detect do we have a reference
frame wirh respect to which laws of motion are simple. Thar’s why “local” is a tricky
word!
Hold on! You just finished saying that the idea o f local gravity is unnecessary. Yet here
you use the “differential gravitational acceleration" to account for relative accelera
tions of test particles and ocean tides near Earth. Is local gravity necessary or not?
Near Earth, two explanations of projectile paths or ocean flow give essentially the
same numerical resulrs. Newron says there is a force of gravity, to be treated like any
other force in analyzing motion. Einstein says gravity differs from all orher forces;
Get rid of gravity locally by climbing into a free-float frame. Near the surface of
Earth both explanations accurately predict relative accelerations of falling particles
toward or away from one another and motions of the tides. In this chapter we use the
more familiar Newtonian analysis to predict relative accelerations.
When tests of gravity are very sensitive, or when gravitational effects are large,
such as near whire dwarfs or neutron stars, then Einstein’s predictions are not the
same as Newton’s. In such cases Einstein’s battle-tested 1915 theory of gravity
(genetal relativity) ptedicts results that are observed; Newton’s theory makes incor
rect predictions. This justifies Einstein’s insistence on getting rid of gravity locally
using free-float frames. All rhat remains of gravity is the relative accelerations of
nearby particles — tidal accelerations.
2.4 REGIONS OF SPACETIME

special relativity is limited to free-float frames
“Region of spacetime.” W hat is the precise meaning of this term? The long narrow
railway coach in Figure 2-5 probes spacetime for a limited stretch of time and in one or
another single direction in space. It can be oriented north-south or east-w est or
2.4 REGIONS OF SPACETIME 35
up-dow n. Whatever its orientation, relative acceleration of the tiny ball bearings
released at the two ends can be measured. For all three directions — and for all
intermediate directions — let it be found by calculation that the telative drift of two
test particles equals half the minimum detectable amount or less. Then throughout a
cube of space 20 meters on an edge and for a lapse of time of 8 seconds (2400 million
meters of light-travel time), test particles moving every which way depart from
straighr-line motion by undetectable amounts. In other words, the reference frame is
free-float in a local region of spacetime with dimensions
(20 meters X 20 meters X 20 meters of space) X 2400 million meters of time
“ Region of spacetim e” is
Notice that this “tegion of spacetime” is four-dimensional: three dimensions of space
four-dimensional
and one of time.
W h y p a y so m u c h a t t e n t i o n to th e s m a ll r e la tiv e a c c e le r a tio n s d e s c r ib e d a b o v e ? W h y n o t
fr o m th e b e g in n in g c o n s id e r a s referen ce fr a m e s o n ly sp a c e sh ip s v e r y f a r fr o m E a r th , f a r
fr o m o u r S u n , a n d f a r fr o m a n y o th e r g r a v i t a t i n g b o d y ? A t th ese d is ta n c e s w e n e e d n o t
w o r r y a t a l l a b o u t a n y r e la tiv e a c c e le r a tio n d u e to a n o n u n ifo r m g r a v i t a t i o n a l f i e l d ,
a n d a f r e e - flo a t f r a m e c a n he h u g e w i t h o u t w o r r y in g a b o u t r e la tiv e a c c e le r a tio n s o f
p a r tic le s a t th e e x tr e m itie s o f th e f r a m e . W h y n o t s t u d y s p e c ia l r e l a t i v i t y i n th ese rem ote
re g io n s o f sp a c e ?
Most of our experiments are carried out near Earth and almost all in our part of the
solar system. Near Earth or Sun we cannot eliminate relative accelerations of test
particles due to nonuniformity of gravitational fields. So we need to know how large
a region of spacetime our experiment can occupy and still follow the simple laws that
apply in free-float frames.
For some experiments local free-float frames are not adequate. For example, a
comet sweeps in from remote distances, swings close to Sun, and returns to deep space.
(Consider only the head of the comet, not its 100-million-kilometer-long tail.)
Particles traveling near the comet during all rhose years move closer together or farther
W hen is g en eral relativity
apart due to tidal forces from Sun (assuming we can neglect effects of the gravitational
required?
field of the comet itself). These relative forces are called tid al, because similat
differential forces from Sun and Moon act on the ocean on opposite sides of Earth to
cause tides (Box 2 -1). A frame large enough to include these particles is not free-float.
So reduce spatial size until relative motion of encompassed particles is undetectable
duting that time. The resulting frame is very much smaller than the head of the comet!
You cannot analyze the motion of a comet in a frame smaller than the comet. So
instead think of a larger free-float frame that surrounds the comet for a limited time
during its orbit, so that the comet passes thtough a series of such frames. Or think of a
whole collection of free-float frames plunging radially toward Sun, through which the
comet passes in sequence. In either case, motion of the comet over a small portion of its
trajectory can be analyzed rigorously with respect to one of these local free-float frames
using special relativity. However, questions about the entire rrajectory cannot be
answered using only one free-float frame; for this we require a series of frames. General
relativity — the theory of gravitation — tells how to describe and predict orbits that
traverse a string of adjacent free-float frames. Only general relativity can describe
motion in unlimited regions of spacetime.
P le a se sto p h e a tin g a r o u n d th e h u s h ! In d e fin in g a fr e e - flo a t f r a m e , y o u s a y t h a t every

te s t p a r tic le a t re st i n su c h a f r a m e r e m a in s a t re st “w i t h i n som e s p e c ifie d a c c u r a c y .”
W h a t a c c u r a c y ? C a n 't y o u he m ore s p e c ific ? W h y d o th ese d e fin itio n s d e p e n d on
w h e th e r or n o t w e a re a b le to p e r c e iv e th e t i n y m o tio n o f som e te s t p a r tic le ? M y e y e s ig h t
g e ts w orse. O r I ta k e m y g la ss e s off. D o e s th e w o r ld s u d d e n ly c h a n g e , a lo n g w i t h th e
s ta n d a r d s f o r “i n e r t i a l f r a m e ”? S u r e ly sc ien ce i s m ore e x a c t, m ore o b je c tiv e t h a n t h a t !
Science can be “exact” only when we agree on acceptable accuracy. A 1000-ton

rocket streaks 1 kilometer in 3 seconds; do you want to measure the sequence of its
positions during that time with an accuracy of 10 centimeters? An astronaut in an
orbiting space station releases a pencil that floats at rest in front of her; do you want to
track its position to 1-millimeter accuracy for 2 hours? Each case places different
demands on the inertial frame from which the observations are made. Specific
figures imply specific requirements for inertial frames, requirements that must be
verified by test particles. The astronaut takes off her glasses; then she can determine
the position of the pencil with only 3-millimeter accuracy. Suddenly— yes! —
requirements on the inertial frame have become less stringent— unless she is willing
to observe the pencil over a longer period of time.
2.5 TEST PARTICLE

ideal tool to probe spacetime without affecting it
Test particle defined “Test particle.” How small must a particle be to qualify as a test particle? It must
have so little mass that, within some specified accuracy, its presence does not affect the
motion of other nearby particles. In terms of Newtonian mechanics, gravitational
attraction of the test particle for other particles must be negligible within the accuracy
specified.
As an example, consider a particle of mass 10 kilograms. A second and less massive
particle placed 10 centimeters from it and initially at rest will, in less than 3 minutes,
be drawn toward it by 1 millimeter (see the exercises for this chapter). For measure
ments of this sensitivity or greater sensitivity, the 10-kilogram object is not a test
particle. A particle counts as a test particle only when it accelerates as a result of
gravitational forces without itself causing measurable gravitational acceleration in
other objects— according to the Newtonian way of speaking.
It would be impossible to define a free-float frame were it not for a remarkable
Free-float frame definable feamre of nature. Test particles of different size, shape, and material in the same
because every substance falls
with same acceleration location all fall with the same acceleration toward Earth. If this were not so, an observer
inside a falling room would notice that an aluminum object and a gold object
accelerate relative to one another, even when placed side by side. At least one of these
test particles, initially at rest, would not remain at rest within the falling room. That is,
the room would not be a free-float frame according to definition.
How sure are we that particles in the same location but of different substances all
fall toward Earth with equal acceleration? John Philoponus of Alexandria argued, in
517 A .D ., that when two bodies “differing greatly in weight” are released simulta
neously to fall, “the difference in their time [of fall] is a very small one.” According to
legend Galileo dropped balls made of different materials from the Leaning Tower of
Pisa in order to verify this assumption. In 1905 Baron Roland von Eotvos checked that
the gravitational acceleration of wood toward Earth is equal to that of platinum within
1 part in 100 million. In the 1960s R. H. Dicke, Peter G. Roll, and Robert V.
Krotkov reduced this upper limit on difference in accelerations— for aluminum and
gold responding to the gravitational field of Sun — to less than 1 part in 100,000
million (less than 1 in 10“ )- This — and a subsequent experiment by Vladimir
Braginsky and colleagues — is one of the most sensitive checks of fundamental
physical principles in all of science: the equality of acceleration produced by gravity on
test particles of every kind.
It follows that a particle made of any material can be used as a test particle to
determine whether a given reference frame is free-float. A frame that is free-float for a
tesr particle of one kind is free-float for test particles of all kinds.
2.6 LOCATING EVENTS WITH A LATTICEWORK OF CLOCKS 37
2.6 LOCATING EVENTS WITH A

LATTICEWORK OF CLOCKS
only the nearest clock records an event
The fundamental concept in physics is event. An event is specified not only by a place
but also by a time of happening. Some examples of events are emission of a particle or a
flash of light (from, say, an explosion), reflection or absorption of a particle or light
flash, a collision.
How can we determine the place and time at which an event occurs in a given
free-float frame? Think of constructing a frame by assembling meter sticks into a
Latticework of rods and clocks
cubical latticework similar to the jungle gym seen on playgrounds (Figure 2-6). At
every intersection of this latticework fix a clock. These clocks are identical. They can be
constructed in any manner, but their readings are in meters of light-travel time
(Section 1.4).
How are the clocks to be set? We want them all to read the “same time” as one
another for observers in this frame. When one clock reads midnight (00.00 hours = 0
meters), all clocks in the same frame should read midnight (zero). That is, we want the
clocks to be synchronized in this frame.
How are the several clocks in the lattice to be synchronized? As follows: Pick one
Synchronizing clocks in lattice
clock in the lattice as the standard and call it the reference clock. Start this reference
FIGURE 2 -6 . L a t t i c e w o r k o f m e te r s tic k s a n d clocks.

clock with its pointer set initially at zero time. At this instant let it send out a flash of
Reference event defined light that spreads out as a spherical wave in all directions. Call the flash emission the
reference event and the spreading spherical wave the referen ce flash.
When the reference flash gets to a slave clock 5 meters away, we want that clock to
read 5 meters of light-travel time. Why? Because it takes light 5 meters of light-travel
time to travel the 5 meters of distance from reference clock to slave clock. So an
assistant sets the slave clock to 5 meters of time long before the experiment begins,
holds it at 5 meters, and releases it only when the reference flash arrives. (The assistant
has zero reaction time or the slave clock is set ahead an additional time equal to the
reaction time.) When assistants at all slave clocks in the lattice follow this prearranged
procedure (each setting his slave clock to a time in meters equal to his own distance
from the reference clock and starting it when the reference light flash arrives), the
lattice clocks are said to be synchronized.
This is an awkward way to synchronize lattice docks with one another. Is there some
simpler and more conventional way to carry out this synchronization?
There are other possible ways to synchronize clocks. For example, an extra portable
clock could be set to the reference clock at the origin and carried around the lattice in
order to set the rest of the clocks. However, this procedure involves a moving clock.
W e saw in Chapter 1 that the time between two events is not necessarily the same as
recorded by clocks in relative motion. The portable clock will not even agree with the
reference clock when it is brought back next to it! (This idea is explored more fully in
Section 4.6.) However, when we use a moving clock traveling at a speed that is a very
small fraction of light speed, its reading is only slightly different from that of clocks
fixed in the lattice. In this case the second method of synchronization gives a result
nearly equal to the first— and standard — method. Moreover, the error can be made
as small as desired by carrying the portable clock around sufficiently slowly.
Use the latticework of synchronized clocks to determine location and time at which
Locate event with latticework
any given event occurs. The space position of the event is taken to be the location of the
clock nearest the event. The location of this clock is measured along three lattice
directions from the reference clock: northward, easrward, and upward. The time of the
event is taken to be the time recorded on the same lattice clock nearest the event. The
spacetime location of an event then consists of four numbers, three numbers that
specify the space position of the clock nearest the event and one number that specifies
the time the event occurs as recorded by that clock.
The clocks, when installed by a foresighted experimenter, will be recording clocks.
Each clock is able to detect the occurrence of an event (collision, passage of light-flash
or particle). Each reads into its memory the nature of the event, the time of the event,
and the location of the clock. The memory of all clocks can rhen be read and analyzed,
perhaps by automatic equipment.
Why a latticework built of rods that are 1 meter long? What is special about 1 meter?
Why not a lattice separation of 100 meters between recording clocks? Or 1 millimeter?
When a clock in the 1-meter lattice records an event, we will not know whether the
event so recorded is 0.4 meters to the left of the clock, for instance, or 0.2 meters to
the right. The location of the event will be uncertain to some substantial fraction of a
meter. The time of the event will also be uncertain with some appreciable fraaion of
a meter of light-travel time, because it may take that long for a light signal from the
event to reach the nearest clock. However, this accuracy of a meter or less is quite
2.7 OBSERVER 39
adequate for observing the passage of a rocket. It is extravagantly good for measure-
menrs on planetary orbits— for a planer it would even be reasonable to increase the
lattice spacing from 1 meter ro hundreds of meters.
Neither 100 meters nor 1 meter is a lattice spacing suitable for studying the tracks
of particles in a high-energy accelerator. There a centimeter or a millimeter would be
more appropriate. The location and time of an event can be determined to whatever
accuracy is desired by constructing a latticework with sufficiently small spacing.
2.7 OBSERVER
ten thousand local witnesses
In relativity we often speak about the observer. Where is this observer? At one place,
or all over the place? Answer: T h e w o rd “ o b serv er” is a sh o rth a n d way o f
O b server defined
sp eaking ab o u t th e w hole collection o f re c o rd in g clocks associated w ith
one free-float fram e. No one real observer could easily do what we ask of the “ideal
observer” in our analysis of relativity. So it is best to think of the observer as a person
who goes around reading out the memories of all recording clocks under his control.
This is the sophisticated sense in which we hereafter use the phrase “the observer
measures such-and-such.”
Location and time of each event is recorded by the clock nearest that event. We
intentionally limit the observer’s report on events to a summary of data collected from
clocks. We do not permit the observer to report on widely separated events that he
himself views by eye. The reason: travel time of light! It can take a long time for light
O b se rv e r limited to clock readings
from a distant event to reach the observer’s eye. Even the order in which events are seen
by eye may be wrong: Light from an event that occurred a million years ago and a
million light-years distant in our frame is just entering our eyes now, after light from an
event that occurred on Moon a few seconds ago. We see these two events in the “wrong
order” compared with observations recorded by our far-flung latticework of recording
clocks. For this reason, we limit the observer to collecting and reporting data from the
recording clocks.
The wise observer pays attention only to clock records. Even so, light speed still
places limits on how soon he can analyze events after they occur. Suppose that events in
a given experiment are widely separated from one another in interstellar space, where a
single free-float frame can cover a large region of spacetime. Let remote events be
recorded instantly on local clocks and transmitted by radio to the observer’s central
control room. This information transfer cannot take place faster than the speed of
light — the same speed at which radio waves travel. Information on dispersed events is
available for analysis at a central location only after light-speed transmission. This
information will be full and accurate and in no need of correction — but it will be late.
Thus all analysis of events must take place after— sometimes long after! — events are
over as recorded in that frame. The same difficulty occurs, in principle, for a free-float
frame of any size.
Nature puts an unbreakable speed limit on signals. This limit has profound
consequences for decision making and control. A space probe descends onto Triton, a Speed limit: c
It's the law!
moon of the planet Neptune. The probe adjusts its rocket thrust to provide a
slow-speed “soft” landing. This probe must carry equipment to detect its distance
from Triton’s surface and use this information to regulate rocket thrust on the spot,
without help from Earth. Earth is never less than 242 light-minutes away from
Neptune, a round-trip radio-signal time of 484 minutes— more than eight hours.
Therefore the probe would crash long before probe-to-surface distance data could be
sent to Earth and commands for rocket thrust returned. This time delay of information
transmission does not prevent a detailed retrospective analysis on Earth of the probe’s
descent onto Triton — but this analysis cannot take place until at least 242 minutes
SAMPLE PROBLEM 2-1

ME T E O R ALERT!
Interstellar Command Center receives word by Center at one quarter light speed. Assume radio
radio that a meteor has just whizzed past an out signals travel with light speed. How long do Com
post situated 100 light-seconds distant (a fifth of mand Center personnel have to take evasive ac
Earth-Sun distance). The report warns that the tion?
meteor is headed directly toward Command
SOLUTION
The warning radio signal and the meteor leave the The meteor takes an additional 100 seconds of
outpost at the same time. The radio signal moves time to move each additional 2 5 light-seconds of
wirh light speed from outpost to Command distance. So it covers the remaining 75 light-se
Center, covering the 100 light-seconds of distance conds of distance in an additional time of 300
in 100 seconds of time. During this 100 seconds seconds.
the meteor also travels roward Command Center. In brief, after receiving the radio warning.
The meteor moves at one quarter light speed, so in Command Center personnel have a relaxed 300
100 seconds it covers one quarter of 100 light-se seconds — or five minutes — to stroll ro their me
conds, or 2 5 light-seconds of distance. Therefore, teor-proof shelter.
when the warning arrives at Command Center, the
meteor is 100 — 25 = 75 light-seconds away.
after the event. Could we gather last-minute information, make a decision, and send
back control instructions? No. Nature rules our micromanagement of the far-away
(Sample Problem 2-1).
2.8 MEASURING PARTICLE SPEED

reference frame clocks and rods put to use
The recording clocks reveal particle motion through the lattice; Each clock that the
particle passes records the time of passage as well as the space location of this event.
How can the path of the particle be described in terms of numbers? By recording
locations of these events along the path. Distances between locations of successive
events and time lapse between them reveal rhe particle speed — speed being space
separation divided by time taken to traverse this separation.
The conventional unit of speed is meters per second. However, when time is
Speed in meters per meter
measured in meters of light-travel time, speed is expressed in meters of distance
covered per meter of time. A flash of light moves one meter of distance in one meter of
light-travel time: its speed has the value unity in units of meters per meter. In contrast,
a particle loping along at half light speed moves one half meter of distance per meter of
time; its speed equals one half in units of meter per meter. More generally, particle
speed in meters per meter is the ratio of its speed to light speed:
_ (meters of distance covered by particle)

(particle speed)
(meters of time required to cover that distance)
_ (particle speed in meters/second)
(speed of light in meters/second)
2.9 ROCKET FRAME 41
In this book we use the letter p to symbolize the speed of a particle in meters of distance
per meter of time, or simply meters per meter. Some authors use the lowercase Greek
letter beta; Let stand for velocity in conventional units (such as meters per
second) and c stand for light speed in the same conventional units. Then
( 2- 1)
From the motion of test particles through a latticework of clocks — or rather from
records of coincidences of these particles with clocks — we determine whether the
Test for free-float frame
latticework constitutes a free-float frame. IF records show (a) that— within some
specified accuracy — a test particle moves consecutively past clocks that lie in a straight
line, (b) that test-particle speed calculated from the same records is constant— again,
within some specified accuracy — and, (c) that the same results are true for as many
test-particle paths as the most industrious observer cares to trace throughout the given
region of space and time, THEN the lattice constitutes a free-float (inertial) frame
throughout that region of spacetime.
P a r tic le sp e e d a s a f r a c ti o n o f l i g h t sp e e d is c e r ta in ly a n u n c o n v e n tio n a l u n i t o f
m e a su re . W h a t a d v a n ta g e s d o es i t h a v e t h a t j u s t i f y th e w o r k n e e d e d to becom e f a m i l i a r
w ith it?
The big advantage is that it is a measure of speed independent of units of space and
time. Suppose that a particle moves with respect to Earth at half light speed. Then it
travels— with respect to Earth— one half meter of distance in one meter of light
travel time. It travels one half light-year of distance in a period of one year. It travels
one half light-second of distance in a time of one second, one half light-minute in one
minute. Units do not matter as long as we use the same units to measure distance and
time; the result always equals the same number: 1/2 . Another way to say this is that
speed is a fraction; same units on top and bottom of the fraction cancel one another.
Fundamentally, v is unit-free. O f course, if we wish we can speak of “meters per
meter."
2.9 ROCKET FRAME

does it move? or is it the one at rest?
Let two reference frames be two different latticeworks of meter sticks and clocks, one
moving uniformly relative to the other, and in such a way that one row of clocks in each
frame coincides along the direction of relative motion of the two frames (Figure 2-7).
Call one of these frames lab o rato ry fram e and the other— moving to the right
Rocket frame defined
relative to the laboratory frame — ro c k e t fram e. The rocket is unpowered and coasts
along with constant velocity relative to the laboratory. Let rocket and laboratory
latticeworks be overlapping in the sense that a region of spacetime exists common to
both frames. Test particles move through this common region of spacetime. From
motion of these test particles as recorded by his own clocks, the laboratory observer
verifies that his frame is free-float (inertial). From motion of the same test particles as
recorded by her own clocks, the rocket observer verifies that her frame is also free-float
(inertial).
Now we can describe the motion of any particle with respect to the laboratory
frame. The same particles and — if they collide — the same collisions may be mea
sured and described with respect to the free-float rocket frame as well. These particles,
their paths through spacetime, and events of their collisions have an existence inde-
FIGURE 2 -7 . L a b o r a to r y a n d r o c k e t f r a m e s . A second ago the tw o latticew o rks were interm eshed.
pendent of any free-float frames in which they are observed, recorded, and described.
However, descriptions of these common paths and events are typically different for
DifFerent frames lead to different free-float frames. For example, laboratory and rocket observers may not
different descriptions
agree on the direction of motion of a given test particle (Figure 2-8). Every track that is
straight as plotted with respect to one reference frame is straight also with respect to the
other frame, because both are free-float frames. This straightness in both frames is
possible only because one free-float frame has uniform velocity relative to any other
LABORATORY ROCKET
FRAME (UNPOWERED)
FRAME
FIGURE 2 -8 . A s e r ie s o f “s n a p s h o ts " o f a t y p i c a l t e s t p a r t i c l e a s m e a s u r e d f r o m la b o r a to r y a n d
r o c k e t f r e e - f l o a t f r a m e s , r e p r e s e n te d b y c u t a w a y c y lin d e r s . S ta r t a t the bottom a n d read u p w a rd
(tim e progresses fro m bottom to top).
2.10 SUMMARY 43
overlapping free-float frame. However, the direction of this path differs from labora
tory to rocket frame, except in the special case in which the particle moves along the
line of relative motion of two frames.
How many different free-float rocket frames can there be in a given region of
spacetime? An unlimited number! Any unpowered rocket moving through that region Many possible free-float frames
in any direction is an acceptable free-float frame from which to make observations.
More: There is nothing unique about any of these frames as long as each of them is
free-float. All “rocket” frames are unpowered, all are equivalent for carrying out
experiments. Even the so-called “laboratory frame” is not unique; you can rename ir No unique free-float frame
“Rocket Frame Six” and no one will ever know the difference! All free-float (inertial)
frames are equivalenr arenas in which to carry out physics experiment. That is the
logical basis for special relativiry, as described more fully in Chapter 3.
A ro c k e t c a r r ie s a fir e c r a c k e r . T h e fir e c r a c k e r e x p lo d es. D o e s t h i s e v e n t — th e e x p lo sio n

— t a k e p la c e i n th e ro c k e t f r a m e o r i n th e la b o r a to r y fr a m e ? W h ic h is th e “h o m e " f r a m e
f o r th e e v e n t? A se c o n d fir e c r a c k e r , o r ig in a lly a t re st i n th e la b o r a to r y f r a m e , e x p lo d es.
D o e s t h i s se c o n d e v e n t o ccu r i n th e la b o r a to r y f r a m e or i n th e ro c k e t f r a m e ?
Events are primary, the essential stuff of Nature. Reference frames are secondary,
devised by humans for locating and comparing events. A given event occurs in both
ftames— and in all possible frames moving in all possible directions and with all
possible constant relative speeds through the region of spacetime in which the event
occurs. The apparatus that “causes” the event may be at rest in one free-float frame;
another apparatus that “causes” a second event may be at rest in a second free-float
frame in motion relative to the first. No matter. Each event has its own unique
existence. Neither is “owned” by any frame at all.
A spark jumps 1 millimeter from the antenna of Mary’s passing spaceship to a pen
in the pocket of John who lounges in the laboratory doorway (Section 1.2). The
“apparatus” that makes the spark has parts riding in different reference frames —
pen in laboratory frame, antenna in rocket frame. The spark jump — in which frame
does this event occur? It is not the property of Mary, not the property of John — not
the property of any other observer in the vicinity, no matter what his or her state of
motion. The spark-jump event provides data for every observer.
Drive a steel surveying stake into the ground to mark the corner of a plot of land.
Is this a “Daytime stake” or a “Nighttime stake”? Neither! It is just a marking
a location in sp a c e, the arena of surveying. Similarly an event is neither a “laboratory
event” nor a “rocket event.” It is just an e v e n t, marking a location in sp a c e tim e , the
arena of science.
Laboratory frame or rocket frame: Which one is the “primary” free-float frame, the
one “really” at rest? There is no way to tell! We apply the names “laboratory” and
“rocker” to two free-float enclosures in interstellar space. Someone switches the
nameplates while we sleep. When we wake up, there is no way to decide which is
which. This realization leads to Einstein’s Principle of Relariviry and proof of the
invariance of the interval, as described in Chapter 3.
2.10 SUMMARY
what a free-float frame is and what it's good for
The free-float fram e (also called the in ertial fram e and the L orentz fram e)
provides a setting in which to carry out experiments without the presence of so-called
‘‘gravitational forces. ’’ In such a frame, a particle released from rest remains at rest and
a particle in motion continues that motion without change in speed or in direction

(Section 2.2), as Newton declared in his First Law of Motion.
Where does that frame of reference sit? Where do the east-west, north-south,
up-down lines run? We might as well ask where on the flat landscape in the state of
Iowa we see the lines that mark the boundaries of the townships. A concrete marker, to
be sure, may show itself as a corner marker at a place where a north-south line meets an
east-west line. Apart from such on-the-spot evidence, those lines are largely invisible.
Nevertheless, they serve their purpose: They define boundaries, settle lawsuits, and fix
taxes. Likewise imaginary for the most part are the clock and rod paraphernalia of the
idealized inertial reference frame. Work of the imagination though they are, they
provide rhe conceptual framework for everything that goes on in the world of particles
and radiation, of masses and motions, of annihilations and creations, of fissions and
fusions in every context where tidal effects of gravity are negligible.
Our ability to define a free-float frame depends on the fact that a test p artic le
made of any material whatsoever experiences the same acceleration in a given gravita
tional field (Section 2.5).
Near a massive (“gravitating”) body, we can still define a free-float frame. How
ever, in such a frame, free test particles typically accelerate toward or away from one
another because of the nonuniform field of the gravitating body (Section 2.3). This
limits — in both space and time — the size of a free-float frame, the domain in which
the laws of motion are simple. The frame will continue to qualify as free-float and
special relativity will continue to apply, provided we reduce the spatial extent, or the
time duration of our experiment, or both, until these relative, or tid al, motions of test
particles cannot be detected in our circumscribed region of spacetime. This is what
makes special relativity "special” or limited (French: relativite restreinte: “restricted
relativity”). General relativity (the theory of gravitation) removes this limitation
(Chapter 9).
So there are three central characteristics of a free-float frame. (1) W e can “get rid of
gravity” by climbing onto (getting into) a free-float frame. (2) The existence of a
free-float frame depends on the equal acceleration of all particles at a given location in
a gravitational field — in Newton’s way of speaking. (3) Every free-float frame is of
limited extent in spacetime. All three characteristics appear in a fuller version of the
quotation by Albert Einstein that began this chapter:
At that moment there came to me the happiest thought of my life . . . for an observer
falling freely from the roof o f a house no gravitational field exists during his fa ll — at least
not in his immediate vicinity. That is, if the observer releases any objects, they remain in a
state of rest or uniform motion relative to him, respectively, independent of their unique
chemical and physical nature. Therefore the observer is entitled to interpret his state as
that of “rest.” -uer"
REFERENCES
Introductory and final quotes: Excerpt from an unpublished manuscript in
Einstein’s handwriting, dating from about 1919. Einstein is referring to the year
1907. Italics represent material underlined in the original. Quoted by Gerald
Holton in Thematic Origins of Scientific Thought, Revised Edition (Harvard
University Press, Cambridge, Mass.,1988), page 382. Photocopy of the original
provided by Professor Holton. Present translation made with the assistance of
Peter von Jagow.
EXERCISE 2-4 SYNCHRONIZATION BY A TRAVELING CLOCK 45
Figure 2-1 and Jules Verne srory: Jules Verne, A Trip From the Earth to the Moon
and A Trip Around the Moon, paperback edition published by Dover Publica
tions, New York. Hardcover edition published in the Great Illustrated Classics
Series by Dodd, Mead and Company, New York, 1962.
Information on Nova Scotia tides in Box 2-1: Guinness Book of World Records
1988 (Bantam Books, New York, 1987), page 125.
Relative acceleration of different materials. Section 2-5: P. G. Roll, R. Krotkov,
and R. H. Dicke, “The equivalence of inertial and passive gravitational mass,”
Annals of Physics (USA), Volume 26, pages 4 4 2 -5 1 7 (1964); V. B. Braginsky
and V. I. Panov, Zhumal Eksperimental’noi i Teoreticheskoi Eiziki, Volume 6 l,
page 873 (1972) [Soviet Physics JETP, Volume 34, page 463 (1972)}.
Experimental proofs of Einstein’s general relativity theory: Clifford Will, Was

Einstein Right? Putting General Relativity to the Test (Basic Books, New York,
1986) .
CHAPTER 2 EXERCISES
different times during the bounces. During what part

PRACTICE of each jump will the scale have zero reading? Ne
glecting air resisrance, whar is the longest part of the
2-1 hum« ■ball cycle during which you mighr consider yourself to be
in a free-float frame?
A person rides in an elevator that is shot upward out
of a cannon. Think of the elevator after it leaves the
cannon and is moving freely in the gravitational field
2-3 practical synchronization of
of Earth. Neglect air resistance.
clocks
a While the elevator is still on the way up, the You are an observer in the laboratory frame stationed
person inside jumps from the “floor” of the elevator. near a clock wirh spatial coordinates x = 6 light-
Will the person (1) fall back to the “floor” of the seconds, y — 8 light-seconds, and z = 0 light-seconds.
elevator? (2) hit the “ceiling” of the elevator? (3) do You wish to synchronize your clock with the one at
something else? If so, what? the origin. Describe in detail and with numbers how
b The person waits to jump until after the eleva to proceed.
tor has passed the top if its trajectory and is falling
back toward Earth. Will your answers to part a be 2-4 synchronization by a
different in this case? traveling clock
C How can the person riding in the elevator tell
Mr. Engelsberg does nor approve of our merhod of
when the elevator reaches the top of its trajectory?
synchronizing clocks by light flashes (Section 2.6).
a “I can synchronize my clocks in any way I
2-2 ffree-float bounce choose!” he exclaims. Is he righr?
Test your skill as an acrobar and contorrionist! Fasren Mr. Engelsberg wishes to synchronize two identical
a weight-measuring barhroom scale under your feet clocks, named Big Ben and Little Ben, which are
and bounce up and down on a trampoline while relatively at rest and separated by one million kilome
reading the scale. Describe readings on rhe scale at ters, which is 10^ meters or approximately three times
46 EXERCISE 2-5 EARTH'S SURFACE AS A FREE-FLOAT FRAME
the distance between Earth and Moon. He uses a third

clock, identical in construction with the first two, that
travels with constant velocity between them. As his
moving clock passes Big Ben, it is set to read the same
time as Big Ben. When the moving clock passes Little
Ben, that outpost clock is set to read the same time as
the traveling clock.
b “Now Big Ben and Little Ben are synchro
nized,” says Mr. Engelsberg. Is he right?
c How much out of synchronism are Big Ben and EXERCISE 2 -6 . Schem atic d ia g ra m o f tw o b a ll bearings fa llin g
Little Ben as measured by a latticework of clocks — at onto E a r th ’s surface. N o t to scale.
rest relative to them both — that has been synchro

nized in the conventional manner using light flashes?
2-6 horizontal extent of free-
Evaluate this lack of synchronism in milliseconds
when the traveling clock that Mr. Engelsberg uses
float frame near Earth
moves at 360,000 kilometers/hour, or 10’ meters/ Consider two ball bearings near the surface of Earth
second. and originally separated horizontally by 20 meters
d Evaluate the lack of synchronism when the (Section 2.3). Demonstrate that when released from
traveling clock moves 100 times as fast. rest (relative to Earth) the particles move closer to
e Is there any earthly reason — aside from mat gether by 1 millimeter as they fall 315 meters, using
ters of personal preference— why we all should not the following method of similar triangles or some
adopt the method of synchronization used by Mr. other method.
Engelsberg? Each particle falls from rest toward the center of
Earth, as indicated by arrows in the figure. Solve the
2-5 Earth's surface as a free- problem using the ratio of sides of similar triangles
ffloat frame abc and a'h'c'. These triangles are upside down with
respect to each other. However, they are similar be
Many experiments involving fast-moving particles
cause their respective sides are parallel: Sides ac and
and light itself are observed in earthbound laborato
a'c' are parallel to each other, as are sides be and b'c'
ries. Typically these laboratories are not in free fall!
and sides ab and a 'b '. W e know the lengths of some
Nevertheless, under many circumstances laboratories
of these sides. Side a'c' = 315 meters is the height of
fixed to the surface of Earth can satisfy the conditions
fall (greatly exaggetated in the diagtam); side ac is
required to be called free-float frames. An example;
effectively equal to the radius of Earth, 6,371,000
a In an earthbound laboratory, an elementary
meters. Side ab = (1/2) (20 meters) equals half the
particle with speed v = 0.96 passes from side to side
original separation of the particles. Side a'b' equals
through a cubical spark chamber one meter wide. For
HALF their CHANGE in separation as they fall onto
what length of laboratory time is this particle in transit
Earth’s surface. Use the ratio of sides of similar trian
through the spark chamber? Therefore for how long a
gles to find this “half-change” and therefore the en
time is the experiment “in progress”? H ow far will a
tire change in separation as two particles initially 20
separate test particle, released from rest, fall in this
meters apart horizontally fall from rest 315 meters
time? [Distance of fall from rest = where
onto the surface of Earth.
g = acceleration of gravity ^ 1 0 meters/second^ and
is the time of free fall in seconds.} Compare your
answer with the diameter of an atomic nucleus (a few
2-7 limit on free-float frame
times 10“ ' ’ meter).
near Earth's Moon
b How wide can the spark chamber be and still Release two ball bearings from rest a horizontal dis
be considered a free-float frame for this experiment? tance 20 meters apart near the surface of Earth’s
Suppose that by using sensitive optical equipment (an Moon. By how much does the separation between
in terfe ro m eter) you can detect a test particle them dectease as they fall 315 meters? How many
change of position as small as one wavelength of seconds elapse during this 315-metet fall? Assume
visible light, say 500 nanometers = 5 X 10~^ meter. that an initial vertical separation of 20 meters is in
How long will it take the test particle to fall this creased by twice the change in horizontal separation in
distance from rest? How far does the fast elementary a fall through the same height. State clearly and com
particle of part a move in that time? Therefore how pletely the dimensions of the tegion of spacetime in
long can an earthbound spark chamber be anti still be which such a freely falling frame constitutes an inertial
considered free-float fot this sensitivity of deteaion? frame (to the given accuracy). Moon radius equals
EXERCISE 2-10 TEST PARTICLE? 47
1738 kilometers. Gravitational acceleration at

Moon’s surface: g = 1.62 meters/second^.
□ □ □ □ □ □ □ □ □ D O
2-8 vertical extent of free-float --------- 0 ^
frame near Earth
N ote: This exercise makes use of elementary calculus
and the Newtonian theory of gravitation.
A paragraph in Section 2.3 says:
As another example, drop the same antique [20-meter-
long] railway coach from rest in a vertical orientation,
EXERCISE 2 -9 . Free-float ra ilw a y coach risin g fro m E a r th ’s s u r
with the lower end of the coach initially 315 meters
fa c e , a s observed in E a rth fra m e . T w o h a ll bearings were ju s t
from the surface of Earth (Figure 2-5, right). Again
released fro m rest w ith respect to the coach. W h a t w i l l be th e ir
release two tiny ball bearings from rest at opposite ends subsequent m otion as observed fro m insid e the coach? F igure not to
of the coach. In this case, during the time of fall [8 scale.
seconds], the ball bearings move a p a rt by a distance of
two millimeters because of the greater gravitational
acceleration of the one nearer Earth, as Newton would the surface of Earth (not shown). Again you release
p u t it. This is twice the change that occurs for horizontal two ball bearings at opposite ends of the coach and at
separation. rest with respect to the coach. Will you observe these
ball bearings to move together or apart?
Demonstrate this 2-millimeter increase in separation.
c In either of the cases described above, can you,
The following outline may be useful. Take the gravi
the rider in the railway coach, distinguish whether the
tational acceleration at the surface of Earth to be^„ =
coach is rising or falling with respect to the surface of
9.8 meters/second^ and the radius of Earth to be r„ =
Earth solely by observing the ball bearings from inside
6.37 X 10® meters. More generally, the gravitational
the coach? W hat do you observe at the moment the
acceleration of a particle of mass m a distance r from
coach stops rising with respect to Earth and begins to
the center of Earth (mass M) is given by the expression
fall?
F GM GM ti
2-10 test particle?
m 4 .2
a Verify the statement in Section 2.5 that a can

a Take the differential of this equation for g to didate test particle of mass 10 kilograms placed 0.1
obtain an approximate algebraic expression for Ag, meter from a less massive particle (initially stationary
the change in g, for a small change A r in height. with respect to it), draws the second toward it by 1
millimeter in less than 3 minutes. If this relative
b Now use A y = to 6od an algebraic motion is detectable by equipment in use at the test
expression for increase in distance A y between ball site, the result disqualifies the 10-kilogram particle as
bearings in a fall that lasts for time t. a “test particle.’’ Assume that both particles are
C Substitute numbers given in the quotation spherically symmetric. Use Newton’s Law of Gravi
above to verify the 2-millimeter change in separation tation:
during fall.
GMm
2-9 the rising railway coach

You are launched upward inside a railway coach in a where the gravitation constant G has the value G =
horizontal position with respect to the surface of 6.673 X 10~“ meterV(kilogram-second^). Assume
Earth, as shown in the figure. After the launch, but that this force does not change appreciably as the
while the coach is still rising, you release two ball particles decrease separation by one millimeter.
bearings at opposite ends of the train and at rest with b Section 2.3 describes two ball bearings re
respect to the train. leased 20 meters apart horizontally in a freely falling
a Riding inside the coach, will you observe the railway coach. They move 1 millimeter closer together
distance between the ball bearings to increase or de during 8 seconds of free fall, showing the limitations
crease with time? on this inertial frame. Verify that these ball bearings
b Now you ride in a second railway coach qualify as test particles by estimating the distance that
launched upward in a vertical position with respect to one will move from rest in 8 seconds under the gravi
48 EXERCISE 2-11 COMMUNICATIONS STORM!
tational attraction of the other, if both were initially at copper that would be consistent with Galileo’s exper
rest in interstellar space far from Earth. Make your imental result.
own estimate of the mass of each ball bearing. b The result of the more modern Dicke experi
ment is that the fraction H^g/g is not greater than 3 X
Assume that the fraction has this more recently
PROBLEMS determined maximum value. Reckon how far behind
the first ball the second one will be when the first
2-11 communications storm! reaches the ground if they are dropped simultaneously
from the top of a 46-meter vacuum chamber. Under
Sun emits a tremendous burst of particles that travels
these same circumstances, how far would balls of
toward Earth. An astronomer on Earth sees the emis
different materials have to fall in a vacuum in a
sion through a solar telescope and issues a warning.
uniform gravitational field of 10 meters/second/se-
The astronomer knows that when the particles arrive,
cond for one ball to lag behind the other one by a
they will wreak havoc with broadcast radio transmis
distance of 1 millimeter? Compare this distance with
sion. Communications systems require three minutes
the Earth-M oon separation (3.8 X 10® meters).
to switch from broadcast to underground cable. What
Clearly the Dicke experiment was not carried out
is the maximum speed of the particle pulse emitted by
using falling balls!
Sun such that the switch can occur in time, between
C A plumb bob of mass m hangs on the end of a
warning and arrival of the pulse? Take Sun to be 500
long line from the ceiling of a closed room, as shown
light-seconds from Earth.
in the first figure (left). A very massive sphere at one
side of the closed room exerts a horizontal gravita
2-12 the Dicke experiment tional force mg, on the plumb bob, where g, = G M /
a The Leaning Tower of Pisa is about 5 5 meters E}, M is the mass of the large sphere, and R the
high. Galileo says, “The variation of speed in air distance between plumb bob and the center of the
between balls of gold, lead, copper, porphyry, and sphere. This horizontal force causes a static deflection
other heavy materials is so slight that in a fall of 100 of the plumb line from the vertical by the small angle
cubits [about 46 meters} a ball of gold would surely £. (Similar practical example: In northern India the
not outstrip one of copper by as much as four fingers. mass of the Himalaya Mountains results in a slight
Having observed this I came to the conclusion that in sideways deflection of plumb lines, causing difficul
a medium totally devoid of resistance all bodies ties in precise surveying.) The sphere is now rolled
would fall with the same speed.” around to a corresponding position on the other side
Taking four fingers to be equal to 7 centimeters, of the room (right), causing a static deflection of the
find the maximum fractional difference in the accel plumb by an angle fi of the same magnitude but in the
eration of gravity ^ g /g between balls of gold and opposite direction.
EXERCISE 2 -1 2 , first figure. L e ft: N earby m assive sphere results sphere to the other side results in s ta tic deflection o f p lu m b lin e in the
in sta tic deflection o f p lu m b lin e fro m vertical. R ig h t: R o llin g the opposite direction.
EXERCISE 2-12 THE DICKE EXPERIMENT 49
Now the angle S is very small. (Deflection due to Dicke experiment? Answer: Our purpose is to detect
the Himalayas is about 5 seconds of arc, which equals the difference— if any— in the gravitational accelet-
0.0014 degrees.) However, as the sphere is rolled ation gj toward Sun for different materials. The cen
around and around outside the closed room, an ob trifugal acceleration v^/R away from Sun is presum
server inside the room can measure the gravitational ably the same for all materials and therefore need not
field gj due to the sphere by measuring with greater entet any comparison of different materials.
and greater precision the total deflection angle 2 e ~ 2 Consider a torsion pendulum suspended from its
sin £ of the plumb line, where fi is measured in ra center by a thin quartz fiber (second figure). A light
dians. Derive the equation that we will need in the rod of length L supports at its ends two bobs of equal
calculation of gj. mass made of different materials — say aluminum
d W e on Earth have a large sphere effectively and gold. Suppose that the gravitational acceleration
rolling around us once every day. It is the most mas gi of the gold due to Sun is slightly greater than the
sive sphere in the solar system: Sun itself! W hat is the acceleration g2 of the aluminum due to Sun. Then
value of the gravitational acceleration g, = GM/R^ there will be a slight net torque on the torsion pendu
due to Sun at the position of Earth? (Some constants lum due to Sun. Fot the position of Sun shown at left
useful in this calculation appear inside the back cover in the figure, show that the net torque is counter
of this book.) clockwise when viewed from above. Show also that
e One additional acceleration must be consid the magnitude of this net torque is given by the
ered that, however, will not enter our final compari expression
son of gravitational acceleration g^ for different mate
rials. This additional acceleration is the centrifugal torque = mgi L /2 — mg^ L /2 = m{g^ — g ^ L /2
acceleration due to the motion of Earth around Sun. = mgsi^g/g) L /2
When you round a corner in a car you are pressed
against the side of the car on the outward side of the g Suppose that the fraction (Ag/gP has the
turn. This outward force— called the centrifugal maximum value 3 X 1 consistent with the results
pseudoforce or the centrifugal inertial force— is due of the final experiment, that L has the value 0.06
to the acceleration of your reference frame (the car) meters, and that each bob has a mass of 0.03 kilo
toward the center of the circular turn. This centrifugal grams. W hat is the magnitude of the net torque?
inertial force has the value m vl^Jr, where is the Compare this to the torque provided by the added
speed of the car in conventional units and r is the weight of a bacterium of mass 10“ ^’ kilogram placed
tadius of the turn. Now Earth moves around Sun in a on the end of a meter stick balanced at its center in the
path that is nearly circular. Sun’s gravitational force gravitational field of Earth.
mg^ acts on a plumb bob in a direction toward Sun; the h Sun moves atound the heavens as seen from
centrifugal inertial force m v^^JR acts in a direction Earth. Twelve hours later Sun is located as shown at
away from Sun. Compare the “centrifugal accelera right in the second figure. Show that under these
tion” position of Earth with the oppo changed citcumstances the net torque will have the
sitely directed gravitational acceleration g, calculated same magnitude as that calculated in part g but now
in part d. W hat is the net acceleration toward or away will be clockwise as viewed from above — in a sense
from Sun of a particle riding on Earth as observed in opposite to that of part g. This change in the sense of
the (accelerated) frame of Earth? the torque every twelve hours allows a small differ
f O f what use is the discussion thus far? A plumb ence Ag = gt ~ g2 in the acceleration of gold and
bob hung near the surface of Earth experiences a aluminum to be detected using the torsion pendu
gravitational acceleration g^ toward Sun— and an lum. As the torsion pendulum jiggles on its fiber
equal but opposite centrifugal acceleration mvl^^/R because of random motion, passing trucks. Earth
away from Sun. Therefore — in the acceletating refer tremors and so forth, one needs to consider only those
ence frame of Earth — the bob experiences no net deflections that keep step with the changing position
force at all due to the presence of Sun. Indeed this is of Sun.
the method by which we constructed an inertial frame i A torque on the rod causes an angular rotation
in the first place (Section 2.2): Let the frame be in free of the quartz fiber of 6 radians given by the formula
fall about the center of gravitational attraction. A
particle at rest on Earth’s surface is in free fall about torque = k9
Sun and therefore experiences no net force due to Sun.
W hat then does all this have to do with measuring the where k is called the to rsio n co n stan t of the fiber.
equality of gtavitational acceleration for particles Show that the maximum angular rotation of the tor
made of different substances — the subject of the sion pendulum from one side to the other during one
50 EXERCISE 2-13 DEFLECTION OF STARLIGHT BY SUN
EXERCISE 2 -1 2 , se co n d figure. Schem atic d ia g ra m o f the o f net torque on torsion p e n d u lu m in the evening com pared w ith the
D icke experim ent. L e ft; H yp o th etica l effect: m orning. R ig h t: H y m orning. The large a lu m in u m h a ll h as the sam e m ass as the sm a ll
p o th e tic a l effect: evening. A n y difference in the g r a v ita tio n a l accel h ig h -d en sity gold ball.
eration o f S u n fo r gold a n d a lu m in u m should result in opposite sense
rotation of Earth is given by the expression

2-13 deflection of starlight by
Sun
mg,L ( Ag''
e„ Esrimate the deflection of starlight by Sun using an
Ss
elementary analysis. D iscussion: Consider first a
j In practice Dicke’s torsion balance can be simpler example of a similar phenomenon. An eleva
thought of as consisting of 0.030-kilogram gold and tor car of width L is released from rest near the surface
aluminum bobs mounted on the ends of a beam 6 X of Earth. At the instant of release a flash of light is
10“ ^ meter in length suspended in a vacuum on a fired horizontally from one wall of the car toward the
quartz fiber of torsion consrant 2 X 10” ® newton other wall. After release the elevator car is an inertial
meter/radian. A statistical analysis of the angular frame. Therefore the light flash crosses the car in a
displacements of this torsion pendulum over long straight line with respect to the car. W ith respect to
periods of time leads to the conclusion that the frac Eatth, however, the flash of light is falling — because
the elevator is falling. Therefore a light flash is de
tion A^/g for gold and aluminum is less than 3 X
flected in a gravitation field, as Newton would phrase
10” *k To what mean maximum angle of rotation
from side to side during one rotation of Earth does this it. (How would Einstein phrase it? See Chapter 9.) As
correspond? Random motions of the torsion another example, a ray of starlight in its passage
pendulum — noise! — are of much greater amplitude tangentially across Earth’s surface receives a gravita
than this; hence the need for the statistical analysis of tional deflection (over and above any refraction by
the results. Earth’s atmosphere). However, the time to cross
Earth is so short, and in consequence the deflection so
References: R. H. Dicke, “The Eocvos Experiment,” Scientific
slight, that this effect has not yet been detected on
American, Volume 205, pages 8 4 —94 (December, 1961). See also
P. G. Roll, R. Krockov, and R. H. Dicke, Annals o f Physics, Volume Earth. At the sutface of Sun, however, the acceleration
26, pages 4 4 2 - 5 1 7 (1964). The first of these articles is a popular of gravity has the much greater value of 275 meters/
exposition written early in the course of the Dicke experiment. The second/second. Moreover, the time of passage across
second article reports the final results of the experiment and takes on the surface is much increased because Sun has a
added interest because of its account of the elaborate precautions
greater diameter, 1.4 X 10'-' meters. In the following,
required to insure that no influence that m ight affect the experiment
was disregarded. Galileo quote from Galileo Galilei, Dialogues Con
assume that the light just grazes rhe surface of Sun in
cerning Two New Sciences, translated by Henry Crew and Alfonso de passing.
Salvio (Northwestern University Press, Evanston, Illinois, 1950). a Determine an “effective time of fall’’ from the
EXERCISE 2-13 DEFLECTION OF STARLIGHT BY SUN 51
diameter of Sun and the speed of light. From this time its forward velocity, deduce the angle of deflection.
of fall deduce the net velocity of fall toward Sun The accurate analysis of special relativity gives the
produced by the end of the whole period of gravita same result. However, Einstein’s 1915 general rela
tional interaction. (The maximum acceleration acting tivity predicted a previously neglected effect, asso
for this “effective time” produces the same net effect ciated with the change of lengths in a gravitational
[calculus proofl} produced by the acmal acceleration field, that produces something like a supplementary
— changing in magnitude and direction along the refraction of the ray of light and doubles the predicted
path — in the entire passage of the ray through Sun’s deflection. [Deflection observed in 1947 eclipse of
field of force.) Sun; (9.8 ± 1.3) X 10“ ^ radian; in the 1952 eclipse:
b Comparing the lateral velocity of the light with (8.2 ± 0.5) X 10-6 radian.}
The name relativity theory was an unfortunate
choice: The relativity of space and time is not the essential
thing, which is the independence of laws of Nature from
the viewpoint of the observer.
Arnold Som m erfeld
3.1 THE PRINCIPLE OF RELATIVITY

fundamental science needs only a closed room
How do you know you are moving? Or at rest? In a car, you pause at a stoplight. You
see the car next to you easing forward. With a shock you suddenly realize that, instead,
your own car is rolling backward. On an international flight you watch a movie with
the cabin shades drawn. Can you tell if the plane is traveling at minimum speed or full
speed? In an elaborate joke, could the plane acmally be sitting still on the runway,
engines running? How would you know?
Everyday observations such as these form the basis for a conjecture that Einstein
Principle of Relativity:
raised to the status of a postulate and set at the center of the theory of special relativity.
With shad es draw n you cannot tell
He called it the P rin cip le o f R elativity. Roughly speaking, the Principle of your speed
Relativity says that without looking out the window you cannot tell which reference
frame you are in or how fast you are moving.
Galileo Galilei made the first known formulation of the Principle of Relativity.
Listen to the characters in his book:
SALVATIUS: Shut yourself up with some friend in the main cabin below decks on some
large ship, and have with you there some flies, butterflies, and other small flying animals.
Have a large bowl of water with some fish in it; hang up a bottle that empties drop by
drop into a wide vessel beneath it. W ith the ship standing still, observe carefully how the
little animals fly with equal speed to all sides of the cabin. The fish swim indifferently in
all directions; the drops fall into the vessel beneath; and, in throwing something to your
G alileo ; First known formulation
friend, you need throw it no more strongly in one direction than another, the distances
of Principle of Relativity
being equal; jumping with your feet together, you pass equal spaces in every direction.
When you have observed all these things carefully (though there is no doubt that when
the ship is standing still everything must happen in this way), have the ship proceed with
any speed you like, so long as the motion is unifotm and not fluctuating this way and that.
You will discover not the least change in all the effects named, nor could you tell from any
53
54 CHAPTER 3 SAME LAWS FOR ALL
of them whether the ship was moving or standing still. In jumping, you will pass on the
floor the same spaces as before, nor will you make larger jumps toward the stern than
toward the prow even though the ship is moving quite rapidly, despite the fact that during
the time that you are in the air the floor under you will be going in a direction opposite to
your jump. In throwing something to your companion, you will need no more force to get
it to him whether he is in the direction of the bow or the stern, with yourself situated
opposite. The droplets will fall as before into the vessel beneath without dropping toward
the stern, although while the drops are in the air the ship runs many spans. The fish in
their water will swim toward the front of their bowl with no more effort than toward the
back, and will go with equal ease to bait placed anywhere around the edges of the bowl.
Finally the butterflies and flies will continue their flights indifferently toward every side,
nor will it ever happen that they are concentrated toward the stern, as if tired out from
keeping up with the course of the ship, from which they will have been separated during
long intervals by keeping themselves in the air . . .
G A LILEO GALILEI
P isa, February 15, 1 5 6 4 — A rcetri, near Florence, J a n u a r y 8 , 1 6 4 2
“My portrait is now finished, a very good likeness, by an excellent hand.”

— September 22, 1635
* * *
“If ever any persons might challenge to be signally distinguished for their intellect from
other men, Ptolemy and Copernicus were they that had the honor to see farthest into and
discourse most profoundly of the World’s systems.”
* * *
“My dear Kepler, what shall we make of all this? Shall we laugh, or shall we cry?”
“When shall I cease from wondering?”

3.1 THE PRINCIPLE OF RELATIVITY 55
SAGREDUS: Although it did not occut to me to put these observations to the test when I
was voyaging, I am sure that they would take place in the way you desaibe. In
confirmation of this I remember having often found myself in my cabin wondering
whether the ship was moving or standing still; and sometimes at a whim I have supposed
it to be going one way when its motion was the opposite . . .
The Galilean Principle of Relativity is simple in this early formulation, yet not as
simple as it might be. In what way is it simple? Physics looks the same in a ship moving
uniformly as in a ship at rest. Relative uniform motion of the two ships does not affect
the laws of motion in either ship. A ball falling straight down onto one ship appears
from the other ship to follow a parabolic course; a ball falling straight down onto that
second ship also appears to follow a parabolic course when observed from the first ship.
The simplicity of the Galilean Principle of Relativity lies in the equivalence of the two
Earthbound frames and the symmetry between them.
In what way is this simplicity not as great as it might be? In Galileo’s account the
Extension of G a lileo 's reasoning
frames of reference are not yet free-float (inertial). To make them so requires only a
from ship to spaceship
small conceptual step; from two uniformly moving sea-going ships to two unpowered
spaceships. Then up and down, north and south, east and west, all become alike. A
ball untouched by force undergoes no acceleration. Its motion with respect to one
spaceship is as uniform as it is with respect to the other. This identity of the law of free
motion in all inertial reference frames is what one means today by the Galilean
Principle of Relativity.
Galileo could not by any stretch of the imagination have asked his hearer to place
himself in a spaceship in the year 1632. Yet he could have described the greater
simplicity of physics when viewed from such a vantage point. Bottles, drops of water,
and all the other test objects float at rest or move at uniform velocity. The zero
acceleration of every nearby object relative to the spaceship would have been intelligi
ble to Galileo of all people. Who had established more clearly than he that relative to
Earth all nearby objects have a common acceleration?
Einstein’s Principle of Relativity is a generalization of such experiments and many
other kinds of experiments, involving not only mechanics but also electromagnetism,
nuclear physics, and so on.
All the laws of physics are the same in every free-float (inertial) reference
Principle of Relativity
frame.
Einstein’s Principle of Relativity says that once the laws of physics have been estab
lished in one free-float frame, they can be applied without modification in any other
free-float frame. Both the mathematical form of the laws of physics and the numerical
values of basic physical constants that these laws contain are the same in every
free-float frame. So far as concerns the laws of physics, all free-float frames are
equivalent.
We can tell where we are on Earth by looking out of the window. Where we are in
the Milky Way we can tell by the configuration of the Big Dipper and other
constellations. How fast and in what direction we are going through the larger
framework of the universe we measure with a set of microwave horns pointed to pick
up the microwave radiation streaming through space from all sides. But now exclude
all information from outside. Screen out all radiation from the heavens. Pull down the
window shade. Then do whatever experiment we will on the movement and collision
of particles and the action of electric and magnetic forces in whatever free-float frame
we please. We find not the slightest difference in the fit to the laws of physics between
measurements made in one free-float frame and those made in another. We arrive at
the Principle of Relativity in its negative form;
No test of the laws of physics provides any way whatsoever to distinguish one Principle of Relativity,
free-float frame from another. negative form
BOX
THE PRINCIPLE OF RELATIVITY RESTS ON
EMPTINESS!
In his paper on special relativity, Einstein says, “We will raise this conjecture
(whose intent will from now on be referred to os the ‘Principle of Relativity’) to
o postulate . . . ” Is the Principle of Relativity just o postulate? All of special
relativity rests on it. How do we know it is true? What lies behind the Principle
of Relativity?
This is a philosophical question, not a scientific one. You will have your own
opinion; here is ours. We think the Principle of Relativity as used in special
relativity rests on one word: emptiness.
Space is empty; there are no kilometer posts or mileposts in space. Do you

want to measure distance and time? Then set up a latticework of meter sticks
and clocks. Pace off the meter sticks, synchronize the clocks. Use the lattice-
work to carry out your measurements. Discover the laws of physics. This
latticework is your construction, not Nature's. Do not ask Nature to choose
your latticework in preference to the similar latticework that I have con
structed. Why not? Because space is empty. Space accommodates both of us
as we go about our constructions and our investigations. But it does not
choose either one of us in preference to the other. How can it? Space is
empty. Nothing whatever can distinguish your latticework from mine. If we
decide in secret to exchange latticeworks. Nature will never be the wiser! It
follows that whatever laws of physics you discover employing your lattice-
work must be the same laws of physics I discover using my latticework. The
same is true even when our lattices move relative to one another. Which one
of us is at rest? There is no way to tell in empty space! This is the Principle of
Relativity.
But is space re a lly empty? “ Definitely not!” says modern quantum physics.
“ Space is a boiling cauldron of virtual particles. To observe this cauldron.
3.2 WHAT IS NOT THE SAME IN

DIFFERENT FRAMES
not the same: space separations,
time separations, velocities,
accelerations, forces, fields
Notice what the Principle of Relativity does say. It does not say that the time
between two events is the same when measured from two different free-float frames.
Sp ace and time separations
Neither does it say that space separation between the two events is the same in the two
not the sam e in different frames
frames. Ordinarily neither time nor space separations are the same in the two frames.
The catalog of differences between readings in the two frames does not end with
labotatory and rocket records of pairs of events. Physics to the Greeks meant the
science of change and so it does to us today. Motion gives us a stream of events, for
example the blinks of a firefly or the pulses of a sparkplug flashing as it moves. These
flashes trace out the sparkplug’s trajectory. Record the positions of two sequential
3.2 WHAT IS NOT THE SAME IN DIFFERENT FRAMES 57
sample regions of space much smaller than the proton. Carry out this sam
pling during times much shorter than the time it takes light to cross the diame
ter of the proton.” These words are familiar or utterly incomprehensible,
depending on the amount of our experience with physics. In either case, we
can avoid dealing with the “ boiling cauldron of virtual particles” by observ
ing events that are far apart compared with the dimensions of the proton,
events separated from one another by times long compared with the time it
takes light to cross the diameter of the proton.
In the realm of classical (nonquantum) physics is space really empty? “ Of

course not!” says modern cosmology. “Space is full of stars and dust and
radiation and neutrinos and white dwarfs and neutron stars and (many be
lieve) black holes. To observe these structures, sample regions of space
comparable in size to that of our galaxy. These structures evolve and move
with respect to one another in times comparable to millions of years.”
So we choose regions far from massive structures, avoid dust, ignore neu
trinos and radiation, and measure events that take place close together in
time compared with a million years.
Notice that for the very small and also for the very large, the “ regions”
described span both space and time — they are regions o f s p a c e tim e . “ Emp
tiness” refers to spacetime. Therefore we should have said from the begin
ning, “ S p a c e tim e is empty” — except for us and our apparatus — with limita
tions described above.
In brief, we can find “ effectively empty” regions of spacetime of spatial

extent quite a few orders of magnitude larger and smaller than dimensions of
our bodies and of time spread quite a few orders of magnitude longer and
shorter than times that describe our reflexes. In spacetime regions of this
general size, empty spacetime can be found. In empty spacetime the Principle
of Relativity applies. Where the Principle of Relativity applies, special relativ
ity correctly describes Nature.
spark emissions in the laboratory frame. Record also the laboratory time between these
sparks. Divide the change in position by the increase in time, yielding the laboratory-
measured velocity of the sparkplug.
Spark events have identities that rise above all differences between reference frames.
These events are recorded not only in the laboratory but also by recording devices and
clocks in the rocket latticework. From the printouts of the recorders in the rocket frame
we read off rocket space and time separations between sequential sparks. W e divide.
The quotient gives the rocket-measured velocity of the sparkplug. But both the space
separation and the time separation between events, respectively, are ordinarily differ
ent for the rocket frame than for the laboratory frame. Therefore the rocket-measured
velocity of the sparkplug is different from the laboratory-measured velocity of that Velocity not the some
sparkplug. Same world. Same motion. Different records of that motion. Figures for
velocity that differ between rocket and laboratory.
Apply force to a moving object: Its velocity changes; it accelerates. Acceleration is Acceleration not the some
the signal that force is being applied. Two events are enough to reveal velocity; three
reveal change in velocity, therefore acceleration, therefore force. The laboratory ob
server reckons velocity between the first and second events, then he reckons velocity
THE SPEED OF LIGHT

A ""fundamental constant of nature""?
O r a mere factor of conversion between two units of measurement?
METERS AND MILES IN THE SECONDS AND METERS IN SPACETIME

PARABLE OF THE SURVEYORS
Meter? Second?
Originally (adopted France, 1799) one ten-millionth of Originally 1/24 of 1/60 of 1/60 of the time from high
the distance along the surface of Earth from its equator noon one day to high noon the next day. Since 1967,
to its pole (in o curved line of latitude passing through ‘ ‘The second is the duration of 9,192,631,770 periods
the center of Paris). of the radiation corresponding to the transition be
tween the two hyperflne levels of the fundamental state
of the atom cesium 133.”
Mile? Meter?
Originally one thousand p a c e s — double step: right to Definition evolved from geographic to platinum meter
left to right— of the Roman soldier. rod to today’s ‘‘One meter is the distance traveled by
light, in vacuum, in the fraction 1/299,792,458 of a
second.”
Modern conversion factor? Modern conversion factor?

1609.344 meters per mile. 299,792,458 meters per second.
Authority for this number? Authority for this number?

Measures of equotor-to-pole distance eventually Meeting of General Conference on Weights and M ea
(1799 to today) lagged in accuracy compared to labo sures, 1983. In the accepted definition of the meter
ratory measurement of distance. So the platinum meter important changes took place over the years, and like
rod at Sevres, Paris, approximating one ten-millionth of wise in the definition of the second. With the 1983 defi
that distance, for awhile became — in and by itself — nition of the meter these two streams of development
the standard of distance. During that time the British have merged. What used to be understood as a mea
Parliament and the United States Congress redefined surement of the speed of light is understood today as
the inch to be e x a c t ly 2.54 centimeters. This decree two ways to measure separation in spacetime.
made the conversion factor (5280 feet/mile) times (12
inches/foot) times (2.54 centimeters/inch) times (1/100
of a meter per centimeter) equal to 1609.344 meters
per mile — exactly!
A fundamental constant of nature? A fundamental constant of nature?

Hardly! Rather, the work of two centuries of commit Hardly! Rather, the work of two centuries of commit
tees. tees.
between the second and third events. Subtracting, he obtains the change in velocity.
From this change he figures the force applied to the object.
The rocket observer also measures the motion; velocity between the first and second
events, velocity between second and third events; from these the change in velocity;
from this the force acting on the object. But the rocket-observed velocities are not
equal to the corresponding laboratory-observed velocities. The change in velocity also
Force not the some
differs in the two frames; therefore the computed force on the object is different for
3.2 WHAT IS NOT THE SAME IN DIFFERENT FRAMES 59
Commentary
Is the distance from Earth’s equator to its pole a fundamental constant of
nature? No. Earth is plastic and ever changing. Is the distance betNveen the
two scratches on the standard meter bar constant? No. Oxidation from
decade to decade slowly changes it. Experts in the art and science of mea
surement move to ever-better techniques. They search out an ever-better
object to serve as benchmark. Via experiment after experiment they move
from old standards of measurement to new. The goals? Accuracy. Availabil
ity. Dependability. Reproducibility.
Make a better measurement of the speed of light. Gain in that way better
knowledge about light? No. Win instead an improved value of the ratio
between one measure of spacetime interval, the meter, and another such
measure, the second — both of accidental and historical origin? Before
1983, yes. Since 1983, no. Today the meter is d e f in e d as the distance light
travels in a vacuum in the fraction 1/299,792,458 of a cesium-defined sec
ond. The two great streams of theory, definition, and experiment concerning
the meter and the second have finally been unified.
What will be the consequence of a future, still better, measuring technique?

Possibly it will shift us from the cesium-atom-based second to a pulsar-based
second or to a still more useful standard for the second. But will that improve
ment in precision change the speed of light? No. Every past International
Committee on Weights and Measures has operated on the principle of mini
mum dislocation of standards; we have to expect that the speed of light will
remain at the decreed figure of 299,792,458 meters per second, just as the
number of meters in the mile will remain at 1609.344. Through the fixity of this
conversion factor c, any substantial improvement in the accuracy of defining
the second will bring with it an identical improvement in the accuracy of
defining the meter.
Is 299,792,458 a fundamental constant of nature? Might as well ask if 5280

is a fundamental constant of nature!
rocket observer and laboratory observer. The Principle of Relativity does not deny that
the force acting on an object is different as reckoned in two frames in relative motion.
An electric field or a magnetic field or some combination of the two, acting on the
electron, is the secret of action of many a device doing its quiet duty day after day in
home, factory, or car. An electromagnetic force acting on an electron changes its Electric and magnetic fields
not the same
velocity as it moves from event P to event Q and from Q to R . Laboratory and rocket
observers do not agree on this change in velocity. Therefore they do not agree on the
value of the force that changes that velocity. Nor, finally, do they agree on the
magnitudes of the electric and magnetic fields from which the force derives.
In brief, figures for electric and magnetic field strengths, for forces, and for
accelerations agree no better between rocket and laboratory observers than do figures
for velocity. The Principle of Relativity does not deny these differences. It celebrates
them. It explains them. It systematizes them.
3.3 WHAT IS THE SAME IN DIFFERENT

FRAMES
the same: physical laws, physical constants in
those laws
Different values of some physical quantities between the two frames? Yes, but
Laws of physics the same
identical physical /a m / For example, the relation between the force acting on a particle
in different frames
and the change in velocity per unit time of that particle follows the same law in the
laboratory frame as in the rocket frame. The force is not the same in the two frames.
Neither is the change in velocity per unit time the same. But the law that relates force
and change of velocity per unit time is the same in each of the two frames. All the laws
of motion are the same in the one free-float frame as in the other.
Not only the laws of motion but also the laws of electromagnetism and all other
laws of physics hold as true in one free-float frame as in any other such frame. This is
what it means to say, “No test of the laws of physics provides any way whatsoever to
distinguish one free-float frame from another.”
Deep in the laws of physics are numerical values of fundamental physical constants,
such as the elementary charge on the electron and the speed of light. The values of
Fundamental constants the same
these constants must be the same as measured in overlapping free-float frames in
relative motion; otherwise these frames could be distinguished from one another and
the Principle of Relativity violated.
One basic physical constant appears in the laws of electromagnetism: the speed of
Speed of light the same
light in a vacuum, c = 299,792,458 meters per second. According to the Principle of
Relativity, this value must be the same in all free-float frames in uniform relative
motion. Has observation checked this conclusion? Yes, many experiments demon
strate it daily and hourly in every particle-accelerating facility on Earth. Nevertheless,
it has taken a long time for people to become accustomed to the apparently absurd
idea that there can be one special speed, the speed of light, that has the same value
measured in each of two overlapping free-float frames in relative motion.
Values of the speed of light as measured by laboratory and by rocket observer turn
out identical. This agreement has cast a new light on light. Its speed rates no longer as a
constant of nature. Instead, today the speed of light ranks as mere conversion factor
between the meter and the second, like the factor of conversion from the centimeter to
the meter. The value of this conversion factor has now been set by decree and the meter
defined in terms of it (Box 3.2). This decree assumes the invariance of the speed of
light. No experimental result contradicts this assumption.
In 1905 the PrincipleofRelativity was a shocking heresy. It offended most people’s
intuition and common-sense way of looking at Nature. Consequences of the Principle
of Relativity are tried out every day in many experiments where it is continually under
severe test. Never has this Principle been verified to lead to a single incorrect experi
mental prediction.
3.3 WHAT IS THE SAME IN DIFFERENT FRAMES 61
E X A M P L E S OF THE P R I N C I P L E OF
RELATIVITY
Two overlapping free-float frames are in uniform must necessarily be the same as measured in the
relative motion. According to the Principle of Rel- two frames? Which quantities are not necessarily
ativity, which of the quantities on the following list the same as measured in the two frames?
a. numerical value of the speed of light in a vacuum
b. speed of an electron
c. value of the charge on the electron
d. kinetic energy of a proton (the nucleus of a hydrogen atom)
e. value of the electric field at a given point
f. time between two events
g. order of elements in the periodic table
Newton’s First Law of Motion (“A particle initially at rest remains at rest,
and . . . ”)
SOLUTION
The speed of light IS necessarily the same in the two frames. This is one of the
central tenets of the Principle of Relativity and a basis of the theory of relativity.
b. The speed of an electron IS N O T necessarily the same in the two frames.
Determining the speed of a particle depends on space and time measurements
between events — such as flashes emitted by the particle. Space and time separa
tions between events, respectively, can be measured to be different for observers
in relative motion. So the speed— ratio of distance covered to time elapsed — can
be different.
The value of the charge on the electron IS necessarily the same in the two frames.
Suppose that the charge had one value for the laboratory frame and progressively
smaller values for rocket frames moving faster and faster relative to the laboratory
frame. Then we could detect the “absolute velocity’’ of the ftame we are in by
measuring the charge on the electron. But this violates the Principle of Relativity.
Therefore the charge on the electron must have the same value in all free-float
frames.
d. The kinetic energy of a proton IS N O T necessarily the same in the two frames.
The value of its kinetic energy depends on the speed of the proton. But speed is
not necessarily the same as measured in the two frames (b).
e. The value of the electric field at a given point IS N O T necessarily the same in the
two frames. The argument is indirect but inescapable: The electric field is
measured by determining the force on a test charge. Force can be measured by
change in velocity that the force imparts to a particle of known mass. But the
velocity— and the change in velocity — of a particle can be different for observers
in relative motion (b). Therefore the electric field may be different for observers in
relative motion.
The time between two events IS N O T necessarily the same in the two frames.
This is a direct result of the invariance of the interval (Chapter 1 and Section 3.7).
S A M P L E P R O B L E M 3 -1
g. The order of elements in the periodic table by atomic number IS necessarily the
same in the two frames. For suppose that the atomic number (the number of
protons in the nucleus) were smaller for helium than for uranium in the labora
tory frame but greater for helium than for uranium in the rocket frame. Then we
could tell which frame we were in by comparing the atomic numbers of helium
and uranium.
h. Newton’s First Law of Motion IS necessarily the same in the two frames.
Newton’s First Law is teally a definition of the inertial (free-float) frame. We
assume that all laboratory and rocket frames are inertial.
3.4 RELATIVITY OF SIMULTANEITY

''same lim e "? ordinarily true for only one
frame!
The Principle of Relativity directly predicts effects that initially seem strange — even
weird. Strange or not, weird or not; logical argument demonstrates them and experi
ment verifies them. One effect has to do with simultaneity: Let two events occur
separated in space along the direction of relative motion between laboratory and rocket
frames. These two events, even if simultaneous as measured by one observer, cannot be
simultaneous as measured by both observers.
Einstein demonstrated the relativity of simultaneity with his famous Train Paradox.
(When Einstein developed the theory of special relativity, the train was the fastest
common cartier.) Lightning strikes the front and back ends of a rapidly moving train,
leaving char marks on the train and on the track and emitting flashes of light that
travel forward and backward along the train (Figure 3-1). An observer standing on the
Train Paradox: Two lightning bolts
ground halfway between the two char marks on the track receives the two light flashes
strike simultaneously
for ground observer at the same time. He therefore concludes that the two lightning bolts struck the ttack
at the same time — with respect to him they fell simultaneously.
A second observer rides in the middle of the train. From the viewpoint of the
observer on the ground, the train observer moves toward the flash coming from the
front of the train and moves away from the flash coming from the rear. Therefore the
train observer receives the flash from the front of the train first.
This is just what the train observer finds: The flash from the front of the train arrives
Two lightning bolts do not
at her position first, the flash from the rear of the train arrives later. But she can verify
strike simultaneously
for train observer that she stands equidistant from the ftont and rear of the train, where she sees char
marks left by the lightning. Moreover, using the Principle of Relativiry, she knows
that the speed of light has the same value in her train frame as for the ground observer
(Section 3.3 and Box 3-2), and is the same for light traveling in both directions in her
frame. Therefore the arrival of the flash first from the front of the train leads her to
conclude that the lightning fell first on the front end of the train. For her the lightning
bolts did not fall simultaneously. (To allow the train observer to make only measure
ments with respect to the train, forcing her to ignote Earth, let the train be a cylinder
without windows — in other words a spaceship!)
Did the two lightning bolts strike the front and the back of the train simulta
neously? Or did they strike at different times? Decide!
Strange as it seems, there is no unique answer to this question. For the situation
described above, the two events are simultaneous as measured in the Earth frame; they
3.5 LORENTZ CONTRACTION OF LENGTH 63
□ □ □ □ □ □ □ □ □ □□ □ □ □ □ □ □ □ □ □ □ a
_______________ 1 C:H-J 1________ L t t ; . C A tJ____________ 0 P
— 0^-------------- 5
ân
-------- m r -m r C^T— -jj.
FIGURE 3 - 1 . Einstein’s T ra in Paradox illu stratin g the relativity o f sim ultaneity. Top: Light
ning strikes the front and hack ends of a moving train, leaving char marks on both track and train. Each
emittedflash spreads out in all directions. Center: Observer riding in the middle of the train concludes that
the two strokes are not simultaneous. Her argument: “(I) I am equidistant from the front and hack char
marks on the train. (2) Light has the standard speed in myframe, and equal speed in both directions. (3) The
flash arrivedfrom thefront of the train first. Therefore, (4) theflash must have left thefront of the train first;
the front lightning holt fell before the rear lightning bolt fell. I conclude that the lightning strokes were not
simultaneous.” Bottom: Observer standing by the tracks halfway between the char marks on the tracks
concludes that the strokes were simultaneous, since the flashes from the strokes reach him at the same time.
are not simultaneous as measured in the train frame. We say that the simultaneity of
Simultaneity is relative
events is, in general, relative, different for different frames. Only in the special case of
two or more events that occur at the same point (or in a plane perpendicular to the line
of relative motion at that point— see Section 3.6) does simultaneity in the laboratory
frame mean simultaneity in the rocket frame. When the events occur at different
locations along the direction of relative motion, thev cannot be simultaneous in both
frames. This conclusion is called the relativity o f sim ultaneity.
The relativity of simultaneity is a difficult concept to understand. Almost without
exception, every puzzle and apparent paradox used to “disprove” the theory of
relativity hinges on some misconception about the relativity of simultaneity, -te r'
3.5 LORENTZ CONTRACTION OF LENGTH

space separation between two length
measuring events? disagreement!
How do we measure the length of a moving rod — the distance between one end and
Length of a rod = separation
the other end? One way is to use our latticework of clocks to mark the location of the
between simultaneous sparks
two ends at the same time. But when the rod lies along the direction of relative motion, at its two ends
someone riding with the rod does not agree that our marking of the positions of the
two ends occurs at the same time (Section 3.4). The relativity of simultaneity tells us
that rocket and laboratory observers disagree about the simultaneity of two events
(firecrackers exploding at the two ends of the rod) that occur at different locations
along the direction of relative motion. Therefore the two observers disagree about
whether or not a valid measurement of length has taken place.
Go back to the Train Paradox. For the observer standing on the ground, the two
lightning bolts strike the front and back of the train at the same time. Therefore for
him the distance between the char marks on the track constitutes a valid measure of the
length of the train. In conrrast, rhe observer riding on the train measures rhe front
lightning bolt to strike first, the rear bolt later. The rider on the train exclaims to her
D isagree about simultaneity?
Earth-based colleague, “See here! Your front mark was made before the back mark
Then d isag ree about length.
— since rhe flash from the front reached me (at the middle of the train) before the flash
from the back reached me. O f course the train moved during the time lapse between
these two lightning strikes. By rhe time the stroke fell at the back of the train, the front
of the train had moved well past the front char mark on the track. Therefore your
measurement of the length of the train is too small. The train is really longer rhan you
measured.”
There are other ways to measure the length of a moving rod. Many of these methods
lead to the same result: the space separation between the ends of the rod is less as
measured in a frame in which rhe rod is moving than as measured in a frame in which
the rod is at rest. This effect is called L orentz co n tractio n . Section 5.8 examines the
Lorentz contraction quantitatively.
Suppose we agree to measure the length of a rod by determining the position of its
two ends at the same time. Then an observer for whom the rod is at rest measures the
rod to be longer than does any other observer. This “rest length” of the rod is often
called its p ro p e r length.
You keep using the word “measure.” Occasionally you say “observe.” You never talk
about th a t most delicate, sensitive, an d refined of our five senses: sight. Why not ju st
look an d see these remarkable relativistic ejfects?
W e have been careful to say that the relativity of simultaneity and the Lorentz
contraction are measured, not seen with the eye. Measurement employs the latticework
of rods and clocks that constitutes a free-float frame. As mentioned in Chapter 2,
seeing with the eye leads to confused images due to the finite speed of light. Stand in
an open field in the southern hemisphere as Sun sets in the west and full Moon rises in
the east: You see Moon as it was 1.3 seconds ago, Sun as it was eight minutes ago, the
star Alpha Centauri (nearest star visible to the naked eye) as it was 4.34 years ago,
the Andromeda nebula as it was 2 million years ago — you see them all now.
Similarly, light from the two separated ends of a speeding rod typically takes
different times to reach your eye. This relative time delay results in visual distortion
that is avoided when the location of each end is recorded locally, with zero or
minimal delay, by the nearest lattice clock. Visual appearance of rapidly moving
objects is itself an interesting study, but for most scientific work it is an unnecessary
distraction. To avoid this kind of confusion we set up the free-float latticework of
synchronized recording clocks and insist on its use — at least in principle!
Aha! Then I have caught you in a contradiction. Figure 3 - 1 shows lightning flashes
an d trains. Is this not a picture of w hat we would see w ith our eyes?
No. Strictly speaking, each of the three “pictures” in Figure 3-1 summarizes where
parts of the train are as recorded by the Earth latticework of clocks at a given instant
of Earth time. The position of each light flash at this instant is also recorded by the
clocks in the lattice. The summary of data is then given to a draftsman, who draws
the picture for that Earth time. To distinguish such a drafted picture from the visual
3.6 INVARIANCE OF TRANSVERSE DIMENSION 65
view, we will often refer to it as a plot. For example. Figure 3-1 (top) is the Earth
plot at the time when lightning bolts strike the two ends of the train.
Actually, all three plots in Figure 3-1 show approximately what you see through a
telescope when you are very far from the scene in a direction perpendicular ro the
direction of motion of the train and at a position centered on the action. At such a
remote location, light from all parts of the scene takes approximately equal times to
reach your eye, so you would see events and objects at approximately the same time
according ro Earth clocks. Of course, you receive this information later than it
actually occurs because of the time it takes light to reach you.
3.6 INVARIANCE OF TRANSVERSE

DIMENSION
''faster" does not mean "thinner" or "fatter"
A rocket ship makes many trips past the laboratory observer, each at successively
higher speed. For each new and greater speed of the rocket, the laboratory observer
measures its length to be shorter than it was on the trip before. This observed
contraction is lo n g itu d in al — along its direction of motion. Does the laboratory
observer also measure contraction in the transverse dimension, perpendicular to the
direction of relative motion? In brief, is the rocket measured to get thinner as well as Transverse dimension sam e for
shorter as it moves faster and faster? laboratory and rocket observers
The answer is No. This is confirmed experimentally by observing the width of
electron and proton beams traveling in high-energy accelerators. It is also easily
demonstrated by simple thought experiments.
S peeding-T rain T h o u g h t E xperim ent: Return to Einstein’s high-speed
railroad train seen end-on (Figure 3-2). Suppose the Earthbound observer measures
the train to get thinner as it moves faster. Then for the Earth observer the right and left
wheels of the train would come closer and closer together as the train speeds up, finally
slipping off between the tracks to cause a tertible wreck. In contrast, the train observer
regards herself as at rest and the tracks as speeding by in the opposite direction. If she
WRONG! WRONG!
in motion -
at rest
— H
ALLEGED "EARTH PLOT" ALLEGED "TRAIN PLOT"

FIGURE 3-2. T wo possible alternatives {both wrong!) i f the moving tra in is m easured to shrink
transverse to its direction of motion. The “E arth p lo t” assumes the speeding train to be measured as
getting thinner with increasing speed. The train’s wheels would slip o/'between the tracks. The “tra in
p lo t” of the same circumstance assumes the speeding rails to he measured as getting closer together. In this
case the wheels would slip offoutside the tracks. But this is a contradiction. Therefore the wheel separation
— and the transverse dimensions of train and track — must he invariant, the samefor allfree-float observers
moving along the track. (If you think that the actual transverse contraction might be too small to cause a
wreckfor the train shown, assume that both the wheels and the track are knife edges; the same argument still
applies.)
measures the speeding tracks to get closer together as they move faster and faster, the
train wheels will slip off outside the tracks, also resulting in a wreck. But this is absurd:
the wheels cannot end up between the tracks and outside the tracks under the same
circumstances. Conclusion: High speed leads to no measured change in transverse
dimensions — no observed thinning or fattening of fast objects. We are left with the
conclusion that high relative speed affects the measuted values of longitudinal dimen
sions but not transverse dimension: a welcome simplification!
Speeding-Pipes T h o u g h t E xperim ent: Start with a long straight pipe. Paint
one end with a checkerboard pattern and the other end with stripes. Cut out and
discard the middle of the pipe, leaving only the painted ends. Now hurl the ends
toward each other, with their cylindrical axes lying along a common line parallel to the
Thought experiments demonstrate
direction of relative motion (Figure 3-3). Suppose that a moving object is measured to
invariance of transverse dimension
be thinner. Then someone riding on the checkerboard pipe will observe the striped
pipe to pass inside her cylinder. All observers — everyone looking from the side — will
see a checkerboard pattern. In contrast, someone riding on the striped pipe will observe
the checkerboard pipe to pass inside his cylinder. In this case, all observers will see a
striped pattern. Again, this is absurd: All observers must see stripes, or all must see
checkerboard. The only tenable conclusion is that speed has no measurable effect on
transverse dimensions and the pipe segments will collide squarely edge on.
A simple question leads to an even more fundamental argument against the differ
ence of transverse dimensions of a speeding object as observed by different free-float
observers in relative motion: About w hat axis does the contraction take place?
We try to define an “axis of shrinkage’’ parallel to the direction of relative motion.
Can we claim that a speeding pipe gets thinner by shrinking uniformly toward an
“axis of shrinkage” lying along its center? Then what happens when two pipe
segments move along their lengths, side by side as a pair? Does each pipe shrink
separately, causing the clear space between them to increase? O r does the combina
tion of both pipes contract toward the line midway between them, causing the clear
space between them to decrease? Is the answer different if one pipe is made of lead
and the other one of paper? Or if one pipe is entirely in our imagination?
There is no logically consistent way to define an “axis of shrinkage.’’ Given the
direction of relative motion of two objects, we cannot select uniquely an “axis of
shrinkage” from the infinite number of lines that lie parallel in this direction. For
each different choice of axis a different pattern of distortions results. But this is
logically intolerable. The only way out is to conclude that there is no transverse
shrinkage at all (and, by a similar argument, no transverse expansion).
The above analysis leads to conclusions about events as well as about objects. A set
of explosions occurs around the perimeter of the checkerboard pipe. More: These
explosions occur simultaneously in this checkerboard frame. Then these events are
simultaneous also in the striped frame. How do we know? By symmetry! For suppose
the explosions were not simultaneous in the striped frame. Then which one of these
FIGURE 3 -3 . T w o id e n tic a l-size p ip e

segm ents h urtle to w a rd each other
along a common centerline. W h a t w ill
happen w hen they m eet? Here are tw o
possible a lte rn a tiv e s (both w rong!) i f a
m oving object is observed to sh rin k
transverse to direction o f motion.
WRONG! WRONG!
W h ich p ip e passes insid e the other?
m m m T he im p ossibility o f a consistent a n
sw er to th is question leads to the con
IJJJJJJJWJJJ clusion th a t n eith er p ip e can he m ea

sured to change transverse dim ension.
"CHECKERBOARD PLOT" "STRIPED PLOT"

3.7 INVARIANCE OF THE INTERVAL PROVED 67
events would occur first in the striped frame? The one on the right side of the pipe or
the one on the left side of the pipe? But “left” and “right” cannot be distinguished by
means of any physical effect: Each pipe is cylindrically symmetric. Moreover, space is
the same in all directions — space is isotropic, the same to right as to left. So neither
“ Sam e time” ag ree d on for
the event on the right side nor the event on the left side can be first. They must be
events sep arated only transverse
simultaneous. The same argument can be made for events at the “top” and “bottom” to relative motion
of the pipe, and for every other pair of events on opposite sides of the pipe. Conclusion:
If the explosions are simultaneous in the checkerboard frame, they must also be
simultaneous in the striped frame.
We make the following summary conclusions about dimensions transverse to the
direction of relative motion:
Dimensions of moving objects transverse to the direction of relative motion
are measured to be the same in laboratory and rocket frames (invariance of
transverse distance).
Two events with separation only transverse to the direction of relative
motion and simultaneous in either laboratory or rocket frame are simulta
neous in both.
3.7 INVARIANCE OF THE INTERVAL

PROVED
laboratory and rocket observers agree on
something important
The Principle of Relativity has a major consequence. It demands that the spacetime
Principle of Relativity lead s to
interval have the same value as measured by observers in every overlapping free-float invariance of spacetim e interval
frame; in brief, it demands “invariance of the interval.” Proof? Plan of attack:
Determine the separation in space and the separation in time between two events, E
and R, in the rocket frame. Then determine the quite different space and time
separations between the same two events as measured in a free-float laboratory frame.
Then look for — and find— what is invariant. It is the “interval.” Now for the details
(Figures 3-4 and 3-5).
Event E we take to be the reference event, the emission of a flash of light from the
central laboratory and rocket reference clocks as they coincide at the zero of time
(Section 2.6). The path of this flash is tracked by the recording clocks in the rocket
lattice. Riding with the rocket, we examine that portion of the flash that flies straight
“up” 3 meters to a mirror. There it reflects straight back down to the photodetector
located at our rocket reference clock, where it is received and recorded. The act of
reception constitutes the second event we consider. This event, R, is located at the
rocket space origin, at the same location as the emission event E. Therefore, for the
rocket observer, the space separation between event E and event R equals zero.
W hat is the time separation between events E and R in the rocket frame? The light
travels 3 meters up to the mirror and 3 meters back down again, a total of 6 meters of
distance. At the “standard” light speed of 1 meter of distance per meter of light-travel
time, the flash takes a total of 6 meters of time to complete the round trip. In
summary, for the rocket observer the event of reception, R, is separated from the event
of emission, £, by zero meters in space and 6 meters in time.
W hat are the space and time separations of events E and R measured in the
free-float laboratory frame? As measured in the laboratory, the rocket moves at high
speed to the right (Figures 3-4 and 3-5). The rocket goes so fast that the simple
FINISH FIGURE 3-4. Plot o f the flash p ath

as recorded in the laboratory
fram e. Time progressesfrom bottom to
top: W ell started: The flash (repre
sented as an asterisk) has been emitted
(event Ej from a moving rocket clock
(shown as a circle) that coincided with
a laboratory clock (shown as a square).
Reaching m irror and Home
stretch: The flash reaches a mirror
and reflects from it. The mirror moves
along in step with the rocket clock.
Finish: The flash is received (event
R J back at the same rocket clock, which
has moved in the laboratory frame to
coincide with a second laboratory
clock. Figure 3-5 shows the trajectory
time
REACHING MIRROR of the same flash in three different
free-float frames.
up-down track of the light in the rocket frame appears in the laboratory to have the
profile of a tent, with its right-hand corner— the place of reception of the light— 8
meters to the right of the starting point.
When does the event of reception, R, take place as registered in the laboratory
frame? Note that it occurs at the time 6 meters in the rocket frame. All we know about
everyday events urges us to say, “Why, obviously it occurs at 6 meters of time in the
laboratory frame too.” But no. More binding than preconceived expectations are the
demands of the Principle of Relativity. Among those demands none ranks higher than
this: The speed of light has the standard value 1 meter of distance in 1 meter of
light-travel time in every free-float frame.
Figure 3-6 punches us in the eye with this point: The light flash travels farther as
G re a te r distance of travel
for light flash: longer time! recorded in the laboratory frame than as recorded in the rocket frame. The perpendic
ular “altitude” of the mirror from the line along which the rocket reference clock
moves has the same value in laboratory frame as in rocket frame no matter how fast the
rocket — as shown in Section 3.6. Therefore on its slanted path toward and away from
the mirror the flash must cover more distance in the laboratory frame than it does in
the rocket frame. More distance covered means more time required at the “standard”
light speed. We conclude that the time between events E and R is greater in the
laboratory frame than in the rocket frame — a staggering result that stood physics on
its ear when first proposed. There is no way out.
In the laboratory frame the flash has to go “up” 3 meters, as before, and “down”
again 3 meters. But in addition it has to go 8 meters to the right: 4 meters to the right
while rising to hit the mirror, and 4 meters more to the right while falling again to the
receptor. The Pythagorean Theorem, applied to the right triangles of Figure 3-6, tells
3.7 INVARIANCE OF THE INTERVAL PROVED 69
LABORATORY PLOT
A
® 0 0 0 4 ^ 0 0 0 0
0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0
0 0 0 0 © 0 0 0 0
__________________________
ROCKET PLOT
A A A A A A A A A AAAAAAAA
A A A A A A A A A A A A A
AA A A A
SUPER-ROCKET PLOT
FIGURE 3-5. Plots of the p a th in space of a reflectedflash of light as m easured in three different
fram es, showing event E, emission of the flash, a n d event R , its reception after reflection. Squares,
circles, and triangles represent latticeworks of recording clocks in laboratory, rocket, and super-rocketframes,
respectively. The super-rocket frame moves to the right with respect to the rocket, and with such relative speed
that the event of reception, R , occurs to the left of the event of emission, E, as measured in the super-rocket
frame. The reflecting mirror is fixed in the rocket, hence appears to movefrom left to right in the laboratory
and from right to left in the super-rocket.
FIGURE 3-6. Laboratory plot of

the p a th o f the light flash. Theflash
rises 3 meters while it moves to the
right 4 meters. Then it falls 3 meters as
it moves an additional 4 meters to the
right. From the Pythagorean Theorem,
the total length of theflash path equals
3 meters plus 5 meters or 10 meters.
Therefore 10 meters of light-travel
time is the separation in time between
emission event E and reception event R
as measured in the laboratory frame.
'• R
70 CHAPTER 3 SAME LAWS FOR AIL
us that each slanted leg of the trip has length 5 meters;
(3 meters)^ + (4 meters)^ = (5 meters)^
Thus the total length of the trip equals 10 meters, definitely longer than the length of
the round trip, 6 meters, as observed in the rocket frame. Moreover, the light can cover
that slanted and greater distance only at the standard rate of 1 meter of distance in 1
meter of light-travel time. Therefore there is no escape from saying that the time of
reception as recorded in the laboratory frame equals 10 meters. Thus there is a great
variance between what is recorded in the two frames (Figure 3-5, Laboratory plot and
Rocket plot); separation in time and in space between the emission £ of a pulse of light
and its reception R after reflection.
In spite of the difference in space separation between events £ and R and the
difference in time lapse between these events as measured in laboratory and rocket
frames, there exists a measure of their separation that has the same value for both
observers. This is the interval calculated from the difference of squares of time and
space separations (Table 3-1). For both observers the interval has the value 6 meters.
The interval is an in v arian t between free-float frames.
Two central results are to be seen here, one of variance, the other of invariance. We
discover first that typically there is not and cannot be an absolute time difference
between two events. The difference in time depends on our choice of the free-ffoat
frame, which inertial frame we use to record events. There is no such thing as a simple
concept of universal and absolute separation in time.
Second, despite variance between the laboratory frame and the rocket frame in the
values recorded for time and space separations individually, the difference between the
Between events: No absolute time,
but invariant interval squares of those separations is identical, that is, invariant with respect to choice of
reference frame. The difference of squares obtained in this way defines the square of
the interval. The invariant interval itself has the value 6 meters in this example.
■ < ; ;] T A B L E 3 f r ^
RECKONING THE SPACETIME INTERVAL FROM

ROCKET AND LABORATORY MEASUREMENTS
Rocket Laboratory
measurements imeasurements
Time from emission 6 meters DIFFERENT! -♦ 10 meters

of the flash to its reception
Distance from the point of
emission of the flash to 0 meters ■<- DIFFERENT! ^ 8 meters
its point of reception
Square of time 36 (meters)^ 100 (meters)^

Square distance and
subttaa —0 (meters)^ -6 4 (meters)^
Result of subtraction 36 (meters)^ 36 (meters)^
This is the square of what 6 meters 6 meters
measurement?
f
SAME SPACETIME
INTERVAL
3.8 INVARIANCE OF THE INTERVAL FOR A ll FREE-FLOAT FRAMES 71
3.8 INVARIANCE OF THE INTERVAL FOR

A ll FREE-FLOAT FRAMES
super-rocket observer joins the agreement
The interval between two events has the same value for a//possible relative speeds of
overlapping free-float frames. As an example of this claim, consider a third free-float
frame moving at a different speed with respect to the laboratory frame— a speed
different from that of the rocket frame.
We now measure the same events of emission and reception from a “super-rocket
Super-rocket: Same interval
ftame’’ moving faster than the tocket (but not faster than light!) along the line
between events
between events E and R (Figure 3-5, Super-rocket plot). For convenience we arrange
that the tefetence clock of this frame also coincides with refetence clocks of the other
two frames at event E.
Events E and R occur at the same place in the rocket frame. Between these two
events the supet-tocket moves to the right with tespect to the tocket. As a result, the
supet-tocket observer records event R as occutring to the left of the emission event.
How far to the left? That depends on the relative speed of the super-tocket frame.
The super-rocket is not super-size; rather it has super-speed. W e adjust this
super-speed so that the reception occurs 20 meters to the left of the emission for the
super-rocket observer. Then the flash of light that rises vertically in the rocket must
ttavel the same 3 meters upward in the super-rocket but also 10 meters to the left as it
slants towatd the mirtor. Hence the distance it travels to the mitror in the supet-tocket
ftame is the length of a hypotenuse, 10.44 meters;
(3 meters)^ + (10 meters)^ = 9 meters^ + 1 0 0 meters^ = 1 0 9 meters^

= (10.44 meters)^
It must travel another 10.44 meters as it slants downwatd and leftwatd to the event of
reception. The total distance ttaveled equals 20.88 meters. It follows that the total
time lapse between E and R equals 20.88 metets of light-travel time for the super
rocket observer.
The speed of the supet-tocket is very high. As a result the space separation between
emission and reception is very great. But then the time separation is also very great.
Moreovet, the magnitude of the time sepatation is petfectly tailored to the size of the
space separation. In consequence, the particulat quantity equal to the difference of
their squares has the value (6 meters)^, no mattet how gteat the space separation and
time separation individually may be. For the super-rocket ftame:
(20.88 meters)^ ~ (20 meters)^ 436 meters^ ~ 400 meters^ — 36 meters^

(6 meters)^
In spite of the difference in space separation observed in the three frames (0 meters
for the rocket, 8 meters for the laboratory, 20 meters for the super-rocket) and the
difference in time separation (6 meters for the rocket, 10 meters for the laboratory,
20.88 meters for the super-rocket), the intetval between the two events has the same
value fot all three observers:
In general; (time sepatation)^ — (space separation)^ = (interval)^
Rocket ftame: (6 metets)^ — (0 meters)^ = (6 meters)^
Laboratory frame: (10 meters)^ — (8 meters)^ = (6 meters)^
Super-rocket ftame: (20.88 meters)^ — (20 metets)^ = (6 meters)^
FIGURE 3-6 (repeated). Labora

tory plot o f the path o f the light
flash.
The laboratory observer clocks the time between the flash and its reception as 10
meters, in total disagreement with the 6 meters of timelike interval he figures between
those two events. The observer in the super-rocket frame marks an even greater
discrepancy, 20.88 meters of her time versus the 6 meters of timelike interval. Only
for the rocket observer does clock time agree with interval. Why? Because only she sees
reception at the same place as emission.
The invariance of the interval can be seen at a glance in Figure 3-6. The hypotenuse
Invariance of interval from
of the first right triangle has a length equal to half the time separation between E and
invariance of transverse dimension
R. Its base has a length equal to half the space separation. To say that (time
separation)^ — (space separation)^ has a standard value, and consequently to state that
(half the time separation)^ — (half the space separation)^ has a standard value, is
simply to say that the altitude of this right triangle has a fixed magnitude (3 meters in
the diagram) for rocket and all super-rocket frames, no matter how fast they move.
And this altitude has a length equal to half the interval between these two events.
SAMPLE PROBLEM 3-2^

THE (C+ M E S O N
A beam of (unstable) mesons, traveling at a first counter records 1000 pulses (1000 passing
speed of t' = 0.868, passes through two counters 9 particles); the second records 250 counts (250
meters apart. The particles suffer negligible loss of passing particles). This decrease arises almost en
speed and energy in passing through the counters tirely from decay of particles in flight. Determine
but give electrical pulses that can be counted. The the half-life of the meson in its own rest frame.
SOLUTION
Unstable particles of different kinds decay at different rates. By definition, the half-life of
unstable particles of a particular species measures the particle wristwatch time during
which — on the average — half of the particles decay. Half of the remaining particles
decay in an additional time lapse equal to the same half-life, and so forth. In this case, one
quarter of the particles remain after passage from counter to counter. Therefore the
particles that survive experience the passage of two half-lives between counter and
counter. We make the interval between those two passages, those two events, the center
of our attention, because it has the same value in the laboratory frame where we do our
measuring as it does in the free-float frame of the representative particle.
3.9 SUMMARY 73
The keystone of the argument establishing the invariance of the interval between
Basis of invariance of interval:
two events for all free-float frames? The Principle of Relativity, according to which
Principle of Relativity
there is no difference in the laws of physics between one free-float frame and another.
This principle showed here in two very different ways. First, it said that distances at
right angles to the direction of relative motion are recorded as of equal magnitude in
the laboratory frame and the rocket frame (Section 3.6). Otherwise one frame could be
distinguished from the other as the one with the shorter perpendicular distances.
Second, the Principle of Relativity demanded that the speed of light be the same in
the laboratory frame as in the rocket frame. The speed being the same, the fact that the
light-travel path in the laboratory frame (the hypotenuse of two triangles) is longer
than the simple round-trip path in the rocket frame (the altitudes of these two
triangles: up 3 meters and down again) directly implies a longer time in the laboratory
frame than in the rocket frame.
In brief, one elementary triangle in Figure 3-6 displays four great ideas that underlie
all of special relarivity: invariance of perpendicular distance, invariance of the speed of
light, dependence of space and time separations upon the frame of reference, and
invariance of the interval.
3.9 SUMMARY
same laws for all; invariant interval for all
The P rin cip le o f R elativity says that the laws of physics are the same in every
inertial (free-float) reference frame (Section 3.1). This simple principle has important
consequences. Specifically:
separation V separation / separation V separation \ ^

( in lab I ( in lab I = I in moving- 1 ( in moving- I
time / position / Vparticle tim e/ particle position /
/ zero separation \ ^
9 meters of distance \ I in space (in
0.868 meters of distance 1 — lo f distance j ~ half-lives)^ — I particle frame)
per meter of time / \ between those
two events /
10.368 meters Y _ / 9 meters y

(2 half-lives)^
= (of light-travel time / \o f distance/
A little arithmetic tells us that two half-lives total 5.15 meters of light-travel time.
Consequently the half-life itself is 2.57 meters of time or (2.57 meters)/(3.00 X 10®
meters/second) = 8.5 X 1 0 ^ second or 8.5 nanoseconds.
1. Two events that lie along the direction of relative motion between two frames
cannot be simultaneous as measured in both frames (relativity o f sim u lta
neity). (Section 3-4)
2 . An object in high-speed motion is measured to be shorter along its direction of
motion than its p ro p e r length, measured in its rest frame (Lorentz co n
traction). (Section 3.5)
3. The dimensions of moving objects transverse to their direction of relative
motion are measured to be the same, whatever the relative speed (invariance
o f transverse distances). (Section 3.6)
4. Two events with separation only transverse to the direction of relative motion
and simultaneous in either frame are simultaneous in both. (Section 3.6)
FASTER THAN LIGHT?

We always want to go faster. Faster than what? Faster than anything has
gone before. What is our greatest possible speed, according to the theory of
relativity? The speed of light in a vacuum! How do we know that this is the
greatest possible speed that we can travel? Many lines of evidence reach this
conclusion. Rocket speed greater than the speed of light would lead to the
destruction of the essential relation between cause and effect, a result ex
plored in Special Topic: Lorentz Transformation (especially Box L-1) and in
Chapter 6. In particular, we could find a frame in which a faster-than-light
object arrives before it starts! Moreover, in particle accelerators built over
several decades we have spent hundreds of millions of dollars effectively
trying to accelerate electrons and protons to the greatest possible speed —
which by experiment never exceeds light speed.
The conclusion that no thing can mave faster than light arises also from the
invariance of the interval. To see this, let a rocket emit two flashes of light a
time t' apart as measured in the rocket frame. (Use a prime to distinguish
rocket measurements from laboratory measurements.) In the rocket frame
the two emissions occur at the same place: the separation x' between them
equals zero. Let f and x be the corresponding separations in time and space
as measured in the laboratary frame. Then the invariance of the interval tells
us that the three quantities f', t, and x are related by the equation
(t')2 - (x')2 = ItV - (0)2 = - x2

whence
(t')2 = f2 - x2 (3-1)
In the laboratory frame the rocket is moving with some speed; give this
speed the symbol v. The distance x between emissions is just the distance that
the rocket moves in time f in the laboratory frame. The relation between
3.9 SUMMARY 75
5. The spacetime interval between two events is invariant— it has the same
value in laboratory and rocket frames (Sections 3-7 and 3.8):
(interval)^ = (Vseparanon/V - f\separation/V
R ocket R ocket
_ / time y _ / space y
Vseparation/ \separation/
6. In any free-float frame, no object moves with a speed greater than the speed of
light (Box 3-3).
distance, time, and speed is
x = Vt (3-2)
Substitute this into equation (3-1) to obtain (t')^ = — (vt)^ = [1 — v^], or
f = t { ] — v^ )''2 (3-3)
Now, V is the speed of the rocket. How large can that speed be? Equation
(3-3) makes sense for any rocket speed less than the speed of light, or when v
has a value less than one.
Suppose we try to force the rocket to move faster than the speed of light. If we
should succeed, v would have a value greater than one. Then v^ also would
have a value greater than one. But in this case the expression 1 — v^ would
have a negative value and its square root would have no physical meaning.
In a formal mathematical sense, the rocket time f would be an imaginary
number for the case of rocket speed greater than the speed of light. But
clocks do not read imaginary time; they read real time—-three hours, for
example. Therefore a rocket speed greater than the speed of light leads to
an impossible consequence.
Equation (3-3) does not forbid a rocket to go as close to the speed of light as
we wish, as long as this speed remains less than the speed of light. For v very
close to the speed of light, equation (3-3) tells us that the rocket time can be
very much smaller than the laboratory time. Now suppose that emission of
the first flash occurs when the rocket passes Earth on its outward trip to a
distont star. Let emission of the second flash occur as the rocket a r r iv e s at that
distant star. No matter how long the laboratory time f between these two
events, we can find a rocket speed, v, such that the rocket time t' is as small as
we wish. This means that in principle we can go to any remote star in as short a
rocket time as we want. In brief, although our speed is limited to less than the
speed of light, the distance we can travel in a lifetime has no limitation. We
can go anywhere! This result is explored further in Chapter 4.
DOES A MOVING CLOCK R E A L L Y

“ RUN SLOW”?
You k e e p s a y in g , " T h e tim e b e t w e e n c lo ck -tic k s is s h o r t e r a s M E A S U R E D in th e
r e s t fra m e o f th e c lo c k th a n a s M E A S U R E D in a fra m e in w h ich th e c lo c k is
m o v in g ." I a m in t e r e s t e d in re a lity , n o t s o m e o n e 's m e a s u r e m e n t s . Tell m e w h a t
re a lly h a p p e n s !
' What is reality? You will have your own opinion and speculations. Here we
pose two related scientific questions whose answers may help you in forming
your opinion.
Are differences in clock rates really verified by experiment?

Different values of the time between two events as observed in different
frames? Absolutely! Energetic particles slam into solid targets in accelerators
all over the world, spraying forward newly created particles, some of which
decay in very short times as measured in their rest frames. But these “ short
lived” particles survive much longer in the laboratory frame as they streak
from target to detector. In consequence, the detector receives a much larger
fraction of the undecayed fast-moving particles than would be predicted
from their decay times measured at rest. This result has been tested thou
sands of times with many different kinds of particles. Such experiments
carried out over decades lead to dependable, consistent, repeatable re
sults. As far as we can tell, they are correct, true, and reliable and cannot
effectively be denied. If that is what you personally mean by "real,” then
these results are “what really happens.”
Does something about a clock really change when it moves, resulting in

the observed change in tick rate?
Absolutely not! Here is why: Whether a free-float clock is at rest or in motion
in the frame of the observer is controlled by the observer. You want the clock
REFERENCES
Introductory quote; A. Sommerfeld, Naturwissenschaftliche Rundschau, Volume
1, pages 9 7 -1 0 0 , reprinted in Gesammelte Schriften (Vieweg, Braunschweig,
1968), Volume IV, pages 6 4 0 -6 4 3 .
Galileo quote. Section 3.1; Galileo Galilei, Dialogue Concerning the Two Chief
World Systems— Ptolemaic and Copemican, first published February 1632; the
translation quoted here is by Stillman Drake (University of California Press,
Berkeley, 1962), pages 186ff. Galileo’s writings, along with those of Dante, by
reason of their strength and aptness, are treasures of human thought, studied
today in Italy by secondary school students as part of a great literary heritage.
Einstein quote. Box 3-1; Albert Einstein, “On the Electrodynamics of Moving
Bodies,’’ Annalen derPhysik, Volume 17, pages 8 9 1 -9 2 1 (1905), translated by
Arthur I. Miller in Albert Einstein’s Special Theory of Relativity (Addison-
Wesley, Reading, Mass., 1981), page 392.
ACKNOWLEDGMENTS 77
to be at rest? Move along with it! Now do you wont the clock to move? Simply
change your own velocity! This is true even when you and the clock are
separated by the diameter of the solar system. The magnitude of the clock's
steady velocity is entirely under your control. Therefore the time between its
ticks as measured in your frame is determined by your actions. How can your
change of motion affect the inner mechanism of a distant clock? It cannot and
does not.
Every time you change your motion on Earth — and even when you sit down,
letting the direction of your velocity change as Earth rotates — you change
the rate at which the planets revolve around Sun, as measured in your frame.
(You also change the shape of planetary orbits, contracting them along the
direction of your motion relative to Sun.) Do you think this change on your
velocity really affects the workings of the “ clock” we call the solar system? If
so, what about a person who sits down on the other side of Earth? That
person moves in the opposite direction around the center of Earth, so the
results are different from yours. Are each of you having a different effect on
the solar system? And are there still different effects — different solar-system
clocks — for observers who could in principle be scattered on other planets?
We conclude that free-float motion does not affect the structure or operation
of clocks (or rods). If this is what you mean by reality, then there are re a lly no
such changes due to uniform motion.
Is there some unity behind these conflicting measurements of time and space?
Yes! The interval: the proper time (wristwatch.time) between ticks of a clock as
measured in a frame in which ticks occur at the same place, in which the clock
is at rest. Proper time can also be calculated by all free-float observers,
whatever their state of motion, and all agree on its value. Behind the confus
ing clutter of conflicting measurements stands the simple, consistent, power
ful view provided by spacetime.
ACKNOWLEDGMENTS
The idea for Box 3-1 was suggested by Kenneth L. Laws. Box 3-4 and the
argument fot Section 3.6, Invariance of Transverse Dimension, is adapted from
material by William A. Shurcliff, private communications. Sample Problem 3-2
is adapted from A. P. French, Special Relativity (W. W. Norton, New York,
1968), page 121,
78 EXERCISE 3-1 RELATIVITY AND SWIMMING
CHAPTER 3 EXERCISES
her. If she runs with nearly the speed of light, will she
PRACTICE be able to see herself in the mirror? Analyze this
question using the Principle of Relativity.
3-1 relativity and swimming
The idea here is to illustrate how remarkable is the 3-3 construction of clocks
invariance of the speed of light (light speed same in all For the measurement of time, we have made no dis
free-float frames) by contrasting it with the case of a tinction among spring clocks, quartz crystal clocks,
swimmer making her way through water. biological clocks (aging), atomic clocks, radioactive
Light goes through space at 3 X 10® meters/sec- clocks, and a clock in which the ticking element is a
ond, and the swimmer goes through the water at 1 pulse of light bouncing back and forth between two
meter/second. “ But how can there otherwise be any mirrors (Figure 1-3). Let all these clocks be adjusted
difference?” one at first asks oneself. by the laboratory observer to run at the same rate
For a light flash to go down the length of a 30- when at rest in the laboratory. Now ler the clocks all
meter spaceship and back again takes be accelerated gently to a high speed in a rocket,
which then turns off irs engines. Make a simple bur
time = (distance)/(speed) powerful argument that the free-float rocket observer
= 2 X (30 meters)/(3 X 10® meters/second) will also measure these different clocks all to run ar
= 2 X 10~^ second the same rate as one another. Does it follow that the
(common) clock rate of these clocks measuted by the
as measured in the spaceship, regardless of whether rocket observer is rhe same as their (common) rate
the ship is stationary at the spaceport or is zooming measured by rhe laboratory observer as they pass by in
past it at high speed. the rocket?
Check how very different the story is for the swim
mer plowing along at 1 meter/second with respect to 3-4 the Principle of Relativity
the water.
Two overlapping free-float frames are in uniform
a How long does it take her to swim down the
relative motion. On the following list, mark with a
length of a 30-meter pool and back again?
“yes” the quantities that must necessarily be the same
b How long does it take her to swim from float A
as measured in the two frames. Mark with a “no” the
to float B and back again when the two floats, A and
quantities that are not necessarily the same as mea
B, are still 30 meters apart, but now are being towed
sured in the two frames.
through a lake at 1/3 meter/second? D iscussion:
a time it takes for light to go one meter of dis
When the swimmer is swimming in the same direc
tance in a vacuum
tion in which the floats are being towed, what is her
b spacetime interval between two events
speed relative to the floats? And how great is the
c kinetic energy of an electron
distance she has to travel expressed in the “frame of
d value of the mass of the electron
reference” of the floats? So how long does it take to
e value of the magnetic field at a given point
travel that leg of her trip? Then consider the same
f distance between two events
three questions for the return trip.
g structute of the DNA molecule
c Is it true that the total time from A to 6 and
h time rate of change of momentum of a neutron
back again is independent of the reference system
(“stationary” pool ends vs. moving floats)?
d Express in the cleanest, clearest, sharpest one-
3-5 many unpowered rockets
sentence formulation you can the difference between In rhe laboratory frame, event 1 occurs at x = 0
what happens for the swimmer and what happens for light-years, / = 0 years. Event 2 occurs at x = 6
a light flash. light-years, / = 10 years. In all rocket frames, event 1
also occurs at the position 0 light-years and the time 0
years. They- and z-coordinates of both events are zero
3-2 Einstein puzzler
in both frames.
When Albert Einstein was a boy of 16, he mulled a In rocker frame A, event 2 occurs ar rime t' =
over the following puzzler: A runner looks at herself 14 years. At what position x ' will event 2 occur in rhis
in a mirror that she holds at arm’s length in front of frame?
EXERCISE 3-7 SPACE WAR 79
b In rocket frame B, event 2 occurs at position x" d ‘‘Relativity postulates that light travels with a
= 5 light-years. At what time f will event 2 occur in standard speed regardless of the free-float frame from
this frame? which its progress is measured. This posmlate is cer
c How fast must rocket frame C move if events 1 tainly wrong. Anybody with common sense knows
and 2 occur at the same place in this rocket frame? that travel at high speed in the direction of a receding
d W hat is the time between events 1 and 2 in light pulse will decrease the speed with which the
rocket frame C of part c? pulse recedes. Hence a flash of light cannot have the
same speed for observers in relative motion. With this
3-6 down with relativity! disproof of the basic postulate, all of relativity col
lapses.”
Mr. Van Dam is an intelligent and reasonable man
e ‘‘There isn’t a single experimental test of the
with a knowledge of high school physics. He has the
results of special relativity.”
following objections to the theory of relativity. An
f ‘‘Relativity offers no way to describe an event
swer each of Mr. Van Dam ’s objections decisively —
without coordinates — and no way to speak about
without criticizing him. If you wish, you may present
coordinates without referring to one or another par
a single connected account of how and why one is
ticular reference frame. However, physical events
driven to relativity, in which these objections are all
have an existence independent of all choice of coordi
answered.
nates and all choice of reference frame. Hence
a ‘‘Observer A says that B’s clock goes slow, and
relativity— with its coordinates and reference frames
observer B says that A’s clock goes slow. This is a
— cannot provide a valid description of these
logical contradiction. Therefore relativity should be
events.”
abandoned.”
g ‘‘Relativityis preoccupied with how we observe
b ‘‘Observer A says that B’s meter sticks are
things, not what is really happening. Hence it is not a
contracted along their direction of relative motion,
scientific theory, since science deals w ith reality.”
and observer B says that A’s meter sticks are con
tracted. This is a logical contradiction. Therefore rela
tivity should be abandoned.”
c ‘‘Relativity does not even have a unique way to
PROBLEMS
define space and time coordinates for the instanta
3-7 space war
neous position of an object. Laboratory and rocket
observers typically record different coordinates for this Two rockets of equal rest length are passing ‘‘head
position and time. Therefore anything relativity says on” at relativistic speeds, as shown in the figure (left).
about the velocity of the object (and hence about its Observer o has a gun in the tail of her rocket pointing
motion) is without meaning.” perpendicular to the direction of relative motion
EXERCISE 3-7. U ff. Two rocket ships passing at high speed. Center: In the frame ofo one expects a bullet
fired when a coincides with a.' to miss the other ship. Right: In theframe of o' one expects a bullet fired when
a coincides with a! to hit the other ship.
80 EXERCISE 3-7 SPACE WAR
(center). She fires the gun when points a and a'

coincide. In her frame the other rocket ship is Lorentz
contracted. Therefore o expects her bullet to miss the
other rocket. But in the frame of the other observer o'
it is the rocket ship of o that is measured to be Lorentz
contracted (right). Therefore when points a and a'
coincide, observer o' should observe a hit.
Does the bullet actually hit or miss? Pinpoint the
looseness of the language used to state the problem EXERCISE 3-8, first figure. Calculation of Cerenkov angle 0 .
and the error in one figure. Show that your argument
is consistent with the results of the Train Paradox EXERCISE 3-8, second figure. Use of Cerenkov radiation for
indirect detection of neutrinos in the Deep Underwater Muon and
(Section 3.4).
Neutrino Detector (DUMAND) 3 0 kilometers offKeahole Point on
the island of Hawaii. Neutrinos have no electric charge and their
mass, if any, has so fa r escaped detection (Box 8-1). Neutrinos
3-8 <£erenk< idialii interact extremely weakly with matter, passing through Earth with
almost no collisions. Indeed, the DUMAND detector array selects
No particle has been observed to travel faster than the for analysis only neutrinos that come upward through Earth. In this
speed of light in a vacuum. However particles have way Earth itself acts as a shield to eliminate all other cosmic-ray
been observed that travel in a material medium faster particles.
What are possible sourcesfor these neutrinos? Theory predicts the
than the speed of light in that medium. When a
emission of very high-energy {greater than 1 0 ’^ electron-volt) neu
charged particle moves through a medium faster than trinosfrom matter plunging toward a black hole. Black holes may be
light moves in that medium, it radiates coherent light the energy sources for extra-bright galactic nuclei and for quasars
in a cone whose axis lies along the path of the particle. — small, distant, enigmatic objects shining with the light of
(Note the rough similarity to waves created by a hundreds of galaxies (Section 9.8). Information about conditions
motorboat speeding across calm water and the more deep within these astronomical structures may be carried by neu
trinos as they pierce Earth and travel upward through the DU
exact similarity to the “cone of sonic boom” created MAND detector array.
by a supersonic aircraft.) This is called Cerenkov radi In a rare event, a neutrino moving through the ocean slams into
ation (Russian C is pronounced as “ch”). Let v be the one of the quarks that make up a proton or a neutron in, say, an
speed of the particle in the medium and be the oxygen nucleus in the water, creating a burst of particles. All of
speed of light in the medium. these particles are quickly absorbed by the surrounding water except
a stable negatively charged muon, 2 0 1 times the mass of the electron
a From this information use the first figure to (thus sometimes called a “fa t electron”). This muon streaks through
show that the half-angle 0 , of the light cone is given the water in the same direction as the neutrino that created it and at
by the expression a speed greater than that of light in water, thus emitting Cerenkov
radiation. The Cerenkov radiation is detected by photomultiplier
cos 0 = tubes in an array anchored to the ocean floor.
Photomultipliers are strung along 9 vertical cables, 8 cables
spaced around a circle 100 meters in diameter on the oceanfloor, the
b Consider the plastic with the trade name Lu- ninth cable rising from the center of the circle. Each cable is 3 3 5
cite, for which v^^^^ = 2 /3 . W hat is the minimum meters long and holds 2 4 glass spheres positioned 10 meters apart on
velocity that a charged particle can have if it is to the top 2 3 0 meters of its length. There are no detectors on the bottom
n o meters, in order to avoid any cloud ofsedimentsfrom the bottom.
produce Cerenkov radiation in Lucite? W hat is the
Above the bottom, the water is so clear and modem photodetectors so
maximum angle 0 at which Cerenkov radiation can sensitive that Cerenkov radiation can he detectedfrom a muon that
be produced in Lucite? Measurement of the angle passes within 4 0 meters of a detector.
provides a good way to measure the velocity of the Photomultipliers in the glass spheres detect Cerenkov radiation
particle. from the passing muons, transmitting this signal through under
water optical fibers to computers on the nearby island of Hawaii.
C In water the speed of light is approximately
The computers selectfor examination only those events in which (I )
flight “ 0. 75. Answer the questions of part b for the several optical sensors detect hursts that are (2) within 4 0 meters or
case of water. See the second figure for an application so of a straight line, (3) spaced in time to show that the particle is
of Cerenkov radiation in water. moving at essentially the speed of light in a vacuum, and (4) from a
particle moving upward through the water. A system of sonar bea
cons and hydrophones tracks the locations of the photomultipliers as
the strings sway with the slow ocean currents. As a result, the
3-9 aberration of starlight direction of motion of the original neutrino can be recorded to an
A star lies in a direction generally perpendicular to accuracy of one degree.
The DUMAND facility is designed to create a new sky map of
Earth’s direction of motion around Sun. Because of neutrino sources to supplement our knowledge of the heavens, sofa r
Earth’s motion, the star appears to an Earth observer obtained primarily from the electromagnetic spectrum (radio, infra
to lie in a slightly different direction than it would red, optical, ultraviolet. X-ray, gamma ray).
EXERCISE 3-9 ABERRATION OF STARLIGHT 81
distance moved by sun

in one meter of light-
travel time
distance moved distance moved

by photon in by photon in
one meter of one meter of
light-travel light-travel
time time
SUN FRAME EARTH FRAME

(In this frame, Earth moves
to right with speed
EXERCISE 3-9. Aberration of starlight. Not to scale.
appear to an observer at rest relative to Sun. This

effect is called a b e rra tio n . Using the diagram, find
this apparent difference of direction.
a Find a trigonometric expression for the aberra
tion angle Xj/ shown in the figure.
b Evaluate your expression using the speed of
Earth around Sun, = 30 kilometers/second.
i- 0) Find the answer in radians and in seconds of arc. (One
-£■6
<D u
E degree equals 60 minutes of arc; one minute equals
o s
8 I 60 seconds of arc.) This change in apparent position
>o 3 . can be detected with sensitive equipment.
c The nonrelativistic answer to this problem —
the answer using nonrelativistic physics— is tan Xj/ =
*'Earth meters/metet). Do you think that the exper
imental difference between relativistic and nonrela
tivistic answers for stellar aberration observed from
Earth can be the basis of a crucial experiment to decide
between the correctness of the two theories?
D iscussion: O f course we cannot climb off Earth
and view the star from the Sun frame. But Earth
reverses direction every six months (with respect to
what?), so light from a “transverse star” viewed in,
o say, July will appear to be shifted through twice the
E aberration angle calculated in part b compared with
>
no
n the light from the same star in January. New ques
tion: Since the background of stars behind the one
under observation also shifts due to aberration, how
can the effect be measured at all?
d A rocket in orbit around Earth suddenly
changes its velocity from a very small fraction of the
tJ L speed of light to t' = 0.5 with respect to Sun, moving
in the same direction as Earth is moving around Sun.
In what direction will the rocket astronaut now see the
star of parts a and b?
82 EXERCISE 3-10 THE EXPANDING UNIVERSE
3-10 the expanding universe part a to describe how the observed period of light in
one spectral line from starlight can be compared to the
a A giant bomb explodes in otherwise empty
proper period of light in the same spectral line from
space. W hat is the nature of the motion of one frag
atoms at rest in the laboratory to give the velocity of
ment relative to another? And how can this relative
recession of the star that emits the light. This observed
motion be detected? D iscussion: Imagine each frag
change in period due to the velocity of the source is
ment equipped with a beacon that gives off flashes of
called the Doppler shift. (For a more detailed treat
light at regular, known intervals A t of time as mea
ment of Doppler shift, see the exercises for Chapters 5
sured in its own frame of reference (proper time!).
and 8.) If the universe began in a gigantic explosion,
Knowing this interval between flashes, what method
how must the observed velocities of recession of dif
of detection can an observer on one fragment employ
ferent stars at different distances compare with one
to determine the velocity v — relative to her — of any
another? Slowing down during expansion — by grav
other fragment? Assume that she uses, in making this
itational attraction or otherwise— is to be neglected
determination, (1) the known proper time A t be
here but is considered in more complete treatments.
tween flashes and (2) the time between the
c The brightest steadily shining objects in the
arrival of consecutive flashes at her position. (This is
heavens are called quasars, which stands for “quasi-
not equal to the time A/ in her frame between the
stellar objects.’’ A single quasar emits more than 100
emission of the two flashes from the receding emitter;
times the light of our entire galaxy. One possible
see the figure.) Derive a formula for v in terms of
source of quasar energy is the gravitational energy
proper time lapse A t and A/^q,n„„. How will the
released as material falls into a black hole (Section
measured recession velocity depend on the distance
9.8). Because they are so bright, quasars can be ob
from one’s own fragment to the fragment at which
served at great distances. As of 1991, the greatest
one is looking? Hint: In any given time in any given
observed quasar red shift A t has the value
frame, fragments evidently travel distances in that
5.9. According to the theory of this exercise, what is
frame from the point of explosion that are in direct
the velocity of recession of this quasar, as a fraction of
proportion to their velocities in that frame.
the speed of light?
b How can observation of the light from stars be
used to verify that the universe is expanding? D is
cussion: Atoms in hot stars give off light of different
3-11 law off addition off
frequencies characteristic of these atoms (“spectral
velocities
lines’’). The observed period of the light in each spec In a spacebus a bullet shoots forwatd with speed 3 /4
tral line from starlight can be measured on Earth. that of light as measured by travelers in the bus. The
From the pattern of spectral lines the kind of atom spacebus moves forward with speed 3/ 4 light speed
emitting the light can be identified. The same kind of as measured by Earth observers. How fast does the
atom can then be excited in the laboratory to emit bullet move as measured by Earth observers: 3 /4 +
light while at rest and the ptoper period of the light in 3 /4 = 6 /4 = 1.5 times the speed of light? No! Why
any spectral line can be measured. Use the results of not? Because (1) special relativity ptedicts that noth-
EXERCISE 3-10. Calculation of the time between arrival at observer of consecutive flashes from
receding emitter. Light moves one meter of distance in one meter of time, so lines showing motion of light are
tilted at 43 ° from the vertical.
EXERCISE 3-11 LAW OF ADDITION OF VELOCITIES 83
ing can travel faster than light, and (2) hundreds of frame and the bullet speed, call it (with a
millions of dollars have been spent accelerating parti prime), as measured in the spacebus frame. The times
cles (“bullets”) to the fastest possible speed without given in parts a, b, and c are of no use to this end.
anyone detecting a single particle that moves faster Worse, we already know that times between events
than light in a vacuum. Then where is the flaw in our are typically different as measured in the spacebus
addition of velocities? And what is the correct law of frame than times between the same events measured
addition of velocities? These questions are answered in in the Earth frame. So get rid of these times! More
this exercise. over, the Lorentz-contracted length L of the spacebus
a First use Earth observers to record the mo itself as measured in the Earth frame will be different
tions of the spacebus (length L measured in the Earth from its rest length measured in the bus frame (Sec
frame, speed and the streaking bullet (speed tion 3.5). So get rid of L as well. Equations (1), (2),
^'bullet)- The bullet starts at the back of the bus. To and (3) can be treated as three equations in the three
give it some competition, let a light flash (speed = unknowns /foârd- Aackward> ^nd L. Substitute equa
1) race the bullet from the back of the bus toward tions for the times (1) and (2) into equation (3).
the front. The light flash wins, of course, reaching Lucky us: The symbol L cancels out of the result.
the front of the bus in time ^forward is also Show that this result can be written
equal to the distance that the light travels in this
time. Show that this distance (measured in the Earth (1 ~ ^^buUet) (1 + y,,i)
frame) equals the length of the bus plus the distance / = (4 )
(1 + i'buUet) (1 “ «"rel)
the bus travels in the same time;
e Now repeat the development of parts a

^forward ^ ^rel ^forward ^forward ( 1) through d for the spacebus frame, with respect to
1
which the spacebus has its rest length L ' and the
bullet has speed t^^buu« (both with primes). Show that
b In order to rub in its advantage over the bul the result is:
let, the light flash reflects from the front of the bus
and moves backward until, after an additional time
^backward! tejoios the forward-plodding bullet. This ( f bullet)
(5 )
/ =
meeting takes place next to the seat occupied by (1 + /''buUet)
Fred, who sits a distance fL behind the front of the
bus, where / is a fraction of the bus length L. Show Discussion: Instead of working hard, work
that for this leg of the trip the Earth-measured dis smart! Why not use the old equations (1) through (4)
tance /backward traveled by the light flash can also be for the spacebus frame? Because there is no relative
expressed as velocity in the spacebus frame; the spacebus is at
rest in its own frame! No problem: Set = 0 in
^backward ^rel ^backward equation (4), replace «^buU«by /^^buoiet ^nd obtain equa
fL tion (5) directly from equation (4). If this is too big a
^backward
\ V,rel
step, carry out the derivation from the beginning in
the spacebus frame.
C The light flash has moved forward and then f Do the two fractions/in equations (4) and (5)
backward with respect to Earth. W hat is the net have the same value? In equation (4) the number /
forwatd distance coveted by the light flash at the locates Fred’s seat in the bus as a fraction of the total
instant it tejoins the bullet? Equate this with the for length of the bus in the Earth frame. In equation (5)
ward distance moved by the bullet (at speed t^buU«) to the number/locates Fred’s seat in the bus as a fraction
obtain the equation of the total length of the bus in the bus frame. But this
fraction must be the same: Fred cannot be halfway
back in the Earth frame and, say, three quarters of the
/'bulletf^forw ard ^backward *forward
way back in the spacebus frame. Equate the two
or expressions for/given in equations (4) and (5) and
solve for to obtain the Law of Addition of
Velocities:
(1 ^bullet^ ^backward ^^ ^bullet) ^forward
d W hat are we after? We want a relation be bullet /'rel

/ ’bullet ( 6)
tween the bullet speed t'buUet measured in the Earth 1+ bullet ^rel
84 EXERCISE 3-12 MKHELSON - MORLEY EXPERIMENT
g Explore some consequences of the Law of Ad b The same airplane now makes a round trip
dition of Velocities. between A and C. The distance between A and C is
the same as the distance from A to 6, but the line from
(1) An express bus on Earth moves at 108
A to C is perpendiculat to the line from A to 6, so that
kilometers/hour (approximately 67 miles/
in moving between A and C the plane flies across the
hour or 30 meters per second). A bullet moves
wind. Show that the round-trip time between A and
forward with speed 600 meters/second with
C under these circumstances is greater by a factor
respect to the bus. W hat are the values of
1/(1 — rd/f2)i/2 than the corresponding round-trip
and t'^buUet in meters/meter? W hat is the value
time in still air.
of their product in the denominator of equa
c Two airplanes with the same air speed c start
tion (6)? Does this product of speeds increase
from A at the same time. One travels from A to B and
the value of the denominator significantly over
back to A, flying first against and then with the wind
the value unity? Therefore what approximate
(wind speed v). The other travels from A to C and
form does equation (6) take for everyday
back to A, flying across the wind. Which one will
speeds? Is this the form you would expect from
arrive home first, and what will be the difference in
your experience?
their arrival times? Using the first two tetms of the
(2) Analyze the example that began this exercise:
binomial theorem.
Speed of bullet with respect to spacebus
t^'buUet ~ 3/4; speed of spacebus with respect
(1 4 -z )” ~ 1 -f nz for |z |« 1
to Earth : re l 3/4. W hat is the speed of the
bullet measured by Earth observers?
show that if v « c, then an approximate expression
(3) Why stop with bullets that saunter along at
for this time difference is A/ ~ {L/2c){v/cY, where L
less than the speed of light? Let the bullet itself
is the round-trip distance between A and B (and
be a flash of light. Then the bullet speed as
between A and C).
measuted in the bus is r'^buUet “ 1 ■For “"rel
d The South Pole Air Station is the supply depot
3 /4 , with what speed does this light flash
for research huts on a circle of 300-kilometer radius
move as measured in the Earth frame? Is this
centered on the air station. Every Monday many sup
what you expect from the Principle of Relativ
ply planes start simultaneously from the station and
ity?
fly radially in all directions at the same altitude. Each
(4) Suppose a light flash is launched from the
plane drops supplies and mail to one of the research
front of the bus directed toward the back
huts and flies directly home. A Fussbudget with a
(j'^bouet ~ ~ W hat is the velocity of this
stopwatch stands on the hill overlooking the air sta
light flash measured in the Earth frame? Is this
tion. She notices that the planes do not all return at the
what you expect from the Principle of Relativ
same time. This discrepancy perplexes her because she
ity?
knows from careful measurement that (1) the dis
Reference: N . David Mermin, American Journal of Physics, Volume
tance from the air station to every research hut is the
51, pages 1 1 3 0 -1 1 3 1 (1983).
same, (2) every plane flies with the same air speed as
every other plane — 3 0 0 k ilo m eters/h o u r — and (3)
every plane travels in a straight line over the ground
3-12 Michelson—Morley from station to hut and back. The Fussbudget finally
experiment decides that the discrepancy is due to the wind at the
a An airplane moves with air speed c (not the high altitude at which the planes fly. W ith her stop
speed of light) from point A to point B on Earth. A watch she measures the time from the return of the
stiff wind of speed p is blowing from B toward A. (In first plane to the return of the last plane to be 4
this exercise only, the symbol v stands for velocity in seconds. W hat is the wind speed at the altitude where
conventional units, for example meters/second.) the planes fly? W hat can the Fussbudget say about
Show that the time for a round trip from A to B and the direction of this wind?
back to A under these circumstances is greater by a e In their famous experiment Michelson and
factor 1/(1 — v'^/c^) than the corresponding round Morley attempted to detect the so-called e th e r d rift
trip time in still air. Paradox: The wind helps on one — the motion of Earth through the “ether,” with
leg of the flight as well as hinders on the other. Why, respect to which light was supposed to have the ve
therefore, is the round-trip time not the same in the locity c. They compared the round-trip times for light
presence of wind as in still air? Give a simple physical to travel equal distances parallel and perpendicular to
reason for this difference. W hat happens when the the direction of motion of Earth around Sun. They
wind speed is nearly equal to the speed of the airplane? reflected the light back and forth between nearly
EXERCISE 3-12 MICHELSON- MORLEY EXPERIMENT 85
parallel mirrors. (This would correspond to part c if traces its path to mirror b. At mirror b parts of each
each airplane made repeated round trips.) By this beam combine to enter telescope / together. The
means they were able to use a total round-trip length transparent piece of glass at c, of the same dimensions
of 22 meters for each path. If the “ether” is at rest as the half-silvered mirror b, is inserted so that both
with respect to Sun, and if Earth moves at 30 X 10^ beams pass the same number of times (three times)
meters/second in its path around Sun, what is the through this thickness of glass on their way to tele
approximate difference in time of return between scope/. Suppose that the perpendicular path lengths
light flashes that are emitted simultaneously and are exactly equal and the instrument is at rest with
travel along the two perpendicular paths? Even with respect to the ether. Then monochromatic light from
the instruments of today, the difference predicted by the two paths that leave mirror h in some relative
the ether-drift hypothesis would be too small to mea phase will return to mirror b in the same phase. Under
sure directly, and the following method was used these circumstances the waves entering telescope/will
instead. add crest to crest and the image in this telescope will
f The original Michelson - Morley interferome be bright. On the other hand, if one of the beams has
ter is diagrammed in the figure. Nearly monochro been delayed a time corresponding to one half period
matic light (light of a single frequency) enters through of the light, then it will arrive at mitror b one half
the lens at a. Some of the light is reflected by the period later and the waves entering the telescope will
half-silvered mirror at b and the rest of the light cancel (crest to trough), so the image in the telescope
continues toward d. Both beams are reflected back will be dark. If one beam is retarded a time corre
and forth until they reach mirrors e and e-^ respectively, sponding to one whole petiod, the telescope image
where each beam is reflected back on itself and re will be bright, and so forth. W hat time corresponds to
EXERCISE 3 -1 2 . M ichelson - M orley interferom eter m ounted on a ro ta tin g m arble slab.

86 EXERCISE 3-13 THE KENNEDY- THORNDIKE EXPERIMENT
one period of the light? Michelson and Morley used Michelson - Morley experiment have for us who do
sodium light of wavelength 589 nanometets (one not believe in the ether theory of light propagation?
nanometer is equal to 10“^ metet). Use the equations Simply this: (1) The round-trip speed of light mea
fX = c and / = 1/ T that relate frequenq^/, period T, sured on earth is the same in every direction — the
wavelength A, and speed c of an electromagnetic speed of light is isotropic. (2) The speed of light is
wave. Show that one period of sodium light corre isotropic not only when Earth moves in one direction
sponds to about 2 X 10“ ' ’ seconds. around Sun in, say, January (call Earth with this
Now thete is no way to “turn o ff’ the alleged ether motion the “laboratory frame”), but also when Earth
drift, adjust the apparatus, and then turn the alleged moves in the opposite direction around Sun six
ether drift on again. Instead of this, Michelson and months later, in July (call Earth with this motion the
Morley floated their interferometer in a pool of mer “rocket frame”). (3) The generalization of this result
cury and rotated it slowly about its center like a to any pair of inertial frames in relative motion is
phonograph record while observing the image in the contained in the statement. The round-trip speed of
telescope (see the figure). In this way if light is delayed light is isotropic both in the laboratory frame and in
on either path when the instrument is oriented in a the rocket frame. This result leaves an important
cettain direction, light on the other path will be de question unanswered: Does the round-ttip speed of
layed by the same amount of time when the insttu- light— which is isotropic in both laboratory and
ment has rotated 90 degrees. Hence the total change rocket frames — also have the same numerical value
in delay time between the two paths observed as the in laboratory and rocket frames? The assumption that
interferometer rotates should be twice the difference this speed has the same numerical value in both
calculated using the expression derived in part c. By frames played a central role in demonstrating the
refinements of this method Michelson and Morley invariance of the interval (Section 3.7). But is this
were able to show that the time change between the assumption valid?
two paths as the instrument rotated corresponded to a An experiment to test the assumption of the
less than one one-hundredth of the shift from one equality of the round-trip speed of light in two inettial
dark image in the telescope to the next dark image. frames in relative motion was conducted in 1932 by
Show that this result implies that the motion of the Roy J. Kennedy and Edward M. Thorndike. The
ethet at the surface of Earth — if it exists at all — is experiment uses an interferometer with atms of un
less than one sixth of the speed of Earth in its orbit. In equal length (see the figute). Assume that one arm of
order to eliminate the possibility that the ether was the interferometer is A / longer than the other arm.
flowing pasr Sun at the same rate as Earth was moving Show that a flash of light entering the apparatus will
its orbit, they tepeated the experiment at intervals of take a time 2A //c longer to complete the round trip
three months, always with negative results. along the longer arm than along the shotter arm. The
g D iscussion question; Does the Michelson- difference in length A/ used by Kennedy and Thorn
Morley experiment, by itself, disprove the theory that dike was approximately 16 centimeters. W hat is the
light is propagated through an ether? Can the ether approximate difference in time for the round trip of a
theory be modified to agree with the results of this light flash along the alternative paths?
experiment? How? W h a t further experim ent can be b Instead of a pulse of light, Kennedy and
used to test the modified theory? Thorndike used continuous monochromatic light of
Reference: A. A. Michelson and E. W . Morley, American Journal of period T = 1.820 X 10“ ' ’ seconds (A = 546.1
Science, Volume 134, pages 3 3 3 - 3 4 5 (1887). nanometers = 546.1 X 10“ ^ meters) from a mercury
source. Light that ttaverses the longer arm of the
interferometer will return approximately how many
3-13 the Kennedy—Thorndike periods n later than light that traverses the shortet
experiment atm? If in the actual experiment the number of pe
N ote: Part d of this exercise uses elementary calculus. riods is an integer, the reunited light from the two
The Michelson - Morley experiment was designed arms will add (crest-to-crest) and the field of view
to detect any motion of Earth relative to a hypotheti seen through the telescope will be bright. In contrast,
cal fluid — the ether— a medium in which light was if in the actual experiment the number of periods is a
supposed to move with characteristic speed c. No half-integer, the reunited light from the two arms will
such relative motion of earth and ether was detected. cancel (crest-to-trough) and the field of view of the
Partly as a result of this experiment the concept of telescope will be dark.
ether has since been discarded. In the modern view, c Earth continues on its path around Sun. Six
light requires no medium for its transmission. What months later Earth has reversed the direction of its
significance idoes the negative result of the velocity relative to the fixed stars. In this new frame of
EXERCISE 3-13 THE KENNEDY- THORNDIKE EXPERIMENT 87
EXERCISE 3-13. Schematic diagram of apparatus used for the jacket is surrounded by an outer water jacket in which the water is
Kennedy- 'Vhomdike experiment. Parts of the interferometer have kept at a temperature that varies less than ± 0 .0 0 1 degrees Celsius.
been labeled with letters corresponding to those used in describing The entire apparatus shown in the figure is enclosed in a small
the Michelson-Morley interferometer (Exercise 3-12). The experi darkroom (not shown) maintained at a temperature constant within
menters went to great lengths to insure the optical and mechanical a few hundredths ofa degree. The small darkroom is in turn enclosed
stability of their apparatus. The interferometer is mounted on a in a larger darkroom whose temperature is constant within a few
plate of quartz, which changes dimension very little when tempera tenths of a degree. The overall size of the apparatus can he judged
ture changes. The interferometer is enclosed in a vacuum jacket so from the fact that the difference in length of the two arms of the
that changes in atmospheric pressure will not alter the effective interferometer (length eb compared with length ejb) is 16
optical path length of the interferometer arms (slightly different centimeters.
speed of light at different atmospheric pressure). The inner vacuum
reference will the round-trip speed of light have the Point out the standards of distance and time used in
same numerical value c as in the original frame of determining this result, as they appear in the equa
reference? One can rewrite the answer to part b for the tion. Quartz has the greatest stability of dimension of
original frame of reference in the form any known material. Atomic time standards have
proved to be the most dependable earth-bound time
f = (2 /« )(A // 7 ) keeping mechanisms.
d In order to carry our the experiment outlined in
where A/ is the difference in length between the two the preceding paragraphs, Kennedy and Thorndike
interferometer arms, T is the time for one period of would have had to keep their interferometer operat
the atomic light source, and n is the number of periods ing perfectly for half a year while continuously ob
that elapse between the return of the light on the serving the field of view through the telescope. Unin
shorter path and the return of the light on the longer terrupted operation for so long a time was not
path. Suppose that as Earth orbits Sun no shift is feasible. The actual durations of their observations
observed in the telescope field of view from, say, light varied from eight days to a month. There were several
toward dark. This means that n is observed to be such periods of observation at three-month time sep
constant. W hat would this hypothetical result tell arations. From the data obtained in these periods,
about the numerical value c of the speed of light? Kennedy and Thorndike were able to estimate that
88 EXERCISE 3-14 THINGS THAT MOVE FASTER THAN LIGHT
over a single six-month observation the number of Einstein theory of equivalence of all inertial reference
periods n of relative delay would vary by less than the frames.
fraction 3/1000 of one period. Take the differential The “sensitivity” of the Kennedy-Thorndike ex
of the equation in part c to find the largest fractional periment depends on the theory under considerarion.
change dc/c of the round-trip speed of light between In the context of theory A the observations set an
the two frames consistent with this estimated change upper limit of about 15 kilometers/second to the
in n (frame 1 — the “labotatory” frame — and frame “speed of Sun through absolute space” (sensitivity
2 — the “rocket” frame— being in the present anal reported in the Kennedy-Thorndike paper). In the
ysis Earth itself at two different times of year, with a context of Einstein’s theory the observations say that
relative velocity twice the speed of Earth in its orbit: the round-trip speed of light has the same numerical
2 X 30 kilometets/second). magnitude— within an error of about 3 meters/
H istorical note: At the time of the Michelson- second — in inertial frames of reference having a rela-
Morley experiment in 1887, no one was ready for the rive velocity of 60 kilometers/second.
idea that physics — including the speed of light— is Reference: R. J. Kennedy and E. M. Thorndike, Physical Review,
the same in every inertial frame of reference. Accord Volume 42, pages 4 0 0 - 4 1 8 (1932).
ing to today’s standard Einstein interptetation it
seems obvious that both the Michelson-Motley and 3-14 things that move faster
the Kennedy-Thorndike experiments should give than light
null results. However, when Kennedy and Thorndike
made their measurements in 1932, two alternatives Can “things” or “messages” move fasrer than light?
to the Einstein theory were open to consideration Does relativity really say “ N o” to this possibility?
(designated here as theory A and theory B). Both A Explore these questions further using the following
and B assumed the old idea of an absolute space, or examples.
“ether,” in which light has the speed c. Both A and B a T h e Scissors P aradox. A very long straight
explained the zero fringe shift in the Michelson- rod, inclined at an angle d to the x-axis, moves down
Motley experiment by saying that all matter that ward with uniform speed as shown in the figure.
moves at a velocity v (expressed as a fraction of light- Eind rhe speed of the point of intersection A of the
speed) relative to “absolute space” undergoes a lower edge of the stick with the x-axis. Can this speed
shrinkage of its space dimensions in the direction of be greater than the speed of light? If so, for what
motion to a new length equal to (1 — times the values of the angle 0 and does this occur? Can the
old length (“Lorentz-FitzGerald contraction hypoth motion of intersection point A be used to transmit a
esis”). The two theories differed as to the effect of message faster rhan lighr from someone at the origin
“ motion through absolute space” on the running rate to someone far out on the x-axis?
of a clock. Theory A said. No effect. Theory B said b T ransm ission o f a H a m m e r Pulse. Sup
that a standard seconds clock moving through abso pose the same rod is inirially at test in the laboratory
lute space at velocity v has a time between ticks of with the point of intersection initially at the origin.
(1 “ seconds. In theory B the ratio A //T in the The region of the rod centered at the origin is struck
equation in part b will not be affected by the velocity sharply with the downward blow of a hammer. The
of the clock, and the Kennedy-Thorndike experi point of intersection moves to the right. Can this
ment will give a null result, as observed (“compli motion of the point of intersection be used to transmit
cated explanation for simple effect”). In theory A the a message faster than the speed of light?
ratio A //T in the equation will be multiplied by the c S earchlight M essenger? A very powerful
factor (1 ~ at a time of year when the “velocity searchlight is rotated rapidly in such a way that its
of Earth relative to absolute space” is v-^ and multi beam sweeps out a flat plane. Observers A and B are
plied by (1 — at a time of year when this at rest on the plane and each the same distance from
velocity is Thus the fringes should shift from one the searchlight but not near each other. How far from
time of year (v^ = t'otbitai *^Sun) to anorher time of the searchlight must A and B be in order that the
year (v^ = “ t'sun) unless by accident Sun searchlight beam will sweep from A to B faster than a
happened to have “zero velocity relative to absolute light signal could travel from A to BP Before they
space” — an accident judged so unlikely as not to took their positions, the two observers were given the
provide an acceptable explanation of the observed following instruction:
null effect. Thus the Kennedy-Thorndike experi To A: “When you see the searchlight beam, fire a bullet
ment ruled out theory A (length contraction alone) at B.”
but allowed theory B (length contraction plus time To B: “When you see the searchlight beam, duck be
contraction) — and also allowed the much simpler cause A has fired a bullet at you.”
EXERCISE 3-15 FOUR TIMES THE SPEED OF LIGHT? 89
EXERCISE 3-14. Can the point of intersection A move with a speed greater than the speed of light?
Under these circumstances, has a warning message surements from our continenr-spanning latticework
traveled from A to B with a speed faster than that of of clocks taken at this moment.
light? Now the rocket continues toward us as we stand in
d O scilloscope W ritin g Speed. The manu New York City. The center figure summarizes data
facturer of an oscilloscope claims a writing speed (the taken as the first flash is about to enter our eye. Flash
speed with which the bright spot moves across the 1 shows us the rocket passing under the Golden Gate
screen) in excess of the speed of light. Is this possible? Bridge. An instant later flash 2 shows us the rocket
passing through the Gateway Arch.
a Answer the following questions using symbols
3-15 four limes the speed of from the first two figures. The images carried by the
light? two flashes show the rocket how far apart in space?
W hat is the time lapse between our reception of these
We look westward across the United States and see
two images? Therefore, what is the apparent speed of
the rocket approaching us at four times the speed of
the approaching rocket we see? For what speed v of
light.
the rocket does the apparent speed of approach equal
four times the speed of light? For what rocket speed
How can this be, since nothing moves faster
do we see the approaching rocket to be moving at 99
than light?
times the speed of light?
b Our friend in San Francisco is deeply disap
pointed. Looking eastward, she sees the retreating
rocket traveling at less than half the speed of light
C. W e did not say the rocket moves faster
(bottom figure). She wails, “Which one of us is
than light; we said only that we see it
wrong?” “Neither one.” we reply. “No matter how
moving faster than light.
high rhe speed v of the rocket, you will never see ir
moving directly away from you at a speed greater than
half the speed of light.”
Here is what happens: The rocket streaks under the Use the bottom figure to derive an expression for
Golden Gate Bridge in San Francisco, emitting a flash the apparent speed of recession of the rocket. When
of light that illuminates the rocket, the bridge, and we in New York see the rocket approaching at four
the surroundings. At time A? later the rocket threads times the speed of light, with what speed does our San
the Gateway Arch in St. Louis that commemorates Francisco friend see it moving away from her? When
the starting point for covered wagons. The arch and we see a faster rocket approaching at 99 times the
the Mississippi riverfront are flooded by a second flash speed of light, what speed of recession does she be
of light. The top figure is a visual summary of mea hold?
90 EXERCISE 3-16 SUPERLUMINAL EXPANSION OF QUASAR 3C273?
SAN FRANCISCO ST. LOUIS NEW YORK
emit emit
flash 1
emit emit
flash flash
1 2
(1-vÂt
vAf-
/ /
ROCKET HEADED EAST
flash 2
emit emit
flash flash
2 3
-At- -vAt-
ROCKET AT NEW YORK

EXERCISE 3-15. Top: Rocket headed east, shown at the instant it Bottom: Rocket headed east, shown at the instant it approaches the
passes under the Gateway Arch in St. Louis and emitsflash 2 . The Empire State Building in New York City and emits flash 3. When
rocket is chasing flash 1, emitted earlier as it passed under the the rocket moves away from the viewer, the distance of rocket travel
Golden Gate Bridge in San Francisco. Center: The two image is added to the separation between flashes. This increases the ap
carrying flashes are close together, so they enter the eye in rapid parent time between flashes, giving the viewer the impression that
succession. This gives the viewer the visual impression that the the rocket movedfrom St. Louis to New York at less than one half
rocket movedfrom San Francisco to St. Louis in a very short time. light-speed.
smaller than any galaxy, the typical quasar manages

3-16 superluminal expansion to put out more than 100 times as much energy as our
off quasar 3C273? own Milky Way, with its hundred billion stars. Qua
The most powerful sources of energy we know or sars, unsurpassed in brilliance and remoteness, we
conceive or see in all the universe are so-called quasi- count today as lighthouses of the heavens.
stellar objects, or quasars, starlike sources of light One of the major problems associated with quasars
located billions of light-years away. Despite being far is that some are composed of two or more components
EXERCISE 3-16 SUPERLUMINAL EXPANSION OF QUASAR 3C273? 91
EXERCISE 3-16, first figure. Left: Bright “knot” ofplasma ejectedfrom a quasar at high speedy emits a
first flash of light toward Earth. Right: The knot emits a second light flash toward Earth a time At later.
This time At is measured locally near the knot using the Earth-linked latticework of rods and clocks (har!
harl).
that appear to be separating from each other with c Now calculate the speed of the rightward
relative velocity greater than the speed of light (“su motion of the knot as seen on Earth. Show that the
perluminal” velocity). One theory that helps explain result is
this effect pictures the quasar as a core that ejects a jet
of plasma at relativistic speed. Disturbances or insta V sin e
.X
-------
bilities in such a jet appear as discrete “knots” of
plasma. The motion and light emission from a knot
1 V cos e
may account for its apparent greater-than-light speed,
d W hat is the value of when the knot is
as shown using the first figure.
emitted in the direction exactly toward Earth? when it
a The first figure shows two Earth-directed light
is emitted perpendicular to this ditection? Find an
flashes emitted from the streaking knot. The time
expression that gives the range of angles Q for which
between emissions is A/ as measured locally near the
is greater than the speed of light. For 0 = 45
knot using the Earth-linked latticework of rods and
degrees, what is the range of knot speeds v such that
clocks. O f course the clock readings on this portion of
is greater than the speed of light?
the Earth-linked latticework are not available to us on
e If you know calculus, find an expression for the
Earth; therefore we cannor measure A/ directly.
angle at which has its maximum value for a
Rather, we see the time separation between the atriv-
given knot speed v. Show that this angle satisfies the
als of the two flashes at Earth. From the figure, show
equation cos B ,^ = v. Whether or not you derive this
that this Earth-seen time separation A t^^ is given by
result, use it to show that the maximum apparent
the expression
transverse speed is seen as
= A /(l V COS Q)
t/'
*^seen, max
b We have another disability in viewing the knot (1 -^4)1/2
from Earth. We do not see the motion of the knot
toward us, only the apparent motion of the knot f W hat is this maximum transverse speed seen
across our field of view. Find an expression for this on Earth when v = 0.99?
transverse motion (call it between emissions of g The second figure shows the pattern of radio
the two light flashes in terms of Ar. emission from the quasar 3C273. The decreased pe-
92 EXERCISE 3-17 CONTRACTION OR ROTATION?
riod of radiation from this source (Exercise 3-10)

shows that it is approximately 2.6 X 10^ light-years
from Earth. A secondary source is apparently moving
away from the central quasar. Take your own mea
surements on the figure. Combine this with data from
the figure caption to show that the apparent speed of
separation is greater than 9 times the speed of light.
N ote: As of 1990, apparent greater-than-light-
speed (“superluminal”) motion has been observed in
approximately 25 different sources.
References: Analysis and first figure adapted from Denise C. Ga-
buzda, Am erican Journal o f Physics, Volume 55, pages 2 1 4 - 2 1 5
(1987). Second figure and data taken from T. J. Pearson, S. C.
Unwin, M. H. Cohen, R. P. Linfield, A. C. S. Readhead, G . A,
Seielstad, R. S. Simon, and R. C. W alker, N ature, Volume 290,
pages 3 6 5 - 3 6 8 (2 April 1981),
3-17 contraction or rotation?

A cube at rest in the rocket frame has an edge of
length 1 meter in that frame. In the laboratory frame
the cube is Lorentz contracted in the direction of
motion, as shown in the figure. Determine this Lor
entz contraction, for example, from locations of four
clocks at rest and synchronized in the laboratory lat
tice with which the four corners of the cube, E, F, G,
H, coincide when all four clocks read the same time.
This latticework measurement eliminates time lags in
the travel of light from different corners of the cube.
Now for a different observing procedure! Stand in
the laboratory frame and look at the cube with one eye
as the cube passes overhead. W hat one sees at any
time is light that enters the eye at that time, even if it
left the different corners of the cube at different times.
Hence, what one sees visually may not be the same as
what one observes using a latticework of clocks. If the
cube is viewed from the bottom then the distance GO
is equal to the distance HO, so light that leaves G and
H simultaneously will arrive ar 0 simultaneously.
Hence, when one sees the cube to be overhead one will
see the Lorentz contraction of the bottom edge.
a Light from E that arrives at 0 simultaneously
with light from G will have to leave E earlier rhan
light from G left G. How much earlier? How far has
the cube moved in this time? W hat is the value of the
distance x in the right top figure?
b Suppose the eye interprets the projection in the
figures as a rotation of a cube that is not Lorentz
contracted. Find an expression for the angle of appar
EXERCISE 3 -1 6 , se co n d figure. Contour lines o f radio emission
fro m the q u a sa r 5 C 2 7 3 sh o w in g a b rig h t “k n o t" o f p la sm a a p par ent rotation (f> of this uncontracted cube. Interpret
en tly m oving a w a y fro m i t a t a speed greater th a n the speed o f light. this expression for the two limiting cases of cube speed
The tim e o f each im age is g iv e n a s ca len dar y e a r a n d decim a l in the laboratory frame: p —* 0 and p —* l .
fra c tio n . H o rizo n ta l scale d iv isio n s are in u n its o f 2 m illi arc-se
C D iscussion question: Is the word “really”
conds. (1 m illi arc-second = lO r ^ jj^ O O degree = 4 .8 5 X 1G~^
ra d ia n )
an appropriate word in the following quotations?
EXERCISE 3-17 CONTRACTION OR ROTATION? 93
Location of cube
E G
(1 - v^)'
EXERCISE 3 -1 7 . h e ft: Position o f eye o f v is u a l observer w a tc h in g cube pass overhead. R i g h t to p : W h a t the

v is u a l observer sees a s she looks u p fro m below. R i g h t b o tto m : H o w the v is u a l observer can interpret the
projection o f the second figure.
(1) An observer using the rocket latticework of W hat can one rightfully say — in a sentence or
clocks says, “The stationary cube is really nei two — to make each observer think it reasonable that
ther rotated nor contracted.” the other observers should come to different conclu
(2) Someone riding in the rocket who looks at the sions?
stationary cube agrees, “The cube is really nei d The analysis of parts b and c assumes that the
ther rotated nor contracted.” visual observer looks with one eye and has no depth
(3) An observer using the laboratory latticework perception. How will the cube passing overhead be
of clocks says, “The passing cube is really Lor- perceived by the viewer with accurate depth percep
entz contracted but not rotated.” tion?
(4) Someone standing in the laboratory frame Reference: For a more complete treatm ent of this topic, see Edwin F.
looking at the passing cube says,' ‘The cube is Taylor, Introductory Mechanics (John Wiley and Sons, New York,
really rotated but not Lorentz contracted.” 1963), pages 3 4 6 - 3 6 0 .
■'■‘'-c-'-i >;^, ;■■
M0
■>:; +:.‘-v^-:
'..■V^sv
... 'h-
'«• -ff-
•v?-
S P E C I A L TOPIC
' -i ■'■ •■
■ ■-^■'jy.^' w > . 5*.-
LORENTZ TRANSFORMATION
L I LORENTZ TRANSFORMATION:
USEFUL OR NOT?
related events or lonely events?
Events, and the intervals between events, define the layout of the physical world. No
Events and intervals only:
latticework of clocks there! Only events and the relation between event and event as
Spacetim e lean and sp are
expressed in the interval. That’s spacetime physics, lean and spare, as it offers itself to
us to meet the needs of industry, science, and understanding.
There’s another way to express the same information and use it for the same
purposes: Set up a free-float latticework of recording clocks, or the essential rudiments
of such a latticework. The space and time coordinates of that Lorentz frame map each
O r isolated events described
event as a lonesome individual, with no mention of any connection, any spacetime
using latticework
interval, to any other event.
This lattice-based method for doing spacetime physics has the advantage that it can
be mechanized and applied to event after event, wholesale. These regimented space
and time coordinates then acquire full usefulness only when we can translate them
from the clock-lattice frame used by one analyst to the clock-lattice frame used by
another.
This scheme of translation has acquired the name “Lorentz transformation.” Its
Lorentz transformation:
usefulness depends on the user. Some never need it because they deal always with
Translate event description
intervals. Others use it frequently because it regiments records and standardizes from lattice to lattice
analysis. For their needs we insert this Special Topic on the Lorentz transformation.
The reader may wish to read it now, or skip it altogether, or defer it until after Chapter
4, 5, or 6. The later the better, in our opinion.
95
96 SPECIAL TOPIC LORENTZ TRANSFORMATION
L.2 FASTER THAN LIGHT?

a reason to examine the Lorentx transformation
No object travels faster than light.
So YOU say, but watch ME: I travel in a rocket that you observe to move at 4/5 light speed. Out
the front of my rocket I fire a bullet that I observe to fly forward at 4/5 light speed. Then you
measure this bullet to streak forward at 4 /5 '\'4 /5 — 8/5 — 1.6 light speed, which is greater
than the speed of light. There!
' No!
Why not? Is it not true that 4/5 + 4/5 — 1-6?
As a mathematical abstraction: always true. As a description of the world: only

sometimes true! Example 1: Add 4 /5 liter of alcohol to 4 /5 liter of water. The result?
Velocities do not odd Less than 8 /5 = 1 .6 liter of liquid! Why? Molecules of water interpenetrate molecules
of alcohol to yield a combined volume less that the sum of the separate volumes.
Example 2: Add the speed you measure for the bullet (4/5) to the speed I measure for
your rocket (4/5). The result? The speed I measure for the bullet is 40/41 = 0.9756.
This remains less than the speed of light.
Why? And where did you get that number 4 0/4l for the bullet speed you measure?
I got the number from the Lorentz transformation, the subject of this Special Topic.
The Lorentz transformation embodies a central feature of relativity: Space and time
separations typically do not have the same values as observed in different frames.
Space an d time separations between w hat?
Between events.
W hat events are we talking about here?
Event 1: You fire the bullet out the front of your rocket. Event 2: The bullet strikes a
target ahead of you.
W hat do these events have to do with speed? We are arguing about speed!
Events define velocities

Let the bullet hit the target four meters in front of you, as measured in your rocket.
Then the space separation between event 1 and event 2 is 4 meters. Suppose the time
of flight is 5 meters as measured by your clocks, the time separation between the two
events. Then your bullet speed measurement is (4 meters of distance)/! 5 meters of
time) = 4 /5 , as you said.
A nd w hat do YOU measure for the space an d time separations in your laboratory fram e?
For that we need the L orentz co o rd in a te tran sfo rm atio n equations.

L.2 FASTER THAN LIGHT? 9 7
Phooey! I know how to reckon spacetime separations in different frames. We have been doing it for
several chapters! From measurements in onefram e we figure the spacetime interval, which has the
same value in a ll frames. End of story.
No, not the end of the story, but at least its beginning. True, the invariant interval has Interval: Only a start in
the same value as derived from measurements in every frame. That allows you to reckoning spacetime separations
predict the time between firing and impact as measured by the passenger riding on the in different frames
bullet — and measured directly by the bullet passenger alone.
Predict how?
You know your space separation x ' = 4 meters (primes for rocket measurements), and
your time separation, t' = 'b meters. You know the space separation for the bullet
rider, x" = 0 (double primes for bullet measurements), since she is present at both the
firing and the impact. From this you can use invariance of rhe interval to determine the
wristwatch time between these events for the bullet rider:
i t ' y - { x " f = {t'Y - {x'Y
or
(/")^ — (0)^ = (5 meters)^ — (4 merers)^ — (3 meters)^
so that t” = 3 meters. This is the proper time, agreed on by all observers but measured
directly only on the wristwatch of the bullet rider.
Fine. C an’t we use the same procedure to determine the space an d time separations between these
events in your laboratory frame, an d thus the bullet speed fo r you?
Unfortunately not. We do reckon the same value for the interval. Use unprimed
Need more to compare velocities
symbols for laboratory measurements. Then f- — xd = {?) meters)^. That, however, is
in different frames
not sufficient to determine x or t separately. Therefore we cannot yet find their ratio
x /t, which determines the bullet’s speed in our frame.
So how can we reckon these x a n d t separations in your laboratory frame, thereby allowing us to
predict the bullet speed you measure?
Use rhe Lorentz transformation. This transformation reports that our laboratory space
Compare velocities using
separation between firing and impact is x = 4 0 /3 meters and the time separation is
Lorentz transformation
slightly greater: t = 4 1 /3 meters. Then bullet speed in my laboratory frame is
predicted to be f = x j t — 40/41 = 0.9756. The results of our analysis in three
reference frames are laid out in Table L-1.
Is the Lorentz transformation generally useful, beyond the specific task o f reckoning speeds as
measured in different frames?
Oh yes! Generally, we insert into the Lorentz transformation the coordinates x ', t' of an
event determined in the rocket frame. The Lorentz transformation then grinds and
whirs, finally spitting out the coordinates x, t of the same event measured in the
laboratory frame. Following are the Lorentz transformation equations. Here is the
relative velocity between rocket and laboratory frames. For our convenience we lay rhe
posirive x-axis along the direction of motion of the rocket as observed in the laboratory
frame and choose a common reference event for the zero of time and space for both
frames.
------C j A B L E L - T ^ ------
HOW FAST THE BULLET?

Bullet fired Bullet hits Speed of bullet
(coordinates (coordinates (computed from
of this event) of this event) frame coordinates)
Rocket frame x' = 0 x' = 4 meters as measured

(moves at = 4/5 l' = 0 t' = 5 meters in rocket frame:
as measured in laboratory) v' = 4/5 = 0.8
Bullet frame x" = 0 x" = 0 as measured
(moves at v' = 4/5 t" = 0 t" = 3 meters in bullet frame:
as measured in rocket) (from invariance r" = 0
of the interval)
Laboratory frame x= 0 X = 40/3 meters as measured
r= 0 t = 41/3 meters in laboratory frame:
(from Lorentz 1- = 40/41 = 0.9756
transformation)
k' rel ‘
(1
Lorentz transformation preview ed x ' + t'
(1
and
Check for yourself that for the impact event of bullet with target (rocket coordi
nates: x ' = A meters, /' = 5 meters; rocket speed in laboratory frame: = 4 /5 ) one
obtains laboratory coordinates x = 4 0 /3 meters and t = 4 1 /3 meters. Hence v = x f t
== 40/41 = 0.9756.
You say the Lorentz transformation is general. I f it is so important, then why is this a special topic
rather than a regular chapter?
The Lorentz transformation is powerful; it brings the technical ability to transform

coordinates from frame to frame. It helps us predict how to add velocities, as outlined
Lorentz transformation: Useful
here. It describes the Doppler shift for light (see the exercises for this chapter). On the
but not fundamental
other hand, the Lorentz transformation is not fundamental; it does not expose deep
new features of spacetime. But no matter! Physics has to get on with the world’s work.
One uses the method of describing separation best suited to the job at hand. On some
occasions the useful fact to give about a luxury yacht is the 50-meter distance between
bow and stern, a distance independent of the direction in which the yacht is headed.
On another occasion it may be much more important to know that the bow is 30
meters east of the stern and 40 meters north of it as observed by its captain, who uses
North-Star north.
W hat does the Lorentz transformation rest on? On w hat foundations is it based?
On two foundations: (1) The equations must be linear. That is, space and time
Two foundations of
coordinates enter the equations to the first power, not squared or cubed. This results
from the requirement that you may choose any event as the zero of space and time.
L.3 FIRST STEPS 99
(2) The spacetime interval between two events must have the same value when
computed from laboratory coordinate separations as when reckoned from rocket
coordinate separations.
All right, I 'll reservejudgment on the validity ofwhat you claim, hut show me the derivation itself.
Read on!
L.3 FIRST STEPS

invariance off the interval gets us started
Recall that the coordinates y and z transverse to the direction of relative motion
between rocket and laboratory have the same values in both frames (Section 3.6):
y-y
z = z' (L-1)
where primes denote rocket coordinates. A second step makes use of the difference in
observed clock rates when the clock is at rest or in motion (Section 1.3 and Box 3-3).
Think of a sparkplug at rest at the origin of a rocket frame that moves with speed
relative to the laboratory. The sparkplug emits a spark at time t' as measured in the
rocket frame. The sparkplug is at the rocket origin, so the spark occurs at x ' = 0.
Where and when (x and t) does this spark occur in the laboratory? That depends on
Derive difference in clock rates
how fast, vî, the rocket moves with respect to the laboratory. The spark must occur at
the location of the sparkplug, whose position in the laboratory frame is given by
X = Vît
Now the invariance of the interval gives us a relation between t and t',
{t'r - ( x y = { t y - {oy = = f - x ^ = f - { v j f = t w - vij>
from which
t' = t { \ - ri,)V2
or
[when x' = 0] (L-2)
The awkward expression 1/(1 — occurs often in what follows. For simplic
ity, this expression is given the symbol Greek lower-case gamma: /.
1
7=
(1 -
Because it gives the ratio of observed clock rates, y is sometimes called the tim e
Time stretch factor defined
stre tc h facto r (Section 5.8). Strictly speaking, we should use the symbol /„i, since
the value of y is determined by For simplicity, however, we omit the subscript in
the hope that this will cause no confusion. With this substitution, equation (L-2)
becomes
yf [when x' = 0] (L-3)
lO O SPECIAL TOPIC LORENTZ TRANSFORMATION
Substitute this into the equation x = v ^ ^ t above to find laboratory position in terms of
rocket measurements:
[when x' = 0] (L-4)
Equations (L-1), (L-3), and (L-4) give the first answer to the question, “If we know
the space and time coordinates of an event in one free-float frame, what are its space
and time coordinates in some other overlapping free-float frame?” These equations are
limited, however, since they apply only to a particular situation: one in which both
events occur at the same place {x' = 0) in the rocket,
L 4 FORM OF THE LORENTZ

TRANSFORMATION
any event can be reference event? then
transformation is linear
W hat general form does the Lorentz transformation have? It has the form that
mathematicians call a lin ear tran sfo rm atio n . This means that laboratory coordi
nates X and t are related to linear (first) power of rocket coordinates x ' and t ' by
equations of the form
Lorentz transformation: / = fix' -f D t'

Linear equations X = Gx' + Ht' (L-5)
where our task is to find expressions for the coefficients B, D, G, and H that do not
depend on either the laboratory or the rocket coordinates of a particular event, though
they do depend on the relative speed
Why must these transfotmations be linear? Because we are free to choose any event
as our reference event, the common origin x = y = z = / = 0 in all reference frames. Let
our rocket sparkplug emit the flashes at = 1 and 2 and 3 meters. These are equally
spaced in rocket time. According to equation (L-3) these three events occur at
laboratory times t = l y and 2y and 37 meters of time. These are equally spaced in
laboratory time. Moving the reference event to the first of these events still leaves them
equally spaced in time for both observers: t' = 0 and 1 and 2 meters in the rocket and t
= 0 and ly and 2y in the laboratory.
In contrast, suppose that equation (L-3) were not linear, reading instead t = Kt'^,
where K is some constant. Rocket times t' = \ and 2 and 3 meters result in laboratory
times t = IK and 4K and 9K meters. These are not equally spaced in time for the
laboratory observer. Moving the reference event to the fitst event would result in
rocket times t' = 0 and 1 and 2 meters as before, but in this case laboratoty times t = 0
and 1K and 4K metets, with a completely different spacing. But the choice of reference
Arbitrary event as reference event?
event is arbitrary: Any event is as qualified to be reference event as any other. A clock
Then Lorentz transformation
must be linear. that runs steadily as observed in one frame must run steadily in the other, independent
of the choice of reference event. We conclude that the relation between t and t' must be
a linear one. A similar argument requires that events equally separated in space in the
rocket must also be equally separated in space as measured in the laboratory. Hence
the Lorentz transformation must be linear in both space and time coordinates.
L.5 COMPLETING THE DERIVATION lO l
L5 COMPLETING THE DERIVATION

invariance off the interval completes the story
Equations (L-3) and (L-4) provide coefficients D and H called for in equation (L-5):
t — Bx' + yt'
X = Gx' + v^yyt' (L-6)
About the two constants B and G we know nothing, for an elementary reason. All
events so far considered occured at point x' = 0 'm the rocket. Therefore the two
coefficients B and G could have any finite values whatever without affecting the
numerical results of the calculation. To determine B and G we turn our attention from
an x' = 0 event to a more general event, one that occurs at a point with arbitrary rocket
Demanding invariance of
coordinates x ' and t ' . Then we demand that the spacetime interval have the same
interval . . .
numerical value in laboratory and rocket frames for any event whatever;
Substitute expressions for t and x from equation (L-6):
(fix' + y t'Y -{ G x ' + p^^yt'y = r'2 - x'^
On the left side, multiply out the squares. This leads to the rather cumbersome result
B2 -b 2B yx't' + y V ^ - G^x'^ - 2Gv^{yx't' - vlyyh'^ =
Group together coefficients of coefficients of x'^, and coefficients of the cross-term

x 't' to obtain
y \\ - + 2y{B - v jG ) x 't' - {G^ - B?)x'^ = t'^ - x'^ a-7)
Now, t' and x ' can each take on any value whatsoever, since they tepresent the . . . between any pair of events
coordinates of an arbitrary event. Under these circumstances, it is impossible to satisfy w hatsoever . . .
equation (L-7) with a single choice of values of B and Gunless they are chosen in a very
special way. The quantities B and G must first be such as to make the coefficient of x 't'
on the left side of equation (L-7) vanish as it does on the tight:
2y{B - VJG) = 0
But 7 can never equal zero. The value of 7 = 1/(1 ~ ^ ) *^^ equals unity when =0
and is greatet than this fot any othet values of Hence the left side of this equation
can be zero only if
(B - = 0 or B — v„,G (L-8)
Second, B and G must be such as to make the coefficient of x equal on the left and
right of equation (L-7); hence
- B2 = 1 (L-9)
Substitute B from equation (L-8) into equation (L-9):

. . . lead s to completed form of
G^ - { v ^ fiY = 1 ot G K \-v l^ )= \ Lorentz transformation.
Divide through by (1 — and take the square root of both sides:
But the right side is just the definition of the time stretch factor y, so that
G=y
Substitute this into equation (L-8) to find B:
B = v ^y
These results plus equations (L-1) and (L-6) yield the Lorentz transformation equa
tions:
t= + yt'
The Lorentz transformation x = yx' + v„iyt' (L-lOa)
y = :/
z^ — z
or, substituting for the value of gamma, y = 1/(1 —
x' + Vg,/
(L-1 Ob)
y-y and z —z
In summary, the Lorentz transformation equations rest fundamentally on the re

quired linearity of the transformation and on the invariance of the spacetime interval.
Invariance of the interval was used twice in the derivation. First, we examined a pair
of events both of which occur at the same fixed location in the rocket, so that rocket
time between these events— proper time, wristwatch time— equals the space-time
interval between them (Section L.3). Second, we demanded that the interval also be
invariant between every possible event and the reference event (the present section).
L.6 INVERSE LORENTZ

TRANSFORMATION
from laboratory event coordinates, reckon
rocket coordinates
Equations (L-10) provide laboratory coordinates of an event when one knows the
rocket coordinates of the same event. But suppose that one already knows the
laboratory coordinates of the event and wishes to predict the coordinates of the event
measured by the rocket observer. W hat equations should be used for this purpose?
An algebraic manipulation of equations (L-10) provides the answer. The first two
of these equations can be thought of as two equations in the two unknowns x ' and t ' .
Solve for these unknowns in terms of the now-knowns x and t. To do this, multiply
both sides of the second equation by and subtract corresponding sides of the
L.7 ADDITION OF VELOCITIES 103
resulting second equation from the first. Terms in x cancel to yield
y ■ t' Long derivation of inverse

t = yt' - vuyt'
2
= nydi -K O f == —
vf' = 3y2 t' = -y Lorentz transformation
Here we have useci the definition — 1/(1 — The equation for t' can then be
written
t' = - ^ r ^ y x + yt
A similar procedure leads to the equation for x '. Multiply the first of equations (L-10)
by i'rel and subtract corresponding sides of the first equation from the second — try it!
The y and z components are respectively equal in both frames, as before. Then the
inverse L orentz tran sfo rm atio n eq u atio n s become
t ' = - v ,,^ y x - V y t
/ - y x - v^^yt (L -lla )
y =y
r_
Z — Z
Or, substituting again for gamma, y = 1/(1 — Inverse Lorentz transformation
(1
X • êl t
(L-llb )
and
Equations (L-11) transform coordinates of an event known in the laboratory frame to

coordinates in the rocket frame.
A simple but powerful argument from symmetry leads to the same result. The symmetry
argument is based on the relative velocity between laboratory and rocket frames. W ith
respect to the laboratory, the rocket by convention moves with known speed in the
positive x-direction. W ith respect to the rocket, the laboratory moves with the same speed
but in the opposite direction, the negative x-direction. This convention about positive and
negative directions — not a law of physics! — is the only difference between laboratory
and rocket frames that can be observed from either frame. Lorentz transformation
equations must reflect this single difference. In consequence, the “inverse” (laboratory-
to-rocket) transformation can be obtained from the “ direct” (rocket-to-laboratory)
Short derivation of inverse
transformation by changing the sign of relative velocity, v ^ , in the equations and
interchanging laboratory and rocket labels (primed and unprimed coordinates). Carrying
out this operation on the Lorentz transformation equations (L-10) yields the inverse
transformation equations (L-11).
L.7 ADDITION OF VELOCITIES

add light velocity to light velocity: get light
velocity!
The Lorentz transformation permits us to answer decisively the apparent contradiction
to special relativity outlined in Section L.2, namely the apparent addition of velocities
to yield a resultant velocity greater than that of light.
I travel in a rocket that you observe to move at 4(5 light speed. Out the front of my rocket I fire a
Return to velocity addition bullet that I observe to fly forward at 4fb light speed. Then you measure this bullet to streak
paradox
forward at4l3'k-4l5 = 8j5 = 1.6 light speed, which is greater than the speed of light. There!
SAMP L E PRO B L E M L - 1
T R A N S F O R M I N G O V E R AND B A C K
A rocket moves with speed = 0.866 (so y = 2) 10 meters, y = 1 meters, z = i meters, and t' =
along the x-direction in the laboratory. In the 20 meters of light-travel time with respect to the
rocket frame an event occurs at coordinates x' = reference event.
a. W hat are the coordinates of the event as observed in the laboratory?
b. Transform the laboratory coordinates back to the rocket frame to verify that the
resulting coordinates are those given above.
SOLUTION
a. We already know from Section 3.6 — as well as from the Lorentz transformation,
equation (L-10) — that coordinates transverse to direction of relative motion are
equal in laboratory and in rocket. Therefore we know immediately that
y — y' — 1 meters
z — z' — i meters
The X and t coordinates of the event as observed in the laboratory make use of the
first two equations (L-10):
t = v^îfx' + yt' = (0 .866)(2)(10 meters) + (2)(20 meters)

= 17.32 + 40 = 57.32 merers
and
X — yx' -h v^^fyt' = 2(10 meters) + (0.866)(2)(20 meters)

= 20 + 34.64 = 54.64 merers
So rhe coordinates of the event in the laboratory are t = 57.32 meters, x = 54.64
meters, y = l meters, and z = 3 meters.
b. Use equarion (L-11) ro rransform back from laboratory to rocket coordinates.
t' = ~ v ^{y x + yt = — (0.866)(2)(54.64 meters) + (2)(57.32 meters)

= —94.64 -b 114.64 = 20.00 meters
and
X = y x — v ^ y t = 2(54.64 merers) — (0.866)(2)(57.32 meters)

= 109.28 - 99.28 = 10.00 meters
as given in rhe original statement of the problem.

To analyze this experiment, convert statements about the bullet to statements about
events, since event coordinates are what the Lorentz transformation transforms. Event
1 is the firing of the gun, event 2 the arrival of the bullet at the target. The Lorentz
transformation equations can give locations x,, and X2, ?2 of these events in the
laboratory frame from their known locations x \ , t \ and x 2 , t' 2 in the rocket frame. In
particular:
X2 = yx2 + v^C/t' 2
Xi = y x / +
Subtract corresponding sides of these two equations:
(X2 — xi) = y (x 2 — x 'l) + v,^{y{t'2 — z'l)
We are inrerested in the differences between the coordinates of the two emissions.
Indicate these differences with the Greek uppercase delta. A, for example Ax. Then
this x-equation and the corresponding /-equation become
A x = y A x ' + /^ „ iy A / Incremental event separations

A / = t^reiyAx' + y A / (L-12) define velocities
The subscript “tel” distinguishes relative speed between laboratory and rocket frames
from other speeds, such as particle speeds in one frame or the other.
Bullet speed in any frame is simply space sepatation between two events on its
trajectory measured in that frame divided by time between them, observed in the same
frame. In the special case chosen, only the x-coordinate needs to be considered, since
the bullet moves along the direction of relative motion. Divide the two sides of the first
equation (L-12) by the corresponding sides of the second equation to obtain labora
tory speed:
Ax yA x^ + v^{yiS.t'
A/ t „ ,y A x ' + y A /
Then the time stretch factor y cancels from the numerator and denominator on the
right. Divide every term in numerator and denominator on the right by A /'.
Ax _ ( A x '/A /) +
At v ^^ fA x '/A t') + 1
Now, A x '/ A t' is just distance covered per unit time by the patticle as observed in
the rocket, its speed — call it v , with a prime. And A x / A t is particle speed in the
laboratory — call it simply v. Then (reversing order of terms in the denominatot to
give the result its usual form) the equation becomes
v' + v„
(1-13) Law of Addition of Velocities
1+ V v„
This is called the Law o f A d d itio n o f V elocities in one dimension. A better name is
the Law o f C o m b ination o f V elocities, since velocities do not “add” in the usual
sense. Using the Law of Combination of Velocities, we can predia bullet speed in the
laboratory. The bullet travels at v' — 4 /5 with respect to the rocket and the rocket
moves at v^^ = 4 /5 with tespect to the laboratory. Therefore, speed v of the bullet
(continued on page 1 10)
S A M P L E P R O B L E M L-2
“ ET TU, S P A C E T I M E I ”
Julius Caesar was murdered on March 15 in the a firecracker: event B. The Enterprise moves along
year 44 B.c. at the age of 55 approximately 2000 a straight line in space that connects it with Earth.
years ago. Is there some way we can use the laws of Andtomeda is 2 million light-years distant in our
relativity to save his life? frame. Compared with this distance, you can ne
Let Caesar’s death be the reference event, la glect the orbit of Earth around Sun. Therefore, in
beled 0: = 0, C ~ 0. Event A is you reading this our frame, event B has the coordinates Xg = 2 X
exercise. In the Earth frame the coordinates of 10^ light-years, tg — 2000 years. Take Caesar’s
event A are x^ = 0 light-years, = 2000 years. murder to be the reference event for the Enterprise
Simultaneous with event A in your frame, Starship too (x / = 0 , r / = 0).
Enterprise cruising the Andromeda galaxy sets off
a. How fast must the Enterprise be going in the Earth frame in order that Caesar’s
murder is happening N O W (that is, = 0) in the Enterprise rest frame? Under
these circumstances is the Enterprise moving toward or away from Earth?
b. If you are acquainted with the spacetime diagram (Chapter 5), draw a spacetime
diagram for the Earth frame that displays event 0 (Caesar’s death), event A (you
reading this exercise), event B (firectacker exploding in Andromeda), your line of
N O W simultaneity, the position of the Enterprise, the worldline of the Enter
prise, and the Enterprise N O W line of simultaneity. The spacetime diagtam need
not be drawn to scale.
c. In the Enterprise frame, what are the x and / coordinates of the firecracker
explosion?
d. Can the Enterprise firecracker explosion warn Caesar, thus changing the course of
Earth history? Justify your answer.
SOLUTION
a. From the statement of the problem.
Xo — x / — 0 = 0 Xg — 2 X 10^ light-years
L = l/ = 0 2000 years /g = 2000 years
W e want the speed of the Entetprise such that tg' — 0. The first two Lorentz
transformation equations (L-10) with tg' = 0 become
^'rel y^B
Xg yX g
W e do not yet know the value of X g '. Solve for by dividing the two sides of the
first equation by the respective sides of the second equation. The unknown Xg'
drops out (along with y), and we are left with in terms of the known quantities
tg and x„:
'B _ 2 X 10^ years

— ---- ^ ----- = 10-3 = 0.001
Xg 22 X 10^
X 1 years
0^ vpar<:
This is the desired speed p^^ between Earth and Enterprise frames. This velocity is
a positive quantity, so the Enterprise moves in the positive x-direction, namely
away from Earth.
Surprised to see a speed given as the ratio of a time separation to a space

separation: Then realize that x^ and /g are not displacements of any
particle. Nothing can travel the distance Xg in the time /g, as discussed in d. The
goal here is to find a frame in which Caesar’s death and the firecracker explosion
are simultaneous. For this limited purpose the rocket speed = /g/xg is correct.
Why is the relative velocity so small compared with the speed of light?
Because of the large denominator Xg in the equation that leads to this value.
Consider the string of Earth clocks stretching toward Andromeda when all Earth
clocks read zero time (Caesar’s death). Enterptise clocks read (from equations
L-11 with / = 0) as follows: / ' = — fx . This is an example of the relativity of
simultaneity (Section 3-4). The farther the x-distance from Earth, the earlier will
Enterprise clock read. With x = 2 million light-years, the relative speed does
not have to be large to carry Enterprise time back 2000 years for Earth.
b.
Earth spacetime diagram, showing events 0, A, and B. Not to scale.
c. We need the value of gamma, y, for the inverse Lorentz transformation equation
(L-11). This value is very close to unity, and from it come tg and Xg'.
1 1 1 10-
y = . 1+
[1 - ii - (10-3)231/2 Q _ 1 0 -6 ]l/ 2
tg = -V ,, J X g -b ytg = }»(- 10“ 3 X 2 X 10^ -f 2 X 103)

= y( - 2 X 103 4- 2 X 103) = 0 years
Xb = yxB - vâ JI b = 7(2 X 10^ - 10-3 X 2 X 103) = 2y(l - 10-«) 10«
10-6 10-
= 2^1 “ 10-6)106 = 2^ 1 106
«= 1.999999 X 106 light-years.
W e chose the relative velocity so that the time of the firecracker explosion as
observed in the rocket is the same as the time of Caesar’s death, namely tg' = 0.
The x-coordinate of this explosion is not much different in the two frames because
their relative velocity is so small.
d. There exists a frame — the rest frame of the Enterprise — in which Caesar’s death
and the firecracker explosion occur at the same time. In this frame a signal
connecting the two events would have to travel at infinite speed. But this is
impossible. Therefore the Enterprise cannot warn Caesar; his death is final. Sorry.
(Note: In the language of Chapter 6 , the relation between the two events is
spacelike, and spacelike events cannot have a cause-effect relationship.)
WHY NO THING TRAVELS FASTER THAN LIGHT

A material object traveling faster than light? No! If one Within four years the Klingons used the Federation
did, we could violate the normal order of cause and Technical Database to develop a faster-than-light pro
effect in a million testable ways, totally contrary to all jectile, the slaughtering Super. On that dark day of
experience. Here we investigate one example, making Great Betrayal (reference event 0), the Klingons
use of Lorentz transformation equations. launched the Super at three times light speed toward
the retreating Federation ship.
The Peace Treaty of Shalimar was signed four years
before the Great Betrayal. So pivotal an event was the
Great Betrayal that it was taken as zero of space and Two Federation space colonies lay between the Klin
time. gons and the point of impact of the Super with the Fed
eration ship. A lonely lookout at the first colony wit
By the Treaty of Shalimar, the murderous Klingons nessed with awe the blinding passage of the Super
agreed to stop attacking Federation outposts in return (event 1). Later many citizens of the second colony
for access to the Federation Technical Database. Fed gaped as the Super demolished one of their communi
eration negotiators left immediately after signing the cation structures (event 2) and zoomed on. Both colo
Shalimar Treaty in a ship moving at 0.6 light speed. nies desperately sent warnings toward the Federation
ship, but to no avail since the Super autran the radio
signals.
Finally, at event 3, the Super overtook and destroyed

the Federation ship. All Federation negotiators were
lost in a terrible flash of light and scattering of debris. A
long dark period of renewed warfare began.
But wait! Look again at events of the Great Betrayal, this

time from the point of view of the Federation rocket
ship. Where and when does the Great Betrayal occur in
this frame? The Great Betrayal is the “ hinge of history,”
the reference event, the zero of space and time coordi
nates for all laboratory and rocket frames.
Where and when does the Super explode (event 3) in

this rocket frame? In the Klingon “ laboratory” frame,
event 3 has coordinates X3 = 3 light-years and = 1
y e a r . U se the in v e rse L o re n tz tra n sfo rm a tio n e q u a tio n s
to find the location of event 3 in the rocket frame of the
Federation negotiators. Calculate the time stretch fac
tor y using speed of the Federation rocket, v,ei = 0.6,
with respect to the Klingon frame:
Klittgon (**iaboratory”) spacetime diagram . The Kltngon worldline is
the vertical time axis. The Treaty of Shalimar is followed four years later by 1 1 1
the Great Betrayal {event 0) at which Klingons launch the Super, which moves ^ [, _ ( o .6 )2]>/2 [i - o .36]''2
at three times light speed. Traveling from left to right, the Super passes one
Federation colony {event 1) and then another {event 2). Finally the Super 1
= 1.25
destroys the retreating ship of Federation negotiators {event [0.64]’'2 0.8
treaty signing at Shalimar (subscript Sh), which has lab

oratory coordinates Xj^ = 0 and tsh ~ years:
f’sh ~ ~ ''reiyxjh -h yfsh

= -(0.6)(1.25)(0 years) -h (1.2 5 )(-4 years)
= —5 years
x'sh = yxsh - Vreiyfsh
= (1.25)(0 years) - (0.6)(1.25)(-4 years)
= -f3 years
In the Federation (rocket) spacetime diagram, the

worldline of Federation negotiators extends from
treaty signing at Shalimar vertically to explosion of the
Super (event 3). The worldline of the Klingons extends
from Shalimar diagonally through the launch of the
Super at event 0.
In the Federation spacetime diagram, the worldline for

the Super tilts downward to the right. In this frame
deaths of Federation negotiators (event 3) occur at a
time f' 3 = minus 1 y e a r , that is, b e f o r e the treacherous
^*Rocket” s p a c e tim e d i a g r a m o f d e p a r t in g F e d e r a tio n n e g o tia to r s . In
this frame their destruction comesfirst {event 3), followed by the passage of the
Klingons launch the Super at the event of Great Be
Super from right to left past Federation colonies in reverse order (event 2 trayal (reference event 0). From the diagram one would
followed by event 1). Finally, the Super enters the Klingon launcher without say that the Super moves with three times light speed
doing further damage {event 0). The Great Betrayal has become the Great fro m Federation ship t o w a r d the Klingons. This seems to
Confusion of Cause and Effect. be verified by the fact that in this frame the Super
passes Federation colonies in reverse order, event 2
followed by event 1 , going in the opposite direction.
Yet Federation negotiators have created no such terri
Substitute these values into equations (L-1 1) to reckon ble weapon and in fact are destroyed by it at the mo
the rocket coordinates of event 3: ment they are supposed to launch it, as proved by the
flying photons and debris. More: Klingons suffer no
f ' 3 = - V re iy x a + yf3 damage from the mighty impact of the slaughtering
= —(0.6H1.25)(3 years) + (1.25)(1 year) Super (event 0). Rather, in this frame it enters their
= —2.25 years + 1.25 years = — 1 year launching cannon mild as a lamb.
x'3 = 7x3 — v„|yt3
= (1.25)(3 years) — (0.6)(1.25)(1 year) What have we here? A confusion of cause and effect, a
= 3.75 years — 0.75 year = 3 years confusion that cannot be straightened out as long as we
assume that the Super — or any other material object
Event 3 is plotted in the rocket diagram and the world — travels faster than light in a vacuum.
line of the Super drawn by connecting event 3 with the
launching of the Super at event 0. Notice that this Why does no signal and no object travel faster than
worldline slopes downward to the right. More about light in a vacuum? Because if either signal or object did
the significance of this in a minute. so, the entire network of cause and effect would be
destroyed, and science as we know it would not be
In a similar manner find the rocket coordinates of the possible.
no SPECIAL TOPIC LORENTZ TRANSFORMATION
relative to the laboratory comes from the expression
Velocity addition p arad o x 4 /5 + 4 /5 8/5 8/5 40

resolved
1 + (4/5X 4/5) 1 + 16/25 41/25 41
Thus the bullet moves in the laboratory at a speed less than light speed.
As a limiting case, suppose that the “bullet” shot out from the front of the rocket is,
in fact, a pulse of light. Guess: W hat is the speed of this light pulse in the laboratory?
Here is the calculated answer. Light moves with respect to the rocket at speed f ' = l
while the rocket continues along at a speed = 4 /5 with respect to the laboratory.
The light then moves with respect to the laboratory at speed p :
Light speed is invariant, 1 + 4 /5 _ 9 /5 _

a s expected.
l+ (l)(4/5) 9/5 ~
So light moves with the same speed in both frames, as required by the Principle of
Relativity. Question: Is this true also when a light pulse is shot out of the rear of the
rocket?
S A M P L E P R O B L E M L-3
THE F I R I N G M E S O N
A K° (pronounced “K-naught”) meson at rest in a respect to which the K° meson travels at a speed of
rocket frame decays into 7T'*' (“pi plus”) meson t'rei ~ 0.9. W hat is the greatest speed that one of
and a 7l~ (“pi minus”) meson, each having a the n mesons can have with respect to the labora
speed of = 0.85 with respect to the rocket. Now tory? W hat is the least speed?
consider this decay as observed in a laboratory with
SOLUTION
Let the speeding fC°-meson move in the positive x-direction in the laboratory. In the
rocket frame, daughter TT-mesons come off in opposite directions. Their common line of
motion can, however, be oriented arbitrarily in this frame. The maximum speed of a
daughter TT-meson in the laboratory results when it is emitted in the forward x-direction.
For such a meson, the law of addition of velocities gives
V + 0.85 + 0.9 1.75

= 0.9915
1 + (0.85)(0.9) 1.765
Thus adding a speed of 0.85 to a speed of 0.9 does not yield a resulting speed greater than
1, light speed.
The slowest laboratory speed for a daughter meson occurs when it is emitted in the
negative x-direction in the rocket frame. In this case the velocity of the daughter meson is
negative and the law of addition of velocities becomes a law of subtraction of velocities:
V + V,rel -0.85 + 0.9 0.05

^min , / 0.2128
\ — V v„ 1 - (0.85K0.9) 0.235
Although the minimum-speed meson moves to the left in the rocket, it moves to the right
in the laboratory because of the very great speed of the original fC°-meson in the
laboratory.
REFERENCE 111
L.8 SUMMARY
Lorentz transformation deals with coordinates,
not invariant quantities
Given the space and time coordinates of an event with respect to the reference event in
one free-float frame, the L orentz co o rd in a te tran sfo rm atio n eq u atio n s tell us
the coordinates of the same event in an overlapping free-float frame in relative motion
with respect to the first. The equations that transform rocket coordinates (primed
coordinates) to laboratory coordinates (unprimed coordinates) have the form
v„,x■' + /
)l/2
(1
(L-lOb)
(1 V rel ) l / 2
and
where stands for relative speed of the two frames (rocket moving in the positive
x-direction in the laboratory). The inverse L orentz tra n sfo rm a tio n eq u atio n s
transform laboratory coordinates to rocket coordinates:
/' =
(L-11b)
(1
■y and
in which is treated as a positive quantity. In both these sets of equations, coordi

nates of events are measured with respect to a reference event. It is really only the
difference in coordinates between events that matter, for example %2 ~ = Ax for any
two events I and 2, not the coordinates themselves. This is important in deriving the
Law of Addition of Velocities.
The Law o f A d d itio n o f V elocities or Law o f C o m b in atio n o f V elocities in
one dimension follows from the Lorentz transformation equations. This law tells us the
velocity of a particle in the laboratory frame if we know its velocity v' with respect to
the tocket and relative speed between rocket and laboratory.
(L-13)
1+ vt
REFERENCE
Sample Problem L-3, The Firing Meson, was adapted from A. P. French, Special
Relativity (W .W . Norton, New York, 1968), page 159.
112 EXERCISE L-1 A SUPER-SPEED SUPER?
SPECIAL TOPIC EXERCISES
L-2 a bad clock

PRACTICE N ote: This exercise uses spacetime diagrams, intro
L-1 a super-speed super? duced in Chapter 5.
Take two more steps in the parable of the Great A pulse of light is reflected back and forth between
Betrayal (Box L-1). mirrors A and B separated by 2 meters of distance in
a Find the speed of a new rocket frame moving the ^-direction in the Earth frame, as shown in the
relative to the Klingon frame such that the Super figure (left). A swindler tells us that this device con
travels at 6 times the speed of light in this new frame. stitutes a clock that “ticks” every time the pulse
Hint: Examine the coordinates x ' and t ' of event 3 in arrives at either mirror.
the new frame. The ratio of these two, x ' / t ' , is the The swindler claims that events 1 through 6 are
speed of the Super in this frame. We know the coor sequential “ticks” of this clock (center). However, we
dinates of event 3 in the Klingon frame. There notice that the ticking of the clock is uneven in a
fore . . . rocket frame moving with speed in the Earth
b Find the speed of yet another rocket frame, frame (right). For example, there is less time between
relative to the Klingon frame, such that the Super events 0 and 1 than between events 1 and 2 as mea
travels with infinite speed in this frame. Hint: W hat sured in the rocket frame.
does infinite speed imply about the time t ' between a W hat is the physical basis for the “bad” be
events 0 and 3 in this new frame? havior of this clock? Use the Lorentz transformation
2 meters
mirror A mirror 6
light
pulse EARTH FRAME ROCKET FRAME
EXERCISE L-2. heft: Horizontal light-pulse clock as observed in the Earth frame. Center: Spacetime diagram showing worldlines ofmirrors
A and B and the “uniformly ticking” light pulse as observed in the Earth frame. Right: Time lapses between sequential ticks ofthe light-pulse
clock are not uniform as observed in the rocket frame.
EXERCISE L-4 LIMITS OF NEWTONIAN MECHANICS 113
equations to account for the uneven ticking of this x' — X— t (Lorentz: v « 1) (5)
clock in the rocket frame.
b Use some of the same events 0 through 4 to t' = — v^^x y t (Lorentz: v « 1) (6)
define a “good” clock that ticks evenly in both the
laboratory frame and the rocket frame. From the W hat is the difference between the time transfor
spacetime diagrams, show qualitatively that your mations for the “Newtonian low-velocity limit” of
good clock “runs slow” as observed from the rocket equation (4) and the “ Lorentz low-velocity limit” of
frame — as it must, since the clock is in motion with equation (6)? How can they both be correct? The term
respect to the rocket frame. does not depend on any time lapse, but only on
C Explain why the clock of Figure 1-3 in the text the separation x of the event from the laboratory
is a “good” clock. origin. This term is due to the difference of synchroni
zation of clocks in the two frames.
e In each of the following cases a laboratory
clock (measuring /) at a distance x from the origin as
L-3 the Galilean transformation measured in the laboratory frame is compared with a
passing rocket clock (measuring / ) . Say whether or
a Use everyday, nonrelativistic Newtonian ar not the time difference t — t' = v^^^x can be detected
guments to derive transformation equations between using wristwatches (accuracy of 10~^ second = 3 X
reference frames moving at low relative velocities. 10^ meters of light-travel time) and using modern
Show that the result is electronic clocks (accuracy of 10“ ^ second = 0.3
meter of time).
{Newtonian: « c) {1)
(1) Sports car traveling at 100 kilometers/hour
(Newtonian: « c) (2) (roughly 30 meters/second) located 1000
kilometers down the road from the origin as
where is time measured in seconds and is measured in the Earth frame.
speed in conventional units (meters/second for exam (2) Moon probe traveling at 30,000 kilometers/
ple). List the assumptions you make in your deriva hour passing Moon, 3.8 X 10’ kilometers
tion. from the origin on Earth as measured in the
b Convert equations (1) and (2) to measure time Earth frame.
t in meters and unitless measure of relative velocity, (3) Distance from origin on Earth at which space
frei “ Non/'"- Show the tesults are: probe traveling at 30,000 kilometers/hour
leads to detectable time difference between
x' = X— t (Newtonian: v « 1) (3) rocket wristwatch and adjacent Earth-linked
latticework clock. Compare with Earth-Sun
t' — t (Newtonian: V « 1) (4) distance of 1.5 X 10“ meters.
Do the new units make these equations correct at f Summarize in a sentence or two the conditions
high relative velocity between frames? under which the regular Galilean transformation
C Use the first two terms in the binomial expan equations (3) and (4) will lead to correct predictions.
sion to find a low-velocity approximation for / in the
Lorentz transformation.
L-4 limits off Newtonian
1
y- (1 - v i y 1/2 1 mechanics
(1 -
Use the particle speed = 1/7 (Exercise L-3) as an
Show that this expression differs from unity by less approximate maximum limit for the validity of
than one percent provided p is less than 1/7. A sports Newtonian mechanics. Determine whether or not
car can accelerate uniformly from rest to 60 miles/ Newtonian mechanics is adequate to analyze motion
hour (about 27 meters/second) in 7 seconds. in each of the following cases, following the example.
Roughly how many days would it take for the sports Exam ple: Satellite circling Earth at 30,000
car to reach y = 1/7 at the same constant accelera kilometers/hour = 18,000 miles/hour. Answer:
tion? Light moves at a speed = (3 X 10’ kilometers/
d Set 7 = 1 in the Lorentz transformation equa second) X (3600 seconds/hour) = 1.08 X 10^
tions. Show that the resulting “low-velocity Lorentz kilometers/hour. Therefore the speed of the satellite
transformation” is in meters/metet is v — Nonv/'" ~ 2.8 X 10~’. This
114 EXERCISE L-5 DOPPLER SHIFT
is much less than = 1/7, so the Newtonian de tory formula for / in terms of x and t to derive the
scription of satellite motion is adequate. simple formula for/in terms of f ' and , the relative
a Earth circling Sun at an orbital speed of 30 speed of laboratory and rocket frames.
kilometers/second.
b Electron circling a proton in the orbit of small (wave moves in
est radius in a hydrogen atom. D iscussion: The positive x-direction]
classical speed of the electron in the inner orbit of an
atom of atomic number Z, where Z is the number of e Now observe a wave moving along the nega
protons in the nucleus, is given, for low velocities, by tive x-direction from the same source at rest in the
the expression v = Z / \ ^ l . For hydrogen, Z = 1. rocket frame. Show that the frequency of the wave
c Electron in the inner orbit of the gold atom, for obsetved in the laboratoty frame is
which Z = 79.
d Electron after acceleration from rest through a
volrage of 5000 volts in a black-and-white television
picture tube. D iscussion: W e say that this electron
/
d F +" ^'rel/
'" ' [wove moves in
negative x-direction]
has a kinetic energy of 5000 electron-volts. One elec d Astronomers define the re d sh ift z of light
tron-volt is equal to 1.6 X 10“ ^^ joule. Try using the from a receding astronomical object by the formula
Newtonian expression for kinetic energy.
e Electron after acceleration from rest through a _fendt /obs
voltage of 25,000 volts in a color television picture fob s
tube.
f A proton ot neutton moving with a kinetic Here/nut is the frequency of the light measured in
energy of 10 MeV (million electron-volts) in a nu the frame in which the emitter is at rest and /^ s the
cleus. frequency observed in another frame in which the
emitter moves directly away from the observet.
The most distant quasar reported as of 1991 has a
PROBLEMS tedshift z = 4.897. W ith what fraction of the speed
of light is this quasar receding from us?
L-5 Doppler shift Reference: D. P. Schneider, M. Schmidt, and J. E. G unn, Astronomi
cal Journal, Volume 102, pages 8 3 7 - 8 4 0 (1991).
A sparkplug at rest in the rocket emits light with a
frequency/'’ pulses or waves per second. W hat is the
frequency / of this light as observed in the laboratory?
L-6 transformation of angles
Let this train of waves (or pulses) of light travel in the a A meter stick lies at rest in the rocket frame
positive x-direction with speed c, so that in the course and makes an angle (/)' with the x'-axis. Laboratory
of one meter of light-travel time, f / c of these pulses observers measure the x- andy-projections of the stick
pass the origin of the laboratory frame. It is under as it streaks past. W hat values do they measure for
stood that the zeroth or “fiducial’’ crest or pulse passes these projections, compared with the x '- and ''-pro
the origin at the zero of time— and that the origin of jections measuted by rocket observers? Therefore
the rocket frame passes the origin of the laboratory what angle (f) does the same meter stick make with
frame at this same time. the x-axis of the laboratory frame? W hat is the length
a Show that the x-coordinate of the «th pulse or of the “meter stick’’ as observed in the laboratory
wave crest is related to the time of observation t (in frame?
meters) by the equation b Make the courageous assumption that the di
rections of electric-field lines around a point charge
n = (//c)(r — x) transform in the same way as the directions of meter
sticks that lie along these lines. (Electric field lines
b The same argument, applied in the rocket around a point charge are assumed to be infinite in
frame, leads to the relation length, so the length transformation of part a does not
apply.) Draw qualitatively the electric-field lines due
n= — x'} to an isolated positive point charge at rest in the rocket
frame as observed in ( 1) the rocket frame and (2) the
Express this rocket formula in laboratory coordi laboratory frame. W hat conclusions follow concern
nates X and t using the Lorentz transformation. ing the time variation of electric forces on nearby
Equate the resulting expression f o r / ' to the labora charges at rest in the laboratory frame?
EXERCISE L-10 THE TILTED METER STICK 115
L>7 transformation of y-velocity b Show that your answer to Exercise L-8 gives
the same result when the velocity v ' is given the value
A particle moves with uniform speed v'^ = ts.y'/ lS.t' unity.
along the;/'-axis of the rocket frame. Transform lS.y' c A particle at rest in the rocket ftame emits light
and A / to laboratory displacements A x , A y , and A / uniformly in all directions. Consider the 50 percent of
using the Lorentz transformation equations. Show this light that goes into the forward hemisphere in the
that the x-component and the y-component of the rocket frame. Show that in the labotatory frame this
velocity of this particle in the laboratory frame are light is concentrated in a narrow forward cone of
given by the expressions half-angle (f)g whose axis lies along the direction of
motion of the particle. The half-angle (j)„ is the solu
‘'rel tion to the following equation:
cos (/)„ =
< (1
This result is called the h ea d lig h t effect.
L-8 transformation of velocity
direction L-10 the tilted meter stick
A particle moves with velocity v' in the x 'y ' plane of N ote: This exercise uses the results of Exercise L-7.
the rocket frame in a direction that makes an angle (f)'
with the x'-axis. Find the angle (f) that the velocity A meter stick lying parallel to the x-axis moves in
vector of this particle makes with the x-axis of the the y-direction in the laboratory frame with speed
laboratory frame. (Hint: Transform space and time as shown in the figure (left).
displacements rather than velocities.) Why does this a In the rocket fram e the stick is tilted upw ard in
angle differ from that found in Exercise L-6 on trans the positive x'-direction as shown in the figure
formation of angles? Contrast the two results when (right). Explain why this is, first without using equa
the relative velocity between the rocket and labora tions.
tory frames is very great. b Let the center of the meter stick pass the point
X = y = x ' = y ' = 0 at time t = t ' = 0. Calculate
the angle <f>' at which the meter stick is inclined to the

L-9 the headlight effect x'-axis as observed in the rocket frame. D iscussion:
A flash of light is emitted at an angle (f)' with respect Where and when does the right end of the meter stick
to the x'-axis of the rocket frame. cross the x-axis as observed in the laboratory frame?
a Show that the angle (f) the direction of motion Where and when does this event of right-end crossing
of this flash makes with respect to the x-axis of the occur as measured in the rocket frame? W hat is the
laboratory frame is given by the equation direction and magnitude of the velocity of the meter
stick in the rocket frame (Exercise L-7)? Therefore
cos (/)' + rel where is the right end of the meter stick at / ' = 0 ,
cos (/) =
1+ (f)' when the center is at the origin? Therefore . . .
EXERCISE L-10. Left: Meter stick moving transverse to its length as observed in the laboratory frame.
Right: Meter stick as observed in rocket frame.
116 EXERCISE L-11 THE RISING MANHOLE
L-11 the rising manhole the holes in the grid are much narrower than to the
stationary man, and she certainly does not expect her
N ote: This exercise uses the results of Exercise L-10. skateboard to fall through them. Which person is
A meter stick lies along the x-axis of the laboratory correct? The answer hinges on the relativity of rigidity.
frame and approaches the origin with velocity .A Idealize the problem as a one-meter rod sliding
very thin plate parallel to the xz laboratory plane lengthwise over a flat table. In its path is a hole one
moves upward in the y-direction with speed Vy as meter wide. If the Lorentz contraction factor is ten,
shown in the figure. The plate has a circular hole with then in the table (laboratory) frame the rod is 10
a diameter of one meter centered on the y-axis. The centimeters long and will easily drop into the one-
center of the meter stick arrives at the laboratory meter-wide hole. Assume that in the laboratory frame
origin at the same time in the laboratory frame as the the meter stick moves fast enough so that it remains
rising plate arrives at the plane y = 0. Since the meter essentially horizontal as it descends into the hole (no
stick is Lorentz-contracted in the laboratory frame it “ tipping’’ in the laboratory frame). Write an equa
will easily pass through the hole in the rising plate. tion in the laboratory frame for the motion of the
Therefore there will be no collision between meter bottom edge of the meter stick assuming that t =
stick and plate as each continues its motion. However, / ' = 0 at the instant that the back end of the meter
someone who objects to this conclusion can make the stick leaves the edge of the hole. Eor small vertical
following argument: “In the rocket frame in which velocities the rod will fall with the usual acceleration
the meter stick is at rest the meter stick is not con g. Note that in the laboratory frame we have assumed
tracted, while in this frame the hole in the plate is that every point along the length of the meter stick
Lorentz-contracted. Hence the full-length meter stick begins to fall simultaneously.
cannot possibly pass through the contracted hole in In the meter stick (rocket) frame the rod is one
the plate. Therefore there must be a collision between meter long whereas the hole is Lorentz-contracted to a
the meter stick and the plate.’’ Resolve this paradox 10-centimeter width so that the rod cannot possibly
using your answer to Exercise L-10. Answer unequiv fit into the hole. Moreover, in the rocket frame differ
ocally the question, Will there be a collision between ent parts along the length of the meter stick begin to
the meter stick and the plate? drop ar different times, due to the relativity of simul
Reference: R. Shaw, Am erican Jo u rn a l o f Physics, Volume 30, page
taneity. Transform rhe laboratory equations into the
72 (1962). rocket frame. Show that the front and back of the rod
will begin to descend at different times in this frame.
The rod will “droop” over the edge of the hole in the
L-12 paradox of the rocket frame — that is, it will not be rigid. Will the
skateboard and the grid rod ultimately descend into the hole in both frames? Is
A girl on a skateboard moves very fast, so fast that the the rod really rigid or nonrigid during the experiment?
relativistic length contraction makes the skateboard Is it possible to derive any physical characteristics of
very short. On the sidewalk she has to pass over a grid. the rod (for example its flexibility or compressibility)
A man standing at the grid fully expects the fast short from the description of its motion provided by rela
skateboard to fall through the holes in the grid. Y et to tivity?
the fast girl her skateboard has its usual length and it Reference: W . Rindler, American Journal o f Physics, Volume 29,
is the grid that has the relativistic contraction. To her page 3 6 5 - 3 6 6 (1961).
EXERCISE L -1 1 . W ill the “m eter s tic k ” pass

through the “one-m eter~diam eter" hole w ith
out collision?
EXERCISE L-13 PARADOX OF THE IDENTICALLY ACCELERATED TWINS 117
L-13 paradox off the identically (3) W hat is Jane’s age at event 7?
accelerated twins (4) W hat is Jane’s age at the same time (in this
frame) as event 3?
N ote: This exercise uses spacetime diagrams, intro
duced in Chapter 5. (5) W hat are the ages of Dick and Jane 20 years
after event 3, assuming that neither moves
Two fraternal twins, Dick and Jane, own identical again with respect to this frame?
spaceships each containing the same amount of fuel.
( 6 How far apart in space are Dick and Jane when
)
Jane’s ship is initially positioned a distance to the right
both have ceased accelerating?
of Dick’s in the Earth frame. On their twentieth
birthday they blast off at the same instant in the Earth (7) Compare this separation with their initial (and
frame and undergo identical accelerations to the right final!) separation measured by Mom and Dad
as measured by Mom and Dad, who remain at home in the Earth frame.
on Earth. Mom and Dad further observe that the
d Extend your results to the general case in which
twins run out of fuel at the same time and move
Mom and Dad on Earth observe a period of identical
thereafter at the same speed v. Mom and Dad also
continuous accelerations of the two twins.
measure the distance between Dick and Jane to be the
same at the end of the trip as at the beginning.
(1) At the two start-acceleration events (the two
Dick and Jane compare the ships’ logs of their
events at which the twins start their rockets),
accelerations and find the entries to be identical.
the twins are the same age as observed in the
However when both have ceased accelerating, Dick
Earth frame. Are rhey the same age at these
and Jane, in their new rest frame, discover that Jane is
events as observed in every rocket frame?
older than Dick! How can this be, since they have an
identical history of accelerations? (2) At the two cease-acceleration events (the two
events at which the rockets run out of fuel), are
a Analyze a simpler trip, in which each spaceship the twins the same age as observed in the Earth
increases speed not continuously but by impulses, as frame? Are they the same age at these events as
shown in the first spacetime diagram and the event observed in every rocket frame?
table. How far apart are Dick and Jane at the begin
(3) The two cease-acceleration events are simulta
ning of their trip, as observed in the Earth frame?
neous in the Earth frame. Are they simulta
How far apart are they at the end of their accelera
neous as observed in every rocket frame? (No!)
tions? W hat is the final speed v (not the average
Whose cease-acceleration event occurs first as
speed) of the two spaceships? How much does each
observed in the final frame in which both twins
astronaut age along the worldline shown in the dia
come to rest? (Recall the Train Paradox, Sec
gram? (The answer is not the Earth time of 12 years.)
tion 3.4.)
b The second spacetime diagram shows the two (4) “ When Dick ceases accelerating, Jane is older
worldlines as recorded in a rocket frame moving with than Dick.” Is this statement true according to
the final velocity of the two astronauts. Copy the the astronauts in their final rest frame? Is the
figure. On your copy extend the worldlines of Dick statement true according to Mom and Dad in
and Jane after each has ceased accelerating. Label your the Earth frame?
figure to show that Jane ceased accelerating before
(5) Criticize the lack of clarity (swindle?) of the
Dick as observed in this frame. Will Dick age the
word when in the statement of the problem:
same between events 0 and 3 in this frame as he aged
‘‘However when both have ceased accelerat
in the Earth frame? Will Jane age the same between
ing, Dick and Jane, in their new rest frame,
events 4 and 7 in this frame as she aged in the Earth
discover that Jane is older than Dick!”
frame?
c Now use the Lorentz transformation to find e Suppose that Dick and J ane both accelerate to
the space and time coordinates of one or two critical the left, so that Dick is in front of Jane, but their
events in this final rest frame of the twins in order to history is otherwise the same. Describe the outcome of
answer the following questions this trip and compare it with the outcome of the
original trip.
(1) How many years earlier than Dick did Jane f Suppose that Dick and J ane both accelerate in
cease accelerating? a direction perpendicular to the direction of their
(2) W hat is Dick’s age at event 3? (not the rocket separation. Describe the outcome of this trip and
time t ' oi this event!) compare it with the outcome of the original trip.
118 EXERCISE L-13 PARADOX OF THE IDENTICALLY ACCELERATED TWINS
Earth Frame Observations
Event x-position Time

number (light years) (years)
0 0 0
1 1 4
2 3 8
3 6 12
4 12 0
5 13 4
6 15 8
7 18 12
ROCKET FRAME
EXERCISE L-13. Top: Worldlines of Dick and Jane as observed in the Earth frame of Mom and Dad.
Bottom: Worldlines of Dick and Jane as observed in the “final" rocketframe in which both Dick and Jane
come to rest after burnout.
D iscussion: Einstein postulated that physics in a in a uniform gravitational field do not remain in
uniform gravitational field is, locally and for small synchronism; rather the higher clock reads a later time
particle speeds, the same as physics in an accelerated (“runs faster”) than the lower clock. General relativ
frame of reference. In this exercise we have found that ity also predicts this result, and experiment verifies it.
two accelerated clocks separated along the direction of (Read about the patrol plane experiment in Section
acceleration do not remain in synchronism as observed 4.10.)
simultaneously in their common frame. Rather, the
Reference: S. P. Boughn, American Journal o f Physics, Volume 57,
forward clock reads a later time (“runs faster”) than pages 7 9 1 - 7 9 3 (September 1989). Reference to general relativity
the rearward clock as so observed. Conclusion from result: Wolfgang Rindler, Essential Relativity (Springer, New York,
Einstein’s postulate: Two clocks one above the other 1977), pages 17 and 117.
EXERCISE L-14 HOW DO RODS LORENTZ-CONTRACT? 119
L-14 how do rods Lorenlz- contraction of a fast-moving rod of proper length L„.
contracl? More: We want a careful inspecror riding on the
fast-moving rod to certify that it has the same proper
N ote: Calculus is used in the solution to this exercise; length L(, as it did when it was at rest in the laboratory
so is the formula for Lorentz contracrion from Section frame. To achieve rhis goal, the inspector insists that
5,8.
the pair of accelerating taps be applied to the front
Laboratory observers measure rhe length of a mov and back rod pieces at the same time in the current rest
ing rod lying along its direction of motion in the
frame of the rod. Otherwise the distance between these
laboratory frame. Then the rod speeds up a little.
pieces would not remain rhe same in the frame of the
Again laboratory observers measure its length, which
rod; the rod would change proper length. [Notice that
they find to be a little shorter than before. They call in Exercise L-13 the taps occur at the same time in the
this shortening of length Lorentz contraction. How laboratory (Earth) frame. This leads to results differ
did this shortening of length come about.^ As happens ent from those of the present exercise.}
so often in relativity, the answer lies in the relativity of
b You are the inspector riding along with the
simultaneity.
front and back pieces of the rod. Consider the two
First, how much shortening takes place when the
events of tapping the front and back pieces. How far
rod changes from speed v to speed v + dv} Let be
apart A x' are these events along the x-axis in your
the proper length of the rod when measured at rest. At
(rocket) frame? How far apart A /' in time are these
speed V its laboratory-measured length L will be
events in your frame? Predict how far apart in time Ar
shorter than this by the Lorentz contraction factor
these events are as measured in the laboratory frame.
(Section 5.8):
Use the Lorentz transformation equation (L-10):
L = (l
b it = V yAx' + y l\t'
a Using calculus, show that when the rod speeds

up from y to a slightly greater speed v + dv, the The relative velocity in equation (L-10) is just v,
change in length dL is given by the expression the current speed of the rod. In the laboratory frame is
the tap on the rear piece earlier or later than the tap on
L^vdv the front piece?
dL = - Your answer to part b predicts how much earlier
(1 - t^2)l/2
the laboratory observer measures the tap to occur on
the back piece than on the front piece of the rod. Let
The negative sign means that the change is a shorten the tap increase the speed of the back end by dv as
ing of the rod. We want to explain this change in measured in the laboratory frame. Then during labo
length. ratory time Ar the back end is moving at a speed dv
How is the rod to be accelerated from v to v dv^. faster than the front end. This relative motion will
Fire a rocket attached to the rear of the rod? No, Why shorten the distance between the back and front ends.
not? Because the rocket pushes only against the rear of After time interval A t the front end receives the iden
the rod; this push is transmitted along the rod to the tical tap, also speeds up by dv, and once again moves
front at the speed of a compression wave — very slow! at the same speed as the back end.
We want the front and back to change speed “at the C Show that the shortening dL predicted by this
same time” (exact meaning of this phrase to be deter analysis is
mined later). How can this be done? Only by
prearrangement! Saw the rod into a thousand equal dL = ~dvlS.t — —ybsx'vdv = —vjL^dv
pieces and tap each piece in the forward direction with
L jjdv
a mallet “at exactly 12 noon” as read off a set of
synchronized clocks. To simplify things for now, set (1 - r ; 2)i/2
aside all but the front and back pieces of the rod. Now
tap the front and back pieces ‘‘at the same time. ’’ The which is identical to the result of part a, which we
change in length of the rod dL is then the change in wanted to explain. QED.
distance between these two pieces as a result of the d Now start with the front and back pieces of the
tapping. So much for how ro accelerare the “rod.” rod at rest in the laboratory frame and a distance L^
Now the central question: W hat does it mean to apart. Tap them repeatedly and identically. As they
tap the front and back pieces of the rod “at the same speed up, be sure these taps take place simultaneously
time”? To answer rhis question, ask another: W hat is in the rocket frame in which the two ends are currently
our final goal? Answer: To account for the Lorentz at rest. (This requires you, the ride-along inspector, to
120 EXERCISE L-15 THE PLACE WHERE BOTH AGREE
resynchronize your rod-rest-frame clocks after each set the binomial expansion
of front-and-back taps.) Make a logically rigorous
argument that after many taps, when the rod is mov (1 + z)” ~ 1 + «z for |z| « 1
ing at high speed relative to the laboratory, the length
to show that for low relative velocity, p,=,' p^ / 2 .
of the rod measured in the laboratory can be reckoned
c W hat isp,=,’ for the extreme relativistic case in
using the first equation given in this exercise.
which f'rei 1? Show that in this case is com
e Now, by stages, put the rod back together.
pletely different from ^rel/2.
The full thousand pieces of the rod, lined up but not
d Suppose we want to go from the laboratory
touching, are all tapped identically and at the same
frame to the rocket frame in two equal velocity jumps.
time in the current rest frame of the rod. One set of
Try a first jump to the plane of equal laboratory and
taps increases the rod’s speed from p to p -h dp in the
rocket times. Now symmetry does work: Viewed
laboratory frame. Describe the time sequence of these
from this plane the laboratory and rocket frames
thousand taps as observed in the laboratory frame. If
move apart with equal and opposite velocities, whose
you have studied Chapter 6 or the equivalent, answer
magnitude is given by the equation in part a. A
the following questions: W hat kind of interval —
second and equal velocity jump should then carry us
timelike, lightlike, or spacelike— separates any pair
to the rocket frame at speed with respect to the
of the thousand taps in this set? Can this pair of taps
laboratory. Verify this directly by using the Law of
be connected by a light flash? by a compression wave
Addition of Velocities (Section L.7) to show that
moving along the rod when the pieces are glued back
together? Regarding the “logic of acceleration,” is
P ,= ,' + P ,= ,'
there any reason why we should not glue these pieces P r .1 —
back together? Done! 1 + P ,= ,'P ,= ,'

f During the acceleration process is the reglued
rod rigid— unchanging in dimensions — as observed
L-16 Fizeau experiment
in the rod frame? As observed in the laboratory frame? Light moves more slowly through a transparent ma
Is the rigidity property of an object an invariant, the terial medium than through a vacuum. Let t'medium
same for all observers in uniform relative motion? represent the reduced speed of light measured in the
Show how an ideal rigid rod could be used to transmit frame of the medium. Idealize to a case in which this
signals instantaneously from one place to another. reduced velocity is independent of the wavelength of
W hat do you conclude about the idea of a “rigid the light. Place the medium at rest in a rocket moving
body” when applied to high-speed phenomena? at velocity p„^, to the right relative to the laboratory
Reference: Edwin F. Taylor and A. P. French, American Journal of frame, and let light travel through the medium, also
Physics, Volume 51, pages 889-893, especially the Appendix to the right. Use the Law of Addition of Velocities
(1983). (Section L.7) to find an expression for the velocity p of
the light in the laboratory frame. Use the first two
L-15 the place where both agree terms of the binomial expansion
At any instant there is just one plane in which both the (1 -f- z)” ~ 1 -b «z for |z| « 1
laboratory and the rocket clocks agree.
a By a symmetry argument, show that this plane to show that for small relative velocity between the
lies perpendicular to the direction of relative motion. rocket and laboratory frames, the velocity p of the
Using the Lorentz transformation equations, show light with respect to the laboratory frame is given
that the velocity of this plane in the laboratory frame approximately by the expression
is equal to
P ^m edium ^ ^ rc lf f ^m edium )
= — [1 - (1 - This expression has been tested by Fizeau using

water flowing in opposite directions in the two arms of
an interferometer similar (but not identical) to the
b Does the expression for p,=,> seem strange? interferometer used later by Michelson and Morley
From our everyday experience we might expect that (Exercise 3-12).
by symmetry the “plane of equal time” would move
Reference: H. Fizeau, Comptes rendus, Volume 33, pages 349-355
in the laboratory at half the speed of the rocket. Verify (1851). A fascinating discussion (in French) of some central themes
that indeed this is correct for the low relative velocities in relativity theory— delivered more than fifty years before Einstein’s
of our everyday experience. Use the first two terms of first relativity paper.
‘k/-
4.1 INVITATION TO CANOPUS

is one lifetime enough?
Approximately ninety-nine light-years from Earth lies the star Canopus. The Space
Agency asks us to visit it, photograph it, and return home with our records.
“But that’s impossible, ”we object. “W e have only a little over forty more years to
live. We can spare at most twenty years for the outward trip, and twenty years for the
return trip. Even if we could travel at the speed of light, we would need ninety-nine
years merely to get there.’’
We are greeted with a smile and a cheery, “Think about our request a little longer,
won’t you?”
4.2 STRIPPED-DOWN FREE-FLOAT

FRAME
throw away most clocks and rods
Troubled thoughts fill us tonight. We dream about invariance of the spacetime
interval (Chapter 3). In our dream we find ourselves aboard the rocket used to
establish that result (Section 3.7). However, the numbers somehow have changed
from meters of distance and meters of light-travel time to light-years of distance and
years of time. Suddenly we see things in a new perspective. Three revelations crowd in
on us.
121
1 22 CHAPTER 4 TRIP TO CANOPUS
The flash of light that got reflected did its work — revelation number one— in
establishing the identity of the spacetime interval as measured in either of the two
frames. We can remember invariance of the interval and forget about the reflected
flash. Eliminating it, we eliminate mirror, photodetector and, most of all, those
upward-extended arrays of printout clocks in rocket and laboratory frames whose only
purpose was ro track the light flash.
The economy goes further. For us aboard the rocket, one reliable calendar clock is
enough. As we start our trip from Earth in our dream, that clock by a happy
coincidence shows noon on the Fourth of July, 2000 A.D. — and so do clocks at the
Space Agency Center on Earth. We celebrate our starr by setting off a firecracker.
Later by 6 years — for us — and with a long shipboard program of research and
study already completed, our rocket clock — still in our dreams — tells us it is again
noon on the Fourth of July and we set off a second firecracker. At that very instant,
thanks to the particular speed we had chosen for our rocket relative to Earth, we are
passing Lookout Station Number 8. Lonely lighthouse, it has in it little more than a
Retain a single string of sentry person and a printout clock, one of a series that we have been passing on our
Earth-linked clocks
trip. They have been stationed out in space, fixed one light-year apart according to
Earth measurements. Each clock is calibrated and synchronized to the reference clock
on Earth using a reference flash as described in Section 2.6. The laboratory latticework
of Figure 2-6 has been reduced to a single rightward-stretching string of lookout
stations and their clocks. That we can thus simplify our vision of what is going on from
three space dimensions to one is our first revelation, -vtsr-
4.3 FASTER THAN LIGHT?

choose your frame, then measure velocity!
Revelation number two strikes us as— still dreaming — we pass Lookout Station
Number 8, 8 light-years from Earth: W hat speed! W e glance out of our window and
see the lookout station clock print out “Fourth of July 2010 A.D.” — 10 years later
than the Earth date of our departure. Our rocket clock reads 6 years. We are not
shocked by the discrepancy in times for, apart from the change in scale from meters of
light-travel time to years, the numbers are numbers we have seen before. Nor are we
astonished at the identity of the spacetime interval as evaluated in the two very
different frames. W hat amazes us is our speed. Have we actually covered a distance of
8 light-years from Earth in a time of 6 years? Can this mean we have traveled faster
than light?
We have often been told that no one and no object can go faster than light. Yet here
we are — in our dream — doing exactly that. Speed, yes, we suddenly say to ourselves,
but speed in which frame? Ha! W hat inconsistency! We took the distance covered, 8
light-years, in the Earth-linked laboratory frame, but the time to cover it, 6 years, in
the rocket frame!
Speed ; M easure distance and time At this point we recognize that we can talk about our speed in one reference frame or
in sam e frame our speed in the other frame, but we get nonsense when we mix together numbers
from two distinct reference frames. So we reform. First we pick for reference frame the
rocket. But then we get nothing very interesting, because we did not go anywhere with
respect to the rocket— we just stayed inside.
distance we cover \
' our speed \
relative to I
( with respect to rocket/ (0 light-years)
= 0
time we take to cover\ (6 years)
, rocket frame /
( it in rocket frame /
4.4 ALL OF SPACE IS OURS! 123
In contrast, our speed relative to the Earth-linked reference frame, the extended
laboratory, equals
/ distance we cover \
' our speed \ \w ith respea to Earth/ (8 light-years)
relative to 1 — 0.8 light-speed
, Earth fram e/ / time we take to cover\ (10 years)
\ it in Earth frame /
In other words we— and the rocket— travel, relative to Earth, at 80 percent of the
maximum possible speed, the speed of light. Revelation number two is our discovery
that speed in the abstract makes no sense, that speed has meaning only when referred
to a clearly stated frame of reference. Relative to such a frame we can approach
arbitrarily close to light speed but never reach it.
4.4 ALL OF SPACE IS OURS!

in one lifetime: go anywhere in the cosmos
Revelation number three strikes us as — dreaming on — we think more about passing
Earth-linked lookout stations. Moving at 80 percent of light speed, we travel 8
light-years in the Earth-linked frame in 6 years of our rocket time. Continuing at the
same rate will get us to Canopus in 74 years of our rocket time. Better than 99 years,
but not good enough.
Let’s use — in imagination — a faster rocket! We suddenly remember the super
rocket discussed in demonstrating the invariance of the spacetime interval (Section
3.8). Converting meters of distance and time to years, we realize that traveling in the
super-rocket would bring us to Earth-linked Lookout Station Number 20, 20 Earth-
frame light-years from Earth, in 6 years of our rocket time. When passing this station,
we can see that this station clock reads 20.88 years. Therefore in the Earth-linked
frame out super-rocket speed amounts to 20/20.88 = 0.958 light speed. Continuing
at the same speed would bring us to Canopus in 29.7 years of our rocket time. This is
nearly short enough to meet our goal of 20 years.
Revelation number three gives us a dizzying new sense of freedom. By going fast
enough we can get to Canopus in five minutes of our rocket time if we want! In fact, no
Five minutes to C a n o p u s-
matter how far away an object lies, and no matter how short the time allotted to us,
or to any star!
nothing in principle stops us from covering the required distance in that time. W e have
only to be quite careful in explaining this new-found freedom to our Space Agency
friends. Yes, we can go any distance the agency requires, however great, provided they
specify the distance in the Earth-linked reference frame. Yes, we can make it in any
nonzero time the agency specifies, however short, provided they agree to measure time
on the rocket clock we carry along with us.
To be sure, the Earth-linked system of lookout stations and printout clocks will
record us as traveling at less than the speed of light. Lookouts will ultimately complain
to the Space Agency how infernally long we take to make the trip. But when our Space
Agency friends quiz the lookouts a bit more, they will have to confess the truth; When
they look through our window as we shoot by station after station, they can see that our
clock reads much less than theirs, and in terms of our own rocket clock we are meeting
the promised time for the trip.
Our dream ends with sunlight streaming through the bedroom window. W e lie
there savoring the three revelations; economy of description of two events in a reference
frame stripped down to one space dimension, speed defined always with respect to a
124 CHAPTER 4 TRIP TO CANOPUS
specified reference frame and thus never exceeding light speed, and freedom to go
arbittarily far in a lifetime.
4.5 FLIGHT PLAN

out and back in 40 years to meet
our remote descendents
Wide awake now, we face yesterday’s question: Shall we go to Canopus, 99 light-years
distant, as the Space Agency asks? Yes. And yes, we shall live to retutn and repott.
We take paper and pencil and sketch our plan. The numbers have to be different
from those we dteamed about. Trial and error gives us the following plan: After a
preliminary run to get up to speed, we will zoom past Earth at 99/101 = 0.9802 light
speed. We will continue at that speed all the 99 light-years to Canopus. We will make
a loop around it and record in those few minutes, by high-speed cameta, the features of
that strange star. We will then retutn at unalteted speed, flashing by our finish line
without any letup, and as we do so, we will toss out our bundle of records to colleagues
on Earth. Then we will slow down, turn, and descend quietly to Earth, our mission
completed.
The fitst long run takes 101 Earth years. W e have already decided to travel at a
speed of 9 9/101, or 99 light-years of distance in 101 years of time. Going at that
speed for 101 Earth years, we will just cover the 99 light-years to Canopus. The return
Round trip: 202 Earth y ears
trip will likewise take 101 Earth years. Thus we will deliver our records to Earth 202
Earth-clock years after the start of our trip.
Even briefer will be the account of our trip as it will be perceived in the free-float
rocket frame. Relative to the ship we will not go anywhere, either on the outbound or
on the return trip. But time will go on ticking away on our shipboard clock. Moreover
our biological clock, by which we age, and all other good clocks carried along will tick
away in concord with it. How much time will that rocket clock rack up on the
outbound trip? Twenty years. How do we know? We reach this answer in three steps.
First, we already know from records in the Earth-linked laboratory frame that the
spacetime interval — the proper time— between departure from Earth and artival at
Round trip: 4 0 astronaut y ears
Canopus will equal 20 years:
= (101 years)^ — (99 years)^
= 10,201 years^ — 9801 years^
= 400 years^ = (20 years)^
Second, as the saying goes, “interval is interval is interval” : The spacetime interval
is invariant between frames. The interval as registered in the rocket frame must
therefore also have this 20-year value. Third, in the rocket frame, separation between
the two events (depatture ftom Earth and atrival at Canopus) lies all in the time
dimension, zero in the space dimension, since we do not leave the rocket. Therefore
separation in rocket time itself between these two events is the proper time and must
likewise be 20 years:
R ocket R ocket
= (time separation)^ — (zero)^
= (rocket time)^ = (proper time)^
= (20 years)^
4.6 TWIN PARADOX 12 5
We boil down our flight plan to bare bones and take it to the Space Agency for
approval: Speed 99/101 = 0.9802 light speed; distance 99 light-years out, 99
light-years back; time of return to Earth 202 years after start; astronaut’s aging during
trip, 40 years. The responsible people greet the plan with enthusiasm. They thank us
for volunteering for a mission so unprecedented. They ask us to take our proposal
before rhe Board of Direcrors for final approval. We agree, not realizing what a
hornets’ nest we are walking into.
The Board of Direcrors consists of people from various walks of life, set up by
Congress to assure that major projects have support of the public at large. The media
have reported widely on our proposal in the weeks before we meet with the board, and
many people with strong objecrions to relativity have written to voice their opinions. A
few have met with board members and talked to them at length. We are unaware of
this as we enter rhe paneled board room.
At the request of the chairman we summarize our plan. The majority appear to
welcome it. Several of their colleagues, however, object.
4.6 TWIN PARADOX

a kink in the path explains the difference
“Your whole plan depends on relativity,’’ stresses James Fastlane, “but telativity is a
swindle. You can see for yourself that it is self-contradictory. It says that the laws of
physics are identical in all free-float frames. Very well, here’s your rocket frame and
here’s Earth frame. You tell me that identical clocks, started near Earth at identical
times, each in one of these free-float frames, will read very different time lapses. You
go away and return only 40 years older, while we and our descendants age 202 years.
But if there’s any justice, if relativity makes any sense at all, it should be equally
Which twin travels?
possible to regard you as the stay-at-home. Relative to you, we speed away in the
opposite direction and return. Hence we should be younger than you when we meet
again. In contrast, you say you will be younger than we are. This is a flat contradiction.
Nothing could show more conclusively that neither result can be right. Aging is aging.
It is impossible to live long enough to cover a distance of 99 light-years twice— going
and coming. Forget the whole idea.”
“Jim ,” we reply, “your description is the basis for the famous Twin Paradox, in
which one twin stays on Earth while the other takes the kind of round rrip we have
been describing. Which twin is older when they come together again? I would like to
leave this question for a minure and consider a similar trip across the United States.
“We all know, Jim, that every July you drive straight north on Interstate Highway
35 from Laredo, Texas, on the Mexican border, to Duluth, Minnesota, near the
Canadian border. Your tires roll along a length of roadway equal to 2000 kilometers
and the odometer on your car shows it.
‘'I too drive from Laredo to Duluth, but last year I had to make a stop in Cincinnati,
Ohio, on the way. I drove northeast as straight as I could from Laredo to Cincinnati,
1400 kilometers, and northwest as straight as I could from Cincinnati to Duluth,
Curved path in space
another 1400 kilometers. Altogether, my tires rolled out 2800 kilometers. When we is a longer path
left Laredo you could have said that my route was deviating from yours, and I could
have said with equal justice that yours was deviating from mine. The great difference
between our travels is this, that my course has a sharp turn in it. That’s why my
kilometerage is greater than yours in the ratio of 2800 to 2000.”
Fastlane interrupts: “Are you telling me that the turn in the rocket trajectory ar
Canopus explains the smaller aging of the rocket traveler? The turn in your trip to
Duluth made your travel distance longer, not shorter.”
“That is the difference between path length in Euclidean space geometry and
wrisrwatch time in Lorentz spacetime geometry,” we reply. “In Euclidean geometry
the shortest path length between two points is achieved by the traveler who does not
change direction. All indirect paths are longer than this minimum. In spacetime the
Astronaut who turns around
greatest aging between two events is experienced by the traveler who does not change
a g e s less . . .
direction. For all travelers who change direction, the total proper time, the total
wristwatch time, the total aging is less than this maximum.
“The distinction between distance in Euclidean geometry and aging in spacetime
comes ditectly from the contrast between plus sign in the expression for distance
between two locations and minus sign in the expression for interval between two
events. In going to Duluth by way of Cincinnati I use the plus sign:
/ northward V / eastward
distance: V
( Laredo to 1 =
separation: 1 _|_ I separation: 1
Laredo to I » 1 Laredo to I
Cincinnati /
\ Cincinnati / T \ Cincinnati /
“Contrast this with motion in spacetime. In analyzing my trip to Canopus, I use the
minus sign:
proper time:' rocket time:' Earth time:' Earth distance:'

. becau se of a minus sign! ( Earth to ( Earth to ( Earth to ( Earth to
Canopus , Canopus , Canopus ,
t Canopus ,
“The contrast between a plus sign and a minus sign: This is the distinction between
distance covered during travel in space and time elapsed — aging— during travel in
spacetime."
4.7 LORENTZ CONTRACTION

go a shorter distance in a shorter time
As James Fastlane ponders this response. Dr. Joanne Short breaks in. “The Twin
Paradox is not the only one you have to explain in order to convince us of the
correctness of your analysis. Look at the outward trip as observed by you yourself, the
rocket traveler. You reach Canopus after just 20 years of your time. Yet we know that
Canopus lies 99 light-years distant. How can you possibly cover 99 light-years in 20
years?”
“That is exactly what I dreamed about, Joanne!” we reply. “ First of all, it is
confusing to combine distances measured in one reference frame with time measured
in another reference frame. The 99-light-year distance to Canopus is measured with
respect to the Earth-linked frame, while the 20 years recorded on the outward
traveler’s clock refers to the rocket frame. No wonder the result appears to imply a rate
of travel faster than light. Why not take what I paid for fuel for my car last week and
divide it by the number of gallons you bought today foryo«rcar, to figure the cost of a
gallon of fuel? A crazy, mixed-up, wrong way to work out cost— but no crazier than
that way to figure speed!
“ But your question about time brings up a similar question about distance: distance
between Earth and Canopus measured in the frame in which they are at rest does not
Canopus much closer for astronaut
agree with the distance between them measured from a rocket that moves along the
line connecting them.
4.8 TIME TRAVELER 12 7
“Any free-float frame is as good as any other for analyzing motion — that is the
Principle of Relativity! So think of the entire outward trip in terms of rocket measure
ments. At the starting gun (or firecracker) Earth is rushing past the rocket at speed
9 9 /101. Twenty years later Canopus arrives at the rocket, Canopus also traveling at
that speed, 99/101 in that rocket frame. This means that for the rocket traveler the
Eatth-Canopus distance is only about 20 light-years. In fact it is just the fraction
(99/101) of 20 light-years, so that at speed 99/101 this distance is covered in exactly
20 years.’’
“O f course. We are dealing with L orentz c o n tra c tio n ,” huffs Professor Bright,
who thinks any objection to relativity is a waste of time. He has no head for politics, so Lorentz contraction
does not appreciate how important it is for the public to accept the expenditures
proposed for this project.
He continues, ‘'Think of a very long stick lying with one end at Eatth, the othet end
at Canopus. Each observer, with the help of colleagues, measures the position of the
two ends of this stick at the same time in his or her frame. By this means the outward
rocket traveler measures a shorter length of the stick — a smaller Earth - Canopus
distance — than does an observer in the Earth-linked frame in which the stick lies at
test.
“The factor by which the stick appears contracted in the rocket frame is just the
same as the ratio of rocket time to Earth time for the outwatd trip. This ratio is (20
years)/(101 years). Hence the rocket observer measures the Earth - Canopus distance
to be (99 light-years)(20/101) = 19.6 light-years — just a bit less than 20 light-
years, as you said.
“Everybody has a satisfactory picture: The astronaut can get to Canopus in 20 years
of rocket time because the astronaut’s measurements show Canopus to be slightly less
than 20 light-years distant. We on Earth agree that the time lapse on the rocket clock is
20 years, but our ‘explanation’ rests on the invariance of the interval between the
events of departure from Earth and arrival at Canopus.” Professor Bright pounds the
table: “Why are you giving this poor astronaut such a hard time, when relativity is so
utterly simple?” He is surprised by the outburst of laughter from other board
members and the audience in the room.
4.8 TIME TRAVELER

visit the future, don't come back.
Laura Long has been thoughtfully following the argument. She comments, “You
know, we have been discussing you as a space travelet. But you are a time traveler as
well. Do you realize that by traveling to Canopus and back at 99/101 of light speed,
you journey six generations forward in time: 202 years at 33 years per generation? So
you will be able to visit your great-great-great-great-great-grandchildren at a cost of
only 40 years of your life.”
“Yes, I did think of that,” we reply. “Time and space are not so different in this
Travel to Earth's future
respect. Just as we can travel to as great an Earth-linked distance as we want in as short
a rocket time as we want, so we can also ttavel as fat forward into Earth’s future as we
wish.
“While I was trying various numbers in making up the proposed plan, I realized
that if we traveled not at 99/101 light speed but at 9999/10,001 light speed, then a
round trip would take not 40 rocket years but only 3.96 rocket years and 198 Earth
years. Ten such round trips will age us 39.6 years and bring us back finally at an Earth
time about two thousand years in the future, or some year in the fortieth century. That
is not six generations ahead, but sixty generations, an additional time equal to one
third of recorded history on Earth.”
“Why stop there?” pursues Laura Long excitedly. “Why not go even faster, make
more round trips, and learn the ultimate fate of Earth and its solar system — or even
the still more remote future of the Universe as a whole? Then you could report back to
us whether the Universe expands forever or ends in a crunch.”
“Sorry, but no report back to our century is possible,” smiles Professor Bright.
“There are differences between travel in time and travel in space. To begin with, we
can stand still on Earth if we choose and go nowhere in space with respect to that frame.
Concerning travel through time, however, we have no such choice! Even when we
stand stock still on Earth, we nevertheless travel gently but inevitably forward in time.
Time proceeds inexorably!
Time travel is one w ay “Second, time travel is one way. You may be able to buy a round-trip ticket to
Canopus, but you can get only a one-way ticket to the fortieth century. You can’t go
backward in time. Time won’t reverse.”
Turning to us he adds, “As for the fate of the solar system and the end of the
Universe, our descendants may meet you there as fellow observers, but we ourselves
will have to bid you a fitm and final ‘good-bye’ as you leave us on any of the trips we
have been discussing. The French au revoir— until we meet again — will not do.”
4.9 RELATIVITY OF SIMULTANEITY

we turn around; our changing colleagues say
Earth's clock flies forward
By this time James Fastlane has gotten his second wind. “I am still stuck in this Twin
Paradox thing. The time for the outward trip is less as measured in the rocket frame
than as measured in the Earth frame. But if relativity is correct, every free-float frame is
equivalent. As you sit on the rocket, you feel yourself to be at rest, stationary,
motionless; you measure our Earth watch-station clocks to be zipping by you at high
velocity. Who cares about labels? For you these Earth clocks are in motion! Therefore
the time for the ourwatd trip should be less as measured on the (‘moving’) Earth clock
than as measured on your (‘stationary’) rocket clock.”
We nod assent and he continues. “ Nothing prevents us from supposing the
Rocket observer:
Fewer Earth-clock ticks
existence of a series of rocket lookout stations moving along in step with your rocket
on outward trip . . . and strung out at separations of one light-year as measured in your rocket frame, all
with clocks synchronized in your rocket frame and running at the same rate as your
rocket clock. Now, as Earth passes each of these rocket lookout stations in turn, won’t
those stations read and record the times on the passing Earth clock to be less than their
own times? Otherwise how can relativity be correct?”
“Yes, your prediction is reasonable,” we reply.
“And on the return trip will not the same be true: Returning-rocket lookout
also few er Earth-clock ticks
on return trip
stations will measure and record time lapses on the passing Earth clock to be less than
on their own clocks?”
“That conclusion is inevitable if relativity is consistent.”
“ Aha!” exclaims Mr. Fastlane, “ Now I’ve got you! If Earth clock is measured by
rocket lookout stations to show smaller time lapses during the outward trip — and also
during the return trip — then obviously total Earth time must be less than rocket
round-trip time. But you claim just the opposite: that total rocket time is less than
Earth time. This is a fundamental contradiction. Your relativity is wrong!” Folding his
arms he glowers at us.
4.9 RELATIVITY OF SIMULTANEITY 129
There is a long silence. Everyone looks at us except Professor Bright, who has his
head down. It is hard to think with all this attention. Yet our mind runs over the trip
again. Going out . . . coming back . . . turning around . . . that’s it!
“ All of us have been thinking the wrong way!’’ we exclaim. “We have been talking
as if there is only a single rocket frame. True, the same vehicle, with its traveler, goes
out and returns. True, a single clock makes the round trip with the traveler. But this
vehicle turns around— reverses its direction of travel — and that changes everything.
“Maybe it’s simpler to think of two rockets, each moving without change of
velocity. We ride on the first rocket going out and on the second rocket coming back.
Each of these two is really a rocket frame: each has its own long train of lookout stations
Astronaut jumps from outgoing
with recording clocks synchronized to its reference clock (Figure 4-1). The traveler can
frame to returning frame
be thought of as 'jumping trains’ at Canopus — from outward-bound rocket frame to
inward-bound rocket frame — carrying the calendar clock.
“Now follow Mr. Fastlane’s prescription to analyze the trip in the rocket frame, but
with this change: make this analysis using two rocket frames — one outward bound,
the other inward bound.
“It is 20 yeats by outward-rocket time when the traveler arrives at Canopus. That is
the reading on all lookour station clocks in that outward-rocket frame. One of rhose
lookout stations is passing Earth when this rocket time arrives. Its clock, synchronized
to the clock of the outward traveler at Canopus, also reads 20 years. W hat time does
that rocket lookout-station guard read on the passing Earth clock? For the rocket
observer Earth clock reads less time by the same factor that rocket clocks read less time
(20 years at arrival at Canopus) for Earth observers (who read 101 years on their own
Outgoing rocket:
clocks). This factor is 2 0/101. Elence for the outward-rocket observer the Earth clock
As it arrives at Canopus,
must read 20/101 times 20 years, or 3.96 years.’’ Earth clock read s 3 .9 6 y ea rs
“W hat!” explodes Fastlane. “According to your plan, the turnaround at Canopus
occurs at 101 years of Earth time. Now you say this time equals less than 4 years on
Earth clock.”
“No sir, I do not say that,” we reply, feeling confident at last. “ I did say that at the
same time as the outgoing rocket arrives at Canopus, Earth clock reads 3.96 years as
measured in that outgoing rocket frame. An equally true statement is that at the same
time as the outgoing rocket arrives at Canopus, Earth clock reads 101 years as
measured in the Earthbound frame. Apparently observers in different reference frames
in relative motion do not agree on what events occur at the same time when these events
occur far apart along the line of relative motion.”
Once again Professor Bright supplies the label. “Yes, that is called relativity o f
sim u ltan eity . Events that occur at the same time— simultaneously— judged from
return-rocket lookout stations
-V ....V --V -.... V..... V --V -.... V - -V-....V -< ^ --V -.....
-A-....A --A --A -> -
outgoing-rocket lookout stations
O .........................Q -------------------- O .......................... - o
Earth Earth lookout stations Canopus
FIGURE 4-1. Schematic plot in the E arth-linked fra m e showing the outgoing rocket a n d the
return rocket used in the round trip between E arth a n d Canopus. The two rockets meet at Canopus,
where the traveler jumps from outgoing rocket to return rocket. Each reference frame has its own string of
lookout stations, at rest and synchronized in that frame, shown by small squares, triangles, and inverted
triangles. In this figure the outgoing and return rocket lines of motion are displaced vertically for purposes of
analysis; tn reality, all motion lies along the single line between Earth and Canopus. The figure is not to
scale!
one free-float frame but far apart along the line of relative motion do not occur
simultaneously as judged from another free-float frame.
“As an example of relativity of simultaneity, consider either chain of lookout
stations strung along the line of relative motion. If all clocks in the lookout stations of
one frame strike exactly at noon in that frame, these strikes are not simultaneous as
measured in another frame in relative motion with respect to the first. This is called
relative synchronization o f clocks.
“Incidentally, most of the so-called ‘paradoxes’ of relativity, one of which we are
considering now, turn on misconceptions about relativity of simultaneity.”
Dr. Short breaks in. “W hat about the returning rocket? W hat time on the Earth
clock will the returning rocket lookout station measure as the traveler starts back?”
“That shouldn’t be too difficult to figure out,” we reply. “W e know that the clock
Returning rocket:
on the returning rocket reads 40 years when we arrive home on Earth. And the Earth
As it leaves Canopus,
Earth clock reads 198.04 years clock reads 202 years on that return. Both of these readings occur ar the same place
(Earth), so we do not need to worry about relativity of simultaneity of that reading.
And during the return trip Earth clock records less elapsed time than rocket clocks’ 20
years by the same factor, 20/101, or a total elapsed time of 20 X 20/101 = 3.96
years according ro return rocket observations. Therefore at the earlier turnaround,
return rocket observers will see Earth clock reading 202 — 3.96 = 198.04 years.”
“Wait a minute!” bellows Eastlane. “First you say that the rocket observer sees the
Earth clock reading 3.96 years at turnaround in the outward-bound frame. Now you
say that the rocket observer sees the Earth clock read 198.04 years at turnaround in the
inward-bound frame. Which one is right?”
“ Both are right,” we reply. “The rwo observations are made from two different
frames. Each of these frames has a duly synchronized system of lookout-station clocks,
as does the Earth-linked frame (Figure 4-1). The so-called Twin Paradox is resolved
by noticing that between the Earth-clock reading of 3.96 years, taken from the
outward rocket lookout station at turnaround and the Earth-clock reading of 198.04
years, taken by the returning-rocket lookout station at turnaround, there is a difference
of 194.08 years.
“This ‘jump’ appears on no single clock but is the result of the traveler changing
Forward “jump” in Earth clock
results from frame change frames at Canopus. Yet this jump, or difference, resolves the paradox: For rhe traveler,
the Earth clock reads small time lapses on the outward leg — and also small time
lapses on the return leg — but it jumps way ahead at turnaround. This jump accounts
for the large value of Earth-aging during the trip: 202 years. In conrrast rhe traveler
ages only 40 years during the trip (Table 4-1).
“And notice that the traveler is unique in the experience of changing frames; only
the traveler suffers the terrible jolt of reversing direction of motion. In contrast, the
OBSERVATIONS OF EVENTS ON CANOPUS TRIP

Earth-clock reading observed by
Time measured outgoing-rocket return-rocket
in Earth-linked Time measured lookout stations lookout stations
Event frame by traveler passing Earth passing Earth
Depart Earth 0 years 0 years 0 years

Arrive Canopus 101 years 20 years 20 years X 20/101
= 3.96 years
Depart Canopus 101 years 20 years 3.96 years 202 - 3.96
= 198.04 years
Arrive Earth 202 years 40 years 202 years
4.10 EXPERIMENTAL EVIDENCE 131
Earth observer stays relaxed and comfortable in the same frame during the astronaut’s
entire trip. Therefore there is no symmetry between rocket traveler and Earth dweller,
so no genuine contradiction in their differing time lapses, and the story of the twins is
not a paradox.
“ In fact, the observer in each of the three frames — Earth-linked, outward-rocket,
and inward-rocket— has a perfectly consistent and nonparadoxical interpretation of
the sequence of events. However, in accounting for disagreements between his or her
readings and those of observers in other free-float frames, each observer infers some
misbehavior of measuring devices in these other frames. Each observes less elapsed
All observers agree on result,
time on clocks in the other frame than on his or her own clocks (time stretching or time disagree on reason
dilation). Each thinks that an object lying along the line of relative motion and at rest
in another frame is contracted (Lorentz contraction). Each thinks that lookout-station
clocks in other frames are not synchronized with one another (relative synchronization
of clocks). As a result, each cannot agree with other observers as to which events far
apart along the line of relative motion occur at the same time (relativity of simultane
ity).’’
“ Boy,” growls Fastlane, “all these different reference frames sure do complicate the
story!’’
“ Exactly!” we exclaim. “These complications arise because observations from any
one frame are limited and parochial. All disagreements can be bypassed by talking
Spacetime interval is
only in the invariant language of spacetime interval, proper time, wristwatch time. universal language
The proper time from takeoff from Earth to arrival at Canopus equals 20 years,
period. The proper time from turnaround at Canopus to rearrival at Earth equals 20
years, period. The sum equals 40 years as experienced by the astronaut, period. On the
Earth clock, the proper time between departure and return is 202 years, period. End of
story. Observers in all free-float frames reckon proper times— spacetime intervals
between these events — using their differing space and time measurements. However,
once the data are translated into the common language of proper time, every observer
agrees. Proper times provide a universal language independent of reference frame.”
4.10 EXPERIMENTAL EVIDENCE

objects large and small, slow and fast:
many witnesses for the Canopus trip
Alfred Missouri has remained silent up to this point. Now he declares, “All this theory
is too much for me. I won’t believe a word you say unless you can show me an
experimental demonstration.”
We reply, “Atomic clocks have been placed on commercial airliners and carried
“Airliner” test of twin effect
around Earth, some in an eastward direction, others in a westward direction. In each
case the airliner clocks were compared with reference clocks at the U.S. Naval
Observatory before and after their trips. These clocks disagreed. Results were consist
ent with the velocity-related predictions of special relativity.
“This verification of special relativity has two minor difficulties and a major one.
Minor difficulties: (1) Each leg of a commercial airliner’s trip may be at a different
speed, not always accurately known and for which the time-stretching effect must be
separately calculated. Also, temperature and pressure effects on airborne clocks are
hard to control in a commercial airliner. (2) More fundamentally. Earth rotates,
cartying the reference Naval Observatory clocks eastward around the center of Earth.
Earth center can be regarded as the inertial point in free-float around Sun. With
1 32 CHAPTER 4 TRIP TO CANOPUS
DO WE NEED GENERAL RELATIVITY? NO!

The group takes a break and mills around the conference room, chatting and
eating refreshments. Joanne Short approaches us juggling coffee, a donut,
and her notes.
‘‘I didn’t want to embarrass you in public,” she says, “ but isn’t your plan
faulty because of the turnaround? You can’t be serious about leaping from
one high-speed rocket to another rocket going in the opposite direction. That
means certain death! Be realistic: You and your rocket will have to slow down
over some time period, come to rest at Canopus, then speed up again, this
time headed back toward Earth. During this change of velocity you will be
thrown against the front of the rocket ship, as I’m thrown when I slam on my
car brakes. Release a test particle from rest and it will hurtle forward! Surely
you are not in an inertial (free-float) frame. Therefore you cannot use special
relativity in your analysis of this time period. What does that do to your
description of the ‘jump ahead’ of Earth clocks as you slow down and speed
up again? Don’t you need general relativity to analyze events in accelerated
reference frames?”
“ Oh yes, general relativity can describe events in the accelerated frame,”

we reply, “ but so can special relativity if we take it in easy steps! I like to think
of a freight yard with trains moving at different speeds along parallel tracks.
Each train has its own string of recording clocks along its length, each string
synchronized in that particular train frame. Each adjacent train is moving at a
slightly different speed from the one next to it. Now we can change frames by
walking a c r o s s the trains, stepping from the top of one freight car to the top of
the freight car rolling next to it at a slightly different speed.
“ Let these trains become rocket trains in space. Each train then has an
observer passing Earth as we step on that train. Each observer, by prearran
gement, reads the Earth clock a t th e s a m e tim e that we step onto his train (‘at
the same time’ as recorded in that frame). When you assemble all these data
later on, you find that the set of observers on the sequence of trains see the
Earth clock jumping forward in time much faster than would be expected. The
net result is similar to the single horrible jerk as you jump from the outgoing
rocket to the incoming rocket.
“ Notice that it takes a whole set of clocks in different frames, all reading the
single Earth clock, to establish this result. So there is never any contradiction
between a single clock in one frame and a single clock in any other frame. In
this case special relativity can do the job just fine.”
The directors reassemble and Joanne Short, smiling, takes her place with
them.
respect to this center, one airborne clock moves even faster eastward than Earth’s
surface, while the other one— heading west with respect to the surface— with respect
to Earth’s center also moves eastward, but more slowly. Taking account of these
various relative velocities adds further complication to analysis of results.
“We overcome these two minor difficulties by having an airplane fly round and
round in circles in the vicinity of a single ground-based reference atomic clock.
4.10 EXPERIMENTAL EVIDENCE 133
Then — to a high accuraq'— only relative motion of these two clocks enters into the
special-relativity analysis.
On N ovember 22, 197 5 ,aU .S .N avy P3 C antisubmarine patrol plane flew back
and forth for 15 hours at an altitude of 25,000 to 35,000 feet (7600 to 10,700
“Circling airplane” test
meters) over Chesapeake Bay in an experiment arranged by Carroll Alley and collabo of twin effect
rators. The plane carried atomic clocks that were compared by laser pulse with
identical clocks on the ground. Traveling at an average speed of 270 knots (140
meters per second), the airborne clocks lost an average of 5.6 nanoseconds = 5.6 X
10“^ seconds due to velocity-related effects in the 15-hour flight. The expected
special-relativity difference in clock readings for this relative speed is 5.7 nanoseconds.
This result is remarkably accurate, considering the low relative velocity of the two
clocks: 4.7 X 10“ ^ light speed.
“The major difficulty with all of these experiments is this: A high-flying airplane is
Trouble: Large frame
significantly farther from Earth’s center than is the ground-based clock. Think of an
is not inertial
observer in a helicopter reading the clocks of passing airplanes and signaling these
readings for comparison to a ground-based clock directly below. These two clocks —
the helicopter clock and the Earthbound clock— are at rest with respect to one
another. Are they in the same inertial (free-float) frame? The answer is No.
“We know that a single inertial reference frame near Earth cannot extend far in a
vertical direction (Section 2.3). Even if the two clocks — helicopter and Earthbound
— were dropped in free fall, they could not both be in the same inertial frame.
Released from rest 30,000 feet one above the other, they would increase this relative
distance by 1 millimeter in only 0.3 second of free fall — too rapid a change to be
ignored. B ut the experim ent w ent on not for 0.3 second b u t for 15 hours!
“Since the helicopter clock and Earthbound clock are not in the same inertial frame,
Solution: Use general relativity
their behavior cannot be analyzed by special relativity. Instead we must use general
relativity — the theory of gravitation. General relativity predicts that during the
15-hour flight the higher-altitude clock in the Chesapeake Bay experiment will record
greater elapsed time by 52.8 nanoseconds due to the slightly reduced gravitational
field at altitudes at which the plane flew. From this must be subtracted the 5.7
nanoseconds by which the airborne clock is predicted to record less elapsed time due to
effects of relative velocity. These velocity effects are predicted by both special relativity
and general relativity and were the only results quoted above. The overall predicted
result equals 52.8 — 5.7 = 47.1 nanoseconds net gain by the high-altitude clock
compared with the clock on the ground. Contrast this with the measured value of 47.2
nanoseconds.
“ Hence for airplanes flying at conventional speeds and conventional altitudes,
tidal-gravitational effects on clocks can be greater than velocity-dependent effects to
which special relativity is limited. In fact, the Chesapeake Bay experiment was
conducted to verify the results of general relativity: The airplane pilot was instructed to
fly as slowly as possible to reduce velocity effects! The P3C patrol plane is likely to stall
below 200 knots, so a speed of 270 knots was chosen.
“In all these experiments the time-stretching effect is small because the speed of an
airplane is small compared to the speed of light, but atomic clocks are now so accurate
that these speed effects are routinely taken into account when such clocks are brought
together for direct comparison.”
Professor Bright chimes in. “W hat the astronaut says is correct: We do not have
large clocks moving fast on Earth. On the other hand, we have a great many small
“High-speed radioactive particle'
clocks moving very fast indeed. When particles collide in high-speed accelerators, test of twin effect
radioactive fragments emerge that decay into other particles after an average lifetime
that is well known when measured in the rest frame of the particle. When the
radioactive particle moves at high speed in the laboratory, its average lifetime is
significantly longer as measured on laboratory clocks than when the patticle is at rest.
The amount of lengthening of this lifetime is easily calculated from the particle speed
in the same way the astronaut calculates time stretching on the way to and from
Canopus. The time-stretch factor can be as great as 10 for some of these particles: the
fast-moving particles are measured to live 10 times longer, on average, than their
measured lifetime when at rest! The experimental results agree with these calculations
in all cases we have tried. Such time stretching is part of the everyday experience of
high-energy particle physicists.
“And for these increased-lifetime experiments there is no problem of principle in
Earth frame: Free-float for
particle experiments
making observations in an inertial, free-float frame. While rhey are decaying, particles
cover at most a few tens of meters of space. Think of the flight of each particle as a
separate experiment. An individual experiment lasts as long as it takes one high-speed
particle to move through the apparatus— a few tens of meters of light-travel time.
Ten meters of light-travel time equals about 33 nanoseconds, or 33 X 10“ ^ seconds.
“Can we construct an inertial frame for such happenings? Two ball bearings
released from rest say 20 meters apart do not move together very far in 33 nanosec
onds! Therefore these increased-lifetime experiments could be done, in principle, in
free-float frames. It follows that special relativity suffices to describe the behavior of
the ‘radioactive-decay clocks’ employed in these experiments. W e do not need the
theory of gravitation provided by general relativity.
“O f course, in none of these high-speed particle experiments do particles move
back and forth the way our astronaut friend proposes to do between Earth and
Canopus. Even that back-and-forth result has been verified for certain radioactive iron
"Oscillating iron nucleus”
nuclei vibrating with thermal agitation in a solid sample of iron. Atoms in a hotter
test of twin effect
sample vibrate back and forth faster, on average, and thus stay younger, on average,
than atoms in a cooler sample. In this case the ‘tick of the clock’ carried by an iron atom
is the period of electromagnetic radiation (‘gamma ray’) given off when its nucleus
makes the transition from a radioactive state to one that is not radioaaive. For detailed
reasons that we need not go into here, this particular ‘clock’ can be read with very high
accuracy. Beyond all such details, the experimental outcome is simply stated: Clocks
that take one or many round trips at higher speed record a smaller elapsed time than
clocks that take one or many round trips at lower speed.
“These various results— plus many others we have not described— combine to
Twin effect verified!
give overwhelming experimental support for the predictions of the astronaut concern
ing the proposed trip to Canopus.’’
Dr. Bright sits back in his chair with a smile, obviously believing that he has
disposed of all objections single-handedly.
“Yes,” we conclude, “about the reality of the effect there is no question. Therefore
if you all approve, and the Space Agency provides that new and very fast rocket, we can
be on our way.”
The meeting votes approval and our little story ends.
REFERENCES
The “airliner check” of time stretching (Section 4.10) is reported in J. C. Hafele
and Richard E. Keating, Science, Volume 177, pages 1 6 6 -1 6 7 and 1 6 8 -1 7 0
(14 July 1972).
The “patrol plane” check of general relativity (Section 4.10) is reported by
Carroll O. Alley in Quantum Optics, Experimental Gravity, and Measurement
Theory, edited by Pierre Meystre and Marian O. Scully (Plenum, New York,
1983). See also 1976 physics Ph.D. theses by Robert A. Reisse and Ralph E.
Williams, University of Maryland.
The “radioactive nuclei” check of time stretching (Section 4.10) is reported in R.
V. Pound and G. A. Rebka, Jr., Physical Review Letters, Volume 4, pages
llA - T V b (1960).
EXERCISE 4-3 A RELATIVISTIC OSCILLATOR 135
CHAPTER 4 EXERCISES
N ote: The following exercises are related to the story 4-2 one-way twin paradox?
line of this chapter. Additional exercises may be se
A worried student writes, “I still cannot believe your
lected from Chapter 3 or the Special Topic on the
solution to the Twin Paradox. During the outward
Lorentz Transformation following Chapter 3.
trip to Canopus, each twin can regard the other as
moving away from him; so how can we say which
twin is younger? The answer is that the twin in the
rocket makes a turn, and in Lorentz spacetime geom
4-1 practical space travel etry, the greatest aging is experienced by the person
who does not turn. This argument is extremely unsat
In 2200 A.D. the fastest available interstellar rocket
isfying. It forces me to ask: W hat if the rocket breaks
moves ^x: V = 0.75 of the speed of light. James Ab
down when I get to Canopus, so that I stop there but
bott is sent in this rocket at full speed to Sirius, the
cannot turn around? Does this mean that it is no
Dog Star (the brightest star in the heavens as seen
longer possible to say that I have aged less than my
from Earth), a distance D = 8.7 light-years as mea
Earthbound twin? But if not, then I would never have
sured in the Earth frame. James stays there for a time
gotten to Canopus alive.” Write a half-page response
T = 7 years as recorded on his clock and then returns
to this student, answering the questions politely and
to Earth with the same speed p — 0. 75. Assume
decisively.
Sirius is at rest relative to Earth. Let the departure
from Earth be the reference event (the zero of time
and space for all observers).
According to Earth-linked observers:
4-3 a relativistic oscillator
a At what time does the rocket arrive at Sirius? In order to test the laws of relativity, an engineer
b At what time does the rocket leave Sirius? decides to construct an oscillator with a very light
c At what time does the rocket arrive back at oscillating bob that can move back and forth very fast.
Earth? The lightest bob known with a mass greater than zero
According to James’s observations: is the electron. The engineer uses a cubical metal box,
d At what time does he arrive at Sirius? whose edge measures one meter, that is warmed
e At what time does he leave Sirius? slightly so that a few electrons “boil off” from its
f At what time does he arrive back at Earth? surfaces (see the figure). A vacuum pump removes air
g As he moves toward Sirius, James is accompa from the box so that electrons may move freely inside
nied by a string of outgoing lookout stations along his without colliding with air molecules. Across the mid
direction of motion, each one with a clock synchro dle of the box — and electrically insulated from it—
nized to his own. W hat is the spatial distance between is a metal screen charged to a high positive voltage by
Earth and Sirius, according to observations made with a power supply. A voltage-control knob on the power
this outgoing string of lookout stations? supply can be turned to change the DC voltage V„
h One of James’s outgoing lookout stations, call between box and screen. Let an electron boiled off
it Q, passes Earth at the same time (in James’s outgo from the inner wall of the box have very small velocity
ing frame) that James reaches Sirius. W hat time does initially (assume that the initial velocity is zero). The
Q's clock read at this event of passing? W hat time electron is attracted to the positive screen, increases
does the clock on Earth read at this same event? speed toward the screen, passes through a hole in the
i As he moves back toward Earth, James is ac screen, slows down as it moves away from the attract
companied by a string of incoming lookout stations ing screen, stops just short of the opposite wall of the
along his direction of motion, each one with a clock box, is pulled back toward the screen; and in this way
synchronized to his own. One of these incoming look oscillates back and forth between the walls of the box.
out stations, call it Z, passes Earth at the same time (in a In how short a time T can the electron be made
James’s incoming frame) that James leaves Sirius to to oscillate back and forth on one round trip between
return home. W hat time does Z ’s clock read at this the walls? The engineer who designed the equipment
event of passing? W hat time does the clock on Earth claims that by turning the voltage control knob high
read at this same event? enough he can obtain as high a frequency of oscilla
To rea/ly understand the contents of Chapter 4, tion / = 1 /T as desired. Is he right?
repeat this exercise many times with new values of p, b For sufficiently low voltages the electron will
D, and T that you choose yourself. be nonrelativistic— and one can use Newtonian me-
136 EXERCISE 4-3 A RELATIVISTIC OSCILLATOR
chanics to analyze its motion. For this case the fre heavily in the region where it is reliable and indicated
quency of oscillation of the electron is increased by by dashes elsewhere. Second curve: the extreme rela-
what factor when the voltage on the screen is doubled? rivisric value from part d, also with dashed lines
D iscussion: At corresponding points of the elec where not reliable. From the resulting graph estimate
tron’s parh before and afrer voltage doubling, how quantitatively the voltage of transition from the
does the Newtonian kinetic energy of rhe electron nonrelativistic to the relativistic region. If possible
compare in the two cases? How does its velocity com give a simple argument explaining why your resulr
pare in rhe two cases? does or does nor make sense as regards order of mag
c W hat is a definite formula for frequency/as a nitude (that is, overlooking factors of 2, 7T, etc.).
funcrion of volrage in the nonrelativistic case? W ait as f Now think of the round-trip “proper period’’
late as possible to substitute numbers for mass of of oscillation T experienced by the electron and logged
elecrron, charge of electron, and so forrh. by its recording wristwatch as it moves back and forth
d W hat is the frequency in the extreme relativis across rhe box. Ar low electron speeds how does rhis
tic case in which over most of its course rhe elecrron is proper period compare wirh rhe laboratory period
moving . . . (rest of sentence suppressed!) . . . ? recorded by the engineer? W hat happens at higher
Call this frequency . electron speeds? At extreme relativistic speeds? How is
e On rhe same graph, plot two curves of the this reflected in the “proper frequency’’ of oscillarion
dimensionless quantity f / f ^ as functions of the di Tproper experienced by rhe elecrron? On rhe graph of
mensionless quantity qW^/{2mc'^), where q is the parr e draw a rough curve in a different color or
charge on the electron and m is its mass. First curve: shading showing qualitatively the dimensionless
the nonrelativistic curve from parr c to be drawn quantity/p„p,^//„„ as a function of qV„/{2mc^).
TREKKING
THROUGH SPACETIME
5.1 TIME? NO. SPACETIME MAP? YES.

no such thing as the unique time off an event!
Events are the sparkling grains of history. They define spacetime. Spacetime, yes.
Time, no.
“ Time” of an event has
‘‘Time, no”? How come? Time here in Tokyo, at this enthronement of the successor no unique meaning
of the Emperor Hirohito? Where is any meter to be seen that shows any such quality of
location as time? Meter to measure the temperature here and now? Yes, this thermom
eter. Meter to measure atmospheric pressure here and now? Yes, this barometer. But
look as we will, nowhere can we see any meter that we can poke into the space
hereabouts to measure its “time.” The time of an event? Impossible! No such thing.
Time is not “meterable.”
Anything with which to compare time? Yes. Odometer reading, whether miles or
kilometers, on the dashboard of our car. There’s no such thing as the odometer reading
of Tokyo. Try every gadget one can, thrust it out into this Tokyo air, not one will
C a r mileage d epends on car's
register anything with the slightest claim to be called the odometer reading of these
path between places
hereabouts.
W hat about looking at the dashboards of the cars in this neighborhood? N ot all of
them; that would be nonsense. Only the cars that were new, with odometer reading
zero, at the time of Hirohito’s own enthronement.
Now at last we are getting into a line of questioning that shows some prospect of
clearing up what we mean by “time.” W e ask our companion, “W hat do all those
day-and-year-counting wristwatches now read that were set to zero at the time of that
earlier ceremony?”
“Sixty-two years, two days,” is her first reply. But then we ask, “W hat about that
team that zoomed out to the nearest eye-catching star. Alpha Centauri, and back with
almost the speed of light? Didn’t they get back ten years younger than we stay-at-
homes?”
137
138 CHAPTER 5 TREKKING THROUGH SPACETIME
“Yes,” she agrees, “surely their wristwatches now read fifty-two years, not sixty-
W ristwatch reading d epends on
two. So let me draw the lesson. There is no such thing as time. There is only totalized
its history of travel between events
interval of time, time as that interval is racked up between the enthronement of
Hirohito and the enthronement of the new Emperor Akihito, between event A and
event B, on a wristwatch that has undergone its own individual history of travel
from A to B.”
“I agree. The concept of time does not apply to location in spacetime. It applies to
individual history of travel through spacetime.”
“How apt the comparison with odometer reading. Each dashboard shows, not the
kilometerage of Akihito, but the kilometers traveled by that particular car between the
one imperial ceremony and the other.”
Yes, it is nonsense to attribute a kilometer reading to Tokyo. However, it is not at
G eo g rap h ic mop assigns
kilometer coordinates to places
all nonsense to make a map showing where Tokyo lies relative to all the towns
roundabout, a map in which kilometers do appear, kilometers north and south,
kilometers easr and west. Likewise the term “the time” of an event is totally without
meaning. However, that event — and every event near it — lends itself to display on a
spacetime diagram (Figure 5-1), with distance (the locator of latticework clock)
running in one direction, and in another direction time (the reading printed out by that
clock on the occasion of that event). Time as employed in this sense acquires meaning
only because it serves as a measure on a latticework-defined map. A different lattice-
Spacetim e map assigns sp ace
work? A different set of clocks, different readings on those clocks, a different map —
and time coordinates to events
but same events, same spacetime, same tools to measure the history-dependent
interval between event and event.
Only on such a spacetime plot does one see at a glance the layout of all nearby
events, and how one history of travel from event A to event B differs from another.
One problem in making our map: Spacerime has four dimensions— three space
dimensions plus time. W e picture our event points most readily when they occupy a
two-dimensional domain and let themselves be dotted in on a two-dimensional page.
Limit attention to one
Therefore for the present we limit attention to time and one space dimension; to
sp a ce dimension plus one time
dimension events, whatever their timing, that occur on one line in space. All events that do not
occur on this line we ignore for now. The space location of each event on this line we
plot along a ho rizo n tal axis on the page. The lattice-clock time at which an event
occurs we plot along a vertical axis, from bottom to top of the page. Space and time
we measure in the same unit, for example meters of distance and meters of time — or
light-years of distance and years of time. We call the result a spacetim e m ap or a
spacetim e diagram . Each spacetime map represents data from a particular reference
frame, for example “the laboratory frame.” Figure 5-1 shows such a spacetime map.
Five sample event points appear on the laboratory spacetime map of Figure 5-1,
events labeled 0, A, B, C, and D.
• E vent 0 is the referen ce event, the firing of the starting gun, which we take
to locate zero position in space and the zero of time. For our own convenience,
we place point 0 at the origin of the spacetime map and measure space and time
locations of all other events with respect to it.
same place FIGURE 5-1. Laboratory space-

tim e map, showing the reference
event O, other events A, B, C, a n d
D, a horizontal dashed line of si
m ultaneity in time, a n d a verti
c al dashed line o f equal position
in space.
space ■
5.2 SAME EVENTS; DIFFERENT FREE-FLOAT FRAMES 139
• E vent B stands on the vertical time axis, directly above reference event 0.
Therefore event B occurs at a later time than event 0. Event B lies neither to the
right of the reference event nor to the left; its horizontal (space) location is zero.
Therefore it occurs at the same place as the reference event 0 in the laboratory
but later in time.
• E vent A lies on the horizontal space axis, directly to the right of reference event
0. Therefore event A occurs at a different space location than event 0. It is
neither above nor below event 0; its vertical (time) location is zero. Therefore it
occurs at the same time as reference event 0 as observed in the laboratory.
• E vent C rests above and to the right of the reference event. Standing higher
than the reference event on the map, event C occurs later in time than 0 in this
frame. Since it lies to the right, event C occurs at a positive space location with
respect to event 0 in this frame.
• E vent D reposes above and to the left of the reference event. It also occurs later
in time than reference event 0 but at a negative space location with respect to
event 0 as observed in the laboratory.
Scatter other event points on the spacetime map. Each event point can represent an
important happening. Then a single glance at the spacetime map gives us, in principle,
a global picture of all significant events that have occurred along one line in space and
as far back in time as we wish to look. The spacetime map puts all this history at our
fingertips!
In exploring history, w e may want to know which events occurred at the same time
as others in the laboratory free-float frame. Two events that occur at the same time
have the same vertical (time) location on the spacetime map. A horizontal line drawn Horizontal line on spacetime
through one event point passes through all events simultaneous with that event in the diagram picks out events that a re
simultaneous in this frame
given frame. In Figure 5 -1, the dashed horizontal line shows that events B and D are
simultaneous as observed in the laboratory frame, although they occur at different
locations in space. Similarly, events 0 and A are simultaneous as observed in this
frame.
When we wish to “retell history,” we draw a sequence of horizontal lines above one
another on the spacetime map. We mimic the advance of time by stepping in
imagination from one horizontal line to the next horizontal line above it, noting which
events occur at each time.
Vertical lines on the spacetime map indicate which events occur at the same place
along the single line in space. Events A and C in Figure 5-1 occur at the same space
location as measured in the laboratory, but at different times as measured in this
frame. Similarly, events 0 and B occur at the same place as one another in the
laboratory.
5.2 SAME EVENTS; DIFFERENT

FREE-FLOAT FRAMES
different frames: different points for an event
on their spacetime maps, but same spacetime
interval between two events
Figure 5-1 demonstrates two great payoffs of the spacetime map: (1) It places space
and time on an equal footing, thus recognizing a basic symmetry of nature. (2) It
allows us to review at a single glance the whole history of events and motions that have
occurred along the given line in space.
140 CHAPTERS TREKKING THROUGH SPACETIME
We want to take advantage of a third payoff of the spacetime map; Plot the same
Sam e events, difFerent frames:
events on two, three, or more spacetime maps based on two, three, or more different
Different spacetim e maps
free-float frames in uniform relative motion. Compare. In this way analyze the various
space and time relations among these events as measured in difFerent frames. Why do
this? In order to find out what is difFerent in the difFerent frames and what remains the
same.
Figure 5-3 shows three spacetime maps — for laboratory, rocket, and super-rocket
free-float frames. The super-rocket moves faster than the rocket with respect to the
laboratory (but not faster than light!). On each of the three spacetime maps we plot the
same two events: the events of emission E and reception f? of a light flash. These are the
two events analyzed in Chapter 3 to derive the expression for the spacetime interval. As
a reminder of the physical phenomena behind events E and R, refer to Figure 5-2.
The light flash is emitted (event E) from a sparkplug attached to the reference clock
of the first rocket. Take event E as the reference event, called event 0 in Figure 5-1. By
prearrangement the sparkplug fires at the instant when both the rocket reference clock
and the super-rocket reference clock pass the laboratory reference clock. All three
LABORATORY PLOT
A
0 0 0 ®
000® 0 0 0 0 ®
0 0 0 ® 0 0 0 0 ®
0 0 0 0 © ® ® ® ®
E R
ROCKET PLOT
A
AAAAAAAAA AAAAAAAAA
AAAAA AAAAAA
AA AA
\1
SUPER-ROCKET PLOT
FIGURE 5-2 (Figure 3-5 repeated). The flash p a th as recorded in three differentfram es, showing
event E, emission of the flash, a n d event R, its reception a fte r reflection. Squares, circles, and
triangles represent the latticework of recording clocks in laboratory, rocket, and super-rocket frames, respec
tively. The super-rocket frame moves to the right with respect to the rocket, so that the event of reception, R,
occurs to the left of the event of emission, E, as measured in the super-rocket frame. The reflecting mirror is
fixed in the rocket, hence appears to move from left to right in the laboratory and from right to left in the
super-rocket.
5.2 SAME EVENTS; DIFFERENT FREE-FLOAT FRAMES 141
laboratory time FIGURE 5-3. Spacetime maps fo r three

fram es, showing emission o f the reference
flash a n d its reception a fte r reflection. The
hyperbola drawn in each map satisfies the equa
tion for the invariant interval (or proper time),
which has the same value in all three frames:
(interval)^ = (time)^ — (space)^.
LABORATORY
SPACETIME MAP
rocket time
SPACETIME MAP
super-rocket time
SUPER-ROCKET
SPACETIME MAP
reference clocks are set to read zero at this reference event, whose event point is placed
at the origin of all three spacetime maps.
Now use the latticework of meter sticks and clocks in each free-float frame (clocks
pictured in Figure 5-2) to measure the position and time of every other event with
respect to the reference event. In particular, record the position and time of the
reception (event R) of the flash in each of the three frames.
The reception of the light ray (event R) occurs at different locations and at different
times as measured in the three frames. In the rocket the reception of the reflected flash
occurs back at the reference clock (the zero of position) and 6 meters of time later, as
seen in Figure 5-2 and more directly in Figure 5-3 (center):
Same events, different frames: R ocket: (position of reception, event R) — 0

Different space and time R ocket: (time of reception, event R) = 6 meters
coordinates
Emission and reception occur at the same place in the rocket frame. Therefore the
rocket time, 6 meters, is just equal to the interval, or proper time, between these two
events:
R ocket R ocket
_ / time of y position o f y
(proper time)'
\ reception / ( reception /
R o ck et R ocket
_ / time of y _
(zero)^ — (6 meters)^
\ reception/
In the laboratory the reception event R occurs at a time greater than 6 meters, as can
be seen from the expression for interval:
/ time of y / p o s .r io n o f y _ ,
\ reception/ \ reception /
In this equation the square of 6 meters results from subtracting a positive quantity
from the square of the laboratory time of reception. Therefore the laboratory time of
reception itself must be greater than 6 meters:
L aboratory: (position of reception, event R) = 8 meters

L aboratory: (time of reception, event /?) = 10 meters
In the laboratory frame, reception appears to the right of the emission, as seen in Figure
5-2. Hence it is plotted to the right of the origin in the laboratory map (Figure 5-3,
top).
In the super-rocket frame, moving faster than the rocket with respect to the
Same events, different frames: laboratory, the event of reception appears to the left of the emission (Figure 5-2).
Sam e spacetim e interval
Therefore the space separation is called negative and plotted to the left of the origin in
the super-rocket map (Figure 5-3, bottom). The time separation in the super-rocket is
greater than 6 meters, by the same argument used for the time of reception in the
laboratory frame:
S u p er-ro ck et S uper-rocket
/ time of y _ / position o f y .
■(6 meters)^
\reception/ \ reception /
In this equation, the space separation is a negative quantity. Nevertheless its square is a
positive quantity. So the equation says that the square of 6 meters results from
subtracting a positive quantity from the square of the super-rocket time of reception.
Therefore the super-rocket time separation must also be greater than 6 meters:
S uper-rocket: (position of reception, event R) — — 20 meters

S uper-rocket: (time of reception, event R) = 20.88 meters
5.4 WORLDLINE 143
5.3 INVARIANT HYPERBOLA

all observers agree: ''event point lies
somewhere on this hyperbola"
Different reception points marked R in different spacetime maps all refer to the same
event. W hat do these different sepatations of the same event from the reference event
have in common? They all satisfy invariance of the interval, reflected in the equation
(time separation)^ — (space sepatation)^ = (interval)^ constant
Constant? Constant w ith respect to w hat?
— W ith respect to free-float frame. Record different space and time measurements in
different frames, but figure out from rhem always the same interval.
Curves drawn on the three maps conform to this equation. This kind of curve, in which
the difference of two squares equals a constant, is called a hyperbola. Somewhere on
this hyperbola is recorded the time and position of one and the same reception event as Invariant hyperbola; Locus of
sam e event in all rocket frames
measured in every possible rocket and supet-tocket frame. Same reception event,
different frames, all summarized in one hyperbola, the in v arian t hyperbola.
Spacetime arrows in all three maps connect the same pair of events. They imply the
identical invariant interval. They embody the same spacetime reality. In a deep sense
these thtee attows on the page tepresent the same artow in spacetime. Spacetime maps
of different observets show different projections — different petspectives — of the
same atrow in spacetime.
The same arrow? The same magnitude fo r the spacetime arrow pictured in a ll three
maps of Figure 5 - 3 ? Then why do the three arrows have obviously different lengths in
the three maps?
Because the paper picture of spacetime is a lie! The length of an arrow on a piece of
paper is Euclidean, related to the sum of squares of the space separations of the
endpoints in two perpendicular directions. Euclidean geometry works fine if what is
being represented is flat space, for example the map of a township. But Euclidean
geometry is the wrong geometry and betrays us when we try to lay out time along one
direction on the page. Instead we need to use Lorentz geometry of spacetime. In
Lorentz geometry, time must be combined with space through a difference of squares
to find the correct magnitude of the resulting spacetime vector— the interval. That
is why the arrows in the different spacetime maps of Figure 5-3 seem to be of
different lengths. The reality that these lengths represent, however— rhe value of
the interval between two events— is the same in all three spacetime maps.
5.4 WORLDLINE
the moving particle traces out a line — its
woHdline— on the spacetime diagram
We describe the world by listing events and showing how they relate to one another.
Until now we have focused on pairs of events and spacetime intervals between them.
Now we turn to a whole chain of events, events that track the passage of a particle
through spacetime. Think of a speeding sparkplug that emits a spark every meter of
time read on its own wrisrwatch. Each spark is an event; the collection of spark events
forms a chain that threads through spacetime, like pearls. String the pearls together.
String of event pearls: Worldline!
The thread connecting the pearl events, tracing out the path of a particle through
spacetime, has a wonderfully evocative name: w o rld lin e. The sparkplug travels
through spacetime trailing its worldline behind it.
The speeding sparkplug is only an example. Every particle has a worldline that
connects events along its spacetime path, events such as collisions or near-collisions
(close calls) with other particles.
Events — pearls in spacerime — exist independent of any reference frame we may
choose to describe them. A worldline strings these event pearls together. The world
line, too, exists independent of any reference frame. A particle traverses spacetime—
follows a worldline — totally oblivious to our poor efforts to describe its motion using
one or another free-float frame. Yet we are accustomed to using a free-float frame and
its associated latticework of rods and clocks. One clock after another records its
encounter with the particle. The worldline of the particle connects this chain of
encounter events.
We can draw this worldline of a particle on the spacetime map for this reference
frame. Such worldlines are shown in Figure 5-5 and in later figures of this chapter.
Worldline versus line on Strictly speaking, the line drawn on the spacetime map is not the worldline itself. It is
spacetime map
an image of the worldline— a strand of ink printed on a piece of paper. When we use a
highway map, we often refer to a line drawn on the paper as “the highway.” Yet is not
the highway itself, but an image. Ordinarily this causes no confusion; no one tries to
drive a car across a highway map! Similarly, we loosely refer to the line drawn on the
spacetime map as the worldline, even though the worldline in spacetime stands above
and beyond all our images of it.
The worldline is seen in no way more clearly than through example. Particle 1 starts
at the laboratory reference clock at zero time and moves to the right with constant
Exam ples of worldlines
speed (Figure 5-4). As particle 1 zooms along a line of laboratory latticework of clocks,
each clock it encounters records the time at which the particle passes. Each clock record
shows where the clock is located and the time at which particle 1 coincides with the
clock. “Where and when” determines an event, the event o f coincidence of particle
and recording clock. Afterwards the chief observer travels throughout the lattice of
clocks, collecting the records of these coincidence events. She plots these events as
points on her spacetime map. She then draws a line through event points in sequence
— the worldline of particle 1 (Figure 5-5).
Particle 1 moves w ith constant speed along a single direction in space. T he distance
it covers is equal for each tick o f the laboratory clocks. T he w orldline o f particle 1
shows equal changes in space during equal lapses o f tim e by being straight on the
spacetim e m ap.
Particle 2 moves to the right faster than particle 1 and so covers a greater distance in
the same time lapse (Figure 5-4). Lattice clocks record their events of coincidence with
particle 2, and the observer collecrs rhese records and plots the worldline of particle 2
on the same spacetime map (worldline shown in Figure 5-5).
And so it goes: Particle 3 is a light flash and moves to the right in space (Figure 5-4)
with maximum speed: one meter of distance per meter of time. W ith horizontal and
vertical axes calibrated in meters, the light-flash worldline rises at an angle of 45
degrees (Figure 5-5).
Particle 4 does not move at all in laboratory space; it rests quietly next to the
laboratory reference clock. Like you sitting in your chair, it moves only along the time
dimension; in the laboratory spacetime map its worldline is vertical (Figure 5-5).
Particle 5 moves not to the right but to the left in space according to the laboratory
observer (Figure 5-4), so its worldline angles up and leftward in the laboratory
spacetime map (Figure 5-5).
Each of these particles moves with constant speed, so each traces out a straight
worldline. After 3 meters of time as measured in the laboratory frame, different
5.4 WORLDLINE 145
-►3 FIGURE 5-4. Trajectories in space

-*►2 {not in spacetime!) o f p articles 1
5 -< - -^ 1
through 3 durin g 3 meters o f time.
-P - ---------<>- ■<> Each particle starts at the reference clock
4
(the square) at zero of time and moves
with a constant velocity.
FIGURE 5-5. W orldlines in space

time o f the particles shown in Fig
ure 3-4, plotted fo r the laboratory
fram e. Only the worldline for particle 1
includes a sample set of event points that
are connected to make up the worldline.
particles have moved different distances from the starting point (Figure 5-4). In the
laboratory spacetime map their space positions after 3 meters of time lie along the
upper horizontal line o f sim ultaneity, shown dashed in Figure 5-5.
Particle 4 is not the only object stationary in space. Every laboratory clock lies at rest
in the laboratory frame; it moves neither right nor left as time passes. Nevertheless
each laboratory clock moves forward in time, tracing out its own vertical worldline in
the laboratory spacetime map. The background vertical lines in Figure 5-5 are
worldlines of rhe row of laboratory clocks.
W hat is the difference between a “p a th in space" a n d a “worldline in spacetime”?
The transcontinental airplane leaves a jet trail in still air. That trail is the plane’s path
Path in space versus
in space. Take a picture of that trail and you have a space map of the motion. From worldline in spacetime
that space map alone you cannot tell how fast the jet is moving at this or that
different point on its path. The space map is an incomplete record of the motion.
The plane moves not only in space but also in time. Its beacon flashes. Plot those
emissions as events on a spacetime map. This spacetime map has not only a
horizontal space axis but also a vertical time axis. Now connect those event points
with a worldline. The worldline gives a complete description of the motion of the jet as
recorded in that frame. For example, from the worldline we can reckon the speed of
the plane at every event along its path.
Worldline gives spacetime map of the journey of the jet. Likewise a worldline
drawn on a spacetime map images the journey of any particle through spacetime. A
worldline is not a physical path, not a trajectory, not a line in space. An object at rest
in your frame has, for you, no path at all through space; it stays always at one space
point. Yet this stationary particle traces out a "vertical” worldline in your spacetime
map (such as line 4 in Figure 5-5). A particle always has a worldline in spacetime. As
you sit quietly in your chair reading this book, you glide through spacetime on your
own unique worldline. Every stationary object lying near you also traces out a
worldline, parallel to your own on your spacetime map.
Not all particles move with constant speed. When a patticle changes speed with
respect to a free-float frame, we know why: A force acts on it. Think of a train moving
on a straight sttetch of track. A force applied by the locomotive speeds up all the cars.
Small speed: small distance covered in a given time lapse; worldline inclined slightly
to the vertical in the spacetime map. Great speed: great distance covered in the same
stretch of time; wotldline inclined at a gteater angle to the vertical in the spacetime
Changing speed means map. Changing speed: changing distances covered in equal time periods; worldline that
curving worldline
changes inclination as it ascends on the spacetime map— a curved worldline!
Wait a minute! The train moves along a straight track. Yet you say its worldline is
curved. Straight or curved? Make up your mind!
Straight in space does not necessarily mean straight in spacetime. Place your finger on
the straight edge of a table near you. Now move your finger rapidly back and forth
along this edge. Clearly this motion lies along a straight line. As your fingertip
changes speed and direction, however, it travels different spans of distance in equal
time periods. During a spell in which it is at rest on the table edge, your fingertip
traces out a vertical portion of its worldline on the spacetime map. When it moves
slowly to the right on the table, it traces out a worldline inclined slightly to the right
of vertical on the map. When it moves rapidly to the left, your fingertip leaves a
spacetime trail inclined significantly to the left on the map. Changing inclination of
the worldline from point to point results in a curved worldline. Your finger moves
straight in space but follows a curved worldline in spacetime!
Figure 5-6 shows a curved worldline, not for a locomotive, but for a particle
constrained to travel down the straight track of a linear accelerator. The particle starts
at the reference clock at the time of the reference event (0 on the map). Initially the
particle moves slowly to the right along the track. As time passes — advancing upward
on the spacetime map — the particle speed increases to a large fraction of the speed of
light. Then the particle slows down again, comes to rest at event Z, with a vertical
tangent to its worldline at that event. Thereafter the particle accelerates to the left in
space until it arrives at event P.
W hat possible worldlines are available to the particle that has arrived at event P? A
Limit on worldline slope:
material particle must move at less than the speed of light. In other words, it travels less
speed of light
than one meter of distance in one meter of time. Its future worldline makes an “angle
with the vertical’’ somewhere between plus 45 degrees and minus 45 degrees when
space and time are measured in the same units and plotted to the same scale along
horizontal and vettical axes on the graph. These limits of slope— which apply to every
point on a particle worldline— are shown as dashed lines emerging from event P in
Figure 5-6 (and also from event 0).
limits on worldline slope FIGURE 5-6. C urved laboratory

w orldline o f a p article th a t
changes speed as it moves back
a n d fo rth along a straight line in
space. Some possible worldlines avail
able to the particle after event P.
5.5 LENGTH ALONG A PATH 147
The worldline gives a complete description of particle motion in spacetime. As

drawn in the spacetime map for any frame, the worldline tells position and velocity of
the particle at every event along its trail. In contrast, the trajectory or orbit or path
shape of a particle in space does not give a complete description of the motion. To
complete the description we need to know when the particle occupies each location on
that trajectory. A worldline in a spacetime map automatically displays all of this
information.
The spacetime map provides a tool for retrospective study of events that have Spacetime map displays only
already taken place and have been reported to the free-float observer who plots them. already detected events
Once she plots these event points, this analyst can trace already plotted worldlines
backward in time. She can examine at a single glance event points that may have
occurred light-years apart in space. These features of the spacetime map do not violate
our experience that time moves only forward or that nothing moves faster than light.
Everything plotted on a spacetime map is history; it can be scanned rapidly back and
forth in the space dimension or the time dimension or both. The spacetime map
supplies a comprehensive tool for recognizing patterns of events and teasing out laws
of nature, but it is useless for influencing the events it represents.
5.5 LENGTH ALONG A PATH

straight line has shortest length between two
given points in spare
Distance is a central idea in all applications of Euclidean geometry. For instance, using
Measure length of curved path
a flexible tape measure it is easy to quantify the total distance along a winding path
with tape measure . . .
that starts at one point (point 0 in Figure 5-7) and ends at another point (point B).
Another way to measure distance along the curved path is to lay a series of short
straight sticks end to end along the path. Provided the straight sticks are short enough
to conform to the gently curving path, total distance along the path equals the sum of
lengths of the sticks.
The length of a short stick laid between any two nearby points on the path — for
. . . or with short straight sticks
instance, points 3 and 4 in Figure 5-7 — can also be calculared using the northward
laid end to end along path
separation and the eastward separation between the two ends of the stick as measured
by a surveyor.
(length)^ = (northward separation)^ -b (eastward separation)^
Distance is invariant for surveyors. Therefore the length of this stick is the same when
calculated by any surveyor, even though the northward and eastward separations
between two ends of the stick have different values, respectively, for different survey
ors. The length of another stick laid elsewhere along the path is also agreed on by all
surveyors despite their use of different northward direcrions. Therefore the sum of the
lengths of all short sticks laid along the path has the same value for all surveyors. This
All surveyors agree on
sum equals the value of the total length of the path, on which all surveyors agree. And
length of path
this total length is just the length measured using the flexible tape.
It is possible to proceed from 0 to B along quite another path — for example along
straight line OB in Figure 5-7. The length of this alternative path is evidently different
from that of the original curved path. This feature of Euclidean geometry is so well
known as to occasion hardly any comment and certainly no surprise: In Euclidean
geometry a curved path between two specified points is longer than a straight path
between them. The existence of this difference of length between two paths violates no
law. No one would claim that a tape measure fails to perform properly when laid along
a curved path.
1211/2
FIGURE 5-7. Length along a w inding p a th startin g a t the town square. Notice that the total length
along the winding path from point O to point B is greater than the length along the straight northward axis
from O /o B.
Among all possible paths between two points in space, the straight-line path is
Straight path in space has unique. All surveyors agree that this path has the shortest length. When we speak of
shortest length
“the distance between two points,” we ordinarily mean the length of this straight
path.
5.6 WRISTWATCH TIME ALONG A

WORLDLINE
straight worldline has longest proper time
between two given events in spaeetime
A curved path in Euclidean space is determined by laying down a flexible tape measure
and recording distance along the path’s length. A curved worldline in Lorentz space-
Measure proper time along time is measured by carrying a wristwatch along the worldline and recording what it
curved worldline with
wristwatch . . . shows for the elapsed time. The summed spacetime interval — the proper time read
directly on the wristwatch — measures the worldline in Lorentz geometry in the same
way that distance measures path length in Euclidean geometry.
A particle moves along the worldline in Figure 5-8. This particle carries a wrist
watch and a sparkplug; the sparkplug fires every meter of time (1, 2 , 3 , 4 , . . . ) as
read off the particle’s wristwatch. The laboratory observer notes which of his clocks the
5.6 WRISTWATCH TIME ALONG A WORLDLINE 14 9
WORLDLINE IN SPACETIME
FIGURE 5-8. Proper time along a curved worldline. Notice that the total proper time along the curved
worldline from event O to event B is smaller than the proper time along the straight line from O to B.
traveling particle is near every time the sparkplug fires. He plots that location and that
lattice clock time on his spacetime map, tracing out the worldline of the particle. He
numbers spark points sequentially on the resulting worldline, as shown in Figure 5-8,
knowing that these numbers register meters of time recorded on the moving wrist-
watch.
Consider the spacetime interval between two sequential numbered flashes of the
sparkplug, for instance those marked 3 and 4 in the figure. In the laboratory frame . . . or as sum of intervals
between adjacent events
these two sparks are separated by a difference in position and also by a difference in
time (the time between them). The squared interval— the proper time squared —
between the sparks is given by the familiar spacetime relation:
(proper time)^ = (difference in time)^ — (difference in position)^
W hat about the proper time between sparks 3 and 4 calculated from measurements
made in the sparkplug frame? In this frame, both sparks occur at the same place,
namely at the position of the sparkplug. The difference in position between the sparks
equals zero in this frame. As a result, the time difference in the sparkplug frame — the
“wristwatch time” — is equal to the proper time between these two events:
(proper time)^ “ (1 meter)^ (zero)^ ~ (1 meter)^ [recorded on traveling wristwatch]
This analysis assumes that sparks are close together in both space and time. For
sparks close enough together, the velocity of the emitting particle does not change
much from one spark to the next; the particle velocity is effectively constant between
sparks; the piece of curved wbrldline can be replaced with a short straight segment.
Along this straight segment the particle acts like a free-float rocket. The proper time is
invariant in free-float rocket and free-float laboratory frames. Thus the laboratory
observer can compute the value of the proper time between events 3 and 4 and predict
the time lapse — one meter— on the traveling wristwatch, which measures the proper
time directly.
Elsewhere along the worldline the particle moves with a different speed. Neverthe
less the proper time between each consecutive pair of sparks must also be independent
of the free-float frame in which that interval is reckoned. For sparks close enough
rogether, this proper time equals the time read directly on the wristwatch.
All observers agree on the proper time between every sequential pair of sparks
All observers a g ree on proper
emitted by the sparkplug. Therefore the sum of of all individual proper times has the
time along worldline
same value for all observers. This sum equals the value of the total proper time, on
which all free-float observers agree. And this total proper time is just the wristwatch
time measured by the traveling sparkplug.
In brief, proper time is the time registered in a rocket by its own clock, or by a person
through her own wristwatch or her own aging. Like aging, proper time is cumulative.
To obtain total proper time racked up along a worldline between some marked
starting event and a designated final event, we first divide up the worldline into
segments so short that each is essentially “straight” or “free-float.” For each segment
we determine the interval, that is, the lapse of proper time, the measurement of aging
experienced on that segment. Then we add up the aging, the proper time for each
segment, to get total aging, total wristwatch time, total lapse of proper time.
An automobile may travel the most complicated route over an entire continent, but
the odometer adds it all up and gives a well-understood number. The traveler through
the greater world of spacetime, no matter how many changes of speed or direction she
undergoes, has the equivalent of the odometer with her on her journey. It is her
wristwatch and her body— her aging. Your own wristwatch and your biological clock
automatically add up the bits of proper time traced out on all successive segments of
your worldline.
It is possible to proceed from event 0 to event B along quite another worldline— for
example, along the straight worldline OB in Figures 5-8 and 5-9 (bottom). The proper
Straight worldline has
time from Ot oB along this new worldline can be measured directly by a flashing clock
longest proper time
that follows this new worldline. It can also be calculated from records of flashes
emitted by the clock as recorded in any laboratory or rocket frame.
Total proper time along this alternative worldline has a different value than total
proper time along the original worldline. In Lorentz geometry a curved worldline
berween two specified events is shorter than the direct worldline between them —
shorter in terms of total proper time, total wristwatch time, total aging.
T otal proper tim e, the aging along any given worldline, straight or curved, is an
invariant: it has the same value as reckoned by observers in all overlapping free-float
frames. This value correctly predicts elapsed time recorded directly on the wristwatch
of the particle that travels this worldline. It correctly predicts the aging of a person or a
mouse that travels this worldline. A different worldline between the same two events
typically leads to a different value of aging — a new value also agreed on by all
free-float observers: Aging is maximal along the straight worldline between two
events. This uniqueness of the straight worldline is also a matter of complete agree
ment among all free-float observers. All agree also on this: The straight worldline is the
one actually followed by a free particle. Conclusion: Between two fixed events, a free
particle follows the worldline of maximal aging. This more general prediction of the
Principle of M aximal Aging
worldline of a free particle is called the P rin cip le o f M axim al Aging. It is true not
predicts motion of free particle
only for “straight” particle worldlines in the limited regions of spacetime described by
special relativiry but also, with minor modification, for the motion of free particles in
wider spacetime regions in the vicinity of gravitating mass. The Principle of Maximal
Aging provides one bridge between special relativity and general relativity.
The stark contrast between Euclidean geometry and Lorentz geometry is shown in
Figure 5-9. In Euclidean geometry distance between nearby points along a curved
5.6 WRISTWATCH TIME ALONG A WORLDLINE 151
FIGURE 5-9. P ath in space: In Euclidean geometry the curved path has greater length. W orldline in
spacetime: In Lorentz geometry the curved worldline is traversed in shorter proper time.
path is always equal to or greater than the northward separation between those two
Stark contrast between Euclidean
points. In contrast, proper time between nearby events along a curved worldline is
and Lorentz geom etries
always equal to or less than the corresponding time along the direct worldline as
measured in that frame.
The difference of proper time between two alternative worldlines in spacetime

violates no law, just as the difference of length between two alternative paths in space
violates no law. There is nothing wrong with a wristwatch that reads different proper
times when carried along different worldlines between events 0 and B in spacetime,
just as there is nothing wrong with a tape measure that records different lengths for
different paths between points 0 and B in space. In both cases the measuring device is
simply giving evidence of the appropriate geometry: Euclidean geometry for space,
Lorentz geometry for spacetime.
In brief, the determination of cumulative interval, proper time, wristwatch time,
Proper times com pare worldlines aging along a worldline between two events is a fundamental method of comparing
different worldlines that connect the same two events.
Among all possible worldlines between two events, the straight worldline is unique.
All observers agree that this worldline is straight and has the longest proper time —
greatest aging — of any possible worldline connecting these events.
5.7 KINKED WORLDLINE

kink in the worldline decreases aging along
that worldline
The change in slope of the worldline from event to event in Figures 5-8 and 5-9
(bottom) means that the clock being carried along this worldline changes velocity: It
accelerates. Different clocks behave differently when accelerated. Typically a clock can
withstand a great acceleration only when it is small and compact. A pendulum clock is
not an accurate timepiece when carried by car through stop-and-go traffic; a wrist
watch is fine. A wristwatch is destroyed by being slammed against a wall; a radioactive
nucleus is fine. Typically, the smaller the clock, the more acceleration it can withstand
Acceleration-proof clocks and still register properly, and the sharper can be the curves and kinks on its worldline.
In all figures like Figures 5-8 and 5-9 (bottom), we assume the ideal limit of small
(acceleration-prooO clocks.
W e are now free to analyze a motion in which particle and clock are subject to a
great acceleration. In particular, consider the simple special case of the worldline of
Figure 5-8. That worldline gradually changes slope as the particle speeds up and slows
down. Now make the period of speeding up shorter and shorter (great driving force!);
Simplify: W orldlines with also make the period of slowing down shorter and shorter. In this way come eventually
straight segments
to the limiting case in which episodes of acceletation and deceleration — curved
portions of the worldline — are too short even to show up on the scale of the spacetime
map (worldline OQB in Figure 5-10). In this simple limiting case the whole history of
motion is specified by (1) initial event 0, (2) final event B, and (3) turnaround event Q,
halfway in time between 0 and B. In this case it is particularly easy to see how the lapse
of proper time between 0 and B depends on the location of the halfway event— and
thus to compare three worldlines, OPB, OQB, and ORB.
Path OPB is the worldline of a particle that does not move in space; it stays next to
the reference-frame clock. Proper time from 0 to B by way of P is evidently equal to
time as measured in the free-float frame of this reference clock:
(total proper time along OPB) = 1 0 meters of time
In contrast, on the way from 0 to B via R, for each segment the space separation equals
the time separation, so the proper time has the value zero:
5.7 KINKED WORLDLINE 153
FIGURE 5-10. Three alternative

worldlines connecting events O
a n d B. The sharp changes of velocity
at events Q and R have been drawn for
the ideal limit of small clocks that tol
erate great acceleration. The bold-face
number j is the proper time along the
segment O Q , reckonedfrom the differ
ence between the squared time separa
tion and the squared space separation:
= 5" - 4^.
(proper time along leg ORY — (time)^ — (space)^

= (5 meters)^ “ (5 meters)^
Z ero p ro p er time for light
= 0
(total proper time along ORB) = 2 X (proper time along OR)
= 0
As far as we know, only three things can travel 5 meters of distance in 5 meters of
time; light (photons), neutrinos, and gravitons (see Box 8-1). No material clock can
travel at light speed. Therefore the worldline ORB is not actually attainable by a
material particle. However, it can be approached arbitrarily closely. One can find a
speed sufficiently close to light speed — and yet less than light speed — so that a trip
with this speed first one way then the other will bring an ideal clock back to the
reference clock with a lapse of proper time that is as short as one pleases. In the same
way we can, in principle, go to the star Canopus and back in as short a round-trip
rocket time as we choose (Section 4.8).
As distinguished from the limiting case ORB, worldline OQB demands an amount
of proper time that is greater than zero but still less than the 10 meters of proper time
along the direct worldline OPB:
(proper time along leg OQ)' = (5 meters)^ — (4 meters)^

= 25 (meters)^ — 16 (meters)^ Reduced proper time along
kinked worldline
= 9 (meters)^
= (3 meters)^
(proper time along leg OQ) = 3 meters

and
(total proper time along both legs OQB) — 2 X (proper time along OQ)
= 6 meters
This is less proper time than (proper time along OPB) = 1 0 meters thar characterized
the “direct” worldline OPB. Our trip to Canopus and back described in Chapter 4
follows a worldline similar to OQB.
SAMPLE PROBLEM 5-1

MORE IS LESS
In the spacetime map shown, time and space are measured in years. A table shows space
and time locations of numbered events in this frame.
SPACE AND TIME

LOCATIONS OF EVENTS
Space Time
(years) (years)
Event 1
Event 2 1
Event 3 - 0 .5
Event 4 2
----- space — ►
T wo alternative worldlines between events I and 4
One traveler moves along the solid straight worldline segments from event 1 to
events 2 ,3 , and 4. Calculate the time increase on her clock between event 1 and
event 2; between event 2 and event 3; between event 3 and event 4. Calculate
total proper time — her aging — along worldline 1, 2, 3, 4.
b. Another traveler, her twin brother, moves along the straight dotted worldline
from event 1 directly to event 4. Calculate the time increase on his clock along the
direct worldline 1, 4.
c. Which twin (solid-line traveler or dotted-line traveler) is younger when they
rejoin at event 4?
SOLUTION
From the table next to the map, space separation between events 1 and 2 equals
0. Time separation equals 1 year. Therefore the interval is reckoned from
(interval)^ = P — 0 ^= 1 ^. Thus the proper time lapse on a clock carried between
events 1 and 2 equals 1 year.
Space separation between event 2 and event 3 equals 1 — (—0.5) = 1.5
light-years. Time separation equals 2 years. Therefore the square of the interval is
2^ — (1.5)^ = 4 — 2.25 = 1.75 (years)^ and the advance of proper time equals
the square root of this, ot 1.32 years.
Between event 3 and event 4 space separation equals 2.5 light-years and time
separation 3 years. The square of the interval has the value 3^ — (2.5)^ = 9 —
6.25 = 2.75 (years)^ and proper time between these two events equals the square
root of this, or 1.66 years.
5.8 STRETCH FACTOR 1 55
Total proper time— aging — along worldline 1, 2, 3, 4 equals the sum of

proper times along individual segments: 1 + 1.32 + 1.66 = 3.98 years.
b. Space separation between events 1 and 4 equals 1 light-year. Time separation is 6
years. The squared interval between them equals 6^ — 1^ = 36 — 1 = 35
(years)^. A traveler who moves along the direct worldline from event 1 to event 4
records a span of proper time equal to the square root of this value, or 5.92 years.
c. The brother who moves along straight worldline 1, 4 ages 5.92 years during the
trip. The sister who moves along segmented worldline 1, 2, 3, 4 ages less: 3.98
years. As always in Lorentz geometry, the ditect worldline (shown dotted) is
longer— that is, it has more elapsed proper time, greater aging — than the
indirect worldline (shown solid).
5.8 STRETCH FACTOR

ratio off frame-clock time to wristwatch time
A speeding beacon emits two flashes, F and S, in quick succession. These two flashes,
as recorded in the rocket that carties the beacon, occur with a 6-meter separation in
time but a zero separation in space. Zero space separation? Then 6 metets is the value
Different reference frames:
of the interval, the proper time, the wristwatch time between F and S. As registered in different times between two events
the laboratory, in contrast, the second flash S occurs 10 meters of time later than the
first flash F. The ratio between this frame time, 10 meters, and the proper time, 6
meters, between the two events we call the time stretch factor, or simply stretch
factor. Some authors use the lowercase Greek letter gamma, y, for the stretch factor,
as we do occasionally. We will also use the Greek letter tau, T , for proper time.
The same two events register in the super-rocket frame that overtakes and passes the
beacon — register with a separation in time of 20.88 meters. In this frame, the time
stretch factor between the two events is (20.88)/6 = 3.48. In the beacon frame the
Time lapse minimum for frame
stretch factor is unity: 6 /6 = 1. Why? Because in this beacon frame flashes F and S in which events occur at sam e
occur at the same place, so beacon-frame clocks record the proper time directly. This place
proper time is less than the time between the two flashes as measuted in either
laboratory or super-rocket frame. The larger value of time observed in laboratory and
super-rocket frames shows up in Figure 5-11 (center and right). Among all conceiv
able frames, the separation in time between the two flashes evidently takes on its
minimum value in the beacon frame itself, the value of the proper time T.
Hold it! In Sections 3.6 an d 5 ■ 7 you insisted th at the time along a straight worldline is
a MAXIM UM . Now you show us a straight worldline along which the time is — you
say — a M INIM UM . Maximum or minimum? Please make up your mind!
— The worldline taken by the beacon wristwatch from F to S is straight. It is straight

^ whether mapped in the beacon frame itself or in the rocket or super-rocket frame.
The beacon racks up 6 meters of proper time regardless of the frame in which we
reckon this time. When we turn from this wristwatch time to what different
rocket time
laboratory time
LABORATORY ROCKET SUPER-ROCKET

f, = 10 meters f', = 6 meters t", = 20.88 meters
FIGURE 5-11. Spacetime maps o f Figure 5-3, modified to show the worldline o f the speeding
beacon (heavy dashed line) and the segment o f this line between emission F o f thefirst flash and
the secondflash S (solid section o f worldline). Emission F is taken as the zero ofspace and time. Time tj
of the second emission S is different as recorded in differentframes. The shortest time is recorded in thatframe
in which the two events occur at the same place— in this case the rocket frame.
free-float frames show for the separation in map time (latticework time, frame time)
between the two flashes, however, the record displays a minimal value for that
separation in time only in the beacon frame itself
In contrast. Figure 5-12 (Figure 5-10 in simplified form) shows two different
worldlines that join events 0 and B mapped in the same reference frame. In this case
we compare two different proper times: a proper rime of 10 meters racked up by a
wristwatch carried along the direct course from 0 to B, and a proper time of 6 meters
recorded by the wristwatch carried along on the kinked worldline OQB. In every such
comparison made in the context of flat spacetime, the direct worldline displays
maximum proper time. Caution: Conditions can be different in curved spacetime
(Chapter 9).
In summary, two points come to the fore in these comparisons of the time between
two events. (1) Are we comparing map time (frame time, latticework time) between
those two events, pure and simple, free of any talk about any wotldline that might
connect those events? Then separation in time between those events is least as
mapped in the free-float frame that shows them happening at the same place. (2) Or
are we directing our arrention to a worldline that connects the two events? More
specifically, to the time racked up by a wrisrwarch toted along that worldline? Then
we have to ask, is that worldline straight? Then it registers maximal passage of
proper time. Or does it have a kink? Then the proper time racked up is not maximal.
When we find ourselves in a free-float frame and see a beacon zooming past in a
straight line with speed v, how much is the factor by which our frame-clock time is
stretched relative to the beacon wristwatch time? Answer: The stretch factor is
1
(stretch factor) = y ■ (5-1)
(1 -r.2)i/2
How can we derive this famous formula? If you do not cover up the following lines
FIGURE 5-12. Figure 5-10 stripped down and derive this answer on your own, here is the reasoning: Start with measurements in
to emphasize total proper time (wrist- the laboratory frame. We know that for this rocket
watch time), pn-inted boldface along two
different worldlines between the same two (advance in proper time)^ = (advance in lab time)^ ~ (lab distance covered)^
events O and B in a given referenceframe.
Among all possible worldlines connecting events
O and B, the straight worldline registers maxi However, we want to compare lapses in laboratory time and proper time; laboratory
mal lapse of proper time. distance covered is not of interest. For the laboratory observer the proper clock moving
5.8 STRETCH FACTOR 157
along a straight worldline covers the distance between the two events in the time
between the events. Therefore this distance and time are related by particle speed:
Stretch factor
(lab distance covered) = (speed) X (advance in lab time)
= frame time/proper time
Substitute this expression into the equation for proper time:
(proper time)^ = (lab time)^ ~ (speed)^ X (lab time)^

= (lab time)^ [1 ~ (speed)^}
This leads to an expression for the square of the stretch factor:
(lab time)^ . . _ 1 1
(stretch factor)^
(proper time)^ 1 — (speed)^ 1
where we use the symbol v — v ^ ^ /c for speed. The equation for the stretch factor
becomes
1 Stretch factor derived

(stretch factor) — y — (5 - 1 )
(1 - «^2) 1/2
The stretch factor has the value unity when p = 0. For all other values of v the stretch
factor is greater than unity. For very high relative speeds, speeds close to th a t o f light
(v —* 1), the value of the stretch factor increases without limit.
The value of the stretch factor does not depend on the direction of motion of the
rocket that moves from first event to second event: The speed is squared in equation
(5-1), so any negative sign is lost.
The stretch factor is the ratio of frame time to proper time between events, where
speed ( = p) is the steady speed necessary for the proper clock to pass along a straight
worldline from one event to the other in that frame.
The stretch factor also describes the Lorentz contraction, the measured shortening of a Lorentz-contraction by
moving object along its direction of motion when the observer determines the distance sam e “ stretch" factor
between the two ends a t the same time. For example, suppose you travel at speed v
between Earth and a star that lies distance L away as measured in the Earth frame. Your
trip takes time t = Ljv in the Earth-linked frame. Proper time T— your wristwatch
tim e— is smaller than this by the stretch factor: T = L /\v X (stretch faaor)} = (L/v) (1
— iqow think of a very long rod that reaches from Earth to star and is at rest in the
Earth frame. How long is that rod in your rocket frame? In your frame the rod is moving at
speed V. One end of the rod, at the position of Earth, passes at speed v. A time T later in
your frame the other end of the rod arrives — along with the star — also moving at speed
V according to your rocket measurements. From these data you calculate that the length of
the rod in your rocket frame— call it L' — is equal to L' = pT = viLjp) (1 — r^)’''^ = L
(1 — r>^)Y2 'This is a valid measure of length. By this method the rod is measured to be
shorter.
Finally, the stretch factor is often used as an alternative measure of particle speed: A Stretch factor as a measure
particle moves with a speed such that the stretch factor is 10. This statement assumes of speed
that the particle is moving with constant speed, so that the separation between any pair
of events on the particle worldline has the same stretch factor as the separation between
any other pair. This way of describing particle speed can be both convenient and
powerful. We will see (Chapter 7) that the total energy of a particle is proportional to
the stretch factor.
SAMPLE PROBLEM 5-2

R O U N D TRIP O B S E R V E D IN
DI F F E RE NT FRAME
Return to the alternative worldlines between time — for worldline OQB as that predicted in the
events 0 and B, shown in Figure 5-10 and the laboratory frame. Similarly show that this
spacetime maps in this sample problem. Measure outward-rocket-frame observer predicts the same
these worldlines from a rocket frame that moves proper time along the direct worldline OPB as does
outward with the particle from 0 to Q and keeps on the laboratory observer. Finally, show that both
going forever at the same constant velocity. Show observers predict the elapsed wristwatch time
that an observer in this outward-rocket frame along OQB to be less than along OPB.
predicts the same proper time— wristwatch
SOLUTION
Here are laboratory and rocket spacetime maps for these round trips, simplified and
drawn to reduced scale.
t
time
10 p \ q
A
c5 ------space — >■ space ■
lABORATORY OUTWARD-ROCKET
SPACETIME MAP SPACETIME MAP
Laboratory a n d outward-rocket spacetime maps, each showing altern ativ e worldlines (direct
OPB a n d indirect OQB) between events O a n d B. Laboratory spacetime map: Figure 3-W ,
redrawn to a different scale. Proper times are shown on the laboratory spacetime map. Outward-rocket
spacetime map: The rocket in which the outgoing particle is at rest. Portions of two invariant hyperbolas
show how events Q and B transform. The direct worldline OPB has longer total proper time— greater
aging— as computed using measurements from either frame.
F ind x ' q a n d t ' q. First compute space and time locations of events Q and B in the
outgoing rocket frame — right-hand map. (Event 0 is the reference event, x = 0 and t =
0 in all frames by convention.) We choose the rocket frame so that the worldline segment
OQ lies vertical and the outbound rocket does not move in this frame. As a result, event Q
occurs at rocket space origin: x ' q — 0. (Primes refer to measurements in the outward-
rocket frame.) The rocket time Cq for this event is just the wristwatch time between 0 and
Q, because the wristwatch is at rest in this frame: Cq = 3 meters.
In summary, using a prime for rocket measurements:
X q= 0
3 meters
5.8 STRETCH FACTOR 159
F in d x ' b a n d In the laboratory frame, the particle moves to the right from event 0 to
event Q, covering 4 meters of distance in 5 meters of time. Therefore its speed is the
fraction t' = 4 /5 = 0.8 of light speed. As measured in the rocket frame, the laboratory
frame moves to the left with speed = 0.8 , by symmetry. Use equation (5-1) with v =
0.8 to compute the value of the stretch factor;
1 1 1 1 1 10
[1 - (0.8)2}*/2 [1 - 0 .6 4 ] V 2 [0.36} i/2 0.6 T
This equals the ratio of rocket time period t'g to proper time along the direct path OPB.
Hence elapsed rocket time t'g = (5 /3 ) X 10 meters = 50/3 meters of time. In this time,
the laboratory moves to the left in the rocket frame by rhe distance x'g = — vt' g =
■“ (4 /5 ) (5 0 /3 ) = ~ 200/15 = ~ 4 0 / 3 meters. In summary for outgoing rocket:
40
—— meters ■ 13~ meters
3
50
I-------meters — 16— meters of time
3 3
Events Q and B ate plotted on the rocket spacetime map.

C o m pare W ristw atch Tim es: Now compute the total proper time — wristwatch
time, aging — along alternative wotldlines OPB and OQB using rocket measurements.
Direct wotldline OB has proper time Tgg given by the regular expression for interval:
2500 1600 900

= ----- = 1 0 0 (meters)^
whence Toe = 1 0 meters computed from rocket measutements. This is the same value as
computed in the laboratory frame (in which proper time equals laboratory time, since
laboratory separation in space is zero).
Worldline OQB has two segments. On the first segment, OQ, proper time lapse is just
equal to the rocket time span, 3 meters, since the space separation equals zero in the rocket
ftame. For the second segment of this wotldline, QB, we need to compute elapsed time in
this ftame:
50 50 9 41
^ QB ^ , = -------3 --- ----------- --- — meters
3 3 3 3
40
—— meters
Therefore,
I^qb)^ 0 qb)^ {x
1681 1600 81
■— = 9 (meters)^
9
whence Tqb = 3 meters. So the total increase in proper time— the total aging— along
worldline OQB sums to 3 + 3 = 6 meters as reckoned from outward-rocket measure
ments. This is the same as figured from laboratory measurements.
How can these weird results be true? In our everyday lives why don’t we have to take
account of clocks that record different elapsed times between events, and rods that we
measure to be contracted as they speed by us?
In answer, consider two events that occur at the same place in our frame. The proper
clock moving in spacetime between these two events has speed zero for us. In this
case the stretch factor has the value unity; the frame clock is the proper clock. The
same is approximately true for events that are much closer together in space (mea
sured in meters) than the time between them (also measured in meters). In these
cases the proper clock moving between them has speed v — measured in meters/
meter— that is very much less than unity. That is, the proper clock moves very
much slower than the speed of light. For such slow speeds, the stretch factor has a
value that approaches unity; the proper clock records very nearly the same time lapse
between two events as frame clocks. This is the situation for all motions on earth that
we can follow by eye. For all such “ordinary-speed” motions, moving clocks and
stationary clocks record essentially the same time lapses. This is the assumption of
Newtonian mechanics: “Absolute, true, and mathematical time, of itself, and from
its own nature, flows equably without relation to anything external . . .”
A similar argument leads to the conclusion that Lorentz contraction is negligible
for objects moving at everyday speeds. Newton’s mechanics— with its unique
measured time between events and its unique measured length for an object whether
or not it moves— gives correct results for objects moving at everyday speeds. In
contrast, for particle speeds approaching light speed (approaching one meter of
distance traveled per meter of elapsed time in the laboratory frame), the denomina
tor on the right of equation (5-1) approaches zero and the stretch factor increases
without limit. Increased without limit, also, is the laboratory time between ticks of
the zooming particle’s wristwatch. This is the case for high-speed particles in
accelerators and for cosmic rays, very high-energy particles (mostly protons) that
continually pour into our atmosphere from space. Newton’s mechanics gives results
wildly in error when applied to these particles and theit interactions; the laws of
relativistic mechanics must be used.
More than one cosmic ray has been detected (indirectly by the resulting shower of
particles in the atmosphere) moving so fast that it could cross our galaxy in 30 seconds
as recorded on its own wristwatch. During this trip a thousand centuries pass as
recorded by clocks on Earth! (See Exercise 7-7.)
5.9 TOURING SPACETIME WITHOUT A

REFERENCE FRAME
all you need is worldlines and events
An explosion is an explosion. Your birth was your birth. An event is an event. Every
event has a concreteness, an existence, a reality independent of any reference frame. So,
Events and worldlines exist too, does a worldline that connects the trail of event points left by a high-speed
independent of sparkplug that flashes as it streaks along. Events mark worldlines, independent of any
any reference frame reference frame.
Worldlines also locate events. The intersection of two worldlines locates an event as
clearly and sharply as the intersection of two straws specifies the place of a dust speck in
agreat barn full of hay (Eigure 5-13). To say that an event marks a collision between
two particles is identification enough. The worldlines of those two particles are rooted
in the past and stretch out into the future. They have a rich texture of connections with
nearby worldlines. The nearby worldlines in turn are linked in a hundred ways with
5.9 TOURING SPACETIME WITHOUT A REFERENCE FRAME 161
FIGURE 5-13. The crossing of straws in a bam full of bay is a symbol for the worldlines that fill up
spacetime. By their crossings and jogs, these worldlines mark events with a uniqueness beyond all need of
reference frames. Straight worldlines track particles with mass; wiggly worldlines trace photons. Typical
events symbolized in the map (black dots) from left to right: absorption of a photon; reemission of a photon;
collision between a particle and a particle; collision between a photon and another particle; another collision
between a photon and a particle; explosion of a firecracker; collision ofa particle from outside with one of the
fragments of that firecracker.
worldlines more remote. How then does one tell the location of an event? Tell first
what worldlines thread the event. Next follow each of these worldlines. Name
additional events that they encounter. These events pick out further worldlines.
Eventually the whole barn of hay is cataloged. Each event is named. One can find one’s
way as surely to a given intersection as the London dweller can pick her path to the
meeting of St. James’s Street and Piccadilly. No numbers giving space and time
location of an event in a given reference frame. No reference frame at all!
Most streets in Japan have no names and most houses no numbers. Yet mail is
Locate house at intersection
delivered just the same. Each house is named after its senior occupant, and everyone
of streets
knows how the streets interconnect these named houses. Now print the map of
Japanese streets on a rubber sheet and stretch the sheet this way and that. The postal
carrier is not fooled. Each house has its unique name and the same interconnections
with neighbor houses as on the unstretched map. So dispense with all maps! Replace
them with a catalog or directory that lists each house by name, notes streets passing the
house, and tabulates the distance to each neighboring house along the streets.
Similarly, the visual pattern of event dots on a spacetime map (spacetime diagram)
Locate event at intersection
and the apparent lengths of worldlines that connect them depend on the reference
of worldlines
frame from which they are observed (for example, compare alternative spacetime
maps of the same worldline shown in the figure in Sample Problem 5-2). However,
each named event is the same for every observer; the event of your birth is unique to
you and to everyone connected with you. Moreover, the segment of a worldline that
connects one event with the next has a unique magnitude— the interval or proper
time — also the same for every observer. Therefore dispense with reference frames
altogether! Replace them with a catalog or directory that lists each event by name,
notes each worldline that threads the event, and tabulates the interval that connects the
event with the next event along each worldline. With this directory in hand we can say
Events and worldlines alone can precisely how all events are interconnected with each other and which events caused
describe Nature
which other evenrs. That is the essence of science; in principle we need no reference
frames.
But reference frames are convenient. We are accustomed to them. Most of us prefer
to live on named streets with numbered houses. Similarly, most of us speak easily of
space separations between events and time separations between the same events as if
space and time separations were unconnected. In this way we enjoy the concreteness of
using our latticework of rods and clocks while suffering the provinciality of a single
reference frame. So be it! Nevertheless, with worldlines Nature gives us power to relate
events — to do science— without reference frames at all.
5.10 SUMMARY
straighter worldline? greater aging!
Events? Yes. Each event endowed with its own location in that great fabric we call
spacetime? Yes. But time? No point in all that fabric displays any trace of anything we
can identify with any such thing as the “time” of that event. Label that event with a
“time” anyway? Sure. No one can stop us. Moreover, such labeling often proves quite
useful. But it is our labeling! A different reference frame, a different wrisrwatch
brought to that event along a different worldline yields a different time label for that
event.
For our own convenience, then, we plot events on a spacetim e m ap (spacetim e
diagram ) for a particular free-float frame and its latticework of rods and clocks. This
map can be printed on the page of a book if events are limited to one line in space.
Distance along this line is plotted horizontally on the spacetime map, with time of the
event plotted vertically (Section 5.1). The time and space values of an event are
measured with respect to a common referen ce event, plotted at the origin of the
spacetime map. The invariance of the interval: (interval)^ = (time)^ — (distance)^
between an event and the reference event corresponds to the equation of a hy p erbola,
the same hyperbola as plotted on the spacetime map of every overlapping free-float
frame. The event point lies somewhere on the same in v arian t h y p erb o la as plotted
on every one of these spacetime maps (Sections 5.2 and 5.3).
Billions of events sparkle like sand grains scattered over the spacetime map. A given
event is unconnected to most other events on the map. Here we pay attention to
particular strings of events that are connected. The w o rld lin e of a particle connects in
sequence events that occur at the particle (Section 5.4). The “ length” of a worldline
between an initial and a final event is the elapsed time measured on a clock carried
along the worldline between the two events (Section 5.6). This is called the proper
time, wrisrwatch time, or aging along this worldline. The lapse of proper time is given
the symbol T, in contrast to the symbol t for the frame time read on the latticework
clocks in a given free-float frame.
Carry a wrisrwatch (or grow old!) along a worldline: This is one way to measure the
total proper time along it from some initial event (such as the birth of a person or a
particle) to some final event (such as death of a person or annihilation of a particle).
This method is direct, experimental, simple. A second method? Calculate the interval
between each pair of adjacent events that make up the worldline, and then add up all
EXERCISE 5-1 MORE IS LESS 163
these intervals, assuming that each tiny segment is short enough to be considered
straight. This method seems more bothersome and detailed, but it can be carried out
by the observer in any free-float frame. All such observers will agree wirh one
another— and with the clock-carrier — on the value of the total proper time from the
initial event to the final event on the worldline (Section 5.6).
Among all possible worldlines between two given events, the straight line is the
worldline of m axim al aging. This is the acmal worldline followed by a free particle
that travels from one of these two events to the other (Section 5.6).
As measured in a given free-float ftame, the stre tc h fa cto r = 1 / ( 1 — equals
the ratio of elapsed frame rime t to elapsed proper time T along a segment of worldline
in which the particle moves with speed v in that frame. The stretch factor is also the
Lorentz contraction factor (Section 5.8): Locate, at the same time, the front and back
ends of an object moving in a given free-float frame. These end locations will be (1 —
t/2)V2 as far apart in that frame as they are in a frame in which the object is at rest.
Worldlines connect events. Like events, they exist independent of any reference
frame. In principle, worldlines allow us to relate events to one another — to do
science— without using reference frames at all (Section 5.9).
REFERENCES
Newton quotation toward the end of Section 5.8: Sir Isaac Newton, Mathemati
cal Principles of Natural Philosophy and His System of the World (Philosophiae
Naturalis Principia Mathematica), Joseph Streater, London, July 5, 1686;
translated from Latin — the scholarly language of Newton’s time— by Andrew
Motte in 1729, revised and edited by Florian Cajori and published in two
paperback volumes (University of California Press, Berkeley, 1962).
Section 5.9 uses slightly modified passages from Charles W. Misner, Kip S.
Thorne, and John Archibald Wheeler, Gravitation (W .H. Freeman, New York,
1973), pages 5 - 8 . Figure 5-13 is taken directly from this reference, its caption
slightly altered from the original.
CHAPTER 5 EXERCISES
a One traveler moves along the solid segmented

PRACTICE worldline from evenr A to events B, C, and D. Calcu
late the time inctease on his wristwatch (proper clock)
5-1 more is less (1) between event A and event B.
The spacerime diagram shows two alternative world (2) between event B and event C
lines from event A to event D. The table shows co (3) between event C and event D.
ordinates of numbered events in this frame. Time (4) Also calculate the total proper time along
and space are measured in years. worldline A, B, C, D.
164 EXERCISE 5-2 TRANSFORMING WORLDLINES
worldline from event A directly to event D. Calculate Event 0 x =0 t= 0
the time increase on her wristwatch between events A Event 1 x = 3 .0 0 0 t= 4 .0 0 0

Event 2 x = 1 .7 5 0 f= 7 .0 0 0
and D.
Event 3 X = 5 .0 0 0 t= 1 1 .0 0 0
c Which twin (solid-line or dotted-line traveler)
is younger when they rejoin at event D? EXERCISE 5 -2 . Two worldlines as recorded in the laboratory
frame. Numbers on the segmented worldline are proper times along
each straight segment.
5-2 transforming worldlines
The laboratory spacetime diagram in the figure shows
two worldlines. One, the vertical line labeled B, is the
worldline of an object that is at rest in this frame. The
5-3 mapmaking in spacetime
other, the segmented line that connects events 0, 1,2, N ote: Recall Exercise 1-6, the corresponding map
and 3, is the worldline of an object that moves at making exercise in Chapter 1.
different speeds at different times in this frame. The Here is a table of timelike intervals between events,
proper time is written on each segment and invariant in meters. The events occur in the time sequence
hyperbolas are drawn through events 1,2, and 3. The ABCD in all frames and along a single line in space in
event table shows the space and time locations in this all frames. (They do not occur along a single line on
frame of the four events 0, 1, 2, and 3. the spacetime map.)
a Trace the axes and hyperbolas onto a blank
piece of paper. Sketch a qualitatively correct space- INTERVAL
time diagram for the same pair of worldlines observed to event A B c D
in a frame in which the particle on the segmented from
worldline has zero velocity between event 1 and event event
2. A 0 1.0 3.161 5.196
b W hat is the velocity, in this new frame, of the B 0 2.0 4.0
particle moving along worldline B? C 0 2.0
D 0
C On each straight portion of the segmented
worldline for this new frame write the numerical value a Use a ruler and the hyperbola graph to con
of the interval between the two connected events. struct a spacetime map of these events. Draw this map
EXERCISE 5-3 MAPMAKING IN SPACETIME 165
----- space —■>-
EXERCISE 5-3. Template of hyperbolas for converting intervals into a spacetime map.
on thin paper so you can lay it over the hyperbola spacetime locations arbitrarily. Then go on to plot
graph and see the hyperbolas. event D.
D iscussion: How to start? W ith three arbitrary Analogy to surveying: In surveying (using Eu
decisions! (1) Choose event A to be at the origin of the clidean geometry) you locate all points a given dis
spacetime map. (2) Choose event B to occur at the tance from some stake by using that stake as origin
same place as event A. That is, event point B is located and drawing a circle of radius equal to the desired
on the positive time axis with respect to event point A. distance. In a spacetime map (using Lorentz geome
After plotting B, use your ruler to draw this straight try) you locate all event points a given interval from
time axis through event points A and B. Keep this line some event by using that event point as origin and
parallel to the vertical lines on the hyperbola graph in drawing a hyperbola with nearest point equal to the
all later constructions. (3) Even with these choices, desired interval.
there are two spacetime locations (x, t) at which you b Now take a new piece of paper and draw a
can locate the event point C; choose either of these two spacetime map for another reference frame. Choose
166 EXERCISE 5-4 THE POLE AND BARN PARADOX
event D to be at the origin of the spacetime map. This Write a reply to the worried student explaining
means that all other events occur before D. Hence clearly and carefully how the pole and barn are treated
turn the hyperbola plot upside down, so that the by relativity without internal contradiaion. Use the
hyperbolas open downward. Choose event B to occur following outline or some other method.
at the same place as D. Now find the locations of A a Make two carefully labeled spacetime dia
and C using the same strategy as in part a. grams, one an xt diagram for the barn rest frame, the
e Find an approximate value for the relative other an x 't ' diagram for the runner rest frame. Re
speed of the two frames for which you have made ferring to the figure, take the event “Q coincides with
spacetime plots. A” to be at the origin of both diagrams. In both plot
d Hold one of your spacetime maps up to the the worldlines of A, B, P, and Q. Pay attention to the
light with the marks on the side of the paper facing scale of both diagrams. Label both diagtams with the
the light. Does the map you see from the back also time (in meters) of the event “ Q coincides with B”
satisfy the table entries? (derived from Lorentz transformation equations or
otherwise). Do the same for the times of events “ P
coincides with A” and “ P coincides with B.”
b D iscussion question: Suppose the barn has
no back door but rather a back wall of steel-reinforced
PROBLEMS concrete. W hat happens after the farmet closes the
front door on the pole?
5-4 the pole and barn paradox c Replace the pole with a line of ten tennis balls
the same length as the pole and moving together with
A worried student writes, “Relativity must be wrong.
the same velocity as the pole. The farmer’s ten chil
Consider a 20-meter pole carried so fast in the direc
dren line up inside the barn, and each catches and
tion of its length that it appears to be only 10 meters
stops one tennis ball at the same time as the farmer
long in the laboratory frame of reference. Let the
closes the front door of the barn. Describe the stop
runner who carries the pole enter a barn 10 meters
ping events as recorded by the observer riding on the
long, as shown in the figure. At some instant the
last tennis ball. Plot them on your two diagrams.
farmer can close the front door and the pole will be
entirely enclosed in the barn. However, look at the
same situation from the frame of reference of the
runner. To him the barn appears to be contraaed to
half its length. How can a 20-meter pole possibly fit 5-5 radar speed trap
into a 5-meter barn? Does not this unbelievable con
A highway patrolman aims a stationary radar trans
clusion prove that relativity contains somewhere a
mitter backward along the highway toward oncom
fundamental logical inconsistency?”
ing traffic. A detector mounted next to the transmitter
analyzes the radar wave reflected from an approach
ing car. An internal computer uses the shift in fre
quency of the reflected wave to reckon and display the
car’s speed. Analyze this shift in frequency as in parrs
a - e or with some other method. Treat the car as a
simple mirror and assume that the radar signals move
back and forth along one line on the highway. Radar
is an electromagnetic wave that moves with the speed
of light.
The figure shows the worldline of the car, world
lines of two adjacent maxima of the radar wave, and
the wavelength A of incident and reflected waves.
a From the 4 5-degree right triangle ABC, show
that
A t = tsAt +
EXERCISE 5-4. Fast runner with “20-meter" pole enclosed in a From the 4 5-degree right triangle DBF, show that
“10-meter" ham. In the next instant he will hurst through the hack
door, which is made of paper. A/ = 4dd«,e - fA i
EXERCISE 5-5 RADAR SPEED TRAP 167
second
EXERCISE 5-5. Worldlines of approaching car and two radar wave maxima that reflect from the car. The speed of the car is greatly
exaggerated.
Eliminate At from these two equations to find an reflected radar is small. Under this assumption, use
expression for in terms of A;„adent and the the first two terms of the binomial expansion
automobile speed v.
b The frequenq? / of radar (in cycles/second) is (1 — z)” ^ 1 — «z for lz| « 1
related to its wavelength A in a vacuum by the for
mula / = c/A, where c is the speed of light (~ the to show that the fractional change of frequency is
speed of radar waves in air). Derive an expression given by the approximate expression
or frequency yrefleaed the reflected radar signal in
terms of f r e q u e n c y o f the incident wave and the
speed V of the oncoming automobile. Show that the — ~ 2v
result is f
Substitute the speed of a car moving at 100 kilome-

ters/hour (= 2 7 .8 meters/second = 60 miles/hour)
fte& ecad f\nad
and show that your assumption about the small frac
tional change is justified.
C For an automobile moving at a speed v = d One radar gun used by the Massachusetts
t'conv/r that is a small fraction of the speed of light, Highway Patrol operates at a frequency of
assume that the fractional change in frequency of 10.525 X 10^ cycles/second. By how many cycles/
168 EXERCISE 5-6 A SUMMER EVENING'S FANTASY
second is the reflected beam shifted in frequency when a Plot EVENTS labeled with the following
reflected from a car approaching at 100 kilometers/ NUMBERS.
hour? 0. your location when the aliens land (at the
e W hat discrimination between different fre origin)
quency shifts must the unit have if it can distinguish 1. Sun explodes
the speed of a car moving at 100 kilometers/hour 2. light from Sun explosion reaches you
from the speed of one moving at 101 kilometers/ 3. Venus’s atmosphere blown away
hour? 4. light from event 3 reaches you
Reference: T. M. Kalotas and A. R. Lee, AmericanJournal ofPhysics, 5. you and aliens depart Earth (you hope!)
Volume 58, pages 187-188 (February 1990). 6. Earth atmosphere blown away
b Plot WORLDLINES labeled with the follow
5-6 a summer evening's fantasy ing CAPITAL LETTERS.
You are standing alone outdoors at dusk on the first A. your worldline
day of summer. You see Sun setting due west and the B. worldline of Earth
planet Venus in the same direction. On the opposite C. aliens’ worldline
horizon the full Moon is rising due east. An alien ship D. worldline of Sun
approaches from the east and lands beside you. The E. worldline of Venus
occupants inform you that they are from Proxima E. worldline of light from Sun’s explosion
Centauri, which lies due east beyond the rising Moon. G. worldline of the “speed-one-half” pulse
They say they have been traveling straight to Earth of particles from Sun’s explosion
and that their reduced approach speed within the H. worldline of light emitted when Venus
solar system was such that the time stretch factor loses atmosphere
gamma during the approach was 5/3. J. terminal part of the worldline of the laser
At the same instant that the aliens land, you see cannon pulse fired at Sun by the aliens
Sun explode. The aliens admit to you that earlier, on
c Write numerical values for the speed v =
their way to Earth, they shot a laser light pulse at Sun,
i'conv/'^ on every segment of all worldlines.
which caused this explosion. They warn that Sun’s
explosion emitted an immense pulse of particles
moving at half the speed of light that will blow away 5-7 the runner on the train
Earth’s atmosphere. In confirmation, shortly after the paradox
aliens land you notice that the planet Venus, lying in
A letter sent to the Massachusetts Institute of Tech
the direction of Sun, suddenly changes color.
nology by Hsien-Yen Tsao of Los Angeles poses the
You grab a passing human of the opposite sex and
following paradox, which he asserts disproves the
plead with the aliens to take you both away from
theory of relativity. The Chairman of the Physics
Earth in order to establish the human gene pool else
Department sends the inquiry along to you, asking
where. They agree and set the dials to flee in an
you to respond to Mr. Tsao. You determine to make
easterly direction away from Sun at top speed, with
the answer clear, concise, decisive, and polite — a
time stretch factor gamma of 25/7. The takeoff is to
personal test of your diplomacy and grasp of relativ
be 7 minutes after the alien landing on Earth.
ity.
Do you make it?
T h e setting: A train travels at high speed. A
Draw a detailed Earth spacetime diagram showing
runner on the train sprints toward the back of the train
the events and worldlines of this story. Use the fol
with the same speed (with respect to the train) as the
lowing information.
train moves forward (with respect to Earth). There
• Sun is 8 light-minutes from Earth. fore the runner is not moving with respect to Earth.
• Venus is 2 light-minutes from Earth. T h e paradox: We know that, crudely speaking,
clocks on the train run “slow” compared to the Earth
• Assume that Sun, Venus, Earth, and Moon all clock. We also know that the runner’s clock runs
lie along a single direction in space and are rela “slow” compared to the train clocks. Therefore the
tively at rest during this short story. The incom runner’s clock should run “doubly slow” with respect
ing and outgoing paths of the alien ship lie along to the Earth clock. But the runner is not moving with
this same line in space. respect to Earth! Therefore the runner’s clock must
• All takeoffs and landings involve instantaneous run at the same rate as the Earth clock. How can it
changes from initial to final speed. possibly be that the runner’s clock runs "doubly slow”
with respect to the Earth clock and also runs at the
• 5" - 3 ' = 42 and (25)2 - {1? = (24)2 same rate as the Earth clock?
EXERCISE 5-8 THE TWIN PARADOX PUT TO REST—A WORKED EXAMPLE 169
5-8 the twin paradox put to events occurring at the same time in the Earth frame.
Totally different— not a horizontal line! — is a line of
rest— a worked example
simultaneity on the Earth spacetime map that con
Motto: The swinging line of nects events simultaneous in the outgoing astronaut
simultaneity tells alH frame. To draw this line of outgoing-astronaut simul
taneity, start with the inverse Lorentz transformation
Combine the Lorentz transformation with the space-
equation for time:
time diagram to clear up — once and for all! — the
solution to the Twin Paradox. An astronaut travels
from Earth to Canopus (Chapter 4) at speed = t' = + yt
99 /1 0 1 , arriving at Canopus t ' = 20 years later ac
cording to her rocket clock, t = 101 years later ac For the outgoing astronaut, = 99/101 and y =
cording to Earth-linked clocks — which means that 101/20. W e want the line of simultaneity that passes
the stretch factor y has the value 101/20. through turnaround event T. So let t ' = 20 years.
The key idea is “lines of simultaneity” (boxed Then:
labels in the figure). A line of simultaneity connects
events that occur “at the same time.” But events 20 = - (99/101)(101/20)x -b (101/20) t
simultaneous in the Earth (“laboratory”) frame are
typically not simultaneous in the rocket frame (Sec Multiply through by 20/101:
tion 3.4). Horizontal is the line of simultaneity on the
Earth (“laboratory”) spacetime map that connects 400/101 = - ( 9 9 /1 0 1 ) x + r
time
EXERCISE 5-8. Earth spacetime map of the trip to Canopus and of simultaneity, which swings to BT. Ar she leaves Canopus, her
hack. A t the astronaut arrives at Canopus, her colleagues in her new colleagues take an Earth-clock reading of 1 9 8 .0 4 years at B.
outgoing reference frame record along line AT events simultaneous At turnaround, the ticks on the Earth clock along worldline segment
with this arrival, including Earth-clock reading of 3-96 years at AB gofrom the outward-moving astronaut's future to the incoming
A. At Canopus the astronaut changesframes, thus changing the line astronaut’s past.
170 EXERCISE 5-8 THE TWIN PARADOX PUT TO REST - A WORKED EXAMPLE
which yields time lapse AS? This is recorded, sure enough, by the
Earth clock plowing forward along worldline OC in its
/ = 0.980 x - f 3.96 comfortable single free-float frame. However, the
story of time AB is quite different for the turn-around
This is the equation for a straight line passing astronaut. Before she reaches turnaround at T, events
through event points A and T in the spacetime dia on line AB are in herfuture. All those Earth clock ticks
gram. It is the line of simultaneity for the outgoing are yet to be recorded by her outgoing colleagues.
astronaut, connecting all events simultaneous with These events lie above her line of simultaneity A T a.s
the arrival of the rocket at Canopus (simultaneous in she arrives at Canopus at T. However, as she turns
that frame). Among these events is event A, the Earth around, her line of simultaneity also slews forward,
clock reading of 3.96 years, which occurs ar Earth swinging from line A T to line BT. Suddenly the
position x = 0. In brief, at the moment the rocket events on line AB— all those intermediate ticks of the
arrives at Canopus, the Earth clock reads 3.96 years as Earth clock— are in the astronaut’s past. These
observed in the outgoing rocket frame. events lie below the line of simultaneity B T as she
Now the astronaut jumps to the incoming rocket starts back at T. Her outward-moving colleague reads
frame. This reverses the velocity of the astronaut with 3.96 years on the Earth clock as she teaches Canopus;
respect to the Earth-linked frame— and so reverses the an instant later on her clock, her new inward-moving
slope of the line of astronaut simultaneity. This new colleague reads 198.04 on the Earth clock.
line of astronaut simultaneity passes through event Shall we say that the Earth clock “jumps ahead’’ as
points B and T in the figure. Event B is the Earth clock the astronaut turns around? No! Utterly ridiculous!
reading of 202 — 3.96 = 198.04 years. Eor what single observer does it jump ahead? Not for
To go back over the astronaut trip while looking at the Earth observer. Not for the outgoing set of clock-
the spacetime map is (finally!) to solve the Twin readers. Not for the returning set of clock readers. For
Paradox. As the astronaut travels outward toward whom then? Nobody! A t the same time as she reaches
Canopus, many colleagues follow her at the same Canopus— old meaning of simultaneous! — the as
speed, with clocks synchronized in her frame. As they tronaut’s outgoing colleague records 3.96 years for
whiz past Earth, each records the reading on the Earth the Earth clock. A t the same time as she leaves
clock. Later analysis leads them to agree that the time Canopus— new meaning of simultaneous! — her new
between ticks of Earth’s clock is longer than the time ingoing colleague records 198.04 years on the Earth
between ticks of their own outward-moving clocks. clock. The astronaut has nobody but herself to blame
(They say, “The Earth clock mns slow.’’) At any for her misperception of a “jump” in the Earth clock
event point on her outward worldline, the astronaut’s reading.
line of simultaneity slopes upward to the right in the The “lost Earth time” AB in the figure makes
Earth spacetime diagram, as shown in the figure. consistent the story each observer tells about the
Simultaneous with astronaut arrival at Canopus clocks. Simple is the story told by the Earth observer:
(event T, when all outward-moving clocks read 20 “ My clock ticked along steadily at the ‘proper’ rate
years), one of her colleagues reads a time 3.96 years on from astronaut departure to astronaut return. In con
the Eanh clock (event A). trast, ticks on the astronaut clock were far apart in
Now the astronaut jumps from the outward-mov time on both the outgoing and incoming legs of her
ing rocket to a returning rocket. She inherits a com trip. W e agree that her total ticks are less than my
pletely new set of colleagues, with a new set of synchro total ticks: she is younger than I when we meet
nized clocks. The astronaut’s new line of simultaneity again.” More complicated is the astronaut account of
slopes upward to the left in the Earth spacetime dia clock behavior: “Ticks on the Earth clock were far
gram. Simultaneous with her deparmre from Cano apart in time as I traveled to Canopus; ticks on the
pus (event T, when all inwatd-moving clocks read 20 Earth clock were also far apart as I traveled home
years), one of her new colleagues reads a time 202 — again. But as I turned around, a whole bunch of Earth
3.96 = 198.04 years on the passing Earth clock clock ticks went from my future to my past. This
(event B). Thereafter new colleague after new col accounts for the larger number of total ticks on the
league streaks past Earth, recording the fact that Earth Earth clock than on my clock during the trip. We
clock ticks are farther apart in time than the ticks on agree that I am younger when we meet again.”
their own clocks. (They say, “The Earth clock runs So saying, the astronaut renounces her profession
slow.’’). and becomes a stand-up comedian.
The analysis so far accounts for the short time Reference: E. Lowry, American Journal o f Physics, Volume 31, page
segments OA and BC recorded by the Earth clock on 59 (1963).
its vertical wotldline AC W hat about the omitted
6.1 LIGHT SPEED: LIMIT ON CAUSAUTY
no signal reaches us faster than light
Nine-year-old Meredith waves her roy magician’s wand and shouts, "Sun is exploding
right now!” Is she right? We have no way on Earth of knowing— at least not for a
while. Sun lies 150,000 million meters from Earth. Therefore it will take 150,000
Signal Sun with super speed?
million meters of light-rravel time for the first light flash from the explosion to reach
us. This equals 500 seconds— 8 minutes and 20 seconds. W e will just have to wait
and see if Meredith is correct . . .
When 8 minutes and 20 seconds pass, we have evidence that Meredith was
mistaken; Looking through our special dark glasses, we see no exploding Sun.
But Meredith’s wand has started us thinking. W hat in the laws of nature prohibirs
rhe wave of her wand from being the signal fot Sun to explode at that same instant?
Or — more reasonably, given the awesome event— what prevents Meredith from
having instanraneous warning, so that she raises her wand simultaneously with Sun’s
explosion in order to give us (in light of later developments) a false impression of her
power?
Both questions have the same answer; “The speed of light.” Whatever her powers, No, just speed of light
Meredith cannot affect Sun in less than 500 seconds; neither can a warning signal reach
us from Sun in less time than that. All during that intervening 500 seconds we would
see the accustomed round shape of Sun, apparently healthy as ever.
More generally, one event cannot cause another when their sparial separation is
greater rhan the distance light can travel in the time between these events. Light speed
sets a limit on causality. No known physical process can overcome this limit; not
gravity, not some other field, not a zooming particle of any kind. "Spacetime interval”
quantifies this limit on causality. Interval between far-away events — unlike distance
between far-away points — can be zero. In this and other ways rhe spacerime geometry
of the real world differs fundamentally from the space geometry of Euclid’s 2300-
year-old rextbook.
171
172 CHAPTER 6 REGIONS OF SPACETIME
6.2 RELATION BETWEEN EVENTS:

TIMELIKE, SPACELIKE, OR LIGHTLIKE
minus sign yields three possible relations
between pairs of events
Using Euclidean geometry, a surveyor reckons the distance between two steel stakes
from the sum of the squates of the northward and eastward separations of these stakes:
uared distance: Positive or zero (distance)^ = (northwatd separation)^ + (eastward separation)^
In consequence, in Euclidean geometry a distance— or its square — always has a

positive value or zero.
In contrast, the spacetime interval between events in Lorentz geometry arises from
the difference of squares of time and space separations:
Squared interval: (interval)^ = (separation in time)^ — (separation in space)^

Positive, zero, or negative
In consequence of the minus sign, this equation yields a number that may be positive,
negative, or zero, depending on whether the time or the space separation predomi
nates. Moreover, whichever of these three descriptions characterizes the interval in one
free-float frame also characterizes the interval in any other free-float frame. Why?
Because the spacetime interval between two events has the same value in all overlap
ping free-float frames. In the threefold possibilities for an interval, nature reveals the
causal relation between events.
An interval between two events earns the name tim e lik e or spacelike or lig h t
like depending on whether the time part predominates, the space part predominates,
or the time and space parts are equal, respectively, as shown in Table 6-1. Eor
convenience, the minus sign is placed so that the resulting squared interval is greater
than or equal to zero.
Timelike interval: T im elik e Interval: W e picture the sequence of sparks emitted by a moving
Time part dominates
sparkplug. Points representing these sparks on the spacetime map trace out the
worldline of the particle (Chapter 5). No material particle has ever been measured to
travel faster than light. Every material particle always travels less than one meter of
distance in one meter of light-travel time. The sparks emitted by the particle have a
greater time separation than their separation in space. In other words, the worldline of
a particle consists of events that have a timelike relation with one another and with the
initial event. We say that a material particle follows a tim e lik e w o rld lin e.
The interval T between two timelike events reveals itself to the observer in any
free-float frame:
(timelike interval)^ = = (time separation)^ — (space separation)^ (6-1)
6 -T ^
CLASSIFICATION OF THE RELATION BETWEEN TWO EVENTS

Description Squared interval is named and reckoned
Time part of interval dominates space part (timelike interval)^ = = (time)^ — (distance)^
Space part of interval dominates time part (spacelike interval)^ = = (distance)^ — (time)^
Time part of interval equals space part (lightlike interval)^ = 0 = (time)^ — (distance)^
6.2 RELATION BETWEEN EVENTS; TIMELIKE, SPACELIKE, OR LIGHTLIKE 173
left-rocket time laboratory time right-rocket time
left-rocket space
LEFT-MOVING ROCKET FRAME LABORATORY FRAME RIGHT-MOVING ROCKET FRAME

FIGURE 6-1. Events A a n d Bform a timelike p a ir (w ith event A a rb itra rily chosen as reference
event), here recorded in the spacetime maps o f three free-float fram es, Point B lies on a hyperbola
opening along the time axis in each frame. The shortest time between events A and B is recorded in the
laboratory frame, the frame in which the two events occur at the same place.
Same two sparks registered in different frames? Different records for the separation
in time between those sparks. Different records for the separation in space. Same figure
for the timelike interval between them!
Nobody can keep us from tracing out on one and the same diagram (Figure 6-1)
the very different records for the separation AB that observers get in different free-float
Timelike interval:
frames. One frame? One point on the diagram. Another frame? Another point on the
Invariant hyperbola opens
diagram. And so on. These many records for the same pair of events AB trace out a along time axis
hyperbola. This hyperbola opens out in the time direction.
The two sparks, A and B — definite locations though they occupy in spacetime—
nevertheless register in different frames of reference as having different separations in
reference-frame time. Among the many conceivable frames, which one records this
separation in time as smallest? Answer: The frame in which spark B occurs at the same
place as spark A. In other words, the frame that happens to move along in sync with
the sparkplug, even if only briefly. In that frame the clock records a separation in time
between A and B identical with the timelike interval AB.
As seen in the left-moving rocket frame in Figure 6-1, spark B lies to the right of
spark A. In contrast, spark B occurs to the left of spark A in the right-moving rocket.
The position of B relative to A depends on the reference frame from which it is
measured. For a pair of events separated by a timelike interval, labels “right” and
“left” have no invariant meaning: they are frame-dependent.
Spacelike interval:
S pacelike Interval: The interval between two events A and D is spacelike when
Space part dominates
the space part predominates over the time part. Such was the case for a possible
explosion of Sun (event A) and Meredith’s wand waving (event D), simultaneous with
A as recorded in the Earth frame (Section 6.1). Events A and D, if they occurred,
would be separated in the Earth-Sun frame by a distance of 150,000 million meters
and separated by a time of zero meters. Clearly the space part predominates over the
time part! Whenever the space part predominates, we call the relation between the two
events spacelike.
The interval s (sometimes called by the Greek letter sigma, (T) between two
spacelike events reveals itself to the observer in any free-float frame:
(sp a celik e interval)^ = = (sp ace separation)^ — (tim e separation)^ (6 - 2 )

Events A and D registered in different frames? Then different records for the separa
tion in time between those events. Also different records for the separation in space.
Same numerical value for rhe spacelike interval between rhem!
We plot on another spacetime diagram (Figure 6-2) all of the very differenr records
for the separation AD that observers get in different free-float frames. One frame? One
Spacelike interval:
point on the diagram. Another frame? Another poinr on the diagram. And so on.
Invariant hyperbola opens
along space axis These many records for the same pair of evenrs AD trace out a hyperbola. This
hyperbola opens out in the space direction.
The two events, A and D — definite locations though they occupy in spacetime —
nevertheless register in different frames of reference as having different separations in
reference-frame space. Among the many conceivable frames, which one records this
separation in space as smallest? Answer: The frame in which spark D occurs at the
same time as spark A. In that frame a long srick records a separation in space between A
and D identical with the spacelike interval, AD. This is called the p ro p e r d istan ce
between the two spacelike events.
In the Earth - laboratory frame in Figure 6-2, Meredith waves her wand (event D)
at the same time as Sun explodes (event A). In the right-moving rocket frame Sun
explodes after Meredirh waves her wand. In the left-moving rocket frame Sun
explodes before the wand wave. For a pair of events separated by a spacelike interval,
labels “before” and “after” have no invariant meaning: they are frame-dependent. To
allow the wand to control Sun would be to scramble cause and effect!
No particle — not even a flash of light — can move between two events connected
by a spacelike interval. To do so would require it to cover a distance greater than the
time available to cover rhis disrance (space separation greater than time separation). In
brief, it would have to travel faster than light. This is alternative evidence that two
events separated by a spacelike interval cannot be causally connected: one of them
cannot “get at” the other one by any possible signal.

FIGURE 6-2. The spacelike p a ir o f events A a n d D {with event A a rb itra rily chosen as reference
event) as recorded in the spacetime maps o f three free-float fram es. Point D lies on a hyperbola
opening along the space axis in every rocket and laboratory frame. The shortest distance between these events
is recorded in the laboratory frame, the frame in which the two events occur at the same time. A heavy line
represents the spacetime separation AD. No particle can travel along this line; the speed would be greater
than light speed— and would be infinitely great as measured in the laboratory frame, since the particle
would have to cover the distance from A /o D in zero time!
6.2 RELATION BETWEEN EVENTS: TIMELIKE, SPACELIKE, OR LIGHTLIKE 175
E PR
RELATIONS BETWEEN EVENTS
Events 1,2, and 3 all have laboratory locations y — z — 0. Their x and t measurements are
plotted on the laboratory spacetime map.
a. Classify the interval between events 1 and 2; timelike, spacelike, or lightlike.
b. Classify the interval between events 1 and 3.
c. Classify the interval between events 2 and 3.
event
2
7
6 event
3,
t '
time
(meters) 3
event
1 2 1,
1 2 3 4 5
----- space (meters) —
SOLUTION
a. For event 1, / = 2 meters and x — I meter. For event 2, t = 7 meters and x = 4
meters. The squared interval between them: (interval)^ — { 7 ~ 2Y — (4 — 1)^ =
5^“ 3^ = 2 5 “ 9== 16 (meters)^. The time part is greater than the space part, so
the interval between these two events is timelike: T = 4 meters.
b. For event 1, t = 2 meters and x = I meter. For event 5, t = 5 meters and x = 6
meters. The squared interval between them: (interval)^ = (5 “ 2)^ — (1 — 6)^ =
3^ — 5^ = 9 ~ 2 5 = “ 16 (meters)^. The space part is greater than the time part,
so the interval is spacelike: s = 4 meters. (For spacelike intervals, we subtract the
squared time part from the squared space parr before taking the square root.)
c. For event 2, / = 7 meters and x = 4 meters. For event 3 ,^ = 5 meters and x = 6
meters. The squared interval between them: (interval)^ ~ { 7 ~ ~ 5)^ — (4 — 6)^ =
2^ — 2^ = 4 — 4 = 0 (meters)^. The time part equals the space part, so the
interval is lightlike-, it is a null interval.
L ightlike In terv al (N ull Interval): Two events stand in a lightlike relation

when the interval between them is zero:
Lightlike interval;
(time separation)^ — (space separation)^ — 0
Time separation equa
sp a ce separation
or
magnitude of (separation in time) — (distance in space) [for lightlike interval] (6-3)

Lightlike interval: An interval that is lightlike? A separation in time between two events, A and G,
Plotted along ±45 degree lines identical to the distance in space between them? W hat does this condition mean? This:
A pulse of light can fly directly from event A and arrive with perfect timing at event G.
How come? Distance in meters between the two locations measures the meters of time
required for light to fly from one place to the other. Separation in time between the two
events represents the time available for the trip. Time available equals time needed?
Guarantee that the pulse from A arrives in coincidence with event G\ More generally,
whenever the influence of one event, spreading out at the speed of light, can directly
affect a second event, then the interval between those two events rates as lightlike, zero,
null.
Only light (“photons”), neutrinos, and gravitons can move directly between two
events connected by a lightlike interval. Only by means of one of these light-speed
particles can the one event in a lightlike pair cause the other.
The spherical out-going pulse of light from an event. A, may trigger two widely
separated events, E and G(Figure 6-3). Does this common genesis imply that E and G
occur at the same time? Yes and no! Yes, there’s always a free-float reference frame in
which the two daughter events appear as simultaneous. That frame — for no good
reason — we call the laboratory frame in Figure 6-3. In other frames of reference — for
example, the left-moving rocket frame in Figure 6-3 — the clocks show that E occurs
before G. There are still other frames — the right-moving rocket frame is one — in
which the clocks register E and G in the opposite order of time. But no frame shows
either £ or G in the past of A.
Hold it! Aren't spacelike separations impossible? I understand timelike a n d lightlike

separations between two events, because a p article— or a t least a light fla sh — can
travel between them. Not even a light flash, however, can travel from one event to a
second event separated from the first by an interval th a t is spacelike. The first event
cannot possibly cause the second event in the spacelike case. Therefore a spacelike
interval cannot arise in nature. So why talk about it?
left-rocket time laboratory time right-rocket time
A A
G .'
laboratory space right-rocket space
FIGURE 6-3. Two lightlike p a irs of events AE a n d AG {with event A a rb itra rily chosen as
reference event) as recorded in spacetime maps o f three free-float fram es. A flash originates at A
and spreads outwardfrom the center ofa rod at rest in the laboratoryframe. Events E and G are receptions of
this flash at the two ends of the rod as recorded by different observers. In the laboratoryframe, reception events
E and G occur at the same time. In the right-moving rocketframe, the rod moves to the left, so event G occurs
sooner than event E. In the left-moving rocket frame, the rod moves to the right, so event E occurs sooner than
event G.
6.3 LIGHT CONE: PARTITION IN SPACETIME 177
Oops! A spacelike interval between two events certainly can and does arise in nature.
Signals from the supernova labeled 1987A reported that event to us in 1987,
which was 150,000 years after the explosion occurred. Yet occur it did! N o astron
omer of Babylonian, Egyptian, or Greek days reported it, nor could they even know
of it. Yet it had already happened for them. That event separated itself from each of
them by a spacelike interval. Only the advance of time to the year 1987 brought
down the interval between that explosion and Earthbound observers from spacelike
to lightlike. In that year a light pulse carried the earliest possible report of that
explosion to our eyes. And look today? See no explosion at that location in the sky.
The light from it has passed us by. Our present relation to that event? Timelike!
6.3 LIGHT CONE: PARTITION IN

SPACETIME
invariance of the interval preserves cause and
effect
Thus far in dealing with the interval between two events, A and B, we have considered
primarily the situation in which these events lie along a single direction in space — on
the reference line where the laboratory and rocket reference clocks are located. In
contrast, the surveyors in our imaginary kingdom made use of two space dimensions
— northward and eastward. We know, however, that Euclidean space is truly three-
dimensional. A surveyor measuring hilly terrain soon appreciates the need for a third
dimension: the direction vertically upward! The measure of distance in three dimen
sions requires a simple extension of the expression for distance in two dimensions: The
square of the distance becomes the sum of the squares of three mutually perpendicular Interval generalized to
three space dimensions
separations:
(distance)^ = (north separation)^ + (east separation)^ + (up separation)^
Euclidean space requires three dimensions. In contrast, spacetime, which includes

the time dimension, demands four. The expression for the square of a timelike interval
now has four terms: a positive term (the square of the time separation) and three
negative terms (the squares of the separations in three space dimensions).
(interval)^ = (time separation)^ — (north separation)^

— (east separation)^ — (up separation)^
The three space terms can be represented by the single distance term in the equation
above, yielding
(timelike interval)^ = (time separation)^ — (distance)^

(spacelike interval)^ = (distance)^ ^ (time separation)^
(lightlike interval)^ = 0 = (time separation)^ — (distance)^
or, for the lightlike interval,
magnitude of (separation in time) = (distance in space) tiightlike interval] (6-3)
For pairs of events with lightlike separation, the interval equals zero. The zero
interval is a unique feature of Lorentz geometry, new and quite different from
p r o b l e m 6-2
E X P L E T I V E DELETED
At 12:00 noon Greenwich Mean Time (GMT) an circuit (event D) temporarily disables the receiving
astronaut on Moon drops a wrench on his toe and amplifier at Mission Control on Earth. Take Earth
shouts “Damn!” into his helmet microphone and Moon to be 3.84 X 10® meters apart in the
(event A), carried by a radio signal toward Earth. Earth frame and assume zero relative motion.
At one second after 12:00 noon GMT a short
a. Does Mission Control on Earrh hear the astronaut’s expletive?
b. Could the astronaut’s strong language have caused the short circuit on Earth?
c. Classify the spacetime separation between events A and D: timelike, spacelike, or
lightlike.
d. Find the proper distance or proper time between events A and D.
e. For all possible rocket frames passing between Earth and Moon, find the shortest
possible distance between events A and D. In the rocket frame for which this
distance is shortest, determine the time between the two events.
SOLUTION
a. In one second, electromagnetic radiation (light and radio waves) travels 3.0 X
10® meters in a vacuum. Therefore the radio signal does not have time to travel
the 3.84 X 10® meters between Moon and Earth in the one second available
between the events A and D as measured in the Earth frame. So Mission Control
does nor hear the exclamation.
b. No signal travels faster than light. So the astronaut’s strong language cannot have
caused the short circuit.
c. The space part of the separation between events (3.84 X 10® meters) dominates
the time part (one second = 3.0 X 10® meters). Therefore the separation is
spacelike.
d. The square of the proper distance s comes from the expression
f2 = (space separation)^ — (time separation)^

— (3.84 X 10® meters)^ — (3.00 X 10* meters)^
= (14.75 - 9.00) X 10*6 (meters)^
= 5.75 X 10*6 (meters)^
The proper distance equals the square root of this value: r = 2.40 X 10® meters
e. The proper distance equals the shortest distance between two spacelike events as
measured in any rocket frame moving between them (Figure 6-2, laboratory
map). Hence 2.40 X 10® meters equals the shortest possible distance between
events A and D. In the particular rocket frame for which the distance is shortest,
the time between the two events has the value zero — events A and D are
simultaneous in this frame.
SAMPLE PROBLEM
SUNSPOT
Bradley grabs his sister’s wand and waves it, be eruption of the sunspot at the surface of Sun
shouting “Sunspot!” At that very instant his fa itself The Eatth-S un distance equals approxi
ther, Lloyd, who is operating a home solar obser mately 1.5 X 10“ meters. Neglect relative motion
vatory, sees a spot appear on the face of Sun. Let between Earth and Sun.
event E be Bradley waving the wand and event A
a. Is it possible that Bradley’s wand waving caused the sunspot to erupt on Sun?
b. Is it possible that the sunspot erupting on Sun caused Bradley to wave his wand?
c. Classify the spacetime separation between events A and E: timelike, spacelike, or
lightlike.
d. Find the value of ptoper distance ot propet time between events A and E.
e. For all possible rocket frames passing between Earth and Sun, find the shortest
possible distance or the shortest possible time between events A and E.
SOLUTION
a. Light travels 1 meter ot distance in 1 meter of time — or 1.5 X 10“ meters of
distance in 1.5 X 10^* meters of time. Hence in the Earth-Sun frame, eruption of
the sunspot (event A) occurted 1.5 X 10“ meters of time before Bradley waved
the wand (event E). So Bradley’s wand waving could not have caused the
eruption on Sun.
b. On the other hand, it is possible that eruption of the sunspot caused Bradley to
wave his wand: He raises the wand in the air, looks over his father’s shoulder, and
waves the wand as the spot appears on the projection screen. (We neglect his
reaction time.)
c. Events A and E are connected by one light pulse; their space and time separations
both have the value 1.5 X 10“ meters in the Eatth frame. Therefore the
spacetime separation between them is lightlike.
d. Space and time separations between events A and E are equal. Therefore the
interval between them has value zero. Hence proper time between them — equal
to ptoper distance between them — also has value zero.
e. The interval is invariant. Thetefore all possible ftee-float rocket frames passing
between Earth and Sun reckon zero interval between events A and E. This means
each of them measures space separation between events A and E equal to the time
separation between these events. The common value of the space and time
separations are not the same for all rocket frames, but they are equal to one
another in every individual rocket frame. We are asked to find the shottest
possible value for this time.
Think of a tocket just passing Sun as the sunspot erupts, the rocket headed
towatd Earth at nearly light speed with respect to Earth. Rocket lattice clocks
record the light flash from the sunspot moving away from the rocket at standard
speed unity. However, these clocks recotd that Earth lies very close to Sun
(Lorentz contraction of distance) and that Earth rushes toward the rocket at nearly
light speed. Therefore light does not travel far to get to Earth in this rocket frame;
neither does it take much time. For a rocket moving arbitrarily close to light
speed, this distance between A and E approaches zero, and so does the time
SA M P L E P R O B L E M 6 -3
between A and E. Hence the shortest possible distance between A and E— equal
to the shortest possible time between A and E — has the value zero. But this
constitutes a limiting case, since rocket speed may approach but cannot equal the
speed of light in any free-float frame.
anything in Euclidean geometry. In Euclidean geometry it is never possible for distance

A G between two points to be zero unless all three of the separations (northward,
eastward, and upward) equal zero. In contrast, interval A G between two events can
vanish even when separation in space and separation in time are individually quite
large. Equation (6-3) describes the separation between lightlike events, but now
separation in space may show up in two or three space dimensions as well as one time
dimension. The distance in space is always positive.
It is interesting to plot on an appropriate map locations of all events, G, Gj, G^,
Light flash traces out light cone
Gj, , that can be connected with one given event A by a single spreading pulse
in spacetime diagram
of light. Every such future event has a distance in space from A identical to its delay in
time after A. Only so can it satisfy the requirement (6-3) for a null interval. For it:
(future time with respect to A) = d- (distance in space from A) [lightlike interval] (6-4)
It is equally interesting to display — and on the same diagram — all the events H, Hj,
H 2 , Hj, . . . that can send a light pulse to A. Every such event fulfills the condition
(past time relative to A) = — (distance in space from A) (for lightlike interval] (6-5)
Both of these equations satisfy the magnitude equation (6-3).

In Figure 6-4 we suppress display of a third space dimension in the interest of
simplicity. We limit attention to future events G, G„ G2 , . . . and past events H, Hj,
H 2 , . . . that lie on a north-south/east-w est plane in space. A flash emitted from
event A expands as a circle on this space plane. As it spreads out from event A, this
circle of light traces out a cone opening upward in the spacetime map of Figure 6-4.
This is called the fu tu re lig h t cone of event A. The cone opening downward traces
the history of an in-coming circular pulse of radiation so perfectly focused that it
converges toward event A, collapsing exactly at event A at time zero. This downward
opening cone has the name past lig h t cone of event A. All the events G, G^,
G2 , . . . lie on the future light cone of event A, all events H, H„ H 2 , . . . on its
past light cone.
Numerous as the events may be that lie on the light cone, typically there are many
more that don’t! Look, for example, at all the events that occur 7 meters of time later
than the zero time of event A. On the spacetime map, these events define a plane 7
meters above the r = 0 plane in which event A lies, and parallel to that plane. The light
cone intersects this plane in a circle (circle in the present map; a sphere in a full
spacetime map with three space dimensions). An event on the plane falls into one or
another of three categories, relative to event A, according as it lies inside the circle (as
does B in Figure 6-4), on it (as does G), or outside it (as does D).
The light cone is unique to Lorentz geometry. It gives nature a structure beyond any
power of Euclidean geometry. The light cone does more than divide events on a single
plane into categories. It classifies every event, everywhere in spacetime, into one or
another of five distinct categories according to the causal relation that event bears to
the chosen event. A:
FIGURE 6-4. Light cone as p artitio n in spacetime; perspective three-dim ensional spacetime map
showing eastw ard, northw ard, a n d time locations of events occurring on a f la t p lan e in space.
Events G, Gj, Gj, and Gj are on the future light cone of event A; events H, H,, H 2 , and Hj are on its past
light cone. See also Figure 6-5.
1. Can a material p artic le emitted at A affect what is going to happen at El

If so, B lies inside the fu tu r e light cone of A and forms a timelike pair with
event A.
2. Can a lig h t ray emitted at A affect— with no time to spare— what is going
to happen at G?
C au se and effect preserved by
If so, d i e s on the fu tu r e light cone of A and forms a lightlike pair with event
light cone
A.
3. Can no effect w h atev er produced at A affect what happens at D?
If so, D lies outside the future and past light cones of A and forms a spacelike
pair with event A. It lies in the absolute elsewhere of A.
4. Can a material p artic le emitted at J affect what is h a p p e n in g at A?
If so, J lies inside the p a st light cone of A and forms a timelike pair with
event A.
5. Can a lig h t ray emitted at H affect — with no time to spare — what is
h ap p e n in g at A?
If so, H lies on the p a st light cone of A and forms a lightlike pair with event
A.
Nature reveals a cause-and-effect structure beyond the vision of Euclidean geome
try. The causal relation between an event B and another event A falls into one or the
FIGURE 6-5. Exploded view o f the regions in

classified w ith respect to a selected event A.
EXERCISE 6-1 RELATIONS BETWEEN EVENTS 183
other of five categories picked out by the light cone of A. That light cone and those
categories have an existence in spacetime quite apart from any space and time
measurements that may be used to describe them. Zero interval between events in one
free-float frame means zero interval between the same events in every overlapping
free-float frame. The light cone is the light cone is the light cone!
Event A appears at the origin of every spacetime map in this chapter. What’s so special
about event A?
1L Nothing whatever is special about event A! On the contrary, we have not captured
the full story of the causal structure of spacetime until for every event A {Aj, A 2 , Aj,
. . . ) we have classified every event B (Bj, B2, Bj, . . . ) into the appropri
ate category— timelike! lightlike! spacelike!— with respect to that event.
Figure 6-5 summarizes the relations between a selected event A and all other events of
spacetime.
CHAPTER 6 EXERCISES
PRACTICE (4) Is it possible to find a rocket frame in which the

temporal order of the two events is reversed?
That is, is it possible to find a rocket frame in
6-1 relations between events which the event that occurs before the other
This is a continuation of Sample Problem 6-1. Events event in the laboratory frame occurs after the
1, 2, and 3 all have the laboratory coordinates y = other event in the rocket frame?
z = 0. Their x- and /-coordinates are plotted on the
laboratory spacetime diagram.
a Answer the following questions three times: event
2
once for the timelike pair of events 1 and 2, once for 7
the spacelike pair of events 1 and 3, and once for the 6 event
lightlike pair of events 2 and 3.
■'t 4'
( 1) W hat is the proper time (or proper distance)
between the two events? time
(meters) ^
(2) Is it possible that one of the events caused the event
other event?
(3) Is it possible to find a rocket frame in which the
I 2
1,
1
spatial order of the two events is teversed? That
is, is it possible to find a rocket frame in which
0 1 2 3 4 5 6
the event that occurs to the right of the other
----- space (meters) —►-
event in the laboratory frame will occur to the
left of the other event in the rocket frame? EXERCISE 6-1. Laboratory spacetime map.
184 EXERCISE 6-2 TIMELIKE, LIGHTLIKE, OR SPACELIKE?
b For the timelike pair of events, find the speed c Find the speed (with respect to the laboratory
and direction of a rocket frame with respect to which frame) of a rocket frame in which evenr 1 and event 2
the two events occurred at the same place. For the in the first table occur at the same place.
spacelike pair of events, find the speed and direction d Find the speed (with respect to the laboratory
of a rocket frame with respect to which the two events frame) of a rocket frame moving along the x-axis in
occurred at the same time. which event 2 and event 3 in the first table occur at the
same time.
6-2 timelike, lightlike, or
spacelike? 6-3 proper time and proper
distance
The first table lists the space and time coordinates of
three events plus the reference event (event 0) as N ote: This exercise uses the Lorentz transformation
observed in rhe laboratory frame. equations.
a Two events P and Q have a spacelike separa
tion. Show in general that a rocket frame can be found
in which the two events occur at the same time. Also
show that in this rocket frame the distance between
LABORATORY COORDINATES OF THREE
the two events is equal to the proper distance between
EVENTS them. (One method: assume that such a rocket frame
t X y exists and then use the Lotentz transformation equa
(years) (years) (years) tions to show that the relative velocity of this rocket
frame is less than the speed of light, thus justifying the
Event 0 0 0 0 assumption made.)
Event 1 3 4 0 b Two events P and R have a timelike separa
Event 2 6 5 0 tion. Show in general that a rocket frame can be found
Event 3 8 8 3 in which the two events occur at the same place. Also
show that in this rocket frame the time between the
two events is equal to the proper time between them.
a Copy the second table. In the top half of each
box in the second table, write the nature of the
interval — timelike, lightlike, or spacelike —
PROBLEMS
between the two corresponding events.
6-4 autobiography of a photon
b In the bottom half of each box in the second
table, write ‘‘yes” if it is possible that one of the events A photon emitted by a star on one side of our galaxy is
caused the other and “no” if it is not possible. absorbed near a star on the other side of our galaxy.
■<j[^[^CERCISE 6 -2
INTERVAL BETWEEN EVENTS: TIMELIKE, LIGHTLIKE, OR SPACELIKE?

Event 1 Event 2 Event 3
Event 0
Event 1
Event 2
EXERCISE 6-5 THE DETONATOR PARADOX 185
100,000 light-years away from its point of origin as not cotrect when viewed ftom the quantum perspec
measured in the frame of the galaxy. How does the tive. Fot more on quantum electrodynamics, read
photon experience its own birth and death? That is to Richard P. Feynman, QED: The Strange Theory of
say, what are the space and time separations between Light and Matter (Princeton, Ptinceton University
the birth and death of the photon in the frame of the Press, 1985).
photon?
D iscussion: We cannot answer this question, be 6-5 the detonator paradox
cause we cannot move along with the photon. No
A U-shaped structure made of the strongest steel
matter how fast the unpowered rocket in which we
contains a detonator switch connected by wire to one
ride, we still measure light to move past us with the
metric ton (1000 kilograms) of the explosive TNT, as
speed of light! Still, we can try to answer the question
shown in the figure. A T-shaped structure made of
as a limiting case in the galaxy frame. Think of ex
the same strong steel fits inside the U, with the long
tremely energetic PROTONS traveling the same
arm of the T not quite long enough to teach the
path. As protons of greater and greater energy are
detonator switch when both structures are at rest in
emitted by the first star and are absorbed near the
the laboratory.
second star at the other side of the galaxy, what
Now the T structure is removed far to the left and
happens to the distance between these two events in
accelerated to high speed. It is Lorentz-contracted
the frame of the proton? W hat happens to the time
along its direction of motion. As a tesult, its long arm
between these events in the frame of the proton?
is not long enough to teach the detonatot switch when
Come in this way to a limiting case in which the
the two collide. Therefote there will be no explosion.
PROTON is moving arbitrarily close to the speed of
light in the galaxy frame. In this limit, what would
you expect the distance and time to be between birth
and death in the frame of a PHOTON traveling the
same path in space?
a You are the photon. Using the above argu
ment, write the first few sentences of your autobiog
raphy.
At the end of the trip, near a star at the fringe of our
galaxy, a galaxy-spanning photon travels 10 kilome
ters vertically through the atmosphere of a planet
before it enters a telescope and is absorbed in the eye
of an astronomer.
The average in dex o f refractio n of the atmo
sphere of this planet is « = 1.00030. The speed of
the photon in such an atmosphere Is v = v ^ ^ /c =
1/n . (The speed of light in a vacuum is unity.)
b W hat is the proper time for this last leg of the
trip — the time in the rest frame of the “slowed-
down” photon? How far apart is the top of the atmo
sphere and the astronomer’s eye in the frame of the
photon?
C Complete your photon autobiography with an
additional couple of sentences.
D iscussion: Relativity is a classical theory — that
is, a nonquantum theory — in which photons are
postulated to move at light speed in a vacuum and at
a speed v = l / « in air, where n is the index of refrac
tion. Q u a n tu m electrodynam ics (QED), the
REST FRAME OF TSTR U aU R E
quantum theory of interactions between light and EXERCISE 6 -5 . B o th a t re st: The leg o f the T alm ost reaches the
matter, tells us that it is incorrect to talk of a single detonator sw itc h w hen both the T a n d the U are a t rest. P oints A
photon moving through air. Rather, one thinks of an a n d B are used in p a r t h o f the exercise. R e s t f r a m e o fX J s t r u c
tu r e : T he leg o f the m oving T is L orentz contracted in the rest fra m e
initial photon being absorbed by an atom in the air
o f the U. Does th is m ean th a t the explosion w ill not ta ke place? R e s t
and a second photon emitted, the second photon then f r a m e o f T s tr u c tu r e : The legs o f the m oving U are L orentz-con-
absorbed by another atom, which emits a third pho tra c te d in the rest fr a m e o f the T . Does th is m ean explosion w ill take
ton, and so forth. The classical relativistic analysis is place?
18 6 EXERCISE 6-6 HOW FAST CAN YOU WALK?
However, look at the same situation in the rest the front foot touching the ground, as observed by all
frame of the T structute. In this frame the arm of the free-float observers. Use the following outline to de
T has its rest length, while the two arms of the U rive the consequences of this definition for the maxi
structure are Lorentz-contracted. Therefore the arm of mum speed of walking.
the T will certainly strike the detonator switch and a Consider the three possible relationships be
there will be a terrible explosion. tween events FrontDown and RearUp: timelike,
a Make a decisive prediction: Will there be an lightlike, and spacelike. For each of these three rela
explosion or not? Your life depends on it! tionships, write down answers to the following three
b The wire from the detonator switch to the questions:
T N T is restrung through point B on the U structure
(1) Will the temporal order of the two events be
when both structures are at rest, and a laser is installed
the same for all observers?
at point A on the T structure. Later, when the two
(2) Does this relationship adequately satisfy the
sttuctures collide at A, the laser fires a pulse at B that
frame-independent definition of walking?
cuts the detonator wire. Does this new apparatus
(3) If so, does this relationship give the maximum
change your prediction about detonation of the TNT?
possible speed for walking?
Acknowledgment: A paper desaibing this paradox crossed the desk
o f one of the authors, but the paper and the name of its author have Show that you answer “yes” to all three questions
been lost. The laser inhibitor device was devised by Gordon Roesler. only for a lightlike relationship between the two
events.
b A lightlike relationship between events Front-
6-6 how fast can you walk? Down and RearUp means that light can just travel
Webster's Eighth says that to “walk” means to “go on from one event to the other with no time left over. Let
foot without lifting one foot clear of the ground before the distance between these events — the length of one
the other touches the ground.” In other words, at least step in the Earth frame — be the unit of distance and
one foot must be on the ground at all times. Use this time. Show that for the limiting speed in this frame,
definition to discover the maximum speed of walking each foot spends two units of time moving forward,
imposed by relativity. then waits one unit while the light signal propagates
We assume advanced technology here! A walking to the other foot, then waits three units while the other
robot moves its free foot forward at nearly the speed of foot goes through the same process. Summary: Out of
light. Then one might argue (ambiguously) as fol six units of time, each foot moves forward at (nearly)
lows: While the free foot is moving forward, the the speed of light for two units. W hat is the average
planted foot is on the ground, ready to be picked up speed of each foot, and therefore the speed of the
when [look out!} the free foot comes down in front. walker, as measured in the Earth frame?
Half the time each foot is in motion at nearly light c Draw a spacetime diagram for the Eatth frame,
speed and half the time it is at rest. Therefore the showing worldlines for each of the robot’s feet and
average speed of each foot, equal to the maximum worldlines for the connecting light flashes. Add a
possible speed of the walking robot, is half the speed worldline showing the averaged motion of the torso,
of light. always located halfway between the two feet in the
Why is this argument ambiguous? Because of the Earth frame. Demonstrate that this torso moves at the
relativity of simultaneity. The word when applied to speed of the walker reckoned above.
separated events should always unfurl a red flag. The d Paul Horwitz says, “W e determined the value
event “front foot down” (label FrontDown) and the of a maximum walking speed by finding a frame-
event “rear foot u p ” (label RearUp) occur at different independent definition of walking. Therefore this
places along the line of motion. Observers in relative walking robot moves at the same speed as observed in
motion will disagree about whether or not events every frame.” Is Paul right?
FrontDown and RearUp occur at the same time. Reference: George B. Rybicki, American Journal o f Physics, Volume
Therefore they will disagree about whethet or not the 59, pages 3 6 8 - 3 6 9 (April 1991).
robot has one foot on the ground at all times in order
to satisfy the dictionary definition of walking.
How to remove the ambiguity in the definition of
6-7 the flickering bulb
walking? One way is to make the conventional defini
paradox: a project
tion frame-independent: One foot must be on the N ote: The following is too long for a regular exercise,
ground at all times as observed in everyfree-float frame but it has many insights worth pursuing as a longer
of reference. W hat limits does this place on the two activity. Therefore we call it a project.
events FrontDown and RearUp? The rear foot must Two long parallel conducting rails are open at one
leave the ground after, or at least simultaneous with. end but connected electrically at the other end
EXERCISE 6-8 THE CONTRACTING TRAIN PARADOX: A PROJECT 187
■4 meters ---- 2 meters •
SLIDER FRAME
EXERCISE 6 -7 . R a i l f r a m e : C onfiguration a t t = 0 in the rest fra m e o f the rails. S lid er C D moves to the
rig h t w ith speed v„i such th a t the L o rentz-contraction fa c to r equals 2 . T he ve rtic a l legs o f the slid er are
conductors; the h o rizo n ta l crosspiece is a n insulator. S l i d e r f r a m e : C onfiguration a t t ' = 0 in the rest
fra m e o f the slider. The ra ils a n d la m p move to the le ft w ith speed \„ i such th a t the L orentz-contraction fa c to r
is 2.
through a lamp and battery, as shown in the figure particular the offset in the upper rail is Lorentz-
(rail frame). One of the rails has a square vertical contracted to a length of one meter. Therefore, in the
offset 2 meters long. Between the rails moves (with slider frame, one or the other of the slider conductors
out friction) an H-shaped slider, whose vertical legs always spans the rails, so the circuit is never broken
are conductors but whose horizontal crosspiece is an and the bulb should never switch off— it should
insulator. (Assume that the vertical legs are not per N O T flicker!
fect conductors so that, with a sufficiently powerful Those trying to disprove relativity shout, “Para
battery, a voltage is maintained between the rails even dox! In the rest frame of the rails the lamp switches off
when they are connected by the vertical legs of the and then on again — it flickers. In contrast, in the rest
slider.) If either vertical leg of the slider connects the frame of the slider the lamp stays on — it does not
two rails, the electrical circuit is completed, permit flicker. Yet all observers must agree: The lamp either
ting the lamp to light. flickers or it does not flicker. Relativity must be
The rest (proper) length of the slider is also 2 wrong!”
meters, but it moves at such a speed that its Lorentz- Analyze the system in sufficient detail either to
contracted length is 1 meter in the rail frame. Hence demonstrate conclusively the correctness of this objec
in the rail frame there is a lapse of time during which tion or to pinpoint its error.
neither leg of the slider is in contaa with the upper Reference: G. P. Sastry, American Journal o f Physics, Volume 55,
rail. Since the circuit is open during this period, the pages 9 4 3 - 9 4 6 (October 1987).
bulb should switch off for a time and then on again
— it should flicker.
6-8 the contracting spaceship
The figure (slider frame) shows the configuration
paradox: a project
at / ' = 0 in the slider frame. In this frame the slider is
at rest, its length is equal to its rest length, 2 meters, N ote: The following is too long for a regular exercise,
while the rails, the lamp, and the battery all move to but it has many insights worth pursuing as a longer
the left with a speed such that their lengths along the activity. Therefore we call it a project.
direction of motion are reduced by a factor of 2. In Kerwin Warnick writes in with the following par-
18 8 EXERCISE 6-8 THE CONTRACTING TRAIN PARADOX: A PROJECT
adox. A spaceship of proper length L„ accelerates length L„ can be arbitrarily large, this average speed
from rest. Its front end travels a distance Xp in time tp can be arbitrarily great, even greater than the speed of
to a final speed at which the ship is contracted to half light. “This disproves relativity!” he exclaims.
its rest length. In the same time t p the rear end moves Analyze this thought experiment in sufficient de
the same distance Xp as the front end plus the distance tail either to demonstrate conclusively the correctness
L„/2 by which the ship has contracted. Distance of Warnick’s objection or to pinpoint its error.
traveled by the rear end Xp -b (L„/2) in time means Reference: Edwin F. Taylor and A. P. French, American Journal of
an average speed {xp -f fLo/2)}//p. Since the proper Physics, Volume 51, pages 889-893 (Ocrober 1983).
7.1 M O M EN ER G Y:TO lAt CONSERVED
IN A COLLISION
momentum conserved,
energy conserved.
momenergy conserved!
Paradoxically, few examples of motion are more complicated than a collision, and few
are simpler. The complication shows nowhere more clearly than in the slow-motion
Smashup complicated?
videotape of the smashup of two automobiles. Millisecond by millisecond the fendet
of one colliding car deforms another fraction of a centimeter. Millisecond by millisec
ond the radiator grille of the other car bends inward a little more on the way to total
collapse: steel against steel, force against force, crumpling surface against crumpling
surface. W hat could be more complex?
For the drivers of the colliding cars the experience is shattering. They are hardly
aware of noise and complicated damage. A single impression overpowers their senses:
Smashup is simplel
the inevitability of the crash. Call it what we will — inertia, momentum, the grip of
spacetime on mass— something is at work that drives the two vehicles together as the
frantic drivers jam their brake pedals down, locking the wheels as the cars slither over
the glassy ice, crash into one another, then slide apart.
Does mass lose its inettia during the collision? No. Inertia does its best to keep each
car going as it was, to keep its m o m e n tu m constant in magnitude and direction.
Momentum: we can think of it loosely as an object’s will to hold its course, to resist
Momentum conservation simplifies
deflection from its appointed way. The higher the object’s momentum, the more
description
violently it hits whatever stands in its way. But the momentum of a single object is not
all-powerful. The two vehicles exchange momentum. But spacetime insists and
demands that whatever momentum one car gains the other car must lose. Regardless
of all complications of detail and regardless of how much the momentum of any one
189
190 CHAPTER/ MOMENERGY
object may change, the combined momentum of the two objects remains constant: the
total is unchanged in the collision. A like statement applies to energy, despite a
Energy too is conserved conversion of energy of motion into heat energy and fender crumpling.
A collision thus manifests a wonderful simplicity: the combination of the motion-
descriptive quantities (momentum and energy) of the two colliding bodies does not
change. That combination is identical before and after the collision. In a word, it is
M om energy is conserved!
conserved. This conserved combination we call m o m e n tu m - energy or, more
briefly, m om energy (defined more carefully in Section 7.2). W e will use the two
terms interchangeably in this book.
A collision cannot be elevated from mere talk to numbers without adopting,
directly or indirectly, the principle of conservation of momentum and energy. In the
enterprise of identifying the right numbers, using them, and understanding them, no
concept is more powerful than what relativity smilingly holds forth: momenergy.
Wait a minute. Apparently you are going to find new expressions for momentum and
energy, then combine them in some way to form a unity: momenergy. But I have three
complaints. (1) What is wrong with what good old-fashioned secondary school physics
textbooks give us, the Newtonian expressionsfor momentum— psmftm ~
kinetic energy = V2 mv^„„,— where v^„„, is expressed in conventional units, say
meters/second? (2) Momentum and energy do not even have the same units, as these
formulas make clear. How can you combine quantities with different units? (3)
Momentum and energy are different things entirely; why try to combine them at all?
Take your questions in order.
1. N ew to n ia n E xpressions: Only for slow-moving particles do we get correct

results when we use Newtonian expressions for momenrum and energy. For
particle speeds approaching that of light, however, total energy and momentum
of an isolated system, as Newton defined momentum and energy, are not
conserved in a collision. In contrast, when momentum and energy are defined
relativistically, then total momentum and total energy of particles in an isolated
system are conserved, no matter what their observed speeds.
2. U nits: It is easy to adopt identical units for momentum and energy. As a start we
adopt identical units for space and time. Then the speed of a particle is expressed
in unit-free form, v, in meters of distance per meter of light-travel time (Section
2.8). This choice of units, which we have already accepted earlier in this book,
gives even Newtonian expressions for momentum — /'Newton ~ kinetic
energy— f^Newton ~ Vtmv^ — the same unit: mass. These are not relativistic
expressions, but they do agree in their units, and agree in units with the correct
relativistic expressions.
3. M o m e n tu m a n d Energy D ifferen t: Yes, of course, momentum and energy
are different. Space and time are different too, but their combination, spacetime,
provides a powerful unification of physics. Space and time are put on an equal
footing, but their separate identities are maintained. Same for momenergy: We
will see that its “space part” is momentum, its “ time part” energy. We will also
discover that its magnitude is the mass of the particle, teckoned using the good
of, ever-lovin’, familiar minus sign: m^ = B3 — f f .
Thus relativity offets us a wonderful unity. Instead of three separate motion-de

scriptive quantities — momentum, energy, and mass — we have a single quantity:
momenergy.
7.2 MOMENERGY ARROW 191
7.2 MOMENERGY ARROW

a spacetime arrow pointing along the worldline
W hat lies behind the name momentum-energy {momenergy}? W hat counts are its
properties. W e most easily uncover three central properties of momenergy by combin
ing everyday observation with momenergy’s essential feature; Total momenergy is
conserved in any collision.
First, think of two pebbles of different sizes moving with the same velocity toward
the windshield of a speeding car. One bounces off the windshield without anyone
noticing; the other startles the occupants and leaves a scratch. Five times the mass? Five Momenergy of particle
proportional to its moss
times the punch-delivering capacity! Five times the momenergy. Momenergy, in other
words, is proportional to mass.
Second, momentum-energy of a particle depends on its direction of travel. A pebble
coming from the front takes a bigger chip out of the windshield than a pebble of equal
mass and identical speed glancing off the windshield from the side. Therefore mo Momenergy a directed quantity
menergy is not measurable by a mere number. It is a directed quantity. Like an arrow
of a certain length, it has magnitude and direction.
Our experience with the unity of spacetime leads us to expect that the momenergy
arrow will have three parts, cotresponding to three space dimensions, plus a fourth
part corresponding to time. In what follows we find that momenergy is indeed a
four-dimensional arrow in spacetime, the m om energy 4-vector (Box 7-1). Its three Momenergy a 4-vector
“space parts” represent the momentum of the object in the three chosen space
directions. Its “time part” represents energy. The unity of momentum and energy
springs from the unity of space and time.
In what direction does the momenergy 4-vector of a particle point? It points in the
“same direction in spacetime” as the worldline of the particle itself (Figure 7-1). There Particle momenergy points along
its worldline
is no other natural direction in which it can point! Spacetime itself has no structure that
indicates or favors one direction rather than another. Only the motion of the particle
itself gives a prefetred direction in spacetime. The particle moves from one event to a
nearby event along its worldline. In so doing, it undergoes a spacetim e d isplace
m en t, small changes in the three space positions along with an accompanying small
advance in the time. The spacetime displacement has four parts: it is a 4-vector. The
momenergy arrow points in the direction of another arrow, the arrow of the particle’s
spacetime 4-vector displacement. Momenergy runs parallel to worldline!
Compare the worldline of an individual particle in spacetime with a single straw in a
great barn filled with hay. This particular straw has a direction, an existence, and a
meaning independent of any measuring method imagined by humans who stack the
hay or by mice that live in it. Similarly, in the rich trelliswork of worldlines that course
through spacetime, the arrowlike momenergy of the particle has an existence and Momenergy independent of
reference frame
definiteness independent of the choice— or even use— of any free-float frame of
reference (Section 5.9).
No frame of reference? Then no clock available to time motion from here to there!
Or rather no clock except one that the particle itself carries, its own wristwatch that
records proper time. Proper time for what? Proper time for spacetime displacement Particle wristwatch logs time for
momenergy
between two adjacent events on the worldline of the particle. Proper time provides the
only natural way to clock the rate of motion of the particle; that is the third and final
feature of momenergy.
In brief, the momenetgy of a particle is a 4-vector; Its magnitude is proportional to
its mass, it points in the direction of the particle’s spacetime displacement, and it is
reckoned using the proper time for that displacement. How are these properties
combined to form momenergy? Simple! Use the recipe for Newtonian momentum:
mass times displacement divided by time lapse for that displacement. Instead of
WHAT IS A 4-VECTOR?
A vector is a mathematical object that has both m a g n i t u d e and d ir e c t i o n . The
meanings of the terms m a g n i t u d e and d ir e c t i o n , however, differ between one
geometry and another. Mathematics offers many geometries. The two ge
ometries important to us in this book are Euclidean geometry and
Lorentz geometry.
Euclidean geometry defines 3-vectors located in 3-dimensional space. Let

a speeding particle emit two sparks. The particle's spatial displacement
from first spark to second spark is a 3-vector. Each of the three compo
nents (northward, eastward, and upward) of this 3-vector displacement
has a value larger or smaller, depending on the orientation of the coordi
nate system chosen. In contrast, the magnitude of the displacement — the
distance traveled (computed as the square root of the sum of the squares of
the three components of displacement) — has the same value in all coordi
nate systems.
Lorentz geometry defines 4-vectors located in 4-dimensional spacetime.

Construct the 4-vector spacetime displacement from the three space com
ponents supplemented by the time component, the time between sparks
emitted by the speeding particle. Each of these four components (including
time) has a value larger or smaller, depending on the choice of free-fioat
frame of reference from which it is measured. The square of the separation
in time between the two sparks as so measured, diminished by the square
of the separation in space in the chosen frame, yields the square of the
spacetime interval between the two events. This interval has the same
value in all free-fioat frames. It is also the proper time, the time between the
two sparks read directly on the particle's wristwatch.
Newtonian mechanics combines (in various ways) time and mass of the
particle with Euclidean 3-vector displacement of the particle to yield addi
tional 3-vectors that describe particle motion; velocity, momentum, acceler
ation. Each 3-vector has magnitude and direction. The values of the three
components of each 3-vector depend on the orientation of the chosen coor
dinate system. But for each 3-vector quantity, the 3-vector itself is the same,
both in magnitude and direction in space, no matter what Euclidean coordi
nate system we choose. Every 3-vector exists even in the absence of any
coordinate system at all! That is why the analysis of Newtonian mechanics
can proceed in all its everyday applications independent of choice of coor
dinate system.
Relativistic mechanics combines (in various ways) proper time and mass of
the particle with Lorentz 4-vector displacement of the particle to yield addi
tional 4-vectors that describe particle motion. Central among these is the
particle's momentum-energy 4-vector, or momenergy. Values of the four
components of the momenergy 4-vector differ as measured in different free-
fioat frames in relative motion. But the momenergy 4-vector itself is the same,
both in magnitude (mass!) and direction in spacetime, no matter what the
frame. The momenergy 4-vector of a particle exists even in the absence of
any reference frame at all! That is why the analysis of relativistic mechanics
can proceed in all its power independent of choice of free-fioat frame of
reference.
7.2 MOMENERGY ARROW 193
FIGURE 7 -1 . M o tio n a n d m o m e n e r g y se e n a s i d e n t i c a l l y d i r e c t e d a r r o w s . T he momenergy arrow is

carried along the w orldline w ith the p a rticle. U nder a ctio n o f a force, the p a rticle traces out a curved
w orldline. The momenergy arrow — its constant m a g n itu d e eq u a l to the mass o f the p a r tic le — co ntinually
alters its t i l t to p o in t in the sam e direction in spacetim e a s the w orldline. (For the special case show n here, the
particle moves in x a n d t, b u t not in y or z.)
Newtonian displacement in space, use Einstein’s displacement in spacetime; instead of

Newton’s “universal time,’’ use Einstein’s proper time.
The result expresses the momenergy 4-vector in terms of the spacetime displace
ment 4-vector:
Particle momenergy
(spacetim e displacem ent) Magnitude: M ass
(m om energy) — (mass) X (7 - 1 )
(proper time for that displacement) Direction: Along spacetime
displacem ent
In any given free-float frame, the momentum of the particle is the three “space parts’’
of the momenergy and the particle’s energy is the “time part.’’ This expression for
momenergy is simple, and it works — works as employed in the law of conservation of
momenergy; Total momenergy before reaction equals total momenergy after reaction.
Investigators have observed and analyzed more than a million collisions, creations,
transformations, decays, and annihilations of particles and radiation. They have failed
to discover a single violation of the relativistic law of conservation of momenergy.
To arrive at a formula as important as (7-1) so painlessly may at first sight create
doubts. These doubts have to be dismissed. Fact is, there is no room for any
alternative — as we see by going step by step through the factors in this equation.
S tatem en t 1: m u n its o f m ass p u rsu in g a given m o tio n carry m tim es th e

m o m en erg y o f one u n it o f mass. Reasoning: m identical objects racing along side
by side carry m times the momentum and m times the energy— and therefore m times
the momenergy— of an object of unit mass.
S tatem en t 2: M om energy p o in ts in th e sam e d irec tio n in spacetim e as
w o rld lin e. Reasoning: Where else can it point? Even the slightest difference in
direction between momenergy and direction of motion along the worldline would bear
witness to some crazy asymmetry in spacetime, for which no experiment in field-free
space has ever given the slightest evidence.
S tatem en t 3: T h e spacetim e disp lacem en t b etw een one event on th e

w o rld lin e an d a nearby event on it specifies th e d irec tio n o f th a t w o rld
line. Reasoning: The very concept of direction implies that there exists a segment, AB,
Momenergy formula justified of the worldline short enough to be considered straight. And to fix the direction of this
spacetime displacement AB, it suffices to know the location of any two events, A and
B, on this short segment.
S tatem en t 4: W o rld lin e d irec tio n — a n d th e re fo re m om energy — is in d e

p e n d e n t o f th e m ag n itu d e o f th e spacetim e disp lacem en t. Reasoning: To
pick an event B' on the worldline half as far from A as B along the short straight
segment — thus to cut in half the spacetime displacement — makes no change in the
direction of the worldline, therefore no change in the direction of the momenergy,
therefore no change in the momenergy itself
S tatem en t 5: T h e u n it 4-v ecto r (spacetim e d isp la c e m e n t)/(p ro p e r tim e fo r

th a t displacem ent) defines an d m easures th e d irec tio n o f th e w o rld lin e
disp lacem en t an d th ere fo re th e d irec tio n o f th e m om energy 4-vector.
Reasoning: W hat matters is not spacetime displacement individually, not proper time
individually, but only their ratio. This ratio is the only directed quantity available to us
to describe the rate of motion of the particle through spacetime.
The spacetime displacement, AB, has a magnitude equal to the interval (or proper
time or wrisrwatch time) the particle requires to pass from A to B. That is why the
ratio in question is a unit 4-vector.
Proper time provides the only natural way, the only frame-independent means, to
clock the particle. If instead we should incorrectly put frame time into the
denominator— frame time measured by the array of clocks in a particular free-float
frame— the value of this time would differ from one frame to another. Divided into
the spacetime displacement, it would typically not yield a unit vector. The vector’s
magnitude would differ from one frame to another. Therefore we must use in the
denominator the proper time to go from A to B, a proper time identical to the
magnitude of the spacetime displacement AB in the numeratot.
S tatem en t 6: T h e m om energy 4-vector o f th e p artic le is
(spacetim e displacem ent)

(m om energy) = (mass) X (7 - 1 )
(proper time for that displacement)
Reasoning: There is no other frame-independent way to construct a 4-vector that lies

along the worldline and has magnitude equal to the mass.
U nits: In this book, as in more and more present-day writing, space and time
appear in the same unit: meter. Numerator and denominator on the right side of
equation (7-1) have the unit of meter. Therefore their quotient is unit-free. As a result,
Unit of momenergy: mass
the right side of the equation has the same unit as the first factor: mass. So the left side,
the momenergy arrow, must also have the unit of mass. As the oneness of spacetime is
emphasized by measuring space and time in the same unit, so the oneness of momen
ergy is clarified by measuring momentum and energy in the same unit: mass. Table
7-1 at the end of the chapter compares expressions for momentum and energy in units
of mass with expressions in conventional units.
7.3 MOMENERGY COMPONENTS AND MAGNITUDE 195
You say that the equation fo r momenergy is
(spacetim e d isp lac em e n t)

(m om energy) — (mass) X
/ thought th at “spacetime displacement’’ was the interval, which is the proper time. I
know, however, th a t I am wrong, because i f spacetime displacement an d proper time
were the same, then the numerator a n d denominator of the fraction would cancel, an d
momenergy would simply equal mass. Surely you would have told us of such simplicity.
W hat have I missed?
— It is easy to confuse a vector— or a 4-vector— with its magnimde.

'* In the expression for momenergy, the spacetime displacement is a 4-vector (Box
7-1). In the laboratory frame this displacement 4-vector has four components, [d t,
dx, dy, d z). In a free-float rocket moving in an arbitrary direction, the displacement
4-vector has four components, { d t', d x ', d y ', dz'}, typically different, respectively,
from those in the laboratory frame.
A vector in space (a 3-vector) has not only a magnitude but also a direction
independent of any coordinate system. (“Which way did they go?’’ “That-a-way!’’
— pointing.) Similarly, the spacetime displacement has a magnitude and direction
in spacetime independent of any reference frame. This sp acetim e d ire c tio n
distinguishes the 4-vector displacement (the numerator above) from its magnitude,
which is the proper time for that displacement (the denominator). This proper time
(interval) can be observed directly: it is the time lapse read off the wristwatch carried
by the particle while it undergoes the spacetime displacement.
In summary the fraction
(spacetim e d isp lacem en t)

Unit 4-vector along worldline
has a numerator that is a 4-vector. This 4-vector numerator has the same magnitude
as the denominator. The resulting fraction is therefore a u n it 4-vector pom tm g along
the worldline of the particle. This unit 4-vector determines the direction of the
particle’s momenergy in spacetime. And the magnitude of the momenergy? It is the
mass of the particle, the first term on the right of the expression at the top of this
page. In brief, the momenergy of a particle is 4-vector of magnitude m pointing
along its worldline in spacetime. This description is independent of reference frame.
7.3 MOMENERGY COMPONENTS AND

MAGNITUDE
space part: itiomentum of the object
time part: energy of the object
magnitude: mass of the object
Accidents o f history have given us not one w ord, m om energy, b u t tw o words,
m o m entum and energy, to describe mass in m otion. Before Einstein, mass and m otion
Break down momenergy for
were described not in the unified context o f spacetime b u t in term s o f space and tim e
examination
separately, as th a t division shows itself in some chosen free-float frame. O ften we still
think in those separated term s. B ut the single concept spacetim e location o f an event
unites the earlier tw o ideas o f its position in space and the tim e o f its happening. In the
same way we combine momentum and energy of a moving object into the single idea
of momenergy arrow. Having assembled it, we now break momenergy down again,
seeking new insight by examining its separate parts.
The unity of momenergy dissolves— in our thinking— into the separateness of
momentum and energy when we choose a free-float frame, say the laboratory. In rhat
laboratory frame the spacetime separation between two nearby events on the worldline
of a particle resolves itself into four different separations: one in laboratory time and
one in each of three perpendicular space directions, such as north, east, and upward.
Wirh each spacetime separation goes a separate part, a separate portion, a separate
c o m p o n e n t of momenergy in the laboratory free-float frame (Figure 7-2).
The “space parts” of momenergy of a particle are its three components of momen
tum relative to a chosen frame. Their general form is not strange to us— mass times a
velocity component. The “ time part,” however, is new to us, foreshadowing impor
tant insights into the nature of energy (Section 7.5). The four components are
/ eastward \ / eastward \
component j _ component 1
of of I
\ momenergy/ \ m om entum /
FIGURE 7-2. Momenergy arrow o f a mov (eastward displacement)
ing object translated into the language of = (mass) X
momentum and energy, shown for the spe (proper time for that displacement)
cial case in which upward momentum
{vertical momentum) equals zero. The mo northward \ / northward \
menergy arrow itself has an existence and direc ( component I _ component 1
tion (in that great haystack of worldlines and of of
events that we call spacetime) independent ofthe
choice, or even presence, of any free-float frame.
In contrast, separate measures of momentum and momenergy/ (m om entum / (northward displacement)
energy do depend on choice offrame. They point = (mass) X
parallel to, that is in the same direction as, the (proper time for that displacement)
corresponding space and time directions of the
chosenframe itself. See Figures 7-3 and 7-4 for ^ upward ^ ( upward \
a still more revealing representation of the pro component = component I
portion between momenergy and its components. of of
(mom energy/ (m om entum /
(upward displacement)
= (
/ time ^
component (time displacement)
= (energy)
( = (mass) X
of (proper time for that displacement)
\ momenergy /
The calculus version of these equations is deliciously brief. Here, as in Section 6.2, tau
( T ) stands for proper time:
dt
E—m—
dx
M om energy components of dx
particle in a given frame p - m - (7-21
dy
py = m —
dx
dz
p ,^ m
dx
7.3 MOMENERGY COMPONENTS AND MAGNITUDE 197
The components of the momenetgy 4-vectot we now have before us in simple form,
but how much is the absolutely-number-one measure of this physical quantity, its
magnitude? This magnitude we reckon as we figure the magnitude of any Lorentz
4-vector: magnitude squared is the difference of squares of the time part and the space
part:
(magnitude of momenetgy arrow)^

= (energy)^ — (east momentum)^ — (north momentum)^ — (up momentum)^
= £2 - (/-J2 _
{dt)^---------------------------------
— (dxY — {dyY ~ {dzY = ^ 2{dxY ------ = m‘-
- m‘-
kdxY KdxY
In brief, the magnitude of the momenetgy 4-vector, or its square,

Magnitude of momenergy
(magnitude of momenetgy arrow)^ = = nY (7-3)
4-vector: mass!
is identical with the particle mass, or its square. Moreover, this mass is a quantity
characteristic of the particle and totally independent of its state of motion.
It’s worthwhile to translate this story into operational language. Begin with a
particle that is at rest. Its 4-vector of energy and momentum points in the pure
timelike direction, all energy, no momentum. Let an accelerator boost that particle.
The particle acquires momentum. The space component of the 4-vector, originally
zero, grows to a greater and greater value. In other words, the momenetgy 4-vector
tilts more and more from the “vertical,” that is, from a purely timelike direction.
However, its magnitude remains totally unchanged, at the fixed value m. In conse-
SAMPLE PROBLEM 7-1 .

MA S S
The energy and momentum components of a particle, measured in the laboratory, are
£ = 6.25 kilograms
p^ = 1.25 kilograms
py = p^ = 2.50 kilograms
W hat is the value of its mass?
SOLUTION
We obtain a value for mass using equation (7-3):
= ipY)^ - {PyY - (/>.)'

= [(6.25)2 - (1.25)2 - (2.50)2 - (2.50)2] (kilograms)2
= [39.06 — 1.56 — 6.25 ~ 6.25} (kilograms)2
= [39.06 — 14.06} (kilograms)2
= 25.00 (kilograms)2
Hence
w = 5.0 kilograms
198 ch apter ; momenergy
quence, the time component of that 4-vector, that is, the energy of the particle,
undergoes a systematic alteration.
If the geometry of spacetime were Euclidean, this ever-growing tilt, this continuing
rotation of the direction of the arrow of momenergy, would cause the vertical or time
component to become ever shorter. However, spacetime is not Euclidean. It is Lo-
rentzian, as appears in the minus sign in the equation for momenergy magnirude m:
m^ — E ^— p^. W ith momenergy magnitude, or particle mass m, being constant, and
momentum p ever growing, Lorentz geomerry itself tells us that the ever-growing tilt,
E= \ 2 mass
0 the ever-larger momentum value, p, causes the time component of the momenergy
52 — rhe energy E— not to shorten, as in a Euclidean spacetime, but to lengthen as the
p = -4 8
acceleration proceeds:
E = f^ 2 0 E= + p^y^^ = an increasing function of momentum, p

25 L ^ m a s s
-15 This marvelously simple relation between energy and momentum, full of geometric as
well as physical content, has by now been tried and verified in so many thousands of
20 experiments of such varied kinds that it counts today as battle-tested.
mass
Energy, momentum, and mass, expressed so far in the language of algebra, let
themselves be displayed even more clearly in the language of pictures. Only one
20
m a s s j^ 25 obstacle stands in the way. The paper is Euclidean and the vertical leg of a right
p = 15 triangle typically is shorter than the hypotenuse. In contrast, spacetime is Lorentzian,
and the timelike dimension (the energy) is typically longer than the “hypotenuse’’ (the
mass). We are indebted to our colleague William A. Shurcliff for a way to have our
cake and eat it too, a device to employ Euclidean paper and yet display Lorentzian
length. How? By laying over the hypotenuse of the Euclidean triangle a fat line or
p = 48
h an d le of length adjusted to the appropriate Lorentzian magnitude (Figure 7-3). The
length of the handle represents the invariant value of the particle mass. This length
remains the same, whatever the values of energy and momentum, values that differ as
energy
20 the particle is observed from one or another frame of reference in relative motion.
mass Figure 7-3 shows a few of the infinitely many different values of energy and
momentum that one and the same particle can have as measured in different free-float
frames. Each arrow, being depicted on a Euclidean sheet of paper, necessarily appears
X-momentum with an apparent length that increases with slope or particle speed. The handle on the
arrow, by contrast, has the length appropriate to Lorentz geometry. This length
FIGURE 7-3. Different views o f one a n d
the same momenergy 4-vector o f a p article
represents particle mass, m = 20, a quantity independent of particle speed. The
in seven different free-float fram es. The y- momenergy 4-vector of a material particle is always timelike. Why timelike? Because
and z-components of momentum are assumed to the momenergy 4-vector lies in the same spacetime direction as the worldline of the
equal zero, and frames are chosen to give integer particle (Section 7-2). The events along the worldline have a timelike relationship;
values for energy and x-momentum components. Time displacement between events is greater than the space displacement. One
The mass of the particle equals 2 0 units as
reckoned in every free-float frame: —
p^. This invariant value of the mass is shown by energy
(single particle: values from FIGURE 7-4. Momenergy 4-vector
the thick “handle" on each vector. For a frame
several frames superposed) fo r the single p article o f Figure
in which the particle is at rest (center diagram),
7-3 as observed in seven free-float
the energy is equal to the mass and the handle
fram es, these plots then super
covers the vector.
posed on a composite momenergy
Does the momenergy 4-vectorfor this particle
diagram . Frames are chosen so that
require for its existence any reference frame? No
y- and z-components of momentum
one would laugh more at such a misapprehension
equal zero. Locus of the tips of the
than the particle! The momenergy 4-vector has
arrows traces out a hyperbola. The
an existence in spacetime independent of any
central short vertical arrow pointing to
clocks and measuring rods. We, however, wish to
the dot labeled m represents momenergy
assign to this 4-vector an energy and momentum.
as measured in the particle rest frame.
For that purpose we do require one or another
In thisframe momentum has value zero
free-float frame.
and energy— “rest energy”— equals
the mass of the particle. For clarity, the
handles have been omittedfrom the 4-
vectors, which all have identical in
variant magnitude m = 20,
7.4 MOMENTUM; "SPACE PART" OF MOMENERGY 199
consequence is that the particle moves at less than the speed of light in every possible
free-floar frame. E, p, m of particle in
The equation E?-— p^ = m^ = (constant) is the formula for a hyperbola. Figure 7-4 different frames related by
generates this hyperbola by superposing on the same figure spacetime vectors that hyperbola
represent energy and momenrum of the same particle in different free-float frames.
For visual clarity the handles are omitred from these 4-vectors. Flowever, each
momenergy 4-vector has the same magnitude, equal to the particle mass, m — 20.
7.4 MOMENTUM: "SPACE PART" OF

MOMENERGY
simply use proper time instead of Newton's
so-called "universal" time
Newton called momentum "quantity of motion.” The expressions for momentum
that spacetime physics gives us, the last three equations in (7-2), seem at first sight to
distinguish themselves by a trivial difference from rhe expressions for momentum
given to us long ago by Newton’s followers:
_ i/x _ dy dz
r X Newton ^ » P y Newton ^ I— [valid for low velocity)
dt dt
That difference? Today, proper rime dT between nearby events on the worldline of the
particle. Laboratory time, in older days, when the concepts of proper time and interval Newtonian versus relativistic
expressions for momentum
were unknown. The percentage difference between the two, trivial or even negligible
under everyday circumstances, becomes enormous when the speed of the object
approaches the speed of light.
We explore most simply the difference between relativistic and Newtonian predic
tions of momentum by analyzing a particle that travels with speed in the x-direction
only. Then the relation between displacement of this particle and its speed is x = vt.
For small displacements, for example between two nearby spark events on the
worldline, this becomes, in the mathematical limit of interest in calculus notation, dx
= vdt.
The proper time between the two nearby sparks is always less rhan rhe laboratory
time:
dT = [(^t ) 2 } ‘/ 2 = [{dty - (dxyy/^ = [{dty - (pdtyy/^

dt
= {dt)(i — (7 - 4 )
y
where gamma, y = 1/(1 ~ is the time stretch factor (Section 5.8). This figure
for the interval, or proper time, between the two nearby sparks we now substitute into
equations (7-2) in order to learn how the relativistic expressions for energy and
momentum depend on particle speed:
dt m
E = m— = --------------- my
dx (1 — (7 - 5 )
dx m {dx/dt) mv^
- mv;y
dx ( 1 - *^2)1/2
200 CHAPTER 7 MOMENERGY
The momentum expression is the same as for Newtonian mechanics — mass m

times velocity (dx/dt) — except for the factor — in the denominator. That
Low sp eed : Newton and Einstein
factor we can call 1 when the speed is small. For example, a commercial airliner moves
a g re e on value of momentum
through the air at approximately one millionth of the speed of light. Then the factor (1
— differs from unity by only five parts in 10*^. Even for an alpha particle
(helium nucleus) ejected from a radioactive nucleus with approximately 5 percent of
the speed of light, the correction to the Newtonian figure for momentum is only a little
more than one part in a thousand. Thus for low speeds the momentum expressed in
equation (7-5) reduces to the Newtonian version.
High sp eed : Relativity reveals
At a speed close to that of light, however, the particle acquires a momentum
much larger momentum
enormous compared with the Newtonian prediction. The unusually energetic cosmic-
ray protons mentioned at the end of Section 5.8 crossed the Milky Way in 30 seconds
of their own time, but a thousand centuries or 3 X 10*^ seconds of Earth time. The
ratio d t/d x between Earth time and proper rime is thus 10". That is also the ratio
between the correct relativistic value of the protons’ momentum and the Newtonian
prediction.
U nits: Both Newtonian and relativistic expressions for momentum contain speed,
a ratio of distance to time. From the beginning we have measured distance and time in
the same unit, for example meter. Therefore the ratio of distance to time is unit-free. In
Secrion 2.8, we expressed speed as a dimensionless quantity, the fraction of light
speed:
(meters of distance covered by particle)

V— ■
(meters of time required to cover that distance)
Unit of momentum: mass
(patticle speed in meters/second) _
(7-6)
(speed of light in meters/second) c
In terms of speed v (called beta, fi, by some authors), Newtonian and relativistic
expressions for rhe magnitude of the momentum have the forms
/'Newton - [valid for low speed] (7-7)

p — m v/{\ — [good at any speed] (7-8)
M ore U nits: In otder to convert momentum in units of mass to momentum in

conventional units, such as kilogram meters/second, multiply expressions (7-6),
(7-7), and (7-8) by the speed of light c and use the subscript “conv” for “conven
tional” :
/'conv Newton '■/'Newton c — m v c - m (v,^Jc) c — mv^ [low speed] (7-9)
Conversion to conventional mvc

__________ '^(Nonv/^)^
momentum units /'conv H -^ 2 y /2 [1 - ( ^ ^ _ / ^ ) 2 ] l / 2
[any speed] (7-10)

a - (Nonv/^)^}'/^
Thus conversion from momentum in units of mass to momentum in conventional

units is always accomplished by multiplying by the conversion factor c. This is true
whether rhe expression for momentum being converted is Newtonian or relativistic.
Table 7-1 at the end of the chapter summarizes these comparisons.
7.5 ENERGY: "TIME PART" OF MOMENERGY 201
7.5 ENERGY: "TIME PART" OF

MOMENERGY
energy has two parts: rest energy (= mass)
plus kinetic energy
W hat about the ‘‘time part” of the m omentum-energy of a particle — the part we
have called its energy? This is certainly a strange-looking beast! As measured in a
particular free-float frame, say the laboratory, this time component as given in
equation (7-5) is
dt m Relativistic expression for energy

E = m — — ---------TTrr ~ m j (7-11)
dx
Compare this with the Newtonian expression for kinetic energy, using K as the symbol
for kinetic energy:
1
^Newton ~ [valid for low speed) (7-12)
How does the relativistic expression for energy, equation (7-11), compare with the
Newtonian expression for kinetic energy (7-12)? To answer this question, first look at
the behavior of these two expressions when particle speed equals zero. The Newtonian
kinetic enetgy goes to zero. In contrast, at zero speed 1/(1 — = 1 and the
relativistic value for energy becomes equal to mass of the particle.
Rest energy of o particle
E^ = m (7-13)
equals its mass
where E ^ is called rest energy o f th e p article. Rest energy of a particle is simply its
mass. So the relativistic expression for energy does not go to zero at zero speed, while
the Newtonian expression for kinetic energy does go to zero.
Is this an irreconcilable difference? The Newtonian formula does not contain an
expression for rest energy, equal to the mass of the particle. But here is the distinction:
The telativistic expression gives the value for total energy of the particle, while the
Newtonian expression describes kinetic energy only (valid for low speed). However, in
Newtonian mechanics any constant potential energy whatever can be added to the
energy of a particle without changing the laws that describe its motion. One may think
of the zero-speed limit of the relativistic expression for energy as providing this
previously undetermined constant.
When we refer to energy of a particle we ordinarily mean total energy of the particle.
As measured in a frame in which the particle is at rest, this total energy equals rest
energy, the mass of the particle. As measured from frames in which the particle moves,
total energy includes not only rest energy but also kinetic energy.
This leads us to define kinetic energy of a particle as energy above and beyond its
rest energy:
(energy) = (rest energy) -f (kinetic energy) Kinetic energy defined
or
E = m ^K (7-14)
S A M P L E P R O B L E M 7- 2
M O T I O N IN THE X - D I R E C T I O N
An object of mass 3 kilograms moves 8 meters W hat value of kinetic energy would Newton pre-
along the x-direction in 10 meters of time as mea- diet for this object? Using relativistic expressions,
sured in the laboratory. W hat is its energy and verify that the velocity of this object equals its
momentum? Its rest energy? Its kinetic energy? momentum divided by its energy.
SOLUTION
From the statement of the problem:
m= i kilograms
r= 10 meters
X= 8 meters
y— 0 meters
2= 0 meters
From this we obtain a value for the speed:
X 8 m e t e r s o f d is t a n c e
v = - = -----------------;------- = 0.8
t 10 meters of time
Use V to calculate the factor 1/(1 — in equation (7-8):
1 1 1 1
( l _ j ,2 ) i/ 2 (i - ( o .8)2)V2 (1 - 0 .6 4 )1 /2 (0.36)1/2 0.6 3
From equation (7-11) the energy is
E - m /{\ — t'2)i/2 = (3 kilograms) (5 /3 ) = 5 kilograms
From equation (7-8) momentum has the magnitude
p = m v/{\ — r'2)V2 = (5 /3 ) X (3 kilograms) X 0.8 = 4 kilograms
Rest energy of the particle just equals its mass:
^cest ~ m = i kilograms
From equation (7-15) kinetic energy K equals total energy minus rest energy:
K — E — m = 5 kilograms — 3 kilograms = 2 kilograms
The Newtonian prediction for kinetic energy is
^Newton = “ 2 ^ ^^ 0 96 kilogram
which is a lot smaller than the correct relativistic result. Even at the speed of light, the
Newtonian prediction would be ^Newton “ 1-5 kilogram, whereas relativistic value would
increase without limit.
7.5 ENERGY: "TIME PART" OF MOMENERGY 203
Equation (7-16) says that velocity equals the ratio (magnitude of momentum)/(en-
ergy):
p 4 kilograms
v = - = ---- — ------= 0.8
E 5 kilograms
This is the same value as reckoned directly from the given quantities.
From this comes the relativistic expression for kinetic energy K:
K= E—E^ = E—m
m r 1
---------------m = m \ ------------------- 1
1 (7-15)
(1 - J
Box 7-2 elaborates the relation between this expression and the Newtonian expres
sion (7-12). Notice that if we divide the respective sides of the momentum equation
(7-8) by corresponding sides of rhe energy equation (7-11), the result gives particle
speed:
P
V ----- (7-16)
E
We could have predicted this direaly from the first figure in this chapter. Figure 7-1.
Speed V is the tilt (slope) of the worldline from the vertical: (space displacement)/(time
for this displacement). Momenergy points along the worldline, with space componentp
and rime component £. Therefore momenergy slope p/E equals worldline slope v.
Still M ore U nits: In order to convert energy in units of mass to energy in

conventional units, such as joules, multiply the expressions above by rhe square of
light speed, c^, and use subscript “conv” :
mc^
f = P f2 — — [good at any speed] {7~ 17)
[1 - (t'ccnvA)^]*/^
'^conv rest [particle at rest] (7-18)
Conversion to conventional
energy units
[,I [1 _ ']
= (£ - E ^)c^ = (^_ /,)2 ]V 2 [good at any speed] (7-19)
_ 1 . . _ 1 , _ 1 mvp.
K,conv Newton ----- m v^c^------m \ ------\ ----- [low speed only] (7-20)
2 2 \ c } 2
Thus conversion from energy in units of mass to energy in conventional units is always
accomplished by multiplying by conversion factor (P. This is true whether rhe expres
sion for energy being converted is Newtonian or relativistic. T able 7-1 at the end of the
chapter summarizes these comparisons.
Equation (7-18) is the most famous equation in all physics. Historically, the factor
(P captured the public imagination because it witnessed to the vast store of energy
available in rhe conversion of even tiny amounts of mass to heat and radiation. The
units of truP are joules; the units of m are kilograms. However, we now recognize that
joules and kilograms are units different only because of historical accidenr. The
SAMPLE PROBLEM 7- 3
MOMENERGY COMPONENTS
For each of the following cases, write down the vector in the given frame in the form [£, p^, p^, p^].
four components of the momentum-energy 4- Each particle has mass m.
a. A particle moves in the positive x-direction in the laboratory with kinetic energy
equal to three times its rest energy.
b. The same particle is observed in a rocket in which its kinetic energy equals its
mass.
c. Another particle moves in the y-direction in the laboratory frame with momen
tum equal to twice its mass.
d. Yet another particle moves in the negative x-direction in the laboratory with total
energy equal to four times its mass.
e. Still another particle moves with equal x, y, and z momentum components in the
laboratory and kinetic energy equal to four rimes its rest energy.
SOLUTION
a. Total energy of the particle equals rest energy m plus kinetic energy 3w. Therefore
its total energy E equals E = m 5m = 4m. The particle moves along the
x-direction, sopy=p^ — 0 and p ^—p, the total momentum. Substitute the value
of E into the equation m^ = E^ — pE to obtain
pE = £2 m2 = {4m)^ - m ^ = \6m^ m-2 =
Hence p^ — {DY^^m.
In summary, the components of the momenergy 4-vector are
[£, p^, py, p j = [4m, (15)*/^w, 0, 0]
O f course the magnitude of this momenergy 4-vector equals the mass of the
particle m — true whatever its speed, its energy, or its momentum.
b. In this rocket frame, total energy— rest energy plus kinetic energy— has the
value £ = 2m. As before, p'^ = E'^ — rrP- — {2m)^ — m2 = 4nP- — rrP- = 5m2 .
Hence p ^ ~ ^ and components of the 4-vector are [£, p,,, py, p ^ = [2m,5
0 , 0].
c. In this caseP x ~ P z ~ ^ andp y= p = 2m . Moreover, E? = -\-p'^ = m^ 4r {2mY
= 'bmP. So, finally, [£, />,, py, pji = [5'^^^m, 0, 2m, 0].
d. We are given directly that £ = 4m, the same as in part a, except here the particle
travels in the negative x-direction so has negative x-momentum. Hence:
[£, p„, py, p 2 = [4m, — (15)*/2«, 0, 0}

e. Total energy equals £ = bm. All momentum components have equal value, say
P .= P y = P z = P
In this case we use the full equation that relates energy, momentum, and mass:
{px)^ + ipy)^ + ipz)^ ~ = E^ — m^ — (5m)^ — m^ = 24mP
or = 8m^ and hence [£, px, py, p ^ = [bm, W-^'^m,

OF MOMENERGY 205
ENERGY IN THE LOW-VELOCITY LIMIT

Energy at relativistic speeds and energy at everyday speeds: How are ex
pressions for these two cases related?
Energy in Terms of Momentum: In the limit of velocities low compared with

the speed of light, the relativistically accurate expression for energy E = (m^
+ reduces to E = m -h p^/(2m) -I- corrections. To see why and how, and to
estimate the corrections, it is convenient to work in dimensionless ratios. Thus
we focus on the accurate expression in the form E/m = [1 + (p/m)^]''^, or even
simpler, y = [1 + x]’'^, and on the approximation to this result, in the form
E/m = 1 -h (1 /2) Ip /m F + corrections, or y = 1 -h (1 /2) x -h corrections
Example: x = 0.21. Then our approximation formula givesy = (1.21 )'^^= 1

-hO. 105 + a correction. The accurate result is y = 1.100, which is the square
root of 1.21. In other words, the correction is negative and extremely
small: correction = —0.005.
Energy in Terms of Velocity: In the limit of velocities low compared with the
speed of light, the relativistically accurate expression for energy E = m/( 1 —
reduces to E = m + ( l / 2 ) m v ^ -t- corrections. It is convenient again to
work in dimensionless ratios. Thus we focus on the accurate expression in the
form E/m = [1 — or even simpler, y = [l — x]~''^, and on the approxi
mation to this result, in the form
E/m = 1 -h (1 /2) v^ -h corrections, or y = 1 + (M2) x + corrections
Example: x = 0.19. Then our approximation formula gives y = 1 -h (1 /2)

0.19 -h a correction == 1.095 + a correction. The accurate result is y = [1 —
0.19]"''^ = (0.81 = (0.9)“ ’ = 1.1 111 . . . In other words, the correc
tion is positive and small: correction = -1-0.0161 1.
Another example: A jet plane. Take its speed to be exactly v = 10“ *. That
speed, according to our approximation, brings with it a fractional aug
mentation of energy, a kinetic energy per unit mass, equal to (1 /2)v^ = 5 X
10“ '^ or 0.000 000 000 000 5. In contrast, the accurate expression E/m =
[ 1 - v2]- ’'2 gives the result E/m = 1.000 000 000 000 500 000 000 009 375
000 000 000 . . . The 5 a little less than halfway down the length of this
string of digits is no trifle, as anyone will testify who has seen the conse
quences of the crash of a jet plane into a skyscraper. However, the 9375
further down the line is approximately a million million times smaller and
totally negligible in its practical consequences.
In brief, low speed gives rise to a kinetic energy which, relative to the mass, is
given to good approximation by (1/2) v^ or by (1/2) (p/m)^. Moreover, the
same one or other unit-free number (a “ fraction” because it is small com
pared to unity) automatically reveals to us the order of magnitude of the
fractional correction we would have had to make in this fraction itself if we
were to have insisted on a perfectly accurate figure for the kinetic energy.
FIGURE 7-5. Kinetic energy as a

function of speed, as predicted by
relativity [equation (7 -1 9 ), valid
fo r a ll speeds] a n d by Newtonian
mechanics [equation (7 -2 0 ), v alid
fo r low speeds only].
^ ~ ''conv
conversion factor c^, like the factot of conversion from seconds to meters or miles to
feet, can today be counted as a detail of convention rather than as a deep new principle.
Central to an understanding of the equation — m ot its equivalent -m r
is the subscript “rest.” Energy is not the same as mass! Energy is only the time part of
the momenergy 4-vector. Mass is the magnitude of that 4-vector. The energy of an
Energy: Time part of momenergy
4-vector
object, expressed in conventional units, has the value mtf only when that object is
M ass: M agnitude of that 4-vector observed from a frame in which it is at rest. Observed from all other free-float frames,
the energy of the object is greater than its rest energy, as shown by equation (7-17).
Eigure 7-5 compares relativistic and Newtonian predictions for kinetic energy per
unit mass as a function of speed. At low speeds the values are indistinguishable (left
side of the graph). When a particle moves with high speed, however, so that the factor
l/( 1 ~ has a value much greater than one, relativistic and Newtonian expres
sions do not yield at all the same value fot kinetic energy (right side of the graph). Then
one must choose which expression to use in analyzing collisions and other high-speed
phenomena. We choose the relativistic expression because it leads to the same value of
the total energy of an isolated system before and after any interaction between particles
in the system — it leads to conservation of total energy of the system.
All this talk of reconciliation at low speeds obscures an immensely powerful feature
of the relativistic expression for total energy of an isolated system of particles. Total
Relativity: All forms of energy
energy is conserved in all interactions among particles in the system: elastic and
automatically conserved
inelastic collisions as well as creations, transformations, decays, and annihilations of
particles. In contrast, total kinetic energy of a system calculated using the Newtonian
formula for low-speed interactions is conserved only for elastic collisions. Elastic
collisions are defined as collisions in which kinetic energy is conserved. In collisions that
are not elastic, kinetic energy transforms into heat energy, chemical energy, potential
energy, or other forms of energy. For Newtonian mechanics of low-speed particles,
each of these forms of energy must be treated separately: Conservation of energy must
be invoked as a separate principle, as something beyond Newtonian analysis of
mechanical energy.
In relativity, all these energies are included automatically in the single time compo
nent of total momenergy of a system — total energy — which is always conserved for
an isolated system. Chapter 8 discusses more fully the momenergy of a system of
particles and the effects of interactions between particles on the energy and mass of the
system.
7.6 CONSERVATION OF MOMENERGY AND ITS CONSEQUENCES 207
7.6 CONSERVATION OF MOMENERGY

AND ITS CONSEQUENCES
total memenergy of an isolated system of
particles is conserved
Momenergy puts us at the heart of mechanics. The relativity concept momenergy gives
us the indispensable tool for mastering every interaction and transformation of parti
cles.
W hat does it mean in practice to say in this language of momenergy components
that the punch given to particle A by particle B in a collision is exactly equal in
magnitude and opposite in spacetime direction to the punch given to B by A? That
gain in momenergy of A is identical to loss of momenergy by B? That the sum of
separate momenergies of A and B — this sum itself regarded as an arrow in spacetime,
the arrow of total momenergy (Figure 7-6) — has the same magnitude and direction
after the encounter that it had before? Or, in brief, how does the p rin c ip le o f
co n serv atio n o f m om energy translate itself into the language of components in a
-6 momentum
M om energy of a system of
particles
(before onrfafter!)
8 12
mass
8
mass
12
mass
mass
Q:= 15/17 o ^= -6/10 o a 16/20

-5/13
FIGURE 7-6. Conservation o f to tal momenergy in a collision. Before: The lighter 8-unit mass,
moving right with V) j 11 light speed, collides with the slower and heavier 12-unit mass moving left (with 5
units of momentum to the left and 13 units ofenergy). System: Arrow oftotal momenergy ofthe system oftwo
particles. Combined momentum of the colliding particles has value —5 -h = JO units rightward.
Combined energy of the two equals 13A -17 = 3 0 units. The total system momenergy is conserved. After: One
of many possible outcomes of this collision: The 8-unit mass bounces back leftward after collision, but the
punch that it provided has reversed the direction of motion and increased the speed of the heavier 12-unit
mass. The handle of the momenergy arrow of each particle gives the true magnitude of that momenergy,
figured in the Lorentz geometry of the real physical world, as contrasted to the length of that 4-vector as it
appears in the Euclidean — and therefore misleading— geometry of this sheet of paper. The scale of
magnitudes in this figure is different from that of Figure 7-3.
given free-float frame? Answer: Each component of the momenergy vector, when added
together for particles A and B, has the same value after the collision as before the
collision. In other words,
'energy of A \ / energy of 6'' to ta l energy\

before the 1 + I before the ( b efo re the 1
, encounter / \ encounter > encounter /
to tal energy\
Energy of system conserved ( afte r the 1
encounter /
energy of A \ / energy of
( after the 1 "I" I ^fter the
encounter / \ encounter y
called conservation of the time part ofmomenergy. Add to this three statements about the
three space components of momenergy, of which the first one reads.
INVARIANT? CONSERVED? CONSTANT?

Is the speed of light a constant? An invariant? Is mass conserved in a collision?
Is it an invariant? A constant? Many terms from everyday speech are taken
over by science and applied to circumstances far beyond the everyday. The
three useful adjectives in v a ria n t, c o n s e r v e d , and c o n s t a n t have distinct mean
ings in relativity .
Invariant
In relativity a quantity is invariant if it has the same value when measured by
observers in different free-float frames — frames in relative motion. First
among relativistic invariants is the speed of light: It has the same value when
reckoned using data from the laboratory latticework of recording clocks as
when figured using data from the rocket latticework. A second central invar
iant is the interval between two events: All inertial observers agree on the
interval (proper time or proper distance). A third mighty invariant is the mass
of a particle. There are many other invariants, every one with its special
usefulness.
Some very important quantities do n o t qualify as invariants. The time between

two events is not an invariant. It differs as measured by observers in relative
motion. Neither is the distance between events an invariant. It too differs
from one frame to another. Neither the energy nor the momentum of a
particle is an invariant.
Conserved
A quantity is conserved if it has the same value before and after some
encounter or does not change during some interaction. The total momenergy
of an isolated system of particles is conserved in an interaction among the
particles. In a given free-float frame this means that the total energy is
conserved. So is each component of total momentum. The magnitude of total
momenergy of a system — the mass of that system — is also conserved in an
interaction. On the other hand, the sum of the individual masses of the
7.6 CONSERVATION OF MOMENERGY AND ITS CONSEQUENCES 209
f eastward component \ / eastward component \

of momentum of 1^1 momentum of B I
^ before the encounter / \ before the encounter /
eastward component \
of total momentum I
before the encounter /
Momentum of system conserved
' eastward component \
of total momentum I
after the encounter /
^ eastward component eastward component''
of momentum of A ( of momentum of B
\ after the encounter after the encounter /
called conservation of the space part of momenergy. Figure 7-6 illustrates the conserva
tion of momenergy in a recoil collision between two particles. Momentum is laid out
constituent particles of a system ordinarily is not conserved in a relativistic

in te ra ctio n . (Fo r e x a m p le s , s e e C h a p t e r 8 .)
Constant
Something that is constant does not change with time. The speed of the Great
Pyramid with respect to the rock plateau of Giza is constant — equal to zero,
or at least less than one millimeter per millennium. This speed may be con
stant, but it is not an invariant: As observed from a passing rocket, the Great
Pyramid moves with blinding speed! Is the speed of the Great Pyramid con
served? Conserved during what encounter? There is no b e f o r e o r a f t e r to
which the term “ conserved” can refer. The term “ conserved” simply does
not apply to the speed of the Great Pyramid.
It is true that the speed of light in a vacuum is constant — it does not change
with time. It is also true, but an entirely different statement, that the speed of
light is an invariant— has the same value measured by different observers in
uniform relative motion. It is true that total momenergy of an isolated system
is constant — does not change with time. It is also true, but an entirely differ
ent statement, that total momenergy of an isolated system is conserved in a
collision or interaction among particles in that system.
When anyone hears the word in v a ria n t, c o n s e r v e d , or c o n s t a n t ,

she is well-advised to listen for the added phrase with r e s p e c t to,
which should always be expressed or implied. Usually (but not
always) c o n s t a n t means with r e s p e c t to the passage of time. C o n
s e r v e d usually (but not always) means with r e s p e c t to a collision or
interaction. I n v a r ia n t can have at least as many meanings as there
are geometries to describe Nature: In Euclidean geometry, d i s
t a n c e is invariant as measured with r e s p e c t to relatively rotated
coordinate axes. In Lorentz geometry, in t e r v a l and m a s s are in
variants as measured with r e s p e c t to free-float frames in relative
motion. The full meaning of the word in v a r ia n t or c o n s e r v e d or
c o n s t a n t depends on the condition under which this property is
invoked.
210 CHAPTER/ M0MENER6Y
right and left on the page; energy is marked off vertically. The left diagram shows two
particles before collision and their momentum-energy vectors. The right diagram
shows the corresponding display after the collision.
The center diagram shows total momenergy of the system of two particles. The
momenergy vectors of the two particles hfore the collision add up to this total; the
M om energy of system conserved!
momenergy vectors of the two particles after the collision add up to the same total.
Total momenergy of the system has the same value after as before: it is conserved in the
collision.
Well, you’ve done it again: You’ve given us a powerful tool that seems impossible to
visualize. How can one think about this momenergy 4-vector, anyway? Can you
personally picture it in your mind’s eye?
We can almost visualize the momenergy arrow, by looking at Figure 7-6 for
example. There momentum and energy components of a given momenergy vector
have their correct relative values. And the direction of the momenergy arrow in
spacetime is correctly represented in the diagram.
However, the magnitude of this arrow— mass of the particle— does not corre
spond to its length in the momenergy diagram. This is because mass is reckoned
from the difference of squares of energy and momentum, whereas length of a line on
the Euclidean page of a book is computed from the sum of squares of horizontal and
vertical dimensions. The handle or thickened region on the typical arrow and the big,
boldface number for mass remind us of the failure— the lie— that results from
trying to represent momenergy on such a page.
To observe a given momenergy 4-vector first from one free-float frame, then from
another, and then from another (Figure 7-3) is to see the apparent direction of the
arrow changing. The change in frame brings with it changes in the energy and
momentum components. However, magnitude does not change. Mass does not
change. To examine the momenergy 4-vector of a particle in different frames is to
gain improved perspective on what momenergy is and does.
See if this analogy helps: The momentum-energy 4-vector is like a tree. The tree
has a location for its base and for its tip whether or not we choose this, that, or the
other way to measure it. The shadow the tree casts on the ground, however, depends
upon the tilt of the tree and the location of Sun in the sky.
Likewise, momenergy of a particle as it passes through a given event on its
worldline has a magnitude and direction, a fixity in spacetime, independent of any
choice we make of free-float frame from which to observe and measure it. No means
of reporting momenergy is more convenient for everyday purposes than separate
specification of momentum and energy of the object in question in some chosen
free-float frame. Those two quantities separately, however, are like the shadow of the
tree on the ground. As Sun rises the shadow shortens. Similarly the momentum of a
car or spaceship depends on the frame in which we see it. In one frame, terrifying. In
another frame, tame. In a comoving frame, zero momentum, as the tree's shadow
disappears when Sun lies in exactly that part of the sky to which the tilted tree points.
In such a special frame of reference, the time component of an object’s momenergy
— that is, its energy — takes on its minimum possible value, which is equal to the
mass itself of that object. However, in whatever free-float frame we observe it, the
arrow of momenergy clings to the same course in spacetime, maintains the same
length, manifests the same mass.
7.7 SUMMARY 211
7 .7 SUMMARY
momenergy of an object unifies energy,
momentum, and mass
The m o m en erg y 4-vector of a particle equals its mass multiplied by the ratio of its
spacetime displacement to proper time— wristwatch time — for that displacement
(Section 7.2):
I spacetim e \
disp lacem en t
/m o m e n e rg y \ ^ 4-vector
(7-1)
V 4 -v e cto r / proper time
for that
\ displacement /
Momenergy of a particle is a 4-vector. It possesses magnitude equal to the particle’s

mass. The momenergy at any given event in the motion of the particle points in the
direction of the worldline at that event (Section 7.2).
The momenergy of a particle has an existence independent of any frame of refer
ence.
The terms momenergy, momentum, and energy, as we deal with them in this book,
all have a common unit: mass. In older times mass, momentum, and energy were all
conceived of as different in nature and therefore were expressed in different units. The
conventional units are compared with mass units in Table 7-1.
The magnitude of the momenergy 4-vector of a particle is reckoned from the
difference of the squares of energy and momentum components in any given frame
(Section 7.3):
iPy? - ip .?
or, more simply.
( E y - ( / '\2
) (7-3)
Mass m of the particle is an invariant, has the same numerical value when computed
using energy and momentum components in the laboratory frame (unprimed compo
nents) as in any rocket frame (primed components).
In a given inertial frame, the momenergy 4-vector of a particle has four co m p o
nents. Three space components describe the momentum of the particle in that frame
(Sections 7.3 and 7.4):
dx
p^ — m —
dx
dy
(7-2)
dx
dz
Pz = m-
dx
The magnitude of the momentum can be expressed as the factor 1/(1 - times
the Newtonian expression for momentum mv. The result is
p — mv/(\ — (7- 8 )
The “ time part” of the momenergy 4-vector in a given inertial frame equals energy of
the particle in that frame (Sections 7.3 and 7.5):
dt m
E= m— = ■ (7-2), (7-11)
dX (l-t^2)l/2
For a particle at rest, the energy of the particle has a value equal to its mass:
E ^, = m (7-13)
For a moving particle, the energy combines two parts: rest energy — equal to mass of
the particle— plus the additional kinetic energy K that the particle has by virtue of its
motion:
E ^. + K- m (7-14)
From these equations comes an expression for k in etic energy:
K-= E — m — m \ ---------------— 1 (7-15)

[(1 - J
The momenergy 4-vector derives from conservation its power to analyze particle
interactions. Conservation states that the total momenergy 4-vecror of an isolated
system of particles is conserved, no matter how particles in the system interact with one
another or transform themselves. This conservation law holds independent of choice of
the free-float frame in which we employ it (Section 7.6).
In any given inertial frame, conservation of total momenergy of an isolated system
breaks apart into four conservation laws:
1. Total energy of the system before an interaction equals total energy of the
system after the interaction.
Velocity from time of flight; Useful in analysis of collisions

energy from conservation when velocity is not of interest
low applied to previous or and attention is focused
subsequent collisions G ives p or V on testing or applying
or E when other conservation law s
two are known
and m is not
of interest
p = m v/(l - v^)
Velocity from time of flight; momentum from

bending of particle track in magnetic field
FIGURE 7-7. Form ulas th a t relate momentum, energy, mass, a n d velocity of a n object, a n d notes
about th eir uses in analyzing experiments. In this diagram, p is the magnitude of the momentum.
ACKNOWLEDGMENT 2 13
< Ij ^ b l e 7 -7 2 ^
QUANTITIES RELATING TO MOMENERGY
In units of mass In conventional units
(for example, E and p (for example, in joules,
both in kilograms; Reference p ^ ^ in kilogram meters/second;
X, y, z, t, T in meters) equations
dt m(P
Energy E = m— ■ (7-2, 5, 11, 17) E = ■
dz (1 - t/2)V2 a -
Rest energy 'rest (7-13, 18) F
*-'cnnv re«r = tnc^
Kinetic energy K = m ( ------ ---------- l ) (7-15, 19) . = m p i---------------------- 1 I

\ t l - /
Momentum p = (1 - t^2)l/2 (7-8, 10)

[1 - (v_A)^]‘/2
dx ”2^«eon»
Momentum components (7-2, 5)
dx (1 - a - (t'convA)^]'/^
dy ______ conv___________
P ,- m — = (7-2, 5) Pyconv
dx (1 - (1 - (*'convA)^)‘^^
dz mv. >”^'eeonv
p, — m — = (7-2, 5) Pz conv
^ dx (1 - t/2)i/2 (1 - (t^_A)^)>/2
Mass nP = E^~ p^ (7-3)
Particle speed _P
V---- (7-16)
E
Newtonian low-speed limit
Kinetic energy f^Newton ^ (7-12, 20) IT __ _ 2
^ c o n v Newcon ,2 ^ ^ ^ conw
Momentum pNewton XnV (7-7, 9) r c o n v Newton " ‘^^conv

Momentum components Px Newton ~ x Pxconv Newton conv
Pyconv Newton fflV y conv

PxNewton VtV^ conv Newton ^ ^ 2 conv
2 . Total x-momentum of the system is the same before and after the interaction.
3. Total y-momentum of the system is the same before and after the interaction.
4. Total z-momentum of the system is the same before and after the interaction.
In this chapter we have developed expressions that relate energy, momentum, mass,
and velocity. Which of these expressions is useful depends upon circumstances and the
system we are trying to analyze. Figure 7-7 summarizes these equations and circum
stances under which they may be useful. Table 7-1 compares energy and momentum
in units of mass and in conventional units,
ACKNOWLEDGMENT
The authors are grateful to William A. Shurcliff for the idea of a “handle” that
displays on a Euclidean page the invariant magnitude of a particle’s momenergy
4-vector — a magnitude equal to the mass of the particle.
214 EXERCISE7-I MOMENERGV 4-VECTOR
CHAPTE3R 7 EXERCISES
PRACTICE 8 there is a lot more discussion about the mass of a

system of particles.)
7-1 momenergy 4-veclor

7-3 much ado about little
For each of the following cases, write down the four
components of the momentum-energy (momen Two freight trains, each of mass 5 X 10® kilograms
(5000 metric tons) travel in opposite directions on the
ergy) 4-vector in the given frame in the form {E,p^,py,
same track with equal speeds of 42 meters/second
p ^. Assume that each particle has mass m .Y o n may
(about 100 miles/hour). They collide head on and
use square roots in your answer.
come to rest.
a A particle moves in the positive x-direction in
the laboratory with total energy equal to five times its a Calculate in milligrams the kinetic energy for
rest energy. each train (1 /2)m v^ before the collision. (Newtonian
b Same particle as observed in a frame in which it expression OK for 100 mph!) (1 milligram = 10~^
gram = 10~® kilogram)
is at rest.
c Another particle moves in the z-direction with b After the collision, the mass of the trains plus
the mass of the track plus the mass of the roadbed has
momentum equal to three times its mass.
increased by what number of milligrams? Neglect
d Yet another particle moves in the negative
energy lost in the forms of sound and light.
y-direction with kinetic energy equal to four times its
mass.
e Still another particle moves with total energy 7-4 fast protons
equal to ten times its mass and x-, y-, and z-compon- Each of the protons described in the table emits a flash
ents of momentum in the ratio 1 to 2 to 3. of light every meter of its own (propet) time dT.
Between successive flash emissions, each proton trav
7-2 system mass els a distance given in the left column. Complete the
Determine the mass of the system of particles shown table. Take the rest energy of the proton to be equal to
in Figure 7-6. Is this system mass equal to the sum of 1 GeV = 10’ eV and express momentum in the
the masses of the individual particles in the system? same units. Hints: Avoid calculating or using the
Does the mass of this system change as a result of the speed V in relativistic particle problems; it is too close
interaction? Does the momenergy 4-vector of the sys to unity to distinguish between protons of radically
tem change as a result of the interaction? (In Chapter different energies. An accuracy of two significant fig-
-< C l^ j^ X E R C IS E 7 -4 ^ ^
FAST PROTONS
Lab distance
Ax traveled Lab time
between flashes Momentum mdx/dx Energy Time stretch between flashes
(meters) (GeV) (GeV) factor y (meters)
0.1
10
103
10^
EXERCISE 7 -7 SUPER COSMIC RAYS 2 15
ures is fine; don’t give more. Recall: = nE 7-6 fast electrons

and E = mdt/d% = my [note tau!}.
The Two-Mile Stanford Linear Accelerator accelerates
electrons to a final kinetic energy of 47 GeV (47 X
10^ electron-volts; one electron-volt = 1.6 X 10“ ^^
PROBLEMS joule). The resulting high-energy electrons are used
for experiments with elementary particles. Electro
magnetic waves produced in large vacuum tubes
7-5 Lorentx transformation for (“klystron tubes’’) accelerate the electrons along a
momenergy components straight pipelike structure 10,000 feet long (approxi
The rocket observer measures energy and momentum mately 3000 meters long). Take the rest energy of an
components of a particle to have the values E' and p j , electron to be w ^ 0.5 MeV = 0.5 X 10^ electron-
py , and p^'. W hat are the corresponding values of volts.
energy and momentum measured by the laboratory a Electrons increase their kinetic energy by ap
observer? The answer comes from the Lorentz trans proximately equal amounts for every merer traveled
formation, equation (L-10) in the Special Topic fol along the accelerator pipe as observed in the labora
lowing Chapter 3. tory frame. W hat is this energy gain in MeV/meter?
The moving particle emits a pair of sparks closely Suppose the Newtonian expression for kinetic energy
spaced in time as measured on its wristwatch. The were correct. In this case how far would the electron
rocket latticework of clocks records these emission travel along the accelerator before its speed were equal
events; so does the laboratory latticework of clocks. to the speed of light?
The rocket observer constmcts components of particle b In reality, of course, even the 47-GeV elec
momentum and energy, equation (7-2), from knowl trons that emerge from the end of the accelerator have
edge of particle mass m, the spacetime displacements a speed v that is less than the speed of light. W hat is
d t ', d x ',d y ', and d z' derived from the event record the value of the difference {I ~ p) between the speed
ings, and the proper time dx computed from these of light and the speed of these electrons as measured in
spacetime components. Laboratory momenergy com the laboratory frame? [Hint: For v very near the value
ponents come from transforming the spacetime dis unity, 1 (1 + v){\ - v ) ^ 2(1 - v).-\ Let
placements. The Lorentz ttansformation, equation a 4 7-GeV electron from this accelerator race a flash of
(L-10), for incremental displacements gives light along an evacuated tube straight through Earth
from one side to the other (Earth diameter 12,740
dt = vydx' + ydt' kilometers). How far ahead of the electron is the light
dx = ydx' -b vydt' flash at the end of this race? Express your answer in
dy = dy' millimeters.
dz = dz' C How long is the “ 3000-meter’’ accelerator
tube as recorded on the latticework of rocket clocks
a Multiply both sides of each equation by the moving along with a 47-GeV electron emerging from
invariant mass m and divide through by the invariant the acceletatot?
proper time dx. Recognizing the components of the
momenergy 4-vector in equation (7-2), show that the
transformation equations for momenergy are 7-7 super cosmic rays
The Haverah Park extensive air shower array near
E = vyp'^ -b yE'
Leeds, England, detects the energy of individual cos
p . = yp'. + mic ray particles indirectly by the resulting shower of
Py = P'y particles this cosmic ray creates in the atmosphere.
A = /. Between 1968 and 1987 the Haverah Park array
detected more than 25,000 cosmic rays with enetgies
b Repeat the process for particle displacements
greater than 4 X 10^^ electron-volts, including 5
dt, dx, dy, and dz recorded in the laboratory frame to
with an energy of approximately 10^° electron-volts,
derive the inverse transformations from laboratory to
(rest energy of the proton = 1 0 ^ electron-volts =
rocket.
1.6 X 10~^“ joule)
E' = — vyp„ + yE a Suppose a cosmic ray is a proton of energy 10^°
p \ = yp. - vyE electron-volts. How long would it take this proton to
P' y=Py cross our galaxy as measured on the proton’s wrist-
watch? The diameter of our galaxy is approximately
216 EXERCISE 7-8 ROCKET NUCLEUS
10’ light-years. How many centuries would this trip

take as observed in our Earth-linked frame?
b The research workers at Haverah Park find no particle A particle 6 particle C
evidence of an upper limit to cosmic ray energies. A (at rest)
proton must have an energy of how many times its

rest energy for the diameter of our galaxy to appear to BEFORE AFTER
it Lorentz-contracted to the diameter of the proton
EXERCISE 7 -9 . Two particles collide to form a third at rest in the
(about 1 femtometer, which is equal to 10“ ^’ laboratory frame.
meters)? How many metric tons of mass would have
to be converted to energy with 100-percent efficiency
in order to give a proton this energy? One metric ton
equals 1000 kilograms. 7-9 sticky collision
Reference: M. A. Lawrence. R. J. O, Reid, and A. A. Watson, An inelastic collision is observed in the laboratory
Journal of Physics G: Nuclear and Particle Physics, Volume 17, pages
frame, as shown in the figure. Suppose that = 2
733-757 (1991).
units, £^ = 6 units, ~ 15 units.
a From the conservation of energy, what is the
energy Eg of particle B?
7-8 rocket nucleus b W hat is the momentum p^ of particle A?
A radioactive decay or “inverse collision” is observed Therefore what is the momentum pg of particle B?
in the laboratory frame, as shown in the figure. c From m^ = E^ — p^ find the mass mg of par
Suppose that = 20 units, m Q = 2 units, and ticle B.
£ ( - = 5 units. d Quick guess: Is the mass of particle C after the
a W hat is the total energy £^ of particle A? collision less than or greater than the sum of the
b From the conservation of energy, find the total masses of particles A and B before the collision? Vali
energy (rest plus kinetic) of particle D. date your guess from the answer to part c.
c Using the expression = ni^ find the
momentum pc of particle C. 7-10 colliding putty balls
d From the conservation of momentum, find the
A ball of putty of mass m and kinetic energy K streaks
momentum p^, of particle D.
across the frozen ice of a pond and hits a second
e W hat is the mass of particle D?
identical ball of putty initially at rest on the ice. The
f Does me + m^, after the collision equal
two stick together and skitter onward as one unit.
before the collision? Explain your answer.
Referring to the figure, find the mass of the combined
g Draw three momenergy diagrams for this re
particle using parts a - e or some other method.
action similar to those of Figure 7-6: BEFORE, SYS
a W hat is the total energy of the system before
TEM, and AFTER. Plot positive and negative mo
the collision? Keep the kinetic energy K explicitly, and
mentum along the positive and negative horizontal
don’t forget the rest energies of both particles A and
direction, respectively, and energy along the vertical
B . Therefore what is the total energy E c of particle C
direction. On the AFTER diagram draw the momen
after the collision?
ergy vectors for particles C and D head to tail so that
b Using the equation m^ — E^ — p^ = {m - \-
they add up to the momenergy vector for the system.
K) 2 — p 2 momentum p^^ of particle A before
Place labeled mass handles on the arrows in all three
the collision. W hat is the total momentum of the
diagrams, including the arrow for the system.
system before the collision? Therefore what is the
momentum pc of particle C after the collision?
particle A particle C
Q
particle D
o—
particle A particle B particle C
(at rest) (at rest)
BEFORE AFTER BEFORE AFTER

EXERCISE 7 -8 . Radioactive decay of a particle. EXERCISE 7 -1 0 . Two putty balls stick together.
EXERCISE 7-12 DERIVATION OF RELATIVISTIC EXPRESSION FOR MOMENTUM 217
c Again use the equation to find is the speed of the particle at this limit? At what
the mass me o f particle C Show that the result satisfies kinetic energy does a proton reach this limit (energy in
the equation MeV)? An electron?
c An electron in a modern color television tube is
accelerated through a voltage as great as 25,000 volts
te — {2my + 2mK = {2mY
(-9 and then directed by a magnetic field to a particular
pixel of luminescent material on the inner face of the
d Examine the result of part c in two limiting tube. Must the designer of color television tubes use
cases. (1) The value of mc 'tn the Newtonian low-ve special relativity in predicting the trajectories of these
locity limit in which kinetic energy is very much less electrons?
than mass: K /m « 1. Is this what one expects from
everyday living? (2) W hat is the value of in the
highly relativistic limit in which K /m » 1? W hat is 7-12 derivation off the
the upper limit on the value of m ^ D iscussion: reiativistic expression ffor
Submicroscopic particles moving at extreme relativis momentum— a worked
tic speeds rarely stick together when they collide. example
Rather, their collision often leads to creation of addi
tional particles. See Chapter 8 for examples. A very fast particle interacts with a very slow particle.
e D iscussion q uestion: Are the results of part If the collision is a glancing one, the slow particle may
c changed if the resulting blob of putty rotates, whir move as slowly after the collision as before. Reckon
ling like a dumbbell about its center as it skitters the momentum of the slow-moving particle using the
along? Newtonian expression. Now demand that momen
tum be conserved in the collision. From this derive the
relativistic expression for momentum of the fast-
7-11 limits off Newtonian moving particle.
mechanics The top figure shows such a glancing collision.
a One electron-volt (eV) is equal to the increase After the collision each particle has the same speed as
of kinetic energy that a singly charged particle experi before the collision, but each particle has changed its
ences when accelerated through a potential difference direction of motion.
of one volt. One electron-volt is equal to 1.60 X Behind this figure is a story. Ten million years ago,
10“ *^ joules. Verify the rest energies of the electron and in another galaxy nearly ten million light-years
and the proton (masses listed inside the back cover) in distant, a supernova explosion launched a proton
units of million electron-volts (MeV). toward Earth. The energy of this proton far exceeded
b The kinetic energy of a particle of a given anything we can give to protons in our earthbound
velocity v is not correctly given by the expression particle accelerators. Indeed, the speed of the proton
1/2 mv^. The error so nearly approached that of light that the proton’s
wristwatch read a time lapse of only one second be
relativistic expressions tween launch and arrival at Earth.
( for kinetic energy / (Newtonian expressions
for kinetic energy / W e on Earth pay no attention to the proton’s
wristwatch. For our latticework of Earth-linked ob
Newtonian expressions
( for kinetic energy I servers, ages have passed since the proton was
launched. Today our remote outposts warn us that
the streaking proton approaches Earth. Exactly one
is one percent when the Newtonian kinetic energy has second on our clocks before the proton is due to arrive,
risen to a certain fraction of the rest energy. W hat we launch our own proton at the slow speed one
fraction? Hint: Apply the first three terms of the meter/second almost perpendicular to the direction
binomial expansion of the incoming proton (BEFORE part of the top
figure). Our proton saunters the one meter to the
(1 + Z ) ’’ - 1 + «z H— «(« 1) + . . . impact point. The two protons meet. So perfect is our
2 aim and timing that after the encounter our proton
simply reverses direction and returns with the same
to the relativistic expression for kinetic energy, an speed we gave it originally (AFTER part of the top
accurate enough approximation if |z| « 1. Let this figure). The incoming proton also does not change
point— where the error is one percent— be arbitrar speed, but it is deflected upward at the same angle at
ily called the “limit of Newtonian mechanics.” W hat which it was originally slanting downward.
218 EXERCISE 7-12 DERIVATION OF RELATIVISTIC EXPRESSION FOR MOMENTUM
i I
Earth Frame: BEFORE Earth Frame: AFTER
EXERCISE 7-12. Top; A symmetric elastic collision between a fast proton and a slow proton in which each
proton changes direction hut not speed as a result of the encounter. Center: Events and separations as
observed in Earth frame before the collision. Here x = 10 million light-years and y = I meter, so these
figures are not to scale! Bottom; Events and separations as observed in the rocket frame before the collision.
How much does ^/-momentum of our slow-mov vertically above one another (botrom figure). For the
ing proton change during this encounter? Newton can rocket observer the transverse y-separations are the
tell us. At a particle speed of one meter/second, his same as for the Earth observer (Section 3.6), soy = 1
expression for momentum, mv, is accurate. Our pro meter in both frames. The order of events 1 and 2,
ton simply reverses its direction. Therefore the change however, is exactly reversed in time: For the racket
in its momentum is just 2 mv, twice its original mo observer, we released our proton at high speed ten
mentum in the y-direction. million years before impact and she releases hers one
W hat is the change in the y-momenmm of the second before the collision. Otherwise the diagrams
incoming proton, moving at extreme relativistic are symmetrical: To make the bottom figure look like
speed? W e demand that the change in j-momentum the center one, exchange event numbers 1 and 2, then
of the fast proton be equal in magnitude and opposite stand on your head!
in direction to the change in y-momentum of our slow Rocket observer and Earth observer do not agree
proton. In brief, y-momentum is conserved. This de on the time between events 1 and 0, but they agree on
mand, plus a symmetry argument, leads to the rela the proper time Tm between them, namely one sec
tivistic expression for momentum. ond. They also agree on the propet time T20 between
Key events in our story are numbered in the center events 2 and 0. Moreover, because of the symmetry
figure. Event 1 is the launching of the proton from the between the center and bottom figures, these two
supernova ten million years (in our frame) before the proper times have the same value: For the case we
impact. Event 2 is the quiet launch of our local proton have chosen, the wristwatch (proper) time for each
one second (in our frame) before the im paa. Event 0 proton is one second between launch and impact.
is the impact itself. The x-direction is chosen so that
y-displacements of both protons have equal magni no > •2 0
tude between launch and impact, namely one meter.

Now view the same events from a rocket moving W e can use these quantities to construct expres
along the x-axis at such a speed that events 1 and 0 are sions for the y-momenta of the two protons. Both are
EXERCISE 7-12 DERIVATION OF RELATIVISTIC EXPRESSION FOR MOMENTUM 219
protons, so their masses m are the same and have the clusion: The left side of this equation yields the rela
same invariant value for both observers. Because of tivistic expression for y-momentum: mass times
the equality in magnitude of they-displacements and displacement divided by proper time for this displace
the equality of Tjo and T,o, we can write ment.
W hat would be wrong with using the Newtonian
y _ y expression for momentum on the left side as well as on
m -------- M --------- [both frames]
the right? That would mean using earth time
"fio '^20
instead of proper time T,o in the denominator of the
The final key idea in the derivation of the relativis left side. But is the time it took the fast proton to
tic expression for momentum is that the slow-moving reach Earth from the distant galaxy as recorded in the
proton travels between events 2 and 0 in an Earth- Earth frame — ten million years or 320 million mil
measured time that is very close in value to the proper lion seconds! W ith this substitution, the equation
time between these events. The vertical separation y would no longer be an equality; the left side would be
between events 2 and 0 is quite small; one meter. In 320 million million times smaller in value than the
the same units, the time between them has a large right side (smaller because r,o would appear in the
value in the Earth frame: one second, or 300 million denominator). Nothing shows more dramatically
meters of light-travel time. Therefore, for such a than this the radical difference between Newtonian
slow-moving proton, the proper time X20 between and relativistic expressions for momentum — and the
events 2 and 0 is very close to the Earth time (2 0 correctness of the relativistic expression that has
between these events; proper time in the denominator.
This derivation of the relativistic expression for
T 20 ^ ^20 [Earth frame only] momentum deals only with its y-component. But the
choice of y-direction is arbitrary. We could have in
Hence rewrite the both-frames equation for the terchanged y and x axes. Also the expression has been
Earth frame: derived for particles moving with constant velocity
before and after the collision. When velocity varies
y _ _ y with time, the momentum is better expressed in terms
M -------- tn -------- [Earth frame only]
of incremental changes in space and time. For a parti
^10 ^20
cle displacement dr between two events a proper time
The right side of this equation gives the y-momenturn dx apart, the expression for the magnitude of the
of the slow proton before the collision, correctly cal momentum is
culated using the Newtonian formula. The change in
momentum of the slow proton during the collision is dr
p = m—
twice this magnimde. Now look at the left side. We ^ dx
claim that the expression on the left side is the y-mo-
mentum of the very fast proton. They-momentum of One-sentence summary; In order to preserve con
the fast proton also reverses in the collision, so the servation of momentum for relativistic collisions,
change is just twice the value of the left side. In brief, simply replace Newton’s “universal time” t in the
this equation embodies the conservation of the y- expression for momentum with Einstein’s invariant
component of total momentum in the collision. Con proper time T .
% ^
''2 > .
'■ 2 • '2 '
V- 'I’g;..'.
S :f- ■Aj'S;: ■, f ^ : ■■■ -
.2?' ^■.'-
•#*' ^
‘■'^••V ■■''V
‘.^'‘W -,'y- f Ai ‘fc*. '

' ■ ) • : ■ ' 'k- ■ ■
*-■ '. ;• ..■ “'v /i It
■V f P - ? >- V
•■■ -'. 4^;;: ■'■'■ ' '■>.■" b! 'i-i’ f ¥'* f
'■' %v^t' ’■■■' ;’ ■' ■,.-' '■-
'V
.-r '
I f : - t-
■-'iX’
■ ■ ,;i:
' '' p y ■'■-.kr'...
t*t
' ' ,2: i' r *’■
rA : • ■%
w 'g' ■ 4*
■ , V*;' .,/ •'> 5-
' 7vs ^ 5s'4. ! ■ ':
r . . .'^ ’4
:i'\ ?i>: !*'.‘V'- r ■.'■■■■. fe,"'8''ss
Hr* s,'ii:S3,
Ssn;g4;Sfef: ^■/
' ...............
- 'f - ' i ’■:g- 4' , A § Ji*'

E .&
w .‘ . ' -
•vfc.\ - .■J,_i ,;■
.. -
'-’-I..''--.."'
. • r 4 '^ ’ ' 2 : ' v ''' 'V ; i'X • . ^ ‘
.;• .- • ' : 4 " 2 ' ”

‘ t y - : . ' - ' ;•_ 2 l ' ■•2-
'■ ■
v l ‘- V ' '. • '.'C •■•■.•-.'.'■ 5 -

— ' 2^ ' ' y
' ’ V ■ ...........v'.-’^'-*- ■’
' / : ‘^ : \ ' X ' / 't ^
.V-.' -
V j,
vp.
.- ■ / ■ ■ ''2 ^ - '> 2 ''' *^‘i A '*1^0 ,
- ' 'i - 2 .-.• -•••-’•■. '• i ,f „ ■ •
;'■ •• f i - ■ .■ _ ■ '. •'■ '-^ -sV .V ^ '' •' '< ■ ■' ., •1 ^ -
' i '“ v ■' '• ' ’ ' ' '
■ . - .' ■ ■ • J t - i ';’ '- ' .• ■ ; 'f c . , , ; ' ' ‘4 v' ^-
S 4 / . ',
■ l i f :. ■ '
' T o -
. .
'■ ■.i.;
--V
• i ’ . V wi^r »• , I*
'%7fw
, ■•': '«• r
- 2^
.. • ^ : 2 2 '. '
• . #
4^V“' ^
--1f'^^^'^^rTl'**’
8.1 THE SYSTEM
an isolated island of violence
Particle physics is one of the gteat adventures of our time. No one can venture into the
heart of it without momenergy as guide and lamp. Particles clash, yes. But however
cataclysmic the encounter, it always displays one great simplicity. It takes place on a
local stage, an island of violence, apart from all happenings in the outside world. In
other isolated arenas of action football players form a team, actors a troupe, soldiers a
platoon; but in a battle of matter and energy, the participants receive the name
system .
W hat the action starts with, what particles there are, what speeds they have, what
directions they take: that’s the story of the system at the start of the action. W e may or
Keep score of momenergy for
may not pursue in all detail every stage of every encounter, as we view the scenes of a
the system
play or watch the episodes of a game. However, nothing that claims to be an account of
the clash, brief though it may be, is worthy of the name unless it reports every
participant that leaves the scene with its speed and its direaion. Departing, they still
belong to the system. Moreover, at every step of the way from entry to departure we
continue to use for the collection of participants the name system.
The child keeps count of who wins and who loses in the shoot-out before he or she
learns to ask questions of right and wrong, of why and wherefore. W e likewise keep
tabs on what goes into an encounter and what comes out only to the extent of
broadcasting the participants’ momenergies before and after the act of violence. We do
not open up in this book the more complex story of the forces, old and new, that
govern the chances for this, that, and the other outcome of a given encounter. W e limit
outselves to the ground mles of momenergy conservation in an isolated system,
221
222 CHAPTER 8 COLLIDE. CREATE. ANNIHILATE.
8.2 THREE MODEST EXPERIMENTS

elastic glass balls; inelastic wads off gum;
weighing heat
A collision does not have to be violent to qualify for attention nor be exotic to make
momenergy scorekeeping interesting. It is fun to begin with momenergy scorekeeping
for three encounters of everyday kinds before strolling out onto the laboratory floor of
high-energy particle physics.
Elastic collision: Momenergy First E xperim ent: Elastic Collision. Suspend two identical glass marbles from
automatically conserved the ceiling by two threads of the same length so that the marbles hang, at rest, just
barely touching. Draw one back with the finger and release it (Figure 8-1). The
released marble gathers speed. The speed peaks just as the first marble collides with the
second. The collision is elastic: Total kinetic energy before the collision equals total
kinetic energy after the collision. The elastic collision brings the first marble to a
complete stop. The impact imparts to the second all the momentum the first one had.
Conservation of momentum could not be clearer:
/ total momentum \ / total momentum \

to the right just to the right just
b efo re the collision, a fte r the collision,
all of it resident on all of it resident on
' the first marble ' 'th e second marble/
And energy? In the collision the two particles exchange roles. The first patticle comes to
a halt. The second particle moves exactly as the first one did before the collision. Hence
energy too is clearly conserved.
Just before the collision and just after: How do conditions compare? Same total
momentum. Same total energy. Therefore same total momenergy.
Inelastic collision: Momenergy Second E xperim ent: Inelastic Collision. Replace the two glass marbles by two
also conserved identical balls of putty, wax, or chewing gum (Figure 8-2). Pull them aside by equal
amounts and release.
Both released balls of chewing gum gather speed, moving toward one anorher. The
equal and opposite velocities peak just before they collide with each other. By
symmetry, the momentum of the right-moving particle has the same magnitude as the
momentum of the left-moving particle. However, these momenta point in opposite
directions. Regarded as vectors, they sum to zero. The momentum of the system
therefore equals zero just before the collision.
Just after rhe collision? The two balls have stuck together. They are both at rest;
each has zero momentum. Their combined momentum is also zero. In other words.
LAB RECORDS, INC.| i|
FIGURE 8-1. One marble collides elastically with another.

8.2 THREE MODEST EXPERIMENTS 223
I I I I I I I I I I I I I I I I I I I I I I I I I I lab r e c o r d s ,
FIGURE 8-2. One ball of chewing gum locks onto the other.
the momentum of the system is zero after the collision. Zero it was also before the
collision. Thus the momentum of the system is conserved.
For system energy the outcome is more perplexing. Just before the collision, each
ball has an energy consisting of its mass m and its kinetic energy K. These energies add
to make the total energy of the system: = 2m + 2K.
After the collision? Both balls of chewing gum are at rest, stuck together as a single
Kinetic energy converted to moss
blob, which now constitutes the entire system. The energy of that stationary blob must
be its rest energy, equal to the mass of the system: Fsystem“ ^rest ~ ■^system- W hat is the
value of that system energy? It must be the same as the energy of the system before the
collision, equal to 2m + 2K, where m is the mass of each ball before the collision.
Hence, if energy is conserved, = 2w + 2K. This is greater than the sum of
masses of the incoming particles.
Where does this extra mass come from? The energy of relative motion of the
incoming particles gets converted, during the collision, into energy of plastic deforma
tion and heat. Each of these forms of locked-in energy yields an increment of mass. In
consequence the mass of the pair of balls, stuck together as one, exceeds the sum of
masses of the two balls before impact.
T h ird E xperim ent: W eighing H eat. If warmed and distorted balls of gum
have more mass than cool and undistorted balls, then maybe we can measure directly
the increased mass simply by heating an object and weighing it. In this case the system
consists of a single large object, such as a tub of water, stationary and therefore with
zero total momentum. System energy consists of the summed individual masses of all
water molecules plus the summed kinetic energies of their random motions. This
summed kinetic energy increases as we add heat to the water; hence its mass should
Con we weigh heat?
increase. Can we detect the corresponding increase in weight as we heat the water in the
Not yet!
tub?
Alas, never yet has anyone succeeded in weighing heat. In 1787 Benjamin Thomp
son, Count Rumford (1 7 5 3 - 1814), tried to detect an increase in weight of barrels of
water, mercury, and alcohol as their temperature rose from 29° E to 6 l ° E (in which
range ice melts). He found no effect. He concluded “that ALL ATTEMPTS TO
DISCOVER ANY EFFECTS OF HEAT UPON THE APPARENT WEIGHTS
OF BODIES WILL BE ERUITLESS” (capital letters his). Professor Vladimir Bra
ginsky of the University of Moscow once described to us a new idea for weighing heat.
Let a tiny quartz pellet hang on the end of a long thin near-horizontal quartz fiber, like
a reeled-in fish at the end of a long supple fishing rod. A fly that settles on the fish
increases its weight; the fishing rod bends a little more. Likewise heat added to the
pellet will increase its mass and will bend the quartz-fiber “fishing rod” a little more.
That is the idea. The sensitivity required to detect a bending so slight unfortunately
surpasses the present limit of technology. Braginsky himself already has invented,
published, and made available to workers all over the world a now widely applied
scheme to measure very small effects. There is real hope that he— or someone
else — will weigh heat and confirm what we already confidently expect.
8.3 MASS OF A SYSTEM OF PARTICLES
energies add. momenta add. masses do not add.
No one with any detective instincts will rest content with the vague thought that heat
has mass. Where within our stuck-together wads of chewing gum or Rumford’s barrel
of water or Braginsky’s quartz pellet is that mass located? In random morions of the
atoms? Nonsense. Each atom has mass, yes. But does an atom acquire additional mass
by virtue of any motion? Does motion have mass? No. Absolutely not. Then where,
and in what form, does the extra mass reside? Answer: Not in any part, but in the
system.
Heat resides not in the particles individually but in the system of particles. Heat
arises not from motion of one particle but from relative motions of two or more
particles. Heat is a system pro p erty .
The mass of a system is greatet when system parts move telative to each other. O f
this central point, no simplet example offers itself than a system composed of a single
pait of masses. Our example? Two identical objects (Figure 8-3). Each has mass 8.
Relative to the laboratory frame of reference each object has momentum 6, but the two
momenta are opposite in direction. The energy of each object is E = (jrP- -f =
(8" + 6^)V2 = 10.
8
mass
8
Tiass The total momentum of the two-object system is / ’system— 6 — 6 = 0. The energy of
a v= 3/5 -3/5
O the system is = 10 + 10 = 20. Therefore the mass of the system is Msystem “
(Esystem^ ~ / ’system^) ~ [(20)^ — 0^]*/^ = 20. Thus the mass of the system exceeds the
sum of the masses of the two parts of the system. The mass of the system does not agree
with the sum of the masses of its parts.
Energy is additive. Momentum is additive. But mass is ttol additive.
Ask where the extra 20 — 16 = 4 units of mass are located? Silly question, any
answer to which is also silly!
Ask where the 20 units of mass are located? Good question, with a good answer.
The 20 units of mass belong to the system as a whole, not to any part individually.
Where is the life of a puppy located? Good question, with a good answet. Life is a
property of the system of atoms we call a puppy, not a property of any part of the
puppy.
Where is the extra ingredient added to atoms to yield a live puppy? Unacceptable
question, any answer to which is also unacceptable. Life is not a property of any of the
20
mass
20 energy individual atoms of which the puppy is constituted. Nor is it a property of the space
between the atoms. Nor is it an ingredient that has to be added to atoms. Life is a
ptoperty of the puppy system.
Life is remarkable, but in one respect the two-object system that we are talking
about is even more remarkable. Life requires organization, but the two-object system
of Figure 8-3 lacks organization. Neither mass interacts with the other. Yet the total
energy of the two-object system, and its total momentum, regarded from first one
frame of reference, then another, then another, take on values identical in every respect
to the values they would have were we dealing throughout with a single object of mass
20 units. Totally unlinked, the two objects, viewed as a system, possess the dynamic
0 momentum attributes — energy, momentum, and mass — of a single object.
This wider idea of mass — the mass of an isolated system composed of disconnected
CONSIDERED AS A SINGLE SYSTEM objects: what right have we to give it the name “mass”? Nature, for whatever reason,
demands conservation of total momenergy in every collision. Each collision, no matter
20 how much it changes the momenergy of each participant, leaves unchanged the sum of
mass
their momenergies, regarded as a directed arrow in spacetime— a 4-vector. Encounter
CO- ■o,) or no encounter, and however complex any encounter, system momenergy does not
alter. Neither in spacetime direction nor in magnitude does it ever change. But the
FIGURE 8-3. Two noninteracting particles, magnitude — the length of the arrow of total momenergy, figured as we figure any
each of mass 8, are in relative motion. Taken spacetime interval — is system mass. Whether the system consists of a single object or
together, they constitute a system of mass 20.
Where does the mass 2 0 reside? In the system!
8.3 MASS OF A SYSTEM OF PARTICLES 225
of many objects, and whether these objects do or do not collide or otherwise interact
with each other, this system mass never changes. That’s why the concept of system
mass makes sense!
An example? Again, two objects of mass 8, again each moving toward a point
midway between them z.tv = (momentum)/(energy) = ( p = 6 ) /( E = 10) = 3 /5 the Different free-float frames.
Same system mass.
speed of light. Now, however, we analyze the two motions in a frame moving with the
right-hand object (Figure 8-4). In this new frame the right-hand object is at rest: mass,
m = 8; momentum, p = 0\ energy, E = [m^ -b p^V^^ = 8. The left-hand object is
approaching with a speed (addition of velocities: Section L.7 of the Special Topic
following Chapter 3; also Exercise 3-11)
3/5 + 3/5 6/5
1 + (3 /5 )(3 /5 ) 34/25 17
It has energy E = m /[l — — 8 / [ l — (15/17)^}*/^ — 17 and momentum p —
vE = 15. So much for the parts of the system! Now for the system itself. For the system
the energy is E^^^^ — 8 + 1 7 — 25 and the momentum is p^ 0 + 15 = 15.
Now for the test! Does the concept of system mass make sense? In other words, does
system mass turn out to have the same value in the new frame as in the original frame?
It does:
^2)1/2 = |-(25)2 - (i5)2]i/2 = i;625 - 225]!/^
^ s y s te m = (^ s,
= I400y/^ -- 20
8
mass
8 energy
25 energy
BEFORE SYSTEM AFTER

(before one/after!)
8
mass
8 20
mass
O -
15/17 '= 3 /5
FIGURE 8-4. System of Figure 8-3 observedfrom a fra m e moving w ith the right-hand object. The
right-hand object is therefore initially at rest. Before: Arrows of momenergyfor two objects before collision.
Each object has a mass of eight units (shaded handles). The upper, vertical, arrow belongs to the particle
originally at rest, the slanted arrow to the incoming particle. System: Addition of the two momenta (one of
them zero!) gives the total momentum before collision. Similarly, addition of the two energies gives the total
energy. Mass of the system — even before the two particles interact!— comes from the expression for the
"hypotenuse” of a spacetimelike triangle. Result: 2 0 units of mass (shaded handle on center 4-vector):
(mass)^ = (energy)^ — (momentum)^ = (25F ~ (D F ~ 625 — 22 5 = 4 0 0 = (20)^
After: The two particles now collide and amalgamate to form one particle. Arrow of total momenergy after
the amalgamation is identical to arrow of total momenergy before the collision. Mass of this two-object system
exceeds the mass of one object plus the mass of the other, not only after the collision but also before. Mass is not
an additive quantity.
S A M P L E P R O B L E M 8-1
MA S S OF A S Y S T E M OF
MATERIAL PARTICLES
Compute Msystemfor each of the following systems. in terms of the unit mass m-, do not use momenta or
The particles that make up these systems do not velocities in your answers. [Note: In the following
interact with one another. Express the system mass diagrams, arrows represent (3-vector) momenta.}
System a
(kinetic energy = K = 3m]
------ ► m (at rest)
System b
(kinetic energy = K = 5m) (kinetic energy = K = 5m)
' O -------------------- ^ 'O -------------- ^

System c
3m
o (energy = E = 7m)
m(3 rest)
System d
o
(E = 6m)
'O
(E = 6m)
SOLUTION
System a: System energy equals the rest energy of the two particles (the sum of their
masses) plus the kinetic energy of the moving particle: = (»z -h w) -f 3»z = 5m.
Squared momentum of the system equals that of the moving particle: ~ P ^~
E?- — m^ = {AmY — mA= \ 5m^. Mass of the system is reckoned from the difference
between the squares of energy and momentum:
^system • ~ = [lOY/^m = 3.162 m
Moreover, if the two objects collide and amalgamate, the system energy remains at the
value 25, the system momentum remains at the value 15, and the system mass
remains 20, as illustrated in Figure 8-4.
In summary, the mass of an isolated system has a value independent of the choice of
frame of reference in which it is figured. System mass remains unchanged by en
counters between the constituents of the system. And why? Because the system mass is
the length (in the sense of spacetime interval) of the arrow of total momentum-energy.
This momenergy total is unaffected by collisions among the parts or by any transfor
mations, decays, or annihilations they may undergo. System mass does make sense!
System! System! You keep talking about “system,” even when the particles do not
interact, as in the system o f Figure 8-3- It seems to me th a t you are totally arbitrary in
the way you define a system. Who chooses which particles are in the system?
8.3 MASS OF A SYSTEM OF PARTICLES 227
System b: System energy equals rest energy of the two particles plus kinetic energy of
the two particles; = 2w + lOm = \2m. Squared momentum of each particle is
^ = E^ — np- ' (<amp vr i'bnP yieldingp — (35)*/^»2. System momentum is
twice this: p = 2 (35)*^^ m. The mass of the system is
= t w - r system - {2 (35)‘/2^}2]V2
= [144 l40V/^m [4}*/^w = 2m
In this one special case the mass of the system equals the sum of masses of the objects
that make up the system. W e could have seen this result immediately by observing the
system from a reference frame that moves along with the particles. In this frame the
particles are at rest and have zero total momenmm; the total energy is identical to the
sum of the individual rest energies (the individual masses). So in this case the mass of
the system is equal to its energy, which is equal to the sum of masses. Moreover, system
mass is an invariant. Thus 2m is the mass of the system as reckoned in ail reference
frames, including the one in which System b is pictured.
System c: Total energy = system energy = £,^5^ = Im ■¥ m = 8w. System
momentum equals the momentum of the moving particle: p^^n ^ = EP ~ m^ —
{Im p — {im p = 49»2^ ~ 9m^ = 40m^. Hence the system mass is
, = \G4m^ - 40»z2}i/2 = \24p/^m = 4.899«
System d: This part of the problem serves as a reminder that momentum is a

Euclidean 3-vector. The squared momentum of each particle is pP = EP — m^ = iEjtrP
— mP = 35w^. Their total momentum is not the algebraic sum of the momenta,
because they are vectors pointing in perpendicular directions. This perpendicular
orientation allows us to equate the squared system momentum to the sum of the
squares of the individual momenta: / ’system^ ~ 35t»^ + iimP — lOm^. System energy is
the sum of the energies (energy is a scalar and adds like a scalar!): E^^^^ = 6m + 6m =
\2m. Hence system mass is
[144«2 - 70»?2]i/2 = {74}i/2y„ == 8.602 w
Compare this result with that of System b, which also contained two particles, each of
total energy 6m.
We do! W e can draw the dashed line around any collection of objects whatever,
subject to this one restriction: no object in our system may interact with any external
object or experience a force from outside the system. Our system must be isolated.
W ith that single limitation, the system we choose is arbitrary, has a conserved total
energy, a conserved total momentum, and a system mass that is invariant — a mass
that has the same value no matter in which free-float frame it is reckoned.
7 can't believe the story you tell. Those two mass-8 objects, you say, may fly past each
other. Then your talk about the system mass isjust talk, terminology. Or they may whang
into each other and amalgamate. Then your talk is all wrong, and for an obvious reason.
As the objects collide they slow and come to rest relative to each other. A t that instant
and in that “rest frame" {the frame of Figure 8-3), each has zero momentum, and
energy equal to its mass. So the total momentum of the system is zero, and its total energy
is 8 4- 8 = 16. That means a mass of 16. Yet you claim 20.
-< JA B L E 8 -T ^
CLEOPATRA’S VASE, HER BATH, AND INTERSTELLAR VACUUM:

ILLUSTRATIVE FRACTIONAL CHANGES IN MASS OF SYSTEMS
F ra a io n a l in a e a s e
in sy stem m ass
S ystem before S ystem afte r (to nearest p o w e r o f 10)
O n e -k ilo g ra m vase V ase sm a sh e d in to so m a n y fra g m e n ts th a t 1 0 0 1 0 ~ '“

centim eters^ o f glass-to -g lass b o n d s are b ro k e n
B a th w a te r a t 1 5 ° C B a th w a te r a t 4 0 ° C io-'2
W a te r ( H , 0 ) A to m ic h y d ro g e n ( H ) a n d oxygen (O ) 1 0 -9
E arth A ll m olecules o f E a rth lifte d a g a in st th e p u ll o f th eir 10^
m u tu a l g ra v ity to infinite se p a ra tio n fro m o n e a n o th e r
H y d ro g e n a to m in low est E lectron w ith d ra w n to infinite se p a ra tio n fro m nucleus io-»

energ y sta te
D e u te ro n D e u te ro n se p a ra te d in to p ro to n a n d n e u tro n 1 0 -’
N e u tr o n sta r W id e ly se p a ra te d iron a to m s a t te s t w ith t e s p e a to 1 0 -'
each o th e r
A v a c u u m , b efo re it is z a p p e d E le c tro n - p o s itr o n p a ir b o u n d as a p o sitro n iu m a to m In fin ite f ra a io n a l

b y co n v erg in g p h o to n s in a e a s e
Slow and come to rest? Yes. But that means force: “elastic,” gravitational, elearo-
magnetic, or nuclear force. That’s the new and valuable point you make here. And
those particles, pushing against that force, store up energy. This energy, too, has to
be put into the bookkeeping. When amalgamating particles come to rest relative to
one another, the energy of interaction “balances the books” — it so happens— and
leads to a final mass of 20, greater than the sum of masses of the original objects. For
the figuring of system mass, however, we really don’t have to get into this detail. It is
enough for us to know that total momentum is c o n s e r v e d , = 0 in Figure 8-3,
and total energy— in whatever way it is apportioned between the objects and the
fields of force that act between them— is also conserved, = 20. The length, in
the sense of interval, of the 4-vector of momenergy for the system remains un
changed: = 20.
W hat about a system that is n o t isolated? A system that has — and keeps— zero
System energy increase?
System mass can increase.
momentum, but receives an increment of energy? Then its mass rises by an amount
exactly equal to that input of energy. The increase in mass is the same whether that
energy goes into altering the relative motion of the parts of the system or increasing the
energy of interaction between them or some combination of motion and interaction.
Supply energy to a system by heating it or setting it into internal vibration or fracturing
the bonds between its parts? Each is a guaranteed way to increase the mass of the
system (Table 8-1)! ■om'
8.4 ENERGY WITHOUT MASS: PHOTON

light moves with zero aging,
photons move with zero mass.
A striking example of the primacy of momenergy over mass is furnished by a
quantum o f light colliding with an electron.
8.4 ENERGY WITHOUT MASS: PHOTON 229
Quantum? A quantum of luminous energy of a given color or, in more technical

terms, light of a given wavelength or frequency of vibration. Max Planck discovered in
1900 that light of a given color comes only in quanta— “hunks” — of energy of a
standard amount, an amount completely determined by the color (Table 8-2). We can
have one quantum, one hunk, one p h o to n , of green light, or two, or fifteen, but never
two and a half.
Nothing did more to raise the light quantum, the hunk of luminous energy, the
photon, to the status of a particle than experiments carried out by 28-year-old Arthur
Compton demonstrates quantum
Holly Compton at Washington University, St. Louis, in 1920. Shining X-rays of
of radiation — photoni
known wavelength (and hence of known frequency and known quantum energy) on a
variety of different substances, he measured the wavelength (and hence the quantum
energy) of the emergent “scattered” X-rays. He got identical changes in wavelength at
identical angles of observation from many kinds of materials. There was no way he
could explain this result except to say that the scattering object was in every case the
same, an electron, whatever the atom in which the electron happened to reside.
But why did the change of wavelength have a unique value, the same for all
materials at a given angle of scattering? Every idea of classical physics failed to fit,
Compton found. “Compton arrived at his revolutionary quantum theory for the
scattering process rather suddenly in late 1922,” a biographer tells us. “He now
treated the interaction as a simple collision between [an X-ray quantum} and a free
electron . . . {He} found that [this hypothesis gave results} which agreed perfectly
with his data . .. When Compton reported his discovery at meetings of the American
Physical Society, it aroused great interest and strong opposition . . .” By 1927,
however, his finding was generally accepted and in that year won him the Nobel Prize.
W hat does it mean to treat a photon on the same footing as a particle? It means this:
attribution to the photon of an energy and a momentum, in other words momenergy.
MOMENTUM AND ENERGY CARRIED BY ONE PHOTON, ONE QUANTUM,

ONE HUNK OF LUMINOUS ENERGY OF VARIOUS “ COLORS"
(Unit of energy used in this table: electron-volt or eV, the amount of energy given
to an electron by accelerating it through an electrical potential difference of one volt)
Source of Momentum (and Frequency in

electromagnetic energy) of a vibrations Wavelength
radiation single quantum per second in meters
KDKA, Pittsburgh: world’s first radio broadcast station 4.22 X lO"’ eV 1.02 X 10^ 294
A sample infrared beam 1.24 X 10-^ eV 3 X 10'2 10-“
Yellow radiation from a sodium arc lamp 2.11 eV 5.09 X lO'-* 5.90 X 10-’
Ultraviolet light from a mercury arc lamp 4.89 eV 1.18 X 10*’ 2.54 X 10-’
Ultraviolet star radiation of just barely sufficient quantum 13.6 eV 3.29 X 10” 0.91 X 10-’
energy to strip a hydrogen atom of its electron
Each of two gamma rays given off in the mutual annihilation 5.11 X 10’ eV 1.23 X 10’“ 2.43 X 10-”
of a slow positron and a slow electron
Each of two gamma rays given out when a neutral pi meson, 6.75 X 10’ eV 1.63 X 10” 1.84 X 10-'“
at rest, decays
Each of two gamma rays given off in the mutual annihilation 0.938 X 10” eV 2.27 X 10” 1.32 X 10-'’
of a slow proton and a slow antiptoton
In what direction in spacetime does the photon’s arrow of momenergy point? In a

Photon momenergy points
lightlike direction, because the photon — a quantum of light— travels with light
in lightlike direction
speed!
When we turn from spacetime to a particular free-float frame of reference and
observe a pulse of light at one event along its worldline and then observe it at a second
event (Figure 8-5), we know in advance something important about the interval
between the two events: It equals zero.
(interval)^ — (distance between two events)^ — (time between two events)^

= (difference between two quantities of identical magnitude)
= 0
A photon in a pulse of light has a momenergy arrow with a tip and a tail, like the
Photon momenergy:
momenergy vector for any other particle. Between the tip and tail there is a magnitude.
magnitude zero
(photon mass = 0 ) The magnitude for the photon, however, has the value zero— zero because this arrow
points in the same direction in spacetime as the worldline of the light pulse (Figure
8-5). For that reason its space component (momentum) and its time component
(energy) are equal. And, of course, we express the square of this magnitude as we
express the square of any interval, as a difference between the squared timelike and
spacelike separations between the two ends of the arrow:
(magnitude of momenergy arrow of photon)^

= (photon energy)^ — (photon momentum)^
= (photon mass)^ = 0
In brief, the lightlike character of the arrow of photon momenergy tells us that (1)
photon mass equals zero and (2) the magnitude of momentum, or punch-delivering
power, of the photon is identical in value with the energy of the photon:
(photon energy) — (magnitude of photon momentum)
and
(photon mass) 0
FIGURE 8-5. W orldline o f a pho

ton. Note its “unit slope in space
time. ” Insets; Unit slope of worldline
means equal space and time separa
tions between events on this worldline,
hence zero interval between them —
and zero aging for the photon. Momen-
ergy of the same photon, also with unit
slope, symbolizing three properties of
the photon: it has zero mass (hence the
bigzero as an invariant “handle"), it
travels with light speed, and it has a
momentum identical in magnitude
with its energy.
8.4 ENERGY WITHOUT MASS: PHOTON 231
2.4 momentum
BEFORE SYSTEM A FER

(before one/after!)
A A A /^ O v= 1 Q 1
O—24/26
^
O =1 1
mass mass mass mass
FIGURE 8-6. Backscattering of a photon by a free electron. The wiggly arrow symbol represents a
photon. Energy, momentum, and mass of all particles are expressed in units of electron mass. Before: The
electron at rest has an energy equal to its mass (vertical arrow)', the photon has an energy (and a momentum)
of 2 electron masses (angled arrow). System: Arrow of total momenergy. (What is the mass of the system?)
After: Arrows of momenergy ofknocked-on electron (labeled 1) and backscattered photon (labeled 0) after the
encounter. Arrow of total momenergy of the system remains the same (is conserved!) during this process.
Figure 8-5 summarizes these features of the elementary quanta of visible light and
other electromagnetic radiations. For a “handle” on the momenergy 4-vector of a
photon — representative of its magnitude— we choose a stylized zero, 0.
Nothing shows these revolutionary features of light to better advantage than the
very collision process studied by Arthur Compton: the encounter between a single
Compton collision analyzed
photon and a single electron. We take the electron, loosely bound though it may be in
one or another outer orbit of an atom, as essentially free and essentially at rest— at rest
compared to the swift motion in which it finds itself after the high-energy photon hits
it (Figure 8-6).
To simplify all numbers, we pick for the photon energy a value typical of gamma
rays, considerably grearer than that of the X-rays with which Compton worked but
easily available today from various sources of radioactivity: 1.022 MeV (million
electron-volts). We pick this number because we want to express all energies in units of
electron mass, 9.11 X 10“ ^' kilograms or 0.511 MeV. Our choice of photon energy
equals exactly two electron masses. Convenient!
Incoming photons of this energy, encountering an electron, are scattered by the
electron sometimes in one direction, sometimes in another, and sometimes straight
backward. In that most extreme of encounters — backward scattering— an inter
change of momentum takes place that nevertheless preserves total momentum and
also total energy, as illustrated in Figure 8-6. The electron is kicked forward with a
momentum of 2 .4 = 1 2 /5 times the electron mass, and the photon bounces backward
with a momentum (and energy) of 0.4 = 2 /5 times the electron mass, much less than
the two-electron masses of momentum (and energy) with which it approached.
S A M P L E P R O B L E M 8-2 m
M A S S OF A S Y S T E M THAT
INCLUDES PHOTONS
A photon has no rest energy — that is, no mass of Find system mass for each of the follow
its own. However, a photon can contribute energy ing systems. The particles that make up these sys
and momentum to a system of objects. Hence the tems do not interact with one another. Express the
presence of one or more photons in a system can system mass in terms of the unit mass m (or the
increase the mass of that system. More: A system unit energy E in the photons-only systems). Use
consisting entirely of zero-mass photons can itself only energy and mass in your answers: no mo-
have nonzero mass! menta or velocities.
System a
m (at rest)
A /W V ^
(photon) (£= 3m)
System b
A /V W ^ A /V ^
(energy = 3E) (energy = E)
System c
A A /W ^
(energy = 3£) (energy = E)
System d
(energy ^ ^
A /V W ^
(energy = 3 E)
SOLUTION
System a: System energy equals the rest energy m of the material particle plus the
energy E = 4 m o i the photon: E^^^^ = m4- 5m = 4m. The momentum of the system is
equal to the momentum of the photon, which is equal to its energy: = 4m. The
mass of the system is reckoned from the difference of the square of energy and
momentum:
^system ~ t£systcm^ ~ ~ \.k4mf '~ — [I6m^ —

— — 2.646m
System b: System energy equals the sum of the energies of the two photons: =
3E 4- E — 4E. System momentum equals sum of momenta of the two photons
— which in this case also equals the sum of the energies of the two photons: =
3£ + £ = 4£. Therefore system mass equals zero:
= t£sv ‘^system
•}i/2 = [(4 £ )2 - ( 4 £ ) 2 ] i/ 2 = 0
We could have predicted this result immediately. Two photons moving along in step
are, as regards momentum and energy, completely equivalent to a single photon of
8.5 PHOTON USED TO CREATE MASS 233
energy equal to the sum of energies of the separate photons. And a single photon has,
of course, zero mass.
System c: Total energy = system energy = = 3E E = 4E. System
momentum equals the difference between the rightward momentum of the first
particle and the leftward momentum of the second particle: = i E ~ E = 2E.
Hence the system mass is
[16£2 - 4£^}>'2 = [12]''2£ = 3.464£
Why can’t we simply make a single photon by adding the energies of the two photons,
as in system b? Because energies add as scalars, and momenta add as 3-vectors. In this
case the total energy is 4£ and the total momentum is 2£. No way to make a single
photon out of this; for a photon, energy and momentum must have equal magnitudes!
System d: This part serves as an additional reminder that momentum is a 3-vector.
The system energy equals E^^^^ = £ -f 3£ = 4£. The squared momentum of the
system equals the sum of squares of the momenta of the separate particles, since they
move in perpendicular directions in this frame: ~ EP (3£)^ = 10£^.
Hence system mass is:
= [16£2 10£ 2}‘/2 = [ 6}*/ 2£ = 2 . 449£
8.5 PHOTON USED TO CREATE MASS

photon hits electron, creates electron —positron
pair
It should not be surprising that a photon can deliver energy without having any mass
of its own. After all, an electron does have mass of its own; yet an electron traveling
sufficiently close to light speed can impart to its target an amount of energy ten, a
hundred, or a thousand times as great as its own mass. Not mass but momentum
governs the size of punch that either photon or electron can deliver.
Incredibly, however, a photon in the presence of an electron can create matter out of M atter is born
empty space. To bring about this process, double the energy of the quantum of radiant
energy shown in Figure 8-6. When a photon with energy equal to four electron masses
hits an electron at rest, the photon most often recoils; in other words, it suffers
backward scattering, an instance of the Compton process. Occasionally, however, the
impacting photon produces out of empty space, near the struck electron, a new pair of
electrons, one with a negative electric charge like all everyday electrons, the other with
an identical amount of positive charge. The electron with positive charge has rhe name
p o sitro n (Box 8-1).
This process goes on all rhe rime high in Earrh’s atmosphere, where cosmic rays pour
o.
FIGURE 8-7. Comparison a n d contrast.
in from outer space. There, however, energies of cosmic-ray photons often far exceed Left: Two protons and an electron forming the
four electron masses. In consequence, the struck electron and the two newly created hydrogen molecule ion of chemistry. (A proton is
electrons go off in slightly different directions and at different speeds. However, when much more massive than an electron but can be
envisioned as occupying less volume.) Right:
the energy of rhe incoming photon is sufficiently finely tuned, in the immediate
Two electrons and a positron, forming a polye
vicinity of an energy of four electron masses, rhe three particles can stick together as a lectron created by impact of a properly tuned
super-light molecule, a po ly electro n , a system analogous to what chemists call the photon {about 2 lAeV of energy) on an electron at
hydrogen molecule ion (Figure 8-7).
BEFORE AFTER
O A A A / ^ 3
mass ■ Q
1 ^=4/5
FIGURE 8-8. Conservation of energy a n d momentum in the process o f creating a p a ir {a positive

a n d a negative electron) in the field o fan electron. Before: A photon that has energy {and momentum)
equal to four electron masses (sloping arrow) strikes an electron essentially at rest (vertical arrow). After:
The photon has ceased to exist, and the two newly created particles have gone off in company with the original
electron at 8 0 percent of light speed— a combined “particle” of three electron masses.
Why does it take a light quantum with an energy of four electron masses to create
(Figure 8-8) a polyelectron, a super-light hydrogen molecule ion, an object with a
mass of three electron masses (in truth, a tiny bit less than three elearon masses
because of the negative binding energy among the three particles)? The question
becomes all the more insistent when we recall that the electron that got hit already
brought to the consummation of the deal a rest energy equal to one electron mass.
In brief, why do we have to put in five electron masses of energy to get out a
three-electron-mass product? Simply asking this question points out where the expla
System momentum means
nation lies. The incident photon brings in a great momentum, and the electron with
not all system energy a vailab le
to create particles which it reacts has no momentum. So all that momentum has to go into the output
product, the polyelectron. Since the polyelectron must have momentum, it must also
have kinetic energy — energy not available for creating additional mass. In conse
quence, that object has so much energy of motion that only a much diminished part of
the energy of the incident photon is available for the creation process itself.
8.6 MATERIAL PARTICLE USED TO

CREATE MASS
proton hits proton, creates proton-antiproton
pair
Particles other than the photon can also create particles. A particle of any type can carry
Any energetic particle can create
other particles
enough energy to create particles similar to or different from itself Each such creation
must not only follow momenergy conservation laws of special relativity, but it is also
subject to the law of conservation of total electric charge and other conservation laws,
as described in elementary particle physics.
8.6 MATERIAL PARTICLE USED TO CREATE MASS 235
BACKYARD ZOO OF PARTICLES

This is not a textbook of particle physics, but our examples include interac
tions between common particles. Here are brief descriptions of some of
them.
Electron
Electrons form the outer structure of every atom and rattle around in approx
imately 99.99999999999 percent of its volume. The mass of the electrons of
an atom, however, accounts for only about one two-thousandth of its mass or
less. The electron carries a negative “ elementary” electrical charge. Every
accepted theory of particle physics treats the electron itself as an elementary
particle — it is not made up of anything more fundamental. The positron is the
antiparticle of the electron, with the same mass but a positive elementary
charge. When positron and electron meet, sooner or later they mutually
annihilate, yielding two or more high-energy photons (gamma rays). This will
be the fate of the positron and one of the electrons in the polyelectron
discussed in Section 8.5 soon after they begin to orbit one another.
Proton
The proton (Greek for “ the first one” ) is, with the neutron, the most massive
constituent of atomic nuclei. The simplest atom, hydrogen, in its most abun
dant form has a single proton as nucleus. The proton has a positive charge
equal in magnitude to that of the electron, but a mass almost two thousand
times as great as that of the electron. As far as we know the proton is stable,-
experiments have shown its lifetime to be greater than 10^' years — very
much longer than the current age of th^ universe (about 10 ’° years). Particle
physicists postulate that protons (and neutrons) are composed of still-more
elementary particles called quarks. The antiproton, antiparticle of the pro
ton, has mass equal to that of the proton but negative unit charge. When it
encounters a proton, the two particles annihilate, sometimes creating gamma
rays but more often other particles not listed in this box.
Neutron
The neutron (from Latin neuter— “ neither” ; neither positively nor negatively
charged) is similar to the proton but has no charge and has slightly greater
mass. It is a constituent of all nuclei except for the most abundant form of
elementary hydrogen. When not in a nucleus, the neutron decays into a
proton, electron, and neutrino with half-life of about 10 minutes.
Photon
The photon, the quantum of light, has zero mass. Its properties are described
in Section 8.4.
Neutrino
There are several kinds of neutrinos, all of which appear to have zero mass
and to move at light speed. The neutrino (Italian for “ little neutral one” ) has
no charge and interacts only weakly with ordinary matter: Neutrinos of
certain energies can pass through a block of lead one light-year thick with
only a 5 0 - 5 0 chance of being absorbed! An immense flux of neutrinos
passes continually through our bodies without injuring us. “ Ten million trillion
[10” ] neutrinos will speed harmlessly through your brain and body in the
time it takes to read this sentence. By the time you have read this sentence,
they will be farther away than the moon.”
BEFORE AFTER
1
mass
1
mass
4 '
massAV
(48) 1/2 ' . +^+ - (48) 1/2
8
FIGURE 8-9. Conservation o f energy a n d momentum in the process o f creation o f a proton -

antiproton p a ir by the im pact of a proton on another proton. Before; The incoming proton (sloping
arrow) moves with a speed (48)'/^ f l = 9 9 percent that of light. The target proton initially stands a t rest
(vertical arrow). After; The resulting three protons and one antiproton are kicked to the right a t (48)'!^ j 8
= 8 1 percent of light speed.
Figure 8-9 shows “the creation of a proton-antiproton pair by a proton in the

presence of another proton.” The antiproton has mass equal to that of the proton but
carries a negative unit charge (Box 8-1). The interaction shown leaves all four resulting
particles moving along together. The resulting particles stay together when the incom
ing particle has the lowest energy that can create the additional pair. This minimum
Threshold energy defined
energy is called the th re sh o ld energy. W e don’t want the four particles to move
apart after the creation. If they did, we would have to supply the incoming particle
with additional kinetic energy. It would have to carry an energy greater than the
threshold energy. We discuss here the threshold energy of the incoming proton.
Magnitudes of the momenergy vectors displayed in Figure 8-9 are expressed in
“natural units” for the proton, namely the mass of the proton itself, 1.67 X 10“ ^^
kilograms or 938.27 MeV. This time the numbers are not all integers; the momentum
of the system has a value equal to the square root of 48, or 6.928 proton masses.
The creation of a proton-andproton pair by a PROTON requires a total of eight
proton units of energy to create two proton units of mass. In contrast the creation of an
electron-antielectron pair by a PH O TO N requires a total of only four electron units
of energy to create two electron units of mass. Why is the photon process so much more
‘Efficiency” of particle production
efficient (in units of mass of the struck particle) than the proton process? Answer; The
photon is annihilated in the creation process. In contrast, the incoming proton is not
annihilated; the bookkeeper must keep the incoming proton on the payroll, providing
momenergy after rhe collision to keep the proton in step with the other three particles.
This after-collision momenergy of the proton is not available to be applied to other
products of the collision. Therefore a proton of given total energy can create less mass
than a photon of the same energy when each strikes a stationary target,
8.7 CONVERTING MASS TO USABLE ENERGY; FISSION, FUSION, ANNIHILATION 237
8.7 CONVERTING MASS TO USABLE

ENERGY: FISSION, FUSION, ANNIHILATION
fission and fusion both slide down the energy
hill toward the minimum, iron, electron and
positron annihilate to yield two energetic
photons.
For a final perspective on the evanescence of mass and the preservation of momenergy,
turn from processes where mass is created to three processes in which mass is destroyed:
fission, fusion, and annihilation.
Anyone who first hears about the splitting of a nucleus (fission) as a source of
energy, and the joining of two nuclei (fusion) also as a source of energy might gain the
mistaken impression that a perpetual motion machine has been invented. Could we
split and join the same nucleus over and over again, each time releasing energy? No.
Here’s why. Fission occurs in the splitting of uranium, for instance when a neutron
strikes a uranium nucleus:
on l^Rh + iîCs
In this equation the lower-left subscript tells the number of protons in the given
nucleus and the upper-left superscript shows number of protons plus neutrons in the Fission and fusion; Both go from
looser to tighter binding
nucleus. The process described by this equation rearranges the 236 nucleons, that is,
92 protons plus 144 neutrons, into a configuration that comes a bit closer to that most
stable of all available nuclear configurations, the iron nucleus:
iiFe
But fusion too, for example the process of uniting two rather light nuclei such as
“heavy hydrogen” or deuterons to form a helium nucleus.
\D + \D ^He
can also be regarded as one step along the way toward rearranging nucleons (protons
and neutrons) to achieve the iron configuration or something like it.
In brief, we can get energy out of nucleon rearrangement processes that move from
looser binding of borh heavier and lighter nuclei roward tighrer binding of the
(intermediate-mass) iron nucleus (Figure 8-10). In neither fission nor fusion, how
ever, is the fraction of mass converted into energy as great as one percent. (For an
example of fusion reacrion in Sun, see Sample Problem 8-5, especially c.)
A n n ih ilatio n is interesting because it can convert 100 percent of matter into
Annihilation converts 100%
radiation. Annihilation is interesting, too, because it has been demonstrated on the of matter into radiation
microscopic scale. A slow positive electron, a positron, joining up by chance to orbit
with an everyday negarive electron, evenrually unites with it to annihilate them both
and produce sometimes two, sometimes three lighr quanta (photons — called
g am m a rays in the case of these high energies):
e*' e —^ 2 or 3 photons
Figure 8-11 displays rhe balance of energy and momentum in the two-quantum
annihilation process.
o
S
FIGURE 8-10. Both the conversion of deute

rium to the more massive helium in fusion and
the conversion of uranium to lighter nuclei in
Mass number of the nucleus fission decrease the mass per nucleon, both
(not a linear scale) toward the most stable of nuclei, iron.
Why 2 or 3 photons? Why can't just a single photon be emitted in this process?
' Brief answer: Conservation of momentum. Fuller answer: Look at Figure 8-11.
Before annihilation, the system has zero total momentum. A single photon remain
ing after the annihilation could not have zero momentum, no matter in which
direction it moved! The presence of a single photon after the collision could not
satisfy conservation of momentum. So annihilation never does and never can end up
giving only a single photon.
A
energy, 1
1
1 momentum, 0
mass
1[sum]
1 energy, 1 i \ 0 j 0
—vmass • mass^,—
momentum, 0
mass 1 enWgy j T Q J 1 energy
- 1 momentum 1 momentum
BEFORE AFTER
mass
1 0 0
mass mass
1 ) ^ < W V \A A /W V > -
v= 1 v= 1
_Meii) 1
^mass
FIGURE 8-11 . Momenergy conservation in the two-photon electron-pmsitron annihilation pro-
cess. Before'. Before annihilation each oppositely charged particle has rest energy and no momentum. After:
The two particles have annihilated, creating two high-energy photons (gamma rays). The two photons fly
apart in opposite directions; total momentum remains zero.
CONVERTING MASS TO USABLE ENERGY: FISSION, FUSION, ANNIHILATION 239
ANALYZING A PARTICLE ENCOUNTER
Conservation of total momenergy! In any given free-float frame that means

conservation of total energy and conservation of each of three components
of total momentum. In no way does the power and scope of this principle
make itself felt more memorably than the analysis of simple encounters of
this, that, and the other kind in an isolated system of particles. “Analyzing an
encounter” means using conservation laws and other relations to find un
known masses, energies, and momenta of particles in terms of known quanti
ties. Sometimes a complete analysis is not possible; the information provided
may be insufficient. Here are suggested steps in analyzing an encounter.
Sample Problems 8-3 and 8-4 illustrate these methods.
1. Draw a diagram of particles before and particles a ft e r the interaction.

Label particles entering with numbers or letters and particles leaving with
different numbers or letters (even if they are the same particles). Use
arrows to show particle directions of motion and label with symbols their
masses, energies, and momenta, whether initially known or unknown.
2. Write down algebraically the conservation of total energy. Do not forget

to include the rest energy — the mass m — of any particle not moving in
the chosen free-fioat frame.
3. Write down algebraically the conservation of total momentum. Do not

forget that momentum is a vector. In general this means demanding con
servation of each of three components of total momentum.
4. Try to solve for unknowns in terms of knowns, still using symbols.
a. Make liberal use of the relation m^ = — p^, where p^ = p„^ + p^^ +

p^^. For a photon or neutrino, mass equals zero and £ = p (in magni
tude: Pay attention to the direction of the momentum vector p — or its
sign if motion is in one space dimension).
b. Do NOT use speed v of a particle unless forced to by requirements of

the problem. Relativistic particles typically move with speeds very
close to light speed, so speed proves to be a poor measure of signifi
cance. Increase by one percent the speed of a particle moving at v =
0.99 and you increase its energy by a factor of almost 10.
c. Substitute numerical values into resulting equations as late as possible.

Before substituting numerical values, check that all values are ex
pressed in concordant units.
5. Check your result. Check units of the solution. Is the order of magnitude of
numerical results reasonable? Substitute limiting values, for example let
ting energy of an incoming particle become very large (and very small). Is
the limiting-case result reasonable?
Is there any general conclusion you can draw from your specific solution?
Does this exercise illustrate a deep principle or lead to an even more inter
esting application of conservation laws?
SYMMETRIC ELASTIC C O L L IS IO N
A proton of mass m and kinetic energy K in the H istorical note; When impact speed is small
laboratory frame strikes a proton initially at rest in compared to the speed of light, this separation of
that frame. The two protons undergo a symmetric directions, 0, is 90 degrees, according to Newto
elastic collision; the outgoing protons move in di nian mechanics. Early cloud-chamber tracks
rections that make equal and opposite angles 0 /2 sometimes showed symmetric collisions with
with the line of motion of the original incoming angles of separation substantially less than 90 de
particle. Find energy and momentum of each out grees, thereby giving evidence for relativistic me
going particle and angle 0 between their outgoing chanics and providing the first reliable measure
directions of motion for rhis symmetric case. ments of impact energy.
SOLUTION/ following steps in Box 8-2
1. D raw a diag ram and label all four particles with letters:
'P c
Symmetry of this diagram implies that the two outgoing particles have equal
energy and equal magnitude of momentum; that is, — Ej and (in magnitude)
Pc = Pd-
2. C onservation o f energy: Energy of each particle equals mass plus kinetic

energy. And the masses don’t change in this reaction. Therefore rotal kinetic
energy after the encounter (divided equally between the two particles) equals the
(known) total kinetic energy before the encounter, all localized on one particle. In
brief; K ^= K j — K J 2 = K /2. Simple answer to one of the three questions we
were asked!
3. C onservation o f m o m en tu m : By symmetry, rhe vertical components of
momenta of the outgoing particles cancel. Horizontal components add, leading
to the relation
A ot Pa ^ Pc cos(0/2) -b p j cos(0/2) = 2pj cos(0/2)
or, in brief.
p ^ — 2 p j C O s (0 / 2 ) [conservation of momentum]
4. Solve fo r th e u n k n o w n angle 0: Along the way find the other requested

quantity, the magnitude p c ~ pd of the momenta after the collision. To that end,
first find the momentum p^ before the collision, using the general formula for the
momentum of an individual particle:
8.7 CONVERTING MASS TO USABLE ENERGY: FISSION, FUSION, ANNIHILATION 241
P = { E ^ - w2]V2 = + m y - »?2}1/2 = (^^2 + 2mK + - m^y/^

= (K^ + 2m K y/^
Therefore
= (K^ + 2/wK)*/2
From conservation of energy, K^ = K j = K /2. Therefore
p , = [{K /2 y + 2m {K /2)y/^
Substitute these expressions for p^ and p j into the equation for conservation of
momentum:
(fC2 + 2« fO '/2 = 2[(fC/2)2 + 2m (K /2)y/^cos(9/2)
Square both sides and solve for cos^(0/2) to obtain
„ K + 2m
Now apply to this result the trigonometric identity
, _ (cos 0 + 1 )
cos2(0/2) = ^
After some manipulation, obtain the desired result:
(K/m)
cos e =
(K/m) + 4
Here K is the kinetic energy of the incoming particle, m the mass of either particle,
and 6 the angle between outgoing particles. This result assumes (1) an elastic
collision (kinetic energy conserved), (2) one particle initially at rest, (3) equal
masses of the two particles, and (4) the symmetry of outgoing paths shown in the
diagram.
5a. L im iting case; Low energy. In the case of low energy (Newtonian limit), the
incoming particle has a kinetic energy K very much less than its rest energy m, so
the ratio K/m approaches zero. In the limit, cos 6 becomes zero and 9 = 90
degrees. This is the accepted Newtonian result for low velocities (except for an
exactly head-on collision, in which case the incoming particle stops dead and the
struck particle moves forward with the same speed and direction as the original
incoming particle).
5b. L im itin g case: H igh energy. For extremely high-energy elastic collisions, the
incident particle has a kinetic energy very much greater than its rest energy, so the
ratio K/m increases without limit. In this case the quantity 4 in the denominator
becomes negligible compared with K/m, so numerator and denominaror both
approach the value K/m, with the result cos 0 ^ 1 and 0 - ^ 0 . This means that
in the special symmetric case discussed here both resulting particles go forward in
the same direction as the incoming particle, sharing equally the kinetic energy of
the incoming particle.
For an incoming particle of very high energy, the elastic collision described here
is only one of several possible outcomes. Alternative processes include creation of
new particles.
ANNIHILATION
A positron of mass m and kinetic energy equal to its with respect to the direction of the incident posi
mass strikes an electron at rest. They annihilate, tron. W hat are the energies of both photons (in
creating two high-energy photons. One photon units of mass of the electron) and direction of
enters a detector placed at an angle of 90 degrees motion of the second photon?
SOLUTION/ following steps in Box 8-2
1. D raw a diagram and label the particles with letters.
Pb = 0
Eb = m Ec, Pc
Ill m
â/ pa
0 b
Ed, Pd
BEFORE AFTER
2. C onservation o f energy expressed ii^ the symbols of the diagram, and includ
ing the rest energy of the initial stationary particle:
m '■ E .+ E,
3. C onservation o f each co m p o n en t o f to tal m o m en tu m :
P x tot Pa P c COS 0 [horizontal momentum]

Pywc = 0 = p , s i n 6 - Pj [vertical momentum]
4. Solve; First of all, the problem states that the kinetic energy K of the incoming
positron equals its rest energy m. Therefore its total energy E^ = m-\~ K = m-\- m
= 2m. Second, the outgoing particles are photons, for w h i c h a n d pj = Ej
in magnitude, respectively. W ith these substitutions, the three conservation
equations become
E^ m — 2m m ^ tn ~ E^ E j [conservation of energy]
Pa = E, cos 6 [conservation of horizontol momentum]
Ej = sin 6 [conservation of vertical momentum]
S A M P L E P R O B L E M 8- 5
C O N V E R S I O N OF M A S S TO
E N E R G Y IN SUN
Luminous energy from Sun pours down on the stant. The radius of Earth equals approximately
outer atmosphere of Earth at a rate of 1372 watts 6.4 X 10^ meters and the Earth-Sun distance
per square meter of area that lies perpendicular to equals 1.5 X 10“ meters. The mass of Sun is
the direction of this radiation. The figure 1372 approximately 2.0 X 10^° kilograms.
watts per square meter has the name so lar co n
243
These are three equations in three unknowns and £ând 6. Square both sides of
the second and third equations, add them, and use a trigonometric identity to get
rid of the angle 9:
p j + E / = £/(cos^ 6 + sin^ 9) = E /
Substitute p / = E / — nE on the left side of this equation and again use E^ = 2m

to obtain a first expression for £ /:
E^ = E ^ — rrp- E / = ArtE — rrp- E / = ^np- + £ /
Now solve the equation of conservation of enetgy for E^ and square it to obtain a
second expression for £ /;
£ / = {5m - EjY = 9m^ - 6mEj + £ /
Equate these two expressions for E / and subtract E / from borh sides to obtain
5rrP = 9m^ — 6mEj
Solve for unknown Ej.
9nP — 5nP 6m^

Ej = ---------------------------- m
Gm Gm
This yields our first unknown. Use this result and conservation of energy to find
an expression for E/.
E^ = 5m — Ej = 5m — m = 2m
Finally, angle 9 comes from conservation of vertical momentum. For a photon p

= £, so
sin 9 '
£^ 2m 2
from which 9 — 50 degrees. W e have now solved for all unknowns: £„ — 2m,
Ej = m, and 9 = 50 degrees.
5. L im iting cases: There is no limiting case here, since the energy of the incoming
positron is specified fully in terms of the mass m. common to electron and positron.
a. How much mass is converted to energy every second in Sun to supply the
luminous energy that falls on Earth?
b. W hat total mass is convened to energy every second in Sun to supply luminous
energy?
c. Most of Sun’s energy comes from burning hydrogen nuclei (mostly protons) into
helium nuclei (mostly a two-proton - two-neutron combination). Mass of the
proton equals 1.67262 X 10“ ^^ kilogram, while the mass of a helium nucleus of
this kind equals 6.64648 X 10“ ^^ kilogram. How many metric tons of hydrogen
must Sun convert to helium every second to supply its luminous output? (One
metric ton is equal to 1000 kilograms, or 2200 pounds.)
d. Estimate how long Sun will continue to warm Earth, neglecting all other processes
in Sun and emissions from Sun.
SOLUTION
a. One watt equals one joule per second = one kilogram meter^/second^. We want
to measure energy in units of mass— in kilograms. Do this by dividing the
number of joules by the square of the speed of light (Section 7.5 and Table 7-1):
1372 joules _ 1.372 X 10^ kilogram metersVsecond^

9.00 X 10^®meters^/second^
= 1.524 X 10“ ''' kilograms
Thus every second 1.524 X 10“ *'' kilogram of luminous energy falls on each
square meter perpendicular to Sun’s rays. The following calculations are based on
a simplified model of Sun (see last paragraph of this solution). Therefore we use
the approximate value 1.5 X 10“ *'' kilogram per second and two-digit accuracy.
W hat total luminous energy falls on Earth per second? It equals the solar
constant (in kilograms per square meter per second) times some area (in square
meters). But what area? Think of a huge movie screen lying behind Earth and
perpendicular to Sun’s rays (see the figure). The shadow of Earth on this screen
forms a circle of radius equal to the radius of Earth. This shadow represents the
zone of radiation removed from that flowing outward from Sun. Call the area of
this circle the cross-sectional area A of Earth. Earth’s radius r = 6.4 X 10^ meters,
so the cross-sectional area A seen by incoming Sunlight equals A = = 1.3 X
10*'' meters^. Hence a total luminous energy equal to (1.5 X 10“ *'' kilograms/
meter^) X (1.3 X 10*'' meters^) = 2.0 kilograms fall on Earth every second. This
equals the mass converted every second in Sun to supply the light incident on
Earth.
light from Sun
8.8 SUMMARY
mass: the magnitude off the 4-vector called
momenergy
Mass can be converted into energy and energy can be converted into mass” — this is a
loose and sometimes misleading way to summarize some consequences of the two
245
b. Assume that Sun delivers sunlight at the same ‘‘solar-constant rate” to every part
of a sphere surrounding Sun of radius equal to the Earth - Sun distance. The area
of this large sphere has the value 471/?^ where i? = 1.5 X 10“ meters, the average
distance of Earth from Sun. This area equals 2.8 X 10^^ meters^. Therefore Sun
converts a total of 2.8X 10^^ meters^ X 1.5 X 10“ kilograms/meter^ (from a)
= 4.2 X 10^ kilograms of mass into luminous energy every second, or about 4
million metric tons per second.
c. Through a series of nuclear processes not described here, four protons transform
into a helium nucleus consisting of two protons and two neutrons. The four
original protons have a mass 4 X 1.67262 X 10“ ^^ = 6.69048 X 10“ ^^
kilogram. The helium nucleus has a mass 6.64648 X 10~^^ kilogram. The
difference, 0.04400 X 10"^^ kilogram, comes out mostly as light. (We cannot
use two-digit accuracy here, because the important result is a difference between
nearly equal numbers.)
The ratio of hydrogen burned to mass converted equals 6.69048/0.04400 =
150 (back to two-digit accuracy!). So for each kilogram of mass converted to
electromagnetic radiation, 150 kilograms of hydrogen burn to helium. In other
words, about 0.7 percent of the rest energy (mass) of the original hydrogen is
converted into radiation. Hence in order to convert 4.2 X 10^ kilograms per
second into radiation. Sun burns 1 5 0 X 4 .2 x 1 0 ^ kilograms per second = 6.3 X
10“ kilograms of hydrogen into helium per second — about 630 million metric
tons each second.
d. W e can reckon Sun’s mass by figuring how much Sun gravity it takes to guide our
planet around in an orbit of 8 light-minute radius and one year time of circuit.
Result: about 2.0 X 10^° kilograms. If Sun were all hydrogen, then the process of
burning to helium at the present rate of 6.3 X 10“ kilograms every second would
take (2.0 X 10^® kilograms)/(6.3 X 10“ kilograms/second) = 3.2 X 10'®
seconds. At 32 million seconds per year, this would last about 10“ years, or 100
billion years.
O f course the evolution of a star is more complicated than the simple conver
sion of hydrogen into helium-plus-radiation. Other nuclear reactions fuse helium
into more massive nuclei on the way to the most stable nucleus, iron-56 (Section
8.7). These other reactions occur at higher temperatures and typically proceed at
faster rates than the hydrogen-to-helium process. Sun emits a flood of neutrinos
(invisible; detected with elaborate apparatus; amount presently uncertain by a
factor of 2, carry away less than 1 percent of Sun’s output). Sun also loses mass as
particles blown away from the surface, called the so lar w ind. And stars do not
convert all their hydrogen to helium and other nuclei — or live for 100 billion
years. According to current theory, the lifetime of a star like Sun equals approxi
mately 10 billion years ( lO'" years). We believe Sun to be 4 to 5 billion years old.
The remaining 6 billion years (6 X 10^ years) or so should be sufficient time for
our descendants to place themselves in the warmth of nearby stars.
principles that ate basic and really accurate: (1) The total momenergy of an isolated
system of particles remains unchanged in a reaction; (2) The invariant magnitude of
the momenergy of any given particle equals the mass of that particle.
How much sound infotmation about physics can be extracted from these basic
principles? W hat troubles sometimes atise from accepting a too loose formulation of
the ‘‘principle of equivalence of mass and energy”? Some answers to these questions
appeat in the dialog that follows, which serves also as a summary of this chaptet.
DIALOG: USE AND ABUSE OF THE

CONCEPT OF MASS
Does an isolated system have the same mass as Yes. Given in terms of energy E and momentum p
observed in every inertial (free-float) reference by ■— in one frame, by rrp- = {E'Y
frame? (p 'Y in another frame. Mass of an isolated system is
thus an invariant.
Does its energy have the same value in every inertial No. Enetgy is given by E = {nY + or
frame?
E = m /{\ —
or
E — (mass) + (kinetic energy) = m-\- K
Value depends on the frame of reference from which

the patticle (or isolated system of particles) is ob
served. Value is lowest in the frame of reference in
which the particle (ot system) has zero momenmm
(zero total momentum in the case of an isolated
system of particles). In that frame, and in that frame
only, energy equals mass.
Does energy equal zero for an object of zeto mass, No. Energy has value £ = (0^ + pYY^^ = p (or in
such as a photon or neutrino or graviton? conventional units = cpôaY- Alternatively one
can say— formally— that the entire energy resides
in the fotm of kinetic energy {K = p in this special
case of zero mass), none at all in the form of rest
energy. Thus,
£ = (mass) + (kinetic energy) = ^ K= K= p
(case of zero mass only!).
Can a photon — that has no mass— give mass to an Yes. Light with energy £ transfers mass m = Eiy=
absorber? Econv/^) to a heavy absorber (Exercise 8.5).
Invariance of mass; Is that feature of nature the same No. It is tme that all elementary particles of the
as the principle that all electrons in the universe have same kind have the same mass. However, that is a
the same mass? fact totally distinct from the principle that the mass
of an isolated system has identical value in whatever
free-float frame it is figured (invariance of system
mass).
Invariance of mass: Is that the same idea as the No. Conservation of momenergy— the principle
conservation of the momenergy of an isolated sys valid for an isolated system — says that the momen-
tem? ergy 4-vector figured before the constiments of a
system have interacted is identical to the momen
ergy 4-vector figured after the constituents have
intetacted. In contrast, invariance of mass — the
magnitude of the momenergy 4-vector— says that
that mass is the same in whatever free-float frame it
is figured.
8.8 SUMMARY: USE AND ABUSE OF THE CONCEPT OF MASS 247
Momenergy: Is that a richer concept than mass? Yes. Momenergy 4-vector reveals mass and more:
the motion of object or system with the mass
Conservation of the momenergy of an isolated sys Yes. Mass of an isolated system, being the magni
tem: Does this imply that collisions and interactions tude of its momenergy 4-vector, can never change
within an isolated system cannot change the sys (as long as the system remains isolated).
tem’s mass?
Conservation of the momenergy of an isolated sys No! The constituents often change in a high-speed
tem: Does this say that the constituents that enter a encounter.
collision are necessarily the same in individual mass Exam ple 1: Collision of two balls of putty that
and in number as the constituents that leave that stick together— after collision hotter and therefore
collision? very slightly more massive than before.
E xam ple 2: Collision of two electrons (e~) with
sufficient violence to create additional mass, a pair
consisting of one ordinary electron and one positive
electron (positron: e'^):
e (fast) + e (at rest) + 3e-.
Exam ple 3: Collision that radiates one or more

photons:
e” (fast) + e~ (at rest) ~
^ electromagnetic \
electrons of \
( intermediate 1 -b
energy (photons) |
emitted in the I
speed /
\ collision process /
In all three examples the system momenergy and
system mass are each the same before as after.
Can I figure the mass of an isolated system composed Ordinarily N O , but yes in one very special case: Two
of a number, n, of freely-moving objects by simply noninteracting objects move freely and in step, side
adding the masses of the individual objects? Exam by side. Then the mass of the system does equal the
ple: Collection of fast-moving molecules. sum of the two individual masses. In the general
case, where the system parts are moving relative to
each other, the relation between system mass and
mass of parts is not additive. The length, in the sense
of interval, of the 4-vector of total momenergy is not
equal to the sum of the lengths of the individual
momenergy 4-vectors, and for a simple reason: In
the general case those vectors do not point in the
same spacetime direction. Energy however, does add
and momentum does add:
^system and P x, system P x ,i

i= 1 i= 1
From these sums the mass of the system can be
evaluated:
M 2 = P 2 _ *2 — /)2 — *2
^ system system r x, system r y, system r z, system
Can we simplify this expression for the mass of an Yes. In this case the mass of the system has a value
isolated system composed of freely moving objects given by the sum of energies of individual particles:
when we observe it from a free-float frame so chosen
M system — '^system
F [in zero-total
as to make the total momentum be zero? momentum frame]
Moreover, rhe energy of each particle can always be

expressed as sum of rest energy m plus kinetic energy
K:
Ei = nii-\- K: a = 1 ,2 ,3 , , «)
So the mass of the system exceeds the sum of the

masses of its individual particles by an amount equal
to the total kinetic energy of all particles (but only as
observed in the frame in which total momentum
equals zero);
= + [in zero-total
momentum fromel
1= 1
For slow particles (Newtonian low-velocity limit)

the kinetic energy term is negligible compared to the
mass term. So it is natural that for years many
thought that the mass of a system is the sum of the
masses of its parts. However, such a belief leads to
incorrect results at high velocities and is wrong as a
matter of principle at all velocities.
W hat’s the meaning of mass for a system in which The energies of interaction have to be taken into
the particles interact as well as move? account. They therefore contribute to the total en
ergy, fisystem , that gives the mass
M
^ ^system = (E^
'•'-'system ■r )l/2
system/
How do we find out the mass of a system of particles Weigh it! Weigh it by conventional means if we are
(Table 8-1) that are held — or stick — together? here on Earth and the system is small enough, other
wise by determining its gravitational pull on a satel
lite in free-float orbit about it.
Does mass measure “amount of matter’’? Nature does not offer us any such concept as
“amount of mattet. ’’ History has struck down every
proposal to define such a term. Even if we could
count number of atoms or by any other counting
method try to evaluate amount of matter, that
number would not equal mass. First, mass of the
specimen changes with its temperature. Second,
atoms tightly bonded in a solid weigh less — are less
massive — than the same atoms free. Third, many
of nature’s atoms undergo radioactive decay, with
still greater changes of mass. Moreover, around us
occasionally, and continually in stars, the number of
atoms and number of particles themselves undergo
change. How then speak honestly? Mass, yes;
“amount of matter,’’ no.
Does the explosion in space of a 20-megaton hydro Yes and no! The question needs to be stated more
gen bomb convert 0.93 kilogram of mass into en carefully. Mass of the system of expanding gases,
ergy (fusion, Section 8.7)? conv/^^ “ fragments, and tadiation has the same value imme
(20 X 10^ tons TNT) X (10^ grams/ton) X (10^ diately after explosion as before; mass M of the
calories/gram of “TN T equivalent’’) X (4.18 system has not changed. However, hydrogen has
joules/calorie)/c^ = (8.36 X 10*^ joules)/(9 X been transmuted to helium and other nuclear trans
10^® meters^/second^) = 0.93 kilogram} formations have taken place. In consequence the
makeup of mass of the system
8.8 SUMMARY: USE AND ABUSE OF THE CONCEPT OF MASS 249
= + (in zero-total
K momentum frame]
has changed. The first term on the right— sum of

masses of individual constituents — has decreased
by 0.93 kilogram:
2 ) ~ ( 2 kilogram
The second term — sum of kinetic energies, includ

ing “kinetic energy’’ of photons and neutrinos
produced — has increased by the same amount:
2) ) “ ( 2 kilogram
/■= 1 / after \ i= 1 / before
The first term on the right side of this equation —

the original heat content of the bomb — is practi
cally zero by comparison with 0.93 kilogram.
Thus part of the mass of constituents has been con
verted into energy; but the mass of the system has not
changed.
The mass of the products of a nuclear fission explo Yes! The key point is the waiting period, which
sion (Section 8.7: fragments of split nuclei of ura allows heat and radiation to flow away until trans
nium, for example) — contained in an underground muted materials have practically the same heat con
cavity, allowed to cool, collected, and weighed — is tent as that of original bomb. In the expression for
this mass less than the mass of the original nuclear the mass of the system
device?
[in zero-total
M.system = 2 momentum frame]
the second term on the right, the kinetic energy of

thermal agitation — whose value rose suddenly at
the time of explosion but dropped during the cool
ing period — has undergone no net alteration as a
consequence of the explosion followed by cooling.
In contrast, the sum of masses
has undergone a permanent decrease, and with it the

mass M of what one weighs (after the cooling period)
has dropped (see the figure).
sum of masses
of individual particles
Em;
Does Einstein’s statement that mass and energy are No. Value of energy depends on the free-float frame
equivalent mean that energy is the same as mass? of reference from which the particle (or isolated
system of particles) is regarded. In contrast, value of
mass is independent of inertial frame. Energy is only
the time component of a momenergy 4-vector,
whereas mass measures entire magnitude of that
4-vector. The time component gives the magnitude
of the momenergy 4-vector only in the special case in
which that 4-vector has no space component; that is,
in a frame in which the momentum of the particle
(or the total momentum of an isolated system of
particles) equals zero. Only as measured in this spe
cial ze ro -m o m en tu m frame does energy have the
same value as mass.
Then what is the meaning of Einstein’s statement Einstein’s statement refers ro the reference frame in
that mass and energy are equivalenr? which the particle is at rest, so that it has zero
momenmm p and zero kinetic energy K. Then E =
m-\- K -* m -{ -0 .ln that case the energy is called rhe
rest energy of the particle:
In this expression, recall, the energy is measured in

units of mass, for example kilograms. Multiply by
the conversion factor to express energy in conven-
rional units, for example joules (Table 7-1). The
result is Einstein’s famous equation:
p = ffKp'
Many treatments of relativity fail to use the sub

script “rest’’— needed to remind us that this equiv
alence of mass and energy refers only to the rest
energy of the particle (for a system, the total energy
in the zero-total-momentum frame).
W ithout delving into all fine points of legalistic The conversion factor c^, like the factor of conver
phraseology, how significant is the conversion factor sion from seconds to meters or miles to feet (Box
in the equation = mc^} 3-2), today counts as a detail of convention, rather
rhan as a deep new principle.
If the factor is not the central feature of the The distinction between mass and energy is this:
relationship between mass and energy, what is cen Mass is the magnitude of the momenergy 4-vector
tral? and energy is the time component of the same 4-vec
tor. Any feature of any discussion that emphasizes
this contrast is an aid to understanding. Any slurring
of terminology thar obscures rhis distinction is a
potential source of error or confusion.
Is the mass of a moving object greater than the mass No. It is the same whether the object is at rest or in
of the same object at rest? motion; the same in all frames.
Really? Isn’t the mass, M, of a system of freely Ouch! The concept of “relativistic mass” is subject
moving particles given, not by the sum of the masses to misunderstanding. That’s why we don’t use it.
W; of the individual constiments, but by the sum of First, it applies the name mass— belonging to the
REFERENCES 25 1
energies £,■ (hut only in a frame in which total mo magnimde of a 4-vector— to a very different con
mentum of the system equals zero)? Then why not give cept, the time component of a 4-vector. Second, it
a new name and call it “relativistic mass’’ of the makes increase of energy of an object with velocity or
individual particle? Why not adopt the notation momentum appear to be connected with some
change in internal strucmre of the object. In reality,
«, rel = E; = m; + K, ? the increase of energy with velocity originates not in
the object but in the geometric properties of space-
W ith this notation, can’t one then write time itself.
[in zero-total
momentum frame]
i = 1
In order to make this point clear, should we call That is what we called it in the first edition of this
invariant mass of a particle its “rest mass’’? book. But a thoughtful student pointed out that the
phrase “rest mass’’ is also subject to misunderstand
ing; W hat happens to the “rest mass’’ of a particle
when the particle moves? In reality mass is mass is
mass. Mass has the same value in all frames, is
invariant, no matter how the particle moves. {Gali
leo: “In questions of science the authority of a thou
sand is not worth the humble reasoning of a single
individual.’’}
Can any simple diagram illustrate this contrast be Yes. The figure shows the momentum-energy 4-
tween mass and energy? vector of the same particle as measured in three
different frames. Energy differs from frame to
frame. Momentum differs from frame to frame.
Mass (magnitude of 4-vector, represented by the
length of handles on the arrows) has the same value,
m = 8, in all frames.
particle
at rest: p ' = - 6
p =0
£=8
8
mass
3 \ \ \£ '= 1 0
mass
LABORATORY ROCKET SUPER-ROCKET

FRAME FRAME FRAME
REFERENCES
Quotation from Count Rumford in Section 8.2: Sanborn C. Brown, Benjamin
Thompson, Count Rumford (MIT Press, Cambridge, Mass., 1979), page 220.
Reference to measurement of very small effects in Section 8.2: Vladimir Bra
ginsky and A. B. Manukin, “Quantum non-demolition,” in Measurement of
Weak Forces in Physics Experiments, edited by David H. Douglas (University of
Chicago Press, 1977).
Quotation from biography of A. H. Compton in Section 8.4: Robert S. Shank-

land, Dictionary of Scientific Biography, edited by Charles Coulston Gillespie,
Volume III (Charles Scribner’s Sons, New York, 1 9 7 1).
Compton scattering reported in A. H. Compton, Physical Review, Volume 2 2 ,
pages 4 0 9 - 4 1 3 (1 9 2 3 ).
The poly electron mentioned in Section 8 .5 has been independently generated,

through interaction of a slow positron with the electrons of a metal surface, by
Alan Mills, Jr., at Bell Telephone Laboratories, as reported in Physical Review
Letters, Volume 4 6 , pages 7 1 7 - 7 2 0 (1 9 8 1 ).
Final quotation in Box 8-1: Timothy Ferris, Coming of Age in the Milky Way
(Anchor Books, Doubleday, New York, 1 9 8 8 ), page 3 4 4 .
Sample Problem 8-5 was suggested by Chet Raymo’s science column in the
Boston Globe, May 2, 1 9 8 8 , page 35.
Galileo quote in final dialog: Galileo Galilei, Dialogo dei due massimi sistemi del
mundo, Landini, Florence. Translation by S. Drake, Galileo Galilei — Dialogue
Concemingthe Two Chief World Systems— Ptolemaic and Copemican, University
of California Press, Berkeley and Los Angeles, 19 5 3.
ACKNOWLEDGMENTS
We thank colleagues old and young for the comments that helped us clarify,
formulate, and describe the concept of mass in this chapter and in the final dialog,
and very specially Academician Lev B. Okun, Institute of Theoretical and Exper
imental Physics, Moscow, for correspondence and personal discussions. W e be
lieve that our approach agrees with that in two of his articles, both entitled “The
Concept of Mass,” which appeared in Physics Today, June 1 9 8 9 , pages 31 - 36,
and Soviet Physics-Uspekhi, Volume 3 2 , pages 6 2 9 - 6 3 8 (July 1 9 8 9).
EXERCISES 253
You now have at your disposal the power of special you continue to enjoy these exercises as your study
relativity to provide physical insight and accurate pre moves on to other subjects. The following table of
dictions about an immense range of phenomena, contents is intended to help organize this ongoing
from nucleus to galaxy. The following exercises give attention.
only a hint of this range. Even so, there are too many R em inder: In these exercises the symbol v (in
to carry out as a single assignment or even several other texts sometimes called f i ) stands for speed as a
assignments. For this reason — and to anchor your fraction of the speed of light c. Let be the speed in
understanding of relativity— we recommend that conventional units; then v = v ^ ^ / c .
CONTENTS
mass and energy 8-21 Speeding light bulb 2 6 4

8-22 D oppler shift at the lim b o f Sun 264
8 -1 Examples o f conversion 2 5 4
8-23 The expanding universe 2 6 4
8-2 Relativistic chemistry 2 5 4
8-24 Twin Paradox using the D oppler
photons shift 2 6 4
8-25 D oppler line broadening 2 6 4
8-3 Pressure o f light 2 5 4 8-26 ^restconv ~ fitom the D oppler
8-4 M easurem ent o f photon energy 2 5 4 shift 2 6 4
8-5 Einstein’s derivation: Equivalence of 8-27 Everything goes forward 2 6 5
energy and mass 2 5 4 8-28 Decay o f 7r°-meson 2 6 7
8-6 Gravitational red shift 2 5 8
8-7 Density o f the com panion of Compton scattering
Sirius 2 5 8
8-29 C om pton scattering 2 6 7
creations, transformations, 8-30 C om pton scattering examples 2 6 8
annihilation 8-31 Energy o f a photon and frequency of
light 2 6 8
8-8 N uclear excitation 2 5 9
8-32 Inverse C om pton scattering 2 6 9
8-9 Photon braking 2 5 9
8-10 Photon integrity 2 5 9
tests of relativity
8-11 Pair production by a lonely
photon? 2 5 9 8-33 Photon energy shift due to recoil of
8-12 Photoproduction o f a pair by two em itter 2 7 0
photons 2 5 9 8-34 Recoilless processes 2 7 0
8-13 Decay o f positronium 2 6 0 8-35 Resonant scattering 2 7 1
8-14 Positron - electron annihilation 1 2 6 0 8-36 M easurem ent o f D oppler shift by
8-15 P ositron-electron annihilation II 260 resonant scattering 2 7 1
8-16 Creation o f p ro to n -a n tip ro to n pair by 8-37 Test o f the gravitational red shift I 272
an electron 2 6 1 8-38 Test o f the gravitational red
8-17 Colliders 2 6 1 shift II 2 7 2
8-39 Test o f the Tw in Paradox 2 7 2
Doppler shift
8-18 D oppler shift along the
free-for-all!
x-direction 2 6 3 8-40 M om entum w ithout mass? 2 7 3
8-19 D oppler equations 2 6 3 8-41 The photon rocket and interstellar
8-20 The physicist and the traffic light 263 travel 2 7 4
25 4 EXERCISE 8-1 EXAMPLES OF CONVERSION
c A spherical Earth satellite has radius r = 1

MASS AND ENERGY meter and mass m = 1 0 0 0 kilograms. Assume that
the satellite absorbs all the sunlight that falls on it.
8-1 examples off conversion W hat is the acceleration of the satellite due to the
a How much mass does a 100-w att bulb dissi force of sunlight, in units of g, the gravitational accel
pate (in heat and light) in one year? eration at Earth’s surface? For a way to reduce this
b The total electrical energy generated on Earth “disturbing” acceleration, see Figure 9-2.
during the year 1 9 9 0 was probably between 1 and d It may be that particles smaller than a certain
size are swept out of the solar system by the pressure of
2 X 10*^ kilowatt-hours. To how much mass is this
energy equivalent? In the acmal production of this sunlight. This certain size is determined by the equal
electrical energy is this much mass converted to en ity of the outward force of sunlight and the inward
ergy? Less mass? More mass? Explain your answer. gravitational attraction of Sun. Estimate this critical
c Eric Berman, pedaling a bicycle at full throttle, particle size, making any assumptions necessary for
produces one-half horsepower of u se fu l power your estimate. List the assumptions with your answer.
(1 horsepower = 7 4 6 watts). The human body is Does your estimated size depend on the particle’s
about 2 5 percent efficient; that is, 75 percent of the distance from Sun?
food burned is converted to heat and only 2 5 percent Reference; For pressure of lighr measuremenr in an elementary labo
is converted to useful work. How long a time will Eric ratory, see Robert Pollock, American Journal of Physics, Volume 31,
pages 901-904 (1963). Pollock’s method of determining the pres
have to ride to lose one kilogram by the conversion of
sure of light makes use of resonance to amplify a small effea to an
mass to energy? How can reducing gymnasiums stay easily measured magnitude. Dr. Pollock developed this experiment
in business? in collaboration with the same group of first-year students at Prince
ton University with whom the authors had the privilege to work out
8-2 relativistic chemistry the presentation of relativity in the first edition of this book.
One kilogram of hydrogen combines chemically with 8-4 measurement off photon
8 kilograms of oxygen to form water; about 10® joules energy
of energy is released.
a Ten metric tons (10'' kilograms) of hydrogen A given radioactive source emits energetic photons
combines with oxygen to produce water. Does the (X-rays) or very energetic photons (gamma rays) with
resulting water have a greater or less mass than the energies characteristic of the particular radioactive
original hydrogen and oxygen? W hat is the magni nucleus in question. Thus a precise energy measure
tude of this difference in mass? ment can often be used to determine the composition
b A smaller amount of hydrogen and oxygen is of even a tiny specimen. In the apparatus dia
weighed, then combined to form water, which is grammed in the figure on page 2 5 5 , only those events
weighed again. A very good chemical balance is able are detected in which a count on detector A
to detect a fractional change in mass of 1 part in 10®. (knocked-on electron) is accompanied by a count on
By what factor is this sensitivity more than enough detector B (scattered photon). W hat is the energy of
— or insufficient— to detect the fractional change in the incoming photons that are detected in this way, in
mass in this reaction? units of the rest energy of the electron?
8-5 Einstein's derivation:

PHOTONS equivalence off energy and
m ass— a worked example
8-3 pressure off light P ro b le m
a Shine a one-watt flashlight beam on the palm From the fact that light exerts pressure and carries
of your hand. Can you feel it? Calculate the total force energy, show that this energy is equivalent to mass
this beam exerts on your palm. Should you be able to and hence— by extension — show the equivalence of
feel it? A particle of what mass exerts the same force all energy to mass.
when you hold it at Earth’s surface? C om m entary: The equivalence of energy and
b From the solar constant (1 .3 7 2 kilowatts/ mass is such an important consequence that Einstein
square meter. Sample Problem 8 -5 ) calculate the very early, after his relativistic derivation of this result,
pressure of sunlight on an Earth satellite. Consider sought and found an alternative elementary physical
both reflecting and absorbing surfaces, and also line of reasoning that leads to the same conclusion. He
“ real” surfaces (partially absorbing). Why does the envisaged a closed box of mass M initially at rest, as
color of the light make no difference? shown in the first figure. A directed burst of electro-
EXERCISE 8-5 EINSTEIN'S DERIVATION: EQUIVALENCE OF ENERGY AND MASS 255
Electron detector
Thin foil
(loosely bound
3 : 4 : 5 triangle
electrons)
\ ''^Collimating slit
Shielding
Photon detector
Source of photons of
characteristic energies
EXERCISE 8-4. Measurement of photon energy.
magnetic energy is emitted from the left wall. It trav tion of the pressure exerted by light on a mirror
els down the length L of the box and is absorbed at the suspended in a vacuum. This measurement had first
other end. The radiation carries an energy E. But it successfully been carried out by E. F. Nichols and G.
also carries momenrum. This one sees from the fol F. Hull between 1901 and 1903. (By now the exper
lowing reasoning. The radiation exerts a pressure on iment has been so simplified and increased in sensitiv
the left wall during the emission. In consequence of ity that it can be carried out in an elementary labora
this pressure the box receives a push to the left, and a tory. See the reference for Exercise 8-3.)
momentum,/). But the momentum of the system as a Thus the radiation carries momentum and energy
whole was zero initially. Therefore the radiation to the right while the box carries momenrum and
carries a momentum p opposite to the momentum of mass to the left. But the center of mass of the system,
the box. How can one use knowledge of the transport box plus radiation, cannot move. So the radiation
of energy and momentum by the radiation to deduce must carry to the right not merely energy but mass.
the mass equivalent of the radiation? Einstein got his How much mass? To discover the answet is the object
answer from the argument that the center of mass of of these questions.
the system was not moving before the transport pro a W hat is the velocity of the box during the time
cess and therefore cannot be in motion during the of ttansit of the radiation?
transport process. But the box obviously carries mass b After the radiation is absorbed in the other end
to the left. Therefore the radiation must carry mass to of the box, the system is once again at rest. How far
the right. So much for Einstein’s reasoning in broad has the box moved during the transit of the radiation?
outline. Now for the details. c Now demand that the center of mass of the
From relativity Einstein knew that the momentum system be at the same location both before and after
/) of a directed beam of radiation is equal to the energy the flight of the radiation. From this argument, what
E of that beam (Section 8.4; both p and E measured in is the mass equivalent of the energy that has been
units of mass). However, this was known before Ein transported from one end of the box to the other?
stein’s relativity theory, both from Maxwell’s theory
of electromagnetic radiation and from direct observa-
S o lu tio n
a During the transit of the radiation the mo
mentum of the box must be equal in magnitude and
opposite in direction to the momentum p of the radi
ation. The box moves with a very low velocity v.
Therefore the Newtonian formula Mv suffices to cal
culate its momentum:
Mv = —p — — E
From this relation we deduce the velocity of the

box,
v = -E IM
EXERCISE 8-5, first figure. Transfer of mass by radiation. b The transit time of the photon is very nearly
256 EXERCISE 8-5 EINSTEIN'S DERIVATION: EQUIVALENCE OF ENERGY AND MASS
/ = L meters of light-travel time. In this time the box cording to the following reasoning. The energy that
moves a distance emerges from the left wall of the box may reside there
originally as heat energy. This thermal energy excites a
A x = pt = — EL/M typical atom of the surface from its lowest energy state
to a higher energy state. The atom returns from this
c If the radiation transported no mass from one higher state to a lower state and in the course of this
end of the box to the other, and if the box were the change sends out the surplus energy in the form of
sole object endowed with mass, then this displace radiation. This radiant energy traverses the box, is
ment A x would result in a net motion of the center of absorbed, and is ultimately converted back into ther
mass of the system to the left. But, Einstein reasoned, mal energy. Whatever the details of the mechanisms
an isolated system with its center of mass originally at by which light is emitted and absorbed, the net effect
rest can never set itself into motion nor experience any is the transfer of heat energy from one end of the box
shift in its center of mass. Therefore, he argued, there to the other. To say that mass has to pass down the
must be some countervailing displacement of a part of length of the box when radiation goes from one wall
the mass of the system. This transport of mass to the to the other therefore implies that mass moves when
right can be understood only as a new feature of the thermal energy changes location. The thermal energy
radiation itself. Consequently, during the time the in turn is derived from chemical energy or the energy
box is moving to the left, the radiation must transport of a nuclear transformation or from electrical energy.
to the right some mass m, as yet of unknown magni Moreover, thermal energy deposited at the far end of
tude, but such as to ensure that the center of mass of the tube can be converted back into one or another of
the system has not moved. The distance of transport is these forms of energy. Therefore these forms of
the full length L of the box diminished by the distance energy — and likewise all other forms of energy— are
A x through which the box has moved to the left in the equivalent in their transport to the transport of mass
meantime. But A x is smaller than L in the ratio EjM. in the amount m = E.
This ratio can be made as small as one pleases for any How can one possibly uphold the idea that a pulse
given transport of tadiant energy E by making the of radiation transports mass? One already knows that
mass M. of the box sufficiently great. Therefore it is a photon has zero mass, by virtue of the relation
legitimate to take the distance moved by the radiation (Section 8.4)
as equal to L itself. Thus, with arbitrarily high preci
sion, the condition that the center of mass shall not (mass)^ = (energy)^ — (momentum)^ = 0
move becomes
Moreover, what is true of the individual photon is
M A x + mL = 0 true of the pulse of radiation made up of many such
photons: The energy and momentum are equal in
Calculate the mass m and find, using A x from part b, magnirude, so that the mass of the radiation necessar
ily vanishes. Is there nor a fundamental inconsistency
m --A x M.i l = -{ -E L /M ){ M ID in saying in the same breath that the mass of the pulse
is zero and that radiation of energy E transports the
or, finally, mass m = E from one place to another?
The source of our difficulty is some confusion be
m= E tween two quite different concepts; (1) energy, rhe
time component of the momentum - energy 4-vector,
In conventional units, we have the famous equation and (2) mass, the magnitude of this 4-vector. When
the system divides itself into two parts (radiation
£conv = going to the right and box recoiling to the left) the
components of the 4-vectors of the radiation and of
W e conclude that the process of emission, rransport, the recoiling box add up to identity with the compo
and reabsorption of radiation of energy E is equivalent nents of the original 4-vector of the system before
to the transport of a mass m = E from one end of the emission, as shown in the second figure. However, rhe
box to the other end. The simplicity of this derivation magnitudes of the 4-vectors (magnirude = mass) are
and the importance of the result makes this analysis not additive. No one dealing with Euclidean geome
one of the most interesting in all of physics. try would expect the length of one side of a triangle to
D iscussion; The mass equivalence of radiant en be equal to the sum of the lengths of the other two
ergy implies the mass equivalence of thermal energy sides. Similarly in Lorentz geometry. The mass of the
and — by exrension — of orher forms of energy, ac system (M ) is not to be considered as equal to the sum
EXERCISE 8-5 EINSTEIN'S DERIVATION: EQUIVALENCE OF ENERGY AND MASS 257
of the mass of the radiation (zero) and the mass of the Even so, will not the distant detector momentarily
recoiling box (less than M ). But components of 4- experience a less-than-normal pull while the radiation
vectors are additive; for example, is in transit down the box? Is not the mass of the
radiation zero, and is not the mass of the recoiling box
/ energy o f\ ^ / energy o f \ _|_ / energy of \ reduced below the original mass M of the system? So
V system / \ radiation / \recoiling box/ is not the total attracting mass less than normal during
the process of transport? No! The mass of the system
Thus we see that the energy of the recoiling box is
— one has to say again — is not equal to the sum of
M — Not only is the energy of the box reduced by
the masses of its several parts. It is instead equal to the
the emission of radiation from the wall; also its mass is
magnimde of the total momentum - energy 4-vector
reduced (see shortened length of 4-vector in dia
of the system. And at no time does either the total
gram). Thus the radiation takes away mass from the
momentum (in our case zero!) or the total energy of
wall of the box even though this radiation has zero
the system change — it is an isolated system. There
mass. The inequality
fore neither is there any change in the magnitude M of
/(mass
mass o f\
f\ ^ / mass of \ mass of the total momentum - energy 4-vectors shown in the
(
\ system / \radiation[zero}/ V^coiling box) second figure. So, finally, there is never any change in
the gravitational attraction.
is as natural in spacetime geometry as is the inequality There is one minor swindle in the way this problem
5 # 3 + 4 for a 3-4-5 triangle in Euclidean geom has been presented: The box cannot in fact move as a
etry. rigid body. If it could, then information about the
W hat about the gravitational attraction exerted by emission of the radiation from one end could be ob
the system on a test object? O f course the redistribu tained from the motion of the other end before the
tion of mass as the radiation moves from left to right arrival of the radiation itself— this information
makes some difference in the attraction. But let the would be transmitted at a speed greater than that of
test object be at a distance r so great that any such light! Instead, the recoil from the emission of the
redistribution has a negligible effect on the attraction. radiation travels along the sides of the box as a vibra
In other words, all that counts for the pull on a unit tional wave, that is, with the speed of sound, so that
test object is the total mass M as it appears in New this wave arrives at the other end long after the radia
ton’s formula for gravitational force: tion does. In the meantime the absorption of the
radiation at the second end causes a second vibrational
force per \ _ GM wave which travels back along the sides of the box.
( unit mass The addition of the vibration of the box to the prob-
Radiation
(zero rest mass)
Energy
Momentum
BEFORE AFTER
EXERCISE 8-5, second figure. Radiation transfers mass from place to place even though the mass of the radiation is zero!
258 EXERCISE 8-6 GRAVITATIONAL RED SHIFT
lem requires a more complicated analysis but does not b W hat fraction of one’s rest energy is converted
change in any essential way the results of the exercise. to potential energy when one climbs a very high lad
References: A. Einstein, Annalen der Physik, Volume 20, pages der that reaches higher than the gravitational influ
6 2 7 - 6 3 3 (1906). For a more careful treatment of the box, see A. P. ence of Earth? Assume that Earth does not rotate and
French, Special Relativity (W . W . Norton, New York, 1968), pages is alone in space. Does the fraction of the energy that is
1 6 - 1 8 and 2 7 - 2 8 .
lost in either part a or part b depend on your original
mass?
8-6 gravitational red shift
c Apply the result of part a to deduce the frac
N ote: Exercises 8-6 and 8-7 assume an acquaintance tional energy change of a photon that rises vertically to
with the following elementary facts of gravitation. a height z in a uniform gravitational field g*. Photons
have zero mass; one can say formally that they have
(1) A very small object— or a spherically sym
only kinetic energy E = K. Thus photons have only
metric object of any radius— with mass M
one purse — the kinetic energy purse— from which
attracts an objea of mass m — also small or
to pay the potential energy tax as they rise in the
spherically symmetric— with a force
gravitational field. Light of frequency / i s composed of
photons of energy E = hf/c^ (see Exercise 8-31).
GMm
F= Show that the fractional energy loss for photons rising
in a gravitational field corresponds to the following
fractional change in frequency:
Here r is the distance between the centers of
the two objects and G is the Newtonian con
stant of gravitation, G = 6 .6 7 X 1 0 “ ** = —g*z [uniform gravitational field]
/
(meter)^/(kilogram-second^).
( 2) The work required to move a test particle of N ote: W e use / for frequency instead of the usual
unit mass from r t o r + dr against the gravita Greek nu, V, to avoid confusion with v for speed.
tional pull of a fixed mass M is GM{dr/r'^). d Apply the result of part b to deduce the frac
Translated from conventional units of energy tional energy loss of a photon escaping to infinity. (To
to units of mass this work is apply b for this purpose is an approximation good to
one percent when this fractional energy loss itself is
GM dr dr less than two percent.) Specifically, let the photon
d W coov = ■ M* — start from a point on the surface of an astronomical
.-2
object of mass M (kilograms) or AI* (meters) =
per unit of mass contained in the test particle. GM/c^ and radius r. From the fractional energy loss,
(3) The symbol M* = GMjc'^ in this formula show that the fractional change of frequency is given
has a simple meaning. It is the mass of the by the expression
center of attraction translated from units of
M*
kilograms to units of meters. For example, the [escape field of spherical object]
mass of Earth = 5.974 X lO^'* kilo f
grams) expressed in length units is =
This decrease in frequency is called the g rav itatio n al
4.44 X 10“ ^ meters, and the mass of Sun
re d shift because, for visible light, the shift is toward
(M s„„= 1.989 X 1030 kg) is M*su„ =
the lower-frequency (red) end of the visible spectmm.
1.48 X 1Q3 meters.
e Calculate the fractional gravitational red shifts
(4) Start the test particle at a distance r from the for light escaping from the surface of Earth and for
center of attraction of mass M and carry it to an
light escaping from the surface of Sun.
infinite distance. The work required is IF =
D iscussion: The results obtained in this exercise
M */r in units of mass per unit of mass con
are approximately correct for light moving near Earth,
tained in the test particle.
Sun, and white dwarf (Exercise 8-7). Only general
So much for the minitutorial. Now to business. relativity correctly describes the motion of light very
a W hat fraction of your rest energy is converted close to neutron star or black hole (Box 9-2).
to potential energy when you climb the Eiffel Tower
(300 meters high) in Paris? Letg* be the acceleration 8-7 density of the companion of
of gravity in meters/meter^ at the surface of Earth: Sirius
N ote: This exercise uses a result of Exercise 8-6.
1 g
= . Sirius (the Dog Star) is the brightest star in the
'Earth Earth heavens. Sirius and a small companion revolve about
EXERCISE 8-12 PHOTOPRODUCTION OF A PAIR BY TWO PHOTONS 259
one another. By analyzing this revolution using New BEFORE ^ O ---------- ^ be found)
tonian mechanics, astronomers have determined that A
the mass of the companion of Sirius is roughly equal
to the mass of our Sun (A1 is about 2 X 10^° kilo
grams; M* is about 1.5 X 10^ meters). Light from AFTER ' " OQ ^
(not known)
the companion of Sirius is analyzed in a spectrometer.
(at rest)
A spectral line from a certain element, identified from
the pattern of lines, is shifted in frequency by a frac EXERCISE 8-9. Stopping a nucleus by emission of a gamma ray.
tion 7 X lO""* compared to the frequency of the same
spectral line from the same element in the laboratory.
(These figures are experimentally accurate to only one
significant figure.) Assuming that this is a gravita
tional red shift (Exercise 8-6), estimate the average
8-10 photon integrity
density of the companion of Sirius in grams/centime- Show that an isolated photon cannot split into two
ter^. This type of star is called a w h ite d w a rf (Box photons going in directions other than the original
9-2). direction. (Hint: Apply the laws of conservation of
momentum and energy and the fact that the third side
of a triangle is shorter than the sum of the other two
sides. W hat triangle?)
CREATIONS,
TRANSFORMATIONS, 8-11 pair production by a
lonely photon?
ANNIHILATIONS A gamma ray (high-energy photon, zero mass) can
carry an energy greater than the rest energy of an
8-8 nuclear excitation electron - positron pair. (Remember that a positron
has the same mass as the electron but opposite
A nucleus of mass m initially at test absorbs a gamma
chatge.) Nevertheless the process
ray (photon) and is excited to a higher energy state
such that its mass is now 1.01 m.
(energetic gamma ray)----- (electron) + (positron)
a Find the energy of the incoming photon
needed to carry out this excitation.
cannot occur in the absence of other matter or radia
b Explain why the required energy of the incom
tion.
ing photon is greater than the change of mass of the
a Prove that this process is incompatible with the
nucleus.
laws of conservation of momentum and enetgy as
employed in the labotatory frame of reference. Ana
lyze the alleged creation in the frame in which electron
BEFORE A / V X A/ V ^ Om® and positron go off at equal but opposite angles ± (f)
with the extended path of the incoming gamma ray.
(at rest)
b Repeat the demonstration— which then be
comes much more impressive— in the center-of-mo-
AFTER c O ^ mentum frame of the alleged pair, the frame of refer
1.01 m ence in which the total momentum of the two
EXERCISE 8-8. Excitation of a nucleus by a gamma ray. resulting particles is zero.
8-12 photoproduction of a pair

8-9 photon braking by two photons
A moving radioactive nucleus of known mass M emits Two gamma rays of different enetgies collide in a
a gamma ray (photon) in the forward direction and vacuum and disappear, bringing into being an
drops to its stable nonradioactive state of known mass electron - positron pair. For what ranges of energies of
m. Find the energy of the incoming nucleus the two gamma tays, and fot what range of angles
(BEFORE diagtam in the figute) such that the result between their initial directions of propagation, can
ing mass m nucleus is at rest (AFTER diagtam). The this reaction occur? (Hint: Start with an analysis of the
unknown energy of the outgoing gamma ray reaction at threshold; at threshold the electron and
should not appear in your answer. positron are relatively at rest.)
260 EXERCISE 8-13 DECAY OF POSITRONIUM
BEFORE mQ-
A BEFORE m O Qm
A B
(at rest)
AFTER A / V W ^
D C
EXERCISE 8-13. Decay of positronium in flight.
8-13 decay off positronium

A moving "atom ” called positronium (an electron
AFTER
and positron orbiting one another) of mass m and
initial energy £ decays into two gamma rays (high-
energy photons) that move in opposite direaions
along the line of motion of the initial atom. Find the
energy of each gamma ray, Ec and £ o , in terms of the
mass m and energy E/^^ of the initial particle. Check
EXERCISE 8-15, first figure. Positron - electron annihilation.
that Ec = £ d in the case that the initial particle is at
rest.
8-15 positron—oloctron
8-14 positron—oloctron annihilation II
annihilation I
A positron e'*’ o f mass m and kinetic energy K is
A positron e"*" of mass m and kinetic energy K is annihilated on a target containing electrons e~ (same
annihilated on a target containing electrons e~ (same mass m) practically at rest in the laboratory frame;
mass m) practically at rest in the laboratory frame:
e‘*'(fast) + e~(at rest)----- * radiation
^■'■(fast) -b e (at rest) ' radiation
The resulting gamma rays go off at different angles
a By considering the collision in the center-of- with respea to the direaion of the incoming positron,
momentum frame (the frame of reference in which as shown in the first figure.
the total momentum of the initial particles is equal to a Derive an expression for the energy of one of
zero), show that it is necessary for at least two gamma the gamma rays in the laboratory frame as a fiinaion
rays (rather than one) to result from the annihilation. of the angle between the direction of emergence of
b Return to the laboratory frame, shown in the that gamma ray and the direction of travel of the
figure. The outgoing photons move on the line along positron before its annihilation. The gamma ray en
which the positron approaches. Find an expression for ergy should be a funaion of only the energy and mass
the energy of each outgoing photon. Let your deriva of the incoming positron and the angle of the outgo
tion be free of any reference to velocity. ing gamma ray. (Hint: Use the law of cosines, as
c Using simple approximations, evaluate the an applied to the second figure.)
swer to part b in the limiting cases (1) very small K
and (2) very large K. (Very small and very large P d ^ = P a ^ + P c ^ ~ 2/>a/»cCos (f)c
compared with what?)
b Show that for outgoing gamma rays moving
along the positive and negative x-direction, the results
Qm of this exercise reduce to the tesults of Exercise
BEFORE mQ.
8-14.
A B
(al rest)
AFTER
D c
LABORATORY FRAME
EXERCISE 8-15, second figure. Conservation of vector momen
EXERCISE 8-14. Positron-electron annihilation. tum means that the momentum triangle is closed.
EXERCISE 8-17 COLLIDERS 261
8-16 creation off proton— b Now transform to a frame in which one of the
antiproton pair by an incoming particles is at rest (transformation given in
electron Exercise 7-5). This would be the situation if we tried
to build an accelerator in which moving antiprotons
W hat is the threshold kinetic energy K,i, of the inci hit a stationary target of, say, liquid hydrogen (made
dent electron for the following process? of protons and electrons). [Simplify: At 0.9 TeV =
9 X 10** eV what is the effective speed v of the
electron (fast) + proton (at rest)----- ^ proton? W hat is its momentum compared with its
electron + antiproton + two protons energy? W hat is the value of the time stretch factor y
= E/mTi If the target protons were at rest, what
energy, in TeV, would the incoming antiproton need
8-17 colliders to have in order to yield the same interaction energy as
How much more violent is a collision of two protons that achieved in the Tevatron?
that are moving toward one another from opposite
directions than a collision of a moving proton with
one at rest? Wait a minute! You keep telling us that
D iscussion: When a moving particle strikes a energy and momentum have different values
stationary one, the energy available for the creation of when measured with respect to different
reference frames. Yet here you assume the
new particles, for heating, and for other interactions
“interaction energy” is the same in the Te
— or, in brief, the available interaaion energy— is vatron laboratory frame as it is in the rest
less than the initial energy (the sum of the rest and frame ofa proton that moves with nearly the
kinetic energies of the initial two particles). Reason: speed of light in the Tevatron frame. Is the
The particles that are left over after the reaction have a energy of a system different in different
net forward motion (law of conservation of momen frames, or is it the same?
tum), the kinetic energy of which is available neither
for giving these particles velocity relative to each other
nor for producing more particles. For this reason
— There is an important distinction between
much of the particle energy produced in accelerators is
the total energy of a system and the
not available for studying interactions because it is “available interaction energy,” just as
carried away in the kinetic energy of the products of there is an important distinction between
the collision. your money in the bank and “ready cash”
However, in the center-of-momenmm frame, the in the bank that you can spend. If some of
frame in which the total momentum of the system is your money in the bank has been p ut in
equal to zero, no momentum need be carried away escrow for payment on a house you are
from the interaction. Therefore the energy available buying, then you cannot spend that part
for interaction is equal to the total energy of the of your bank money to buy a new car.
incoming particles. Similarly, the total energy of the proton-
antiproton system is much smaller in the
Is there some way that the laboratory frame can be
Tevatron laboratory frame than in the
made also the center-of-momentum frame? One way
frame in which the proton is initially at
is to build two particle accelerators and have the two rest, but all of the Tevatron laboratory-
beams collide head on. If the energy and masses of the frame energy can be spent— used to
particles in each beam are respectively the same, then create new particles, for example. In con
the laboratory frame is the center-of-momentum trast, only a minute fraction of the energy
frame and all the energy in each collision is available in the frame in which the proton is initially
interaaion energy. It is easier and cheaper to achieve at rest can be spent to create new particles,
the same efficiency by arranging to have particles since total momentum must be conserved;
moving in opposite directions in the same accelerator. most of the total energy is kept “in
A magnetic field keeps the particles in a circular path, escrow” for this purpose. The number
and kinds of new particles created must be
“storing” them at their maximum energy for re
the same for all observers! Therefore the
peated tries at interaction. Such a facility is called a
“ available interaction energy” must be
co llid er. The figure on page 262 gives some details the same for all observers. The central
of a particular collider. point here is that the Tevatron collider
a W hat is the total available interaction energy design makes all of the energy in the
for each encounter in the laboratory frame of the proton-antiproton system “available”
Tevatron shown on page 262? for use in the laboratory.
262 EXERCISE 8-17 COLLIDERS
EXERCISE 8-17. Top: Aerial view of the Tevatron

ring at Fermi National Accelerator Laboratory in
Batavia, Illinois. The ring is 6 .3 kilometers in cir
cumference. Bottom: View along the tunnel of the
Tevatron. Protons (positive charge) and antiprotons
(antiparticle of the proton: same mass, negative
charge) circulate in separate beams in opposite direc
tions in the same vacuum chamber in the lower ring of
superconducting magnets shown in the photo. The
upper ring of regular magnets accelerates protons from
8 CeV to 1 5 0 GeV. Some of these protons are injected
into the lower set of magnets directly, rotating clock
wise. Other protons strike a copper target and create
antiprotons at a lower energy that are accumulated
over approximately 15 hours in a separate ring (not
shown) and then reaccelerated to 15 0 GeV and in
serted into the lower ring, circulating counterclock
wise. (Opposite charge, opposite motion yields same
magnetic force toward the center, hence counterrota
tion around the same circle.) Then particles in both
beams in the lower ring of magnets are accelerated at
the same time from 15 0 GeV to a final energy of 0 .9
TeVper particle. (1 teraelectron-volt = 1 0 ‘^ electron-
volts, or approximately 1 0 0 0 times the rest energy of
the proton or antiproton.) After acceleration, the
beams are switched magnetically so that they cross
each other at multiple intersection points around the
ring, allowing protons and antiprotons to collide in the
laboratory center-of-momentum frame. Detectors at
the points of intersection monitor products of the colli
sions. Protons and antiprotons that do not interact at
one intersection are not wasted; they may interact at
another intersection point or on subsequent trips
around the ring. The particles are allowed to coast
around and around at full energy for as long as 2 4
hours as they interact. Question: Approximately how
many revolutions around the ring does a given proton
or antiproton make in 2 4 hours? Photographs cour
tesy of Fermi Laboratory.
EXERCISE 8-20 THE PHYSICIST AND THE TRAFFIC LIGHT 263
laboratory x-direction, the photon has an energy E'

DOPPLER SHIFT given by the equation
E' = E y (l — cos (/))

8-18 Doppler shift along the
x-direclion
and moves in a direction that makes an angle (f)' with
N ote: Recall Exercise L-5 in the Special Topic on the x'-axis given by the equation
Lorentz Transformation, following Chapter 3.
Apply the momenergy transformation equations COS 0 “
(Exercise 7-5) to light moving in the positive x-direc- cos 0 ' =
1- cos 0
tion for which p ^ = p = E.
a Show that the relation between photon energy b Derive the inverse equations for E and cos 0 as
E' in the rocket frame and photon energy E in the functions of E', cos 0 \ and Show that the results
laboratory frame is given by the equation
are
(1 + v) E' E = E 'y (l + v„^ cos (f)')

E -= y (l+ p )E ' =
(1 -
, COS <f>' + t',
(1 + v) E' cos 0 - — ---------- ^
1+ cos 0
[(1 - t.)(l +
[photon moves along c If the frequency of the light in the laboratory
positive x-direction]
frame is / what is the frequency f ' of the light in the
rocket frame? Use the Einstein relation between pho
b Use the Einstein relation between photon en
ton energy E and classical wave frequency / namely
ergy E and classical wave frequency / , namely =
Econv “ ^ /o t E = bf/c^ and E' = h f'/c^, to derive the
h f or E = hf/c^ and E' = h f /c^, to derive the trans
transformations for frequency
formation for frequency
/' =fy{\- cos 0 )
[wove motion along f = f 'y ( X + cos 0 ')
/ positive x-direction]
This difference in frequency due to relative motion

This is the Doppler shift equation for light waves
is called the D o p p le r shift.
moving along the positive x-direction.
N ote: We use/ for frequency instead of the usual
N ote: W e u se/fo r frequency instead of the usual
Greek nu, V, ro avoid confusion with v for speed.
Greek nu, V, to avoid confusion with v for speed.
d For wave motion along the positive and nega
C Show that for a wave moving along the nega
tive x-direction, show that the results of this exercise
tive x-direction, the equation becomes
reduce to the results of Exercise 8-18.
r1- [wave motion along
e D iscussion question: Do the Doppler
equations enable one to determine the rest frame of
/' negative x-direction]
the source that emits the photons?
d Derive the corresponding equations that con
vert laboratory-measured frequency/to rocket-mea 8-20 the physicist and the
sured frequency f for waves moving along both posi traffic light
tive and negative x-directions.
A physicist is arrested for going through a red light. In
court he pleads that he approached the intersection at
8-19 Doppler equations such a speed that the red light looked green to him.
A photon moves in the xy laboratory plane in a direc The judge, a graduate of a physics class, changes the
tion that makes an angle (f) with the x-axis, so that its charge to speeding and fines the defendant one dollar
components of momentum are p^ —p cos (f) and py = for every kilometer/hour he exceeded the local speed
p sin (f) and p^ = 0. limit of 30 kilometers/hour. W hat is the fine? Take
a Use the Lorentz transformation equations for the wavelength of green light to be 5 30 nanometers =
the momentum - energy 4-vector (Exercise 7-5) and 530 X 10“^ meter) and the wavelength of red light to
the relation — /»^ = 0 for a photon to show that in be 650 nanometers. The relation between wavelength
the rocket frame, moving with speed along the A and frequency/for light is/A = c. Notice that the
264 EXERCISE 8-21 SPEEDING LIGHT BULB
light propagates in the negative x-direction {(f) = c Will allowance for the past effect of gravitation
(f)' = n). in slowing the expansion increase or decrease the esti
mated time back to the start of this expansion?
8-21 speeding light bulb Reference: E. Hubble, Proceedings of the U. S. National Academy of
Sciences, Volume 15, pages 168-173 (1929).
A bulb that emits spectrally pure red light uniformly
in all directions in its rest frame approaches the ob
server from a very great distance moving with nearly 8-24 twin paradox using the
the speed of light along a straight-line path whose Doppler shifft
perpendicular distance from the observer is A Both
The Twin Paradox (Chapter 4 and Exercises 4-1 and
the color and the number of photons that reach the
5-8) can be resolved elegantly using the Doppler shift
observer per second from the light bulb vary with
as follows. Paul remains on Earth. His twin sister
time. Describe these changes qualitatively at several
Penny travels at a high speed, v, to a distant star and
stages as the light bulb passes the observer. Consider
returns to Earth at the same speed. Both Penny and
both the Doppler shift and the headlight effect (Exer
Paul observe a distant variable star whose light gets
cises 8-19 and L-9).
alternately dimmer and then brighter with a fre
quency/in the Earth frame { /' in the rocket frame).
8-22 Doppler shift at the limb This variable star is very much farther away than the
off Sun length of Penny’s path and is in a direction perpen
Sun rotates once in about 25.4 days. The radius of dicular to this path in the Earth frame. Both observers
Sun is about 7.0X10® meters. Calculate the Doppler will count the same total number of pulsations of the
shift that we should observe for light of wavelength variable star during Penny’s round trip. Use this fact
500 nanometers = 500 X 10“ ^ meter) from the edge and the expression for the Doppler shift at the 90-de
of Sun’s disk (the lim b) near the equator. Is this shift gree laboratory angle of observation (Exercise 8-19)
toward the red end or toward the blue end of the to verify that at the end of the trip described in
visible spectrum? Compare the magnitude of this Chapter 4, Penny will be only 20 years older while
Doppler shift with that of the gravitational red shift of Paul will have aged 202 years.
light from Sun (Exercise 8-6). Reference: E. Feenberg, American Journal of Physics, Volume 27,
page 190 (1959).
8-23 the expanding universe
N ote; Recall Exercise 3-10. 8-25 Doppler line broadening
a Light from a distant galaxy is analyzed by a The average kinetic energy of a molecule in a gas at
spectrometer. A spectral line of wavelength 730 nano temperature T degrees Kelvin is {i/2 )k T . (The con
meters = 730 X 10“ ^ meters is identified (from the stant k is called the Boltzmann constant and has the
pattern of other lines) to be one of the lines of hydro value 1.38 X 10“ ^^joules/degree Kelvin). Molecules
gen that, for hydrogen in the laboratory, has the of gas move in random directions. Calculate the aver
wavelength 487 nanometers. If the shift in wave age speed from the low-velocity approximation of
length is a Doppler shift, how fast is the observed Newtonian mechanics. Estimate the fractional change
galaxy moving relative to Earth? Notice that the light in frequency due to the Doppler shift that will be
propagates in a direction opposite to the direction of observed in light emitted from a molecule in a gas at
motion of the galaxy {(f) = <f)' — 7 l ) . temperature T. Will this shift increase or decrease the
b There is independent evidence that the ob observed frequency of the emitted light? This effect,
served galaxy is 5 X 10® light years away. Estimate called Doppler broadening of spectral lines, is one
the time when that galaxy parted company from our reason why a given spectral line from a gas excited in
own galaxy — the Milky Way — using the simplify an electric discharge contains a range of frequencies
ing assumption that the speed of recession was the around a central frequency.
same throughout the past (that is, not slowed down
by the gravitational attractions between one galaxy
and another). The astronomer Edwin Hubble discov
8-26 £>•■1 con v = ntc^ ffrom the
ered in 1929 that this time— whose reciprocal is
Doppler shifft
called the Hubble constant, and which may itself Einstein’s famous equation in conventional units,
therefore appropriately be called the Hubble time — Eratconv ~ ^F'd the telativistic expression for en
has about the same value for all galaxies whose dis ergy can be derived from (1) the relativistic expression
tances and speeds can be measured. Hence the con for momentum (derived separately, for example in
cept of the expanding universe. Exercise 7-12), (2) the conservation laws, and (3) the
EXERCISE 8-27 EVERYTHING GOES FORWARD 265
Doppler shift (Exercise 8-18). In conventional units, units result by to convert to conventional units and
a photon has energy — hf, where h is Planck’s the equation in the well-known form
constant and / i s the frequency of the corresponding
classical wave. (We use/ for frequency instead of the energy released (conventional units) = {Am)c^
usual Greek nu, V, to avoid confusion with v for
speed.) Divide by to convert to units of mass: E = b Second result: = m . Now add the con
hf/c^. Expressed in units of mass, a photon has equal dition that energy is conserved in the laboratory
energy and momentum. Therefore the momentum of frame:
a photon is also given by the equation p = hf/c^.
Momentum does differ from energy, however, in that ^ b efo re - -Eafter + ( 2)
it is a 3-vector. In one dimensional motion, the sign of
the momentum (positive for motion to the right, Compare equations (1) and (2). These two equa
negative for motion to the left) is important, as in the tions both describe a particle at rest. Show that they
analysis below. are consistent if Etefore = "^before and E ^ = and
A _ '“'de of mass emits two photons in that therefore in general
opposite directions while remaining at rest in the lab
oratory frame. Conservation of momentum requires
these two photons to have equal and opposite mo
menta and therefore to correspond to the same classi or, in conventional units.
cal frequency / In consequence, they also have the
same energy. ^ re stco n v =
a First resu lt: Energy released = A w . Now
view this process from a rocket frame moving at speed c T h ird result: At any speed, E = my. Next
V — t'c o n v / ^ along the direction of flight of the two add the condition that energy be conserved in the
photons. The particle moves in this frame, but does rocket frame. Place primes on expressions for rocket-
not change velocity on emitting the photons. The measured energy of the particle and use the Doppler
photon emitted in the same direction as the rocket equations to transform the classical frequency back to
motion will be upshifted in energy (and in corre the laboratory value f. Show that the result is
sponding classical frequency) as compared with the
energy observed in the laboratory; the other back ■^before E'after + K^hf/ (3)
ward-moving photon will be downshifted. We can
calculate this frequency shift using the Doppler for The salient difference between equations (2) and
mulas (Exercise 8-18). Use the expression m jv for (3) is that in the rocket frame the particle is in motion.
momentum of a particle, equation (7-8), to state the Deduce that the general expression for energy of a
conservation of momentum (notice the minus sign particle includes the stretch factor gamma:
before the second photon term, representing the pho
ton moving to the left): E = my
1/2 or, in conventional units,

hf
= m ^ v y -b
C‘
[^ 1 ficonv = m yc^
h f\ 1 Reference: Fritz Rohrlich, American Journal o f Physics, Volume 58,
1+ pages 3 4 8 - 3 4 9 (April 1990).
Simplify this expression to

8-27 everything goes forward
»«before = "âfter + ( 1) “ Everything goes forward’’ is a good mle of thumb
for interactions between highly relativistic particles
or and stationary targets. In the laboratory frame, many
particles and gamma rays resulting from collisions
m\before = Am = 2hf/c^ = energy released continue in essentially the same direction as the in
coming particles.
Conservation of momentum in both frames im The first figure (top) shows schematically the colli
plies a change in particle mass equal to the total sion of two protons in the center-of-momentum
energy of the emitted photons. Multiply the mass- frame, the frame in which the system has zero total
266 EXERCISE 8-28 DECAY OF 7T°-MESON
momentum. A great many different particles are cre C In order to solve equation (1) for (^, we need to
ated in the collision, including a gamma ray (the know the value of Equation (2) is a quadratic in
fastest possible particle) that by chance moves per Show that the solution is
pendicular to the line of motion of the incoming
particles: (f)' = 7T/2 radians.
The first figure (bottom) shows the same interac (3)
•^proton
tion in the laboratory frame, in which one proton is
initially at rest. At what angle (f) does the product
Here yp„K,n the stretch factor y using the proton
gamma ray move in this frame?
velocity
a From the Doppler equations (Exercise 8-19), d We are interested in finding the angle (f) when
show that the outgoing angle <f) for the gamma ray in
the incoming proton is highly telativistic. In this case
the laboratory frame is given by the expression
•^proton 1. From the approximation for small angles
( 0 expressed in radians)
cos 0 = ( 1)
cos 0 ~ 1 — 0 ^ /2 101 « 1
b W hat is the speed t'proton of the rightward-
moving proton in the laboratory frame? We define
show that the angle 0 is given approximately by the
the laboratory frame by riding at speed on the expression
leftward-moving proton in the center-of-momentum
frame. Therefore the rightward-moving proton also
moves with speed v„^ in the center-of-momentum (4)
frame. Use the law of addition of velocities to find the
speed of the rightwatd-moving proton in the labora
e W hat is the value of 0 in radians and in de
tory frame (Section L.7 and Exercise 3.11).
grees for incident protons of energy £ = 200 GeV?
For incident protons of energy 2 X 10“*GeV? (1 GeV
2v.rel
V
*^pro«Ki — ( 2) = 10’ electron-volts. Mass of the proton is approxi
1+ mately 1 GeV.)
CENTER-OF-MOMENTUM FRAME
EXERCISE 8-27, first figure. In the center-of-momentum frame two incoming protons collide,
creating many particles, among them a gamma ray that movesperpendicular to the original line ofmotion. In
the laboratory frame, in which oneproton is initially at rest, in what direction does the gamma ray move?
EXERCISE 8-29 COMPTON SCATTERING 267
8-28 decay of ;i*-meson rocket frame in which the meson is at rest these pho
tons are emitted in the positive and negative ^'-direc
A 7T meson (neutral pi-meson) moving in the x-dir-
tions, as shown in the figure. Find the energies of the
ection with a kinetic energy in the laboratory frame
two photons in the rocket frame (in units of the mass
equal to its mass m decays into two photons. In the
of the meson) and the energies and directions of prop
agation of the two photons in the laboratory frame.
COMPTON SCATTERING
8-29 Compton scattering
Analyze Compton scattering of an incident photon
that collides with and recoils from an electron that is
initially at rest. Compton scattering in one dimension
was discussed in Section 8.4. Here we analyze Comp
ROCKET FRAME LABORATORY FRAME ton scattering in rwo dimensions. The goal is ro deter
EXERCISE 8 -2 8 . Two photons resulting from the decay of a 71° mine the reduced energy of the photon that has been
meson, as observed in rocket and laboratory frames. scattered with a change of direction measured by the
EXERCISE 8 -2 7 , se co n d figure. Forward spray of particles created in collisions near the middle of the
picture. An incident particle, probably a charged %-meson, enters from the left with energy approximately
100 to 2 0 0 times its rest energy and strikes a nucleus of neon or hydrogen. Curving paths in the imposed
magnetic field are probably knock-on electrons. These and the cascade of other particles move initially in the
same direction as the incoming 71-meson: “Everything goes forward!" Photograph courtesy of Fermi
Laboratory.
268
A A A /V ^ O
A 6
BEFORE
AFTER
EXERCISE 8-29, first figure. Compton scattering ofa photonfrom
an electron initially at rest. The angle (f) is called the scattering
angle.
EXERCISE 8-29, second figure. Conservation of vector momen

tum means that the momentum triangle is closed.
angle (f>. The angle 0 is called the scatterin g angle.

Use the notation in the first figure. Do not use fre
quency or wavelength or Planck’s constant or speed in
your analysis — only the laws of conservation of mo
mentum and energy plus equations:
F - [for an electron]
£2 • / = 0 [for a photon]
D iscussion: The conservation of momentum is a

vector conservation law. This means that the vector
sum of the momenta after the collision equals the
momentum of the photon before the collision. In
other words, the vectors form a triangle, as shown in
the second figure. Apply the law of cosines to this
----------- Energy of scattered photons ------- ► -
figure:
EXERCISE 8-29, third figure. Results of the Compton experiment
in which photons were scattered from the electrons in a graphite
pD —pA^ + Pc 2/»a/>cCOS0
target. At each angle of the detector except (f} = 0 there are some
photons scattered with loss of energy (electron recoils by itself) and
a Now replace all momenta with energies (easy other photons scattered with little or no loss of energy (electron and
for photons, more awkward for the electron), com- atom recoil as a unit).
EXERCISE 8-32 INVERSE COMPTON SCATTERING 269
bine with the conservation of energy, and derive the a A photon moves along the positive x-axis. Re
Compton scattering formula: sults of Exercise 8-18 show the relation between the
energy of this photon measuted in the rocket frame
"'incident and its energy measured in the laboratory frame. A
^scattered p classical electromagnetic wave moves along the posi
^ _j_ încident ^ ^
COS </>) tive x-axis. Results of Exercise L-5 (at the end of the
m
Special Topic following Chapter 3) show the telation
between the frequency of this wave measured in the
Exercise 8-30 gives some examples of this result.
rocket frame and its enetgy measured in the labora
b Compton’s original experiments showed that
tory frame. Compare these two results to show that if
some photons were scattered without a measurable
we associate photons with a light wave in one coordi
change of energy. These photons were scattered by
nate system, this association will hold in all coordinate
electrons that did not leave the atom in which they
systems.
were bound, so that the entire atom recoiled as a unit.
b The theory of relativity does not tell us the
Assume that the energy of the incoming photon is at
value of Planck’s constant b in the formula E =
most a few times the rest energy of the electron. In this
(^/c^)/that relates photon energy (in units of mass) to
case, show that the enetgy change is negligible for
classical wave frequency. Experiment shows the con
photons scattered by electrons tightly bound to an
stant b to have the value 6.63 X 10” ^“*joule-second.
atom of average mass (say 10 X 2000 X mass of an
Show that if energy is measured in conventional units,
electron). See the third figure.
the relation between energy and frequency has the
Reference; A. H. Compton, Physical Review, Volume 22, pages form
409-413 (1923).
Econv ~ [energy in conventional units]
8-30 cempton scattering
examples C Show that the formula for Compron scattering
(Exercise 8-29) becomes
a A gamma ray photon of energy equal to twice
the mass of the electron scatters from an electron .Anddent
initially at test. Provide the following answets in units /s o
of MeV. (Mass of the electron is 0.511 MeV.) From 1+ (1 - cos 0 )

the Compton scattering formula find the enetgy of the
scattered photon for scattering angles 0, 90, and 180
In the 1920s there was great resistance to the idea
degrees. If you have access to a computer, calculate
that when the electron is “ shaken ” by the electric field
this energy at 10-degree increments between zero and
of wave at one frequency it should scatter (reemit) this
180 degrees and plot the resulting curve of energy vs.
radiation at a lower frequency.
angle.
b In a new set of experiments, the incident
gamma ray has energy equal to five times the test 8-32 Inverse Compton
energy of the electron. Repeat the calculations of part scattering
a for this case.
In Compton’s original experiment an X-ray photon
scattered with reduced energy from an electron ini
tially at rest. In contrast, a photon scattered from a
8-31 energy of a photon and moving electron can increase the energy of the pho
frequency of light ton. Such an interaction is called inverse C o m p to n
Planck found himself forced in 1900 to recognize that scattering. The figure (page 270) shows an exam
light of frequency/(vibrations/second) is composed ple.
of quanta (Planck’s word) or photons (Einstein’s later When a high-energy electron collides head on with
word), each endowed with an energy E = bf/c^ (en a low-energy photon, what is the energy of the outgo
ergy in units of mass) where is a universal constant of ing photon? Answer this question using parts a - e or
propottionality called Planck’s constant. How can by some other method.
Planck’s formula possibly make sense when — as we a Write down equations of conservation of en
now know — not only E but also/depend upon the ergy and momenrum, using subscripts A through D
frame of reference in which the light is observed? (We from the figute.
use/for frequency instead of the usual Greek nu, V, to b Recall that the energy of a photon is equal to
avoid confusion with v for speed.) the magnitude of its momentum. Use this to simplify
270 EXERCISE 8-33 PHOTON ENERGY SHIFT DUE TO RECOIL OF EMITTER
m (y a Write down the conservation laws in a form

A that makes no reference to velocity. Consider the case
in which the fractional change in mass in the emission
BEFORE process is very small compared to unity. Show that for
this special case the photon has an energy E„ = m^ —
m. For the general case show that
m (y
c D
AFTER
EXERCISE 8-32. Inverse Compton scattering. A low-energy photon or
is scattered by a high-energy electron.
A£ E
the conservation equations, taking leftward momen
tum to be negative.
X 2m„
c We are not interested in the energy or the

momentum of outgoing electron C. Therefore solve b Show that this shift in energy for visible light
the energy equation for and the momentum equa (£„ ~ 3 eV) emitted from atoms (mc'^ ~ 10 X 10®
tion for Pc, square and subtract the two sides, and use eV) in a gas is very much less than the Doppler shift
Ec ~ Pc ~ W hat happens to Ej^ and p^ on the due to thermal motion (Exercise 8-25) even for tem
other side of the resulting equation? For now keep peratures as low as room temperature {kT ~ 1/4 0
terms in the first power o f w i t h o u t substituting the eV).
awkward equivalent p^ = {Ej^ +
d Solve the resulting equation for the energy of 8-34 recoilless processes
the outgoing photon. a A free atom of iron ’^Fe — formed in a so-
e Now consider an important special case in called “excited state” by the radioactive decay of
which the incoming electron is extremely energetic, cobalt ’^Co— emits from its nucleus a gamma ray
with an energy of, say, thousands of times its rest (high-energy photon) of energy 14.4 keV and trans
energy as measured in the laboratory. Show that this forms to a “normal” ’^Fe atom. By what fraction is
case the incoming electron behaves in essential re the energy of the emitted ray shifted because of the
spects as a photon; Pa ^ - \ - E^. Simplify your equation recoil of the atom? The mass of the ’^Fe atom is about
of part d to show that under these circumstances the equal to that of 57 protons.
outgoing photon has the energy of the incoming elec b That not all emitted gamma rays experience
tron no matter what the energy of the incoming photon. this kind of frequency shift was the important discov
ery made in 1958 by R. L. Mossbauer at the age of
TESTS OF RELATIVITY 29. He showed that when radioactive nuclei embed
ded in a solid emit gamma rays, some significant
N ote: Exercises 8-33 through 8-39 form a connected fraction of these atoms fail to recoil as free atoms.
tutorial on tests of relativity. Some of these exercises Instead they behave as if locked rigidly to the rest of
depend on each other and on earlier exercises, espe the solid. The recoil in these cases is communicated to
cially Exercise 8-6. the solid as a whole. The solid being heavier than one
atom by many powers of 10, these events are called
8-33 photon energy shift due recoilless processes. For gamma rays emitted in
to recoil of emitter recoilless processes, the m^ in Exercise 8-33 is the mass
of the entire chunk in which the iron atoms are em
N ote: This exercise uses the results of Exercise 8-25.
bedded. When this chunk has a mass of one gram, by
2„ and initially at rest
A free particle of initial mass W
what fraction is the frequency of the emitted ray
emits a photon of energy E. The particle (now of mass
shifted in this “recoilless” process?
m) recoils with velocity v, as shown in the figure.
c The gamma rays emitted from excited ’^Fe
BEFORE O (attest)
atoms do not have a precisely defined energy but are
spread over a narrow energy range— or frequency
range — or natural line width, shown as a bell-shaped
curve in the figure. (The physical basis for this curve is
AFTER O --- ^ ^ explained by quantum physics.) The full width of this
E m
curve at half maximum is denoted by Av. R. V.
EXERCISE 8-33. Recoil of a particle that emits a photon. Pound and G. A. Rebka selected ’^Fe for experiments
EXERCISE 8-36 MEASUREMENT OF DOPPLER SHIFT BY RESONANT SCATTERING 271
sheet containing ^^Fe will be less at the 14.4 keV

resonance energy than at any nearby energy. This
process is called reso n an t scattering.
a Show that when a gamma ray of the resonant
energy £„ is incident on a free iron atom initially at rest
then the free nucleus cannot absorb the gamma ray at
its resonant energy, because the process cannot satisfy
both the law of conservation of momentum and the
law of conservation of energy.
b Show that both conservation laws are satisfied
when an iron atom embedded in a one-gram crystal
EXERCISE 8-34. Natural line width ofphotons emittedfrom^^Fe. absorbs such a gamma ray by a recoilless process, in
which the entire crystal absorbs the momentum of the
incident gamma ray. (“Satisfied”? For momentum,
with recoilless processes because the fractional ratio yes; for energy, no. However, the fractional discrep
A///o has the very small value 6 X for the ancy in energy— equivalent to the fractional discrep
14.4 keV gamma ray from ’^Fe. How much is the ancy in frequency— is less than 6 parts in 10^^ and
natural line width, A/, o f’^Fe expressed in cycles/sec- therefore small enough so that the iron nucleus is
ond? Compare the fractional natural line width with “unable to notice” the discrepancy and therefore ab
the fractional shift due to recoil of a free iron atom. sorbs the gamma ray.)
And compare it with the fractional shift of a gamma
ray from a recoilless process.
Reference: For a more detailed account of Mossbauer’s discovery —
for which the German scientist was awarded the Nobel prize in
1961 — see S. DeBenedetti, “The Mossbauer Effect,” Scientific 8-36 measurement of Doppler
American, Volume 202, pages 7 2 - 8 0 (April I960). For the selec
tion o f ’^Fe, see R. V. Pound and G. A. Rebka, Jr., Physical Review
shift by resonant
Letters, Volume 3, pages 4 3 9 - 4 4 1 (1959).
scattering
Pound and Rebka's application of recoilless processes thus
In the experimental arrangement shown in the figure,
put into one’s hands a resonance phenomenon sharp in a source containing excited ’^Fe nuclei emits (among
frequency to the fantastic precision of 6 parts in 10‘^. other radiations) gamma rays of energy £„ by a recoil
Exercise 8-35 deals with detection of this radiation. less process. An absorber containing ’^Fe nuclei in the
Exercise 8-36 uses motion (Doppler shift) as a means for normal state absorbs some of these gamma rays by
producing controlled changes of a few parts in 10^^ — or another recoilless process and reemits this energy in
much larger changes— in the effective frequency of source various forms in all directions. Thus the counting rate
or detector or both. To what uses can radiation of precisely on a gamma ray counter placed as shown is less for an
defined frequency he put? There are many uses. For in absorber containing normal ’^Fe than for an equiva
stance, the effect is the basis of important techniques in
lent absorber without normal ^^Fe. Now the source is
solid-state physics, molecularphysics, and biophysics. One
moved toward the absorber with speed v.
can detect the change in the natural frequency of radia
tion from ^^Fe atoms caused by other atoms in the
a W hat must be the velocity of the source if the
neighborhood— and by external magnetic fields— and gamma rays are to arrive at the absorber shifted in
in this way analyze the interaction between the iron atom frequency by 6 parts in 10*^? Express your answer in
its surroundings. Here we aim at detection of various centimeters/second.
effects predicted by relativity.
8-35 resonant scattering

The nucleus of normal ^^Fe absorbs gamma rays at
the resonant energy of 14.4 keV much more strongly
than it absorbs gamma rays of any nearby energy. The
\ /
energy absorbed in this way is converted to internal !\f\S\r*- AAA/- AAA/-
energy of the nucleus and transmutes the ^^Fe to the
“excited state.” After a time this excited nucleus Source
/
Absorber Counter
drops back to the “normal state,” emitting the excess
energy in various forms in all directions. Therefore the
number of gamma rays transmitted through a thin EXERCISE 8-36. Resonant scattering of photons.
272 EXERCISE 8-37 TEST OF THE GRAVITATIONAL RED SHIFT
b Will the counting rate of the counter increase Modify the analysis of Exercise 8-6 to make a
or decrease under these circumstances? prediction about this experiment and compare your
c W hat will happen to this counting rate if the prediction with the results of the Scout D rocket
source is moved away from the absorber with the experiment.
same speed? References: Description of experiment and preliminary results:
d Make a rough plot of counting rate of the R. F. C. Vessot and M. W. Levine, General Relativity and Gravita
counter as a function of the source velocity toward the tion, Volume 10, Number 3, pages 181 -204 (1979). Final results:
absorber (positive velocity) and away from rhe ab R. F. C. Vessot, M. W. Levine, and others. Physical Review Letters,
sorber (negative velocity). Volume 45, pages 2081-2084 (1980), Popular explanation: Clif
ford M. Will, Was Einstein Right? (Basic Books, New York, 1986),
e D iscussion question: Does this method pages 42-64.
allow one to measure the “absolute velocity” of the
source, in violation of the Principle of Relativity
(Chapter 3)?
8-39 test off the twin paradox
8-37 test off the gravitational For Penny to leave her twin brother Paul behind in the
red shifft I laboratory, go away at high speed, return, and find
herself younger than stay-at-home Paul is so contrary
A 14.4-keV gamma ray emitted from ^^Fe without to everyday experience that it is astonishing to find
recoil travels verrically upward in a uniform gravita- that the experiment has already been done and the
rional field. By what fraction will the energy of this prediction upheld! Chalmers Sherwin pointed out
photon be reduced in rising to a height z (Exercise that the twins can be identical iron atoms just as well
8-6)? An absorber located at this height must move as living beings. Let one iron atom remain at rest. Let
with what speed and in what direction in order to the other make one forth-and-back trip. Or many
absorb such gamma rays by recoilless processes? Cal
round trips. The percentage difference in aging of the
culate this velocity when the height is 22.5 meters.
twin atoms is the same after a million round trips as
Plot the counting rate as a function of absorber veloc
after one round trip — and it is easiet to measure.
ity expected if (a) the gravitational red shift exists, and
How does one get the second atom to make many
(b) thete is no gravitational red shift. A frequency round trips? By embedding it in a hot piece of iron, so
shift: of A///o = (2.56 + 0.03) X 10“ ^^ was deter that it vibrates back and forth about a position of
mined in an experiment conducted by R. V. Pound equilibrium (thermal agitation!). How does one mea
and J. L. Snider. You will notice that this shift is vety sure the difference in aging? In the case of Penny and
much smaller than the natural line widrh A ///, = Paul the number of birthday firecrackers rhat each
6 X 10~^^ (see the figure for Exercise 8-34). There sets off during their separation are counted. In the
fore the result depended on a careful exploration of experiment with iron atoms one compares not the
the shape of this line and was derived sratistically number of flashes of firecrackers up to the time of
from a large number of photon counts. meeting but the frequency of the photons emitted by
References: Original experiment: R. V. Pound and G. A. Rebka, Jr., recoilless processes, and thus — in effect— the num
Physical Review Letters, Volume 4, pages 337-341 (I960). Im ber of ticks from two identical nuclear clocks in the
proved experiment: R. V. Pound and J. L. Snider, Physical Review,
course of one laboratory second. In other words, one
Volume 140, pages B788-B803 (1965).
compares the effective frequency of INTERNAL nu
clear vibrations (not to be confused with the back-
8-38 test off the gravitational
and-forth vibration of the iron atom as a whole!) as
red shifft II
observed in the laboratory for (a) an iron nucleus at
On June 18, 1976, a Scout D rocket was launched rest and (b) an iron nucleus in a hot specimen.
from Wallops Island, Virginia, carrying an atomic It is difficult to obtain an iron nucleus at rest.
hydrogen-maser clock as the payload. It achieved a Therefore the actual experiment compared the effec
maximum altitude of 10^ meters. By means of mi tive internal nuclear frequency for two crystals of iron
crowave signals, its clock was compared with an iden with a difference of temperarure AT. R. V. Pound
tical clock at the surface of Earth. The experiment and G. A. Rebka, Jr., measured that a sample
used continuous comparison of these two clocks as the warmed up by the amount A T = 1 degree Kelvin
payload rose and fell. Simplifying (and somewhat underwent a fractional change in effective frequency
misrepresenting) the experiment, we report their re of Ay/yj, = (—2.09 ± 0.24) X 10“ *’ (fewer vibra
sult as a fractional frequency red shift at the top of the tions; fewer clock ticks; fewer birthdays; more youth
trajectory due ro gravitational effects of A j / / = ful!). (We use / for frequency instead of the usual
0.945 X 10“ ‘o ± 6.6 X IQ-^’. Greek nu, V, to avoid confusion with v for speed.)
EXERCISE 8-40 MOMENTUM WITHOUT MASS? 273
To simplify thinking about the experiment, go crepancy factor 1 — (1 — How much is the
back to the idea that one iron atom is at rest and the deficit in number of “ticks” (for hot atom versus
other is in thermal agitation at temperature T; predict atom at rest) in the lapse of laboratory time dt? Show
the fractional lowering in number of internal vibra- that the cumulative deficit in number of “ticks” from
rions in the hot sample per laboratory second; and the hot atom in one second is/,(t^/2)j^g(l second)
compare with experiment. where (t'^)avg means “the time average value of the
D iscussion; The figure compares the effective square of rhe atomic speed” (relative to the speed of
“ticks” of the two “internal nuclear clocks” in the light). Note that the mean kinetic energy of thermal
laboratory time dt. Note that the speed of thermal agitation of a hot iron atom (mass = 5 7 mp,otoJ is
agitation is about the speed of light. W hat given by the classical kinetic theory of gases:
algebraic approximation suggests itself for rhe dis
(1/2) = (3/2) k T
Here k is Boltzmann’s factor of conversion be

tween two units of energy, degrees and joules (or
degrees and ergs); k = 1.38 X 10“ ^^ joule/degree
Kelvin {k = 1.38 X 10“ *^ erg/degree Kelvin). How
does the experimental result of Pound and Rebka
compare with the result of your calculation?
References: Chalmers W. Sherwin, Physical Review, Volume 120,
pages 17-21 (I960). R. V. Pound and G. A. Rebka, Jr., Physical
Review Letters, Volume 4, pages 274-275 (I960).
FREE-FOR-ALL!
8-40 momentum without mass?
A small motor mounted on a board is powered by a
battery mounted on top of it, as shown in the figure
on page 274. By means of a belt the motor drives a
paddlewheel that stirs a puddle of water. The paddle-
wheel mechanism is mounted on the same board as
the motor but a distance x away. The motor performs
work at a rate dE/dt.
a How much mass is being transferred per sec
ond from the motor end of the board to the paddle-
wheel end of the board?
b Mass is being transferred over a distance x at a
rate given by your answer to part a. W hat is the
momentum associated with this transfer of mass?
Since this momentum is small, Newtonian momen
tum concepts are adequate.
c Let the mounting board be inirially at rest and
supported by frictionless rollers on a horizontal table.
The board will move! In which direction? W hat hap
pens to this motion when the battery mns down? How
far will the board have moved in this time?
d Show that an observer on the board sees the
energy being transferred by the belt; an observer on
the table sees the energy being transferred partly by
the belt and partly by the board; an observer riding
one way on the belt sees the energy being transferred
EXERCISE 8-39. Comparison of nuclear dock at rest with nuclear partly by the belt moving in the other direction and
clock in thermal motion. partly by the board. Evidently it is not always possible
274 EXERCISE 8-41 THE PHOTON ROCKET AND INTERSTELLAR TRAVEL
to make a statement satisfactory to all observers about (y2 — \ y / 2 hence that the equation of part c can
the path by which energy travels from one place to be written in the form
another or about the speed at which this energy moves
from one place to another! P - 2yf+ 1 = 0
8-41 the photon rocket and e W hat is the value of the fraction / = (final
interstellar travel spaceship mass)/(initial spaceship mass) for a time
stretch factor y = 10? In your opinion, is it possible to
The “perfect” rocket engine combines matter and construct a spaceship whose shell and payload is this
antimatter in a controlled way to yield photons (high- small a fraction of takeoff mass?
energy gamma rays), all of which are directed out the f Substitute the result of part e into the conser
rear of the rocket. Suppose we start with a spaceship vation of energy equation in part a. Show that the
of initial mass Al„, initially at rest. At burnout the total energy of emitted radiation is less than the mass
remaining spaceship moves with speed v and has a of fuel consumed. Why?
mass equal to the fraction/of the original mass. For a g Does your analysis apply to takeoff from
given fraction / , we want to know the final rocket Earth’s surface? From Earth orbit? From somewhere
speed V or, better yet, the time stretch factor y = else? W hat safety precautions apply to the backward
1/(1 — (N ote: H e re ,/is not frequency.) blast of gamma rays?
a W hat is the total energy of the system initially? h You are the astronaut assigned to this space
Let stand for the total energy of radiation after ship. Do you want to stop at your distant destination
burnout. Find an expression for the total energy of the star or fly past at high speed? Do you want to return to
system after burnout and set up the conservation of Earth? Do you want to stop at Earth on your return or
energy equation. merely wave in passing? Must all fuel for the entire
b Similarly, set up the conservation of momen trip be on board at takeoff or can you refuel at your
tum equation. W hat is the total momentum of the destination star? From your answers to these ques
system initially? The momentum of the radiation at tions, plan your trip and find the resulting fractions of
burnout? The momentum of the spaceship at burn spaceship mass to initial mass for different stages of
out? your trip.
c Eliminate ^rad between the two conservation i D iscussion question: From your results for
equations. Show that the result can be written this exercise, what are your conclusions about the
technical possibilities of human flight to the stars?
7 /+ y v f^ 1
References: Adapted from A. P. French, Special Relativity (W.W.
Norton, New York, 1968), pages 183-184. See alsoj. R. Pierce,
d From the definition of y, show that yv = Proceedings of the IRE, Volume 47, pages 1053- 1061 (1959).
f"'’ ’- i>_,' .r * 2 - < ■■■■'■'■=■- ; ■•; - 2 T ■
■- Sk'"--'?-. -'i ' ■' " ■

■ ■
ITY: CURVED SPACETIME

IN ACTION
9.1 GRAVITY IN BRIEF

the mutual grip e l mass and spacetime
Gravity, as we see it today, does not count as a foreign force transmitted through space
and time. Gravity manifests the curvature o/spacetime.
Ten years after his special relativity, Einstein gave us his 1915 battle-tested and still
standard theory of gravitation. Its message comes in a single simple sentence: Space-
Spacetime tells mass how to move
time grips mass, telling it how to move-, and mass grips spacetime, telling it how to curve.
The grip of spacetime on mass enforces a central principle of special relativity:
conservation of energy and momentum in a smash (Figure 9-1). The coupling of mass
and spacetime geometry, far from being the weakest force in nature, is the strongest.
Now for the back-reaction, the grip mass exerts on spacetime! W hat curvature does Mass to spacetime: "Curve!'
that grip impose on spacetime? And how does that curvature give an account of gravity
unrivaled for scope and accuracy?
9.2 GALILEO, NEW TON, AND EINSTEIN

Only historical judgment liberates the spirit
from the pressure of the past; it maintains its
neutrality and seeks only to furnish light.
— Benedetto Croce
Galileo and Newton viewed motion as properly described with respect to a rigid
Euclidean reference frame that extends through all space and endures for all time. This
275
276 CHAPTER 9 GRAVITY: CURVED SPACETIME IN ACTION
BEFORE
FIGURE 9-1. spacetime grips mass, keeping an object moving straight when free. By its power, it enforces
conservation of energy and momentum in a smash.
supposed reference frame stands high above the battles of matter and energy. Within
Newton: One global frame.
Einstein: M any local fram es. this ideal space of Galileo and Newton there acts a mysterious force of gravity, an
interloper from the world of physics, a foreign influence not described by geometry.
In contrast, Einstein says that there exists no mysterious “force of gravity,’’ only the
structure of spacetime itself. Climb into an unpowered spaceship, he says, and see for
yourself that there is no gravity there. Physics is locally gravity-free (Chapter 2). Every
free particle moves in a straight line at uniform speed. In a free-float (inertial) frame,
physics looks simple. But such a frame rates as free-float in only a limited region of
spacetime (Section 2.3) — a fact emphasized here by repeated use of the word “local”
in describing a free-float frame.
Complications arise in describing the relation between (1) the direction of motion of
a particle in one local frame and (2) the direction of motion of the same particle as
observed from a nearby local frame. Any difference between the two directions is
described in terms of the “curvature of spacetime,” Einstein tells us. The existence of
this curvature destroys the possibility of describing motion with respect to a single
ideal Euclidean reference frame that pervades all space. W hat is simple is the geometry
in a region small enough to look flat.
How did the views of Galileo, Newton, and Einstein develop? And what is the
concrete substance of the strange phrase “curvature of spacetime”?
9.3 LOCAL MOVING ORDERS FOR MASS 277
9.3 LOCAL MOVING ORDERS FOR MASS

moving orders front the local commander,
spacetime!
Navigation satellites near Earth drift away from “perfect” orbits because thin air and
solar radiation pressure affect their motion. Figure 9-2 shows an experimental satellite
that carries a “conscience” designed to assure that the same motion will be maintained
proof moss
upper boom
upper fuel tank
lower fuel tank
lower boom
FIGURE 9-2. "Conscience-guided" satellite. A satellite in orbit around Earth is subject to small
accelerations due to solar radiation pressure and residual atmospheric drag. Uncorrected, these accelerations
are between 1 0 ~ ‘g and 10~^g, where g is the acceleration of gravity at Earth’s surface. The acceleration
was reduced to 5 ^ 10 ~ ’^g for more than a year in orbit by use of a conscience or proof mass and the
Disturbance Compensation System (DISCOS) mounted on a TRIAD U.S. Navy satellite. The conscience, a
gold-platinum sphere 2 .2 centimeters in diameter, floatsfreely inside a spherical housing. Any nongravita-
tionalforce results in an incremental velocity change. Thefloating proofmass continues in its original state of
motion in an idealfriction-free environment. Observing the proof mass through capacitor sensing devices, the
satellite becomes aware that it is not keeping up with the motion demanded by the proof mass. An opposite
vernier rocket fires long enough to bring the spaceship hack into concord with its proof mass— its conscience.
To reduce gravitational effects of the satellite itself on the proof mass, fuelfor the vernier rockets is stored in
donut-shaped tanks placed symmetrically above and below the proof mass; power supply and radio transmit
ter are each held at the end of a boom 2 .7 meters long on either side of the control unit. For an Earth-based
microgravity environment, recall Figure 2-3. (Used with permission of AIAA. Journal of Spacecraft.)
ISAAC N E W T O N
Woolsthorpe, December 25, 1642—Kensington (London), March 20, 1727
“The marble index of a mind forever

Voyaging through strange seas of thought, -uXon^.''— Wordsworth
■k -k -k
“I do not know what 1 may appear to the world; but to myself I seem to have been only
like a boy, playing on the sea-shore, and diverting myself, in now and then finding a
smoother pebble or a prettier shell than ordinary, whilst the great ocean of truth lay all
undiscovered before me.”—Newton
★ ★ ★
“ Why do I call him a magician? Because he looked on the whole universe and all that
is in it as a riddle, as a secret which could be read by applying thought to certain
evidence, certain mystic clues which God had laid about the world to allow a sort of
philosopher’s treasure hunt to the esoteric brotherhood. He believed that these clues
were to be found partly in the evidence of the heavens and in the constitution of
elements (and that is what gives the false suggestion of his being an experimental
natural philosopher), but also partly in certain papers and traditions handed down by
the brethren in an unbroken chain back to the original cryptic revelation in Babylonia.
He regarded the universe as a cryptogram set by the Almighty—just as he himself
wrapt the discovery of the calculus in a cryptogram when he communicated with
Leibnitz. By pure thought, by concentration of mind, the riddle, he believed, would be
revealed to the initiate.” —A'ervreit
(Reprinted by permission of the publisher. Horizon Press, from Essays in Biography by

John Maynard Keynes. Copyright 1951.
9.3 LOCAL MOVING ORDERS FOR MASS 279
when it encounters these disturbances as when it moves through perfect emptiness.

The “conscience” — called a p ro o f m ass — is a separate sphere that floats inside the
larger ship. The proof mass undergoes no acceleration relative to the ship as long as the
ship moves freely. When relative motion does occur, the error in the tracking must be
due to the satellite. By small rockets the satellite gives itself a brief spurt of acceleration
"Conscience-guided” satellite.
and comes back into step with the inner proof mass — the satellite’s conscience.
What guides the conscience?
Though resistance is present, the rocket thrust overcomes it. The satellite takes the
same course it would have taken had both resistance and thrust been absent.
As satellite and proof mass come to empty space, they fly through it in perfect step,
without use of rockets or sensing devices. W hat a remarkable harmony they present!
The inner proof mass does not see outer space. It does not touch, feel, or see the ship
that surrounds it on every side. Yet it faithfully tracks the ship’s route through
spacetime. Moreover, this tracking is as perfect when the proof mass is made of
aluminum as when it is made of gold. How do proof masses — of whatever atomic
constitution and whatever construction — know enough to follow a standard world
line? Where does mass get its moving orders?
Locally, answers Einstein. From a distance, answers Newton.
Einstein says that the proof mass gets its information in the simplest way possible. It
responds to the structure of spacetime in its immediate vicinity. It moves on a straight
line in the local free-float frame. No simpler motion and no straighter motion can be
imagined.
Newton says that the inner proof mass gets its information about how to move from
a distance, via a “force of gravity.” Motion telative to what? Motion relative to an
ideal, God-given, never-changing Euclidean reference frame that spans all of space
and endures for all time. He tells us that the proof mass would have moved along an
ideal straight line in this global frame had not Earth deflected it. How can this ideal
line be seen? How sad! There is nothing, absolutely nothing, that ever moves along this
ideal line. It is an entirely imaginary line. But it nevertheless has a simple status,
Newton tells us, in this respect: Every satellite and every proof mass, going at whatever
speed, is deflected away from this ideal line at the same acceleration (Figure 9-3).
Einstein says: Face it; there is no ideal background Euclidean reference frame that Physics is simple only when
extends over all space. And why say there is, when even according to Newton no analyzed locally
particle, not even a light ray, ever moves along a straight line in that ideal reference
frame. Why say spacetime is Euclidean on a large scale when no evidence directly
supports that hypothesis? To try to set up an all-encompassing Euclidean reference
frame and attempt to refer motion to it is the wrong way to do physics. Don’t try to
describe motion relative to faraway objects. Physics is simple only when analyzed
locally. And locally the wotldline that a satellite follows is already as straight as any
wotldline can be. Forget all this talk about “deflection” and “ force of gravitation.”
I’m inside a spaceship. Or I’m floating outside and near it. Do I feel any “force of
gravitation?” Not at all. Does the spaceship “feel” such a force? No. Then why talk
about it? Recognize that the spaceship and I are traversing a region of spacetime free of
all force. Acknowledge that the motion through that region is already ideally straight.
How can one display the straightness of the motion? Set up a local lattice of meter
sticks and clocks, a local free-float (inertial) reference frame — also called a Lorentz
reference frame (Chapter 2). How does one know the frame is free-float? Watch every
particle, check every light ray, test that they all move in straight lines at uniform speed
relative to this frame. And having thus verified that the frame is free-float, note that
the proof mass too moves at a constant speed in a straight line— or remains at
rest — relative to this local free-float frame. W hat could be simpler than the moving
orders for mass: “Follow a straight line in the local free-float reference frame.” Does a
proof mass have to know the location of Earth and Moon and Sun before it knows how
to move? N ot at all! Surrounded on all sides by the black walls of a satellite, it has only
to sense the local structure of spacetime— right where it is — in order to follow the
correct track.
FIGURE 9 -3 . In N e w to n ia n m echanics d ifferent p a rticles going a t differen t speeds are a ll deflected a w a y

fro m the id e a l stra ig h t lin e w ith e q u a l acceleration. In th is respect there is no difference in princ iple between
the f a l l o f a projectile a n d the m otion o f a sa tellite. In th is p ic tu re o f N e w to n 's p u b lish e d in 16 8 6 , cannon o f
successively greater pow er m ounted on a m ountaintop fire out th e ir balls horizontally. T he more p o w e rfu l
cannon launches a sa tellite. The outer tw o curves show other possible sa te llite orbits. In brief, N e w to n h as one
global reference fra m e , b u t w ith in th is reference fra m e no sa te llite is ever grav ity -free, a n d no p a rticle ever
moves in a stra ig h t line a t constant speed. E instein, in contrast, m akes use o f m a n y local regions in each o f
w h ich the geometry is L o ren tzia n (as in special re la tiv ity ); the la w s o f g ra v ita tio n arise fro m the la ck o f
id e a lity in the relation between one local region a n d the n ext (g ra vita tio n ; spacetim e curvature; general
re la tiv ity ).
9.4 SPACETIME CURVATURE

not one but two particles witness to gravitation
Splendid! And also simple! But isn’t Einstein’s view of motion too simple? We started
out interested in the motion of a spaceship around Earth and in “gravitation.” We
seem to have ended up talking only about the motion of the satellite— or the proof
mass — relative to a strictly local inertial reference frame, a trivially simple straight-
line Inotion. Where is there any evidence of ‘‘gravitation” to be seen in that? Nowhere.
9.5 PARABLE OF THE TWO TRAVELERS 281
This is the great lesson of Einstein: Spacetime is always and everywhere locally
Lorentzian. No evidence of gravitation whatsoever is to be seen by following the
motion of a single particle in a free-float frame.
One has to observe the relative acceleration of two particles slightly separated from
each other to have any proper measure of a gravitational effect. Separated by how
much? That depends on the region of spacetime and the sensitivity of the measuring
equipment. Two ball bearings with a horizontal separation of 20 meters, dropped
from a height of 315 meters above Earth’s surface with 0 initial relative velocity, hit
the ground 8 seconds later (24 X 10® meters of light-travel time later) with a
separation that has been reduced by 10~^ meter (Section 2.3). Two ball bearings with a
vertical separation of 20 meters, dropped from a height of 315 meters with 0 initial
relative velocity, in the same 8 seconds increase their separation by 2 X 10“ ^ meter. To
measuring equipment unable to detect such small relative displacements the ball
bearings count as moving in one and the same free-float reference frame. No evidence
for gravitation is to be seen. More sensitive apparams detects the tid e-p ro d u c in g
actio n of gravity — the accelerated shortening of horizontal separations parallel to
Earth’s surface, the accelerated lengthening of vertical separations. Each tiny ball
bearing still moves in a straight line in its own local free-float reference frame. But
now — with the new precision — the region of validity of the one free-float reference
frame does not reach out far enough to give a proper account of the motion of the other
steel ball. The millimeter or two discrepancy is the way “gravity” manifests itself.
Tidal acceleration displays gravity as a local phenomenon. No mention here of the
distance of the steel balls from the center of Earth! No mention here of acceleration
relative to that center! The only accelerations that come into consideration are those of
nearby particles relative to each other, the tidal accelerations described in the preceding
paragraph.
These relative accelerations double when the separations are doubled. The true
measure of the tide-producing effect has therefore the character of an acceleration per
unit of separation. Let the acceleration be measured in meters of distance per meter of
light-travel time per meter of light-travel time; that is, in units meters/meter^ or
1/m eter [x = (1 /T)af-, soa = I x j f Y Then the measure of the tide-producing effect
(different for different directions) has the units (acceleration/distance) or (1/meter^).
In the example, in the two horizontal directions this quantity has the value [2(— 0.001
m eter)/(24X 10® meter)^]/20 meter = — 17.36X 10“ ^‘*meter~^ and in the vertical
direction twice the value and the opposite sign: -b 34.72 X 10“ ^'* meter"^. The
tide-producing effect is small but it is real and it is observable. Further, it is a locally
defined quantity. And Einstein tells us that we must focus our attention on locally
defined quantities if we want a simple description of nature. Einstein’s railw ay coach in free fa ll.
Einstein says more: This tide-producing effect does not require for its explanation
some mysterious force of gravitation, propagated through spacetime and additional to
the structure of spacetime. Instead, it can and should be described in terms of the
geometry of spacetime itself as the cu rv a tu re o f spacetim e.
Though Einstein speaks of four-dimensional spacetime, his concepts of curvature
can be illustrated in terms of two-dimensional geometry on the surface of a sphere.
9.5 PARABLE OF THE TWO TRAVELERS

space curvature on a sphere accounts for
relative acceleration of travelers
One traveler. A, stands at the equator, ready to travel straight north. A’s companion B,
standing against him shoulder to shoulder, wheels 90 degrees and marches straight
FIGURE 9 -4 . Travelers A and B,

sta r tin g out p a ra lle l a n d d e v ia tin g nei
ther to the le ft nor to the rig h t, neverthe
less f i n d themselves approaching each
other a fte r they have traveled some d is
tance. Interpretation I: Some m ysterious
force o f " g r a v ita tio n " is a t work. Inter
p re ta tio n 2: They are tra v e lin g on a
curved surface. Figure not d ra w n to
scale.
east. She paces off 20 kilometers along the equator. There she again turns a sharp 90
degrees and faces straight north. Both travelers now start north and travel 200
kilometers (Figure 9-4). In the beginning their tracks are strictly parallel. Moreover,
no travelers could be more conscientious than they are in continuing precisely in their
original directions. Each of them deviates neither to the right nor to the left. Yet an
umpire sent out to measure their separation after their 200-kilometer treks finds it to
be less than the original 20 kilometers. Why? We know perfectly well: The surface of
the globe is curved. If they continue north, their paths will meet at the north pole.
Already at this early stage of their trip the travelers are approaching each other,
although they had started out not approaching at all. Initially their velocity relative to
one another was zero; now they move toward one another with a small relative
velocity. In this sense they are slowly accelerating toward each other.
The travelers accelerate toward each other as surely as two tiny ball bearings in a
free-fall horizontal railway coach accelerate toward each other (Figure 9-5). We
Curvature of Earth dem onstrated
ascribe the relative acceleration of ball bearings in the railway coach to the “tidal”
by change in separation of two
originally parallel paths
effects of nonuniform gravitation near Earth. To be sure, the relevant picture for the
travelers is the two-dimensional curved space of the surface of Earth, whereas what
counts for the ball bearings is curvature of spacetime. This parallelism between the
geometrical concept of curvature and the gravitational concept of tide-producing
effect foreshadows Einstein’s geometrical interpretation of gravity.
The two travelers, who started out so conscientiously on parallel tracks and deviated
neither to the left nor to the right, have been told by the umpire of distances that
despite all precautions they are now slowly accelerating toward one another. They
blame this development on the existence of some mysterious ‘‘gravitational force” that
deflects their paths. They explore the nature of this “gravitational force.” Repeating
the travel with bicycles, motorcycles, light cars, and heavy trucks all moving north
ward with the same speed, they find always the same relative acceleration toward one
another. They conclude that the “gravitational force” leads to the same acceleration of
all objects, no matter what they are made of or how massive they are.
Learned would-be pundits analyze the motion of travelers. They say, in words
utterly mysterious to us, “See here. You find the same acceleration for every vehicle
9.5 PARABLE OF THE TWO TRAVELERS 283
VELOCITY TOWARD
MIDPOINT:
10 meter/second or
I 10 “ %
3 X light-speed
INCLINATION
TOWARD
>k MIDPOINT
2 x 1 0 -* ^ km
1 km ^ ^ 1 km
2 0 0 km 2 0 0 km
------------------ > -
location to right of
center line
TRAVELERS HEADED "NORTH" ON EARTH

FIGURE 9 -5 . C o m p a r is o n o f t h e p a t h s o f n o r t h w a r d t r a v e l e r s o n E a r t h ’s s u r fa c e w i t h th e
w o r l d l i n e s o f b a l l b e a r in g s r e le a s e d s i d e b y s i d e f r o m r e s t n e a r E a r t h ’s s u r fa c e . In both cases the
" p a th ” o f each “tra v e le r” sta rts p a ra lle l w ith th a t o f the second traveler (zero i n i t i a l relative velocity). In
both cases th is “p a t h ” g ra d u a lly inclines to w a rd the centerline ( “re lative acceleration”). In both cases the
p a th s can be accounted fo r in term s o f the local cu rva tu re o f geometry (curvature o f E a r th ’s surface f o r the
travelers; cu rva tu re o f spacetim e geom etry— g r a v ita tio n ! — f o r the b a ll bearings). In each dia g ra m ,
ve rtic a l distances are d r a w n — fo r v iv id n e s s— to a d ifferen t scale th a n h o rizo n ta l distances. B oth d ia
gram s suffer fro m th is a d d itio n a l im perfection: they a tte m p t to show, on the f l a t E uclidean surface o f th is
page, trajectories th a t can be correctly represented only in term s o f a curved geometry.
you try. This means that the ratio of gravitational mass to inertial mass is the same for
all sorts of objects. You have made a great discovery abour mass.”
All this time we and our space-traveler friends are looking down from on high. We
see the many treks. We watch the many measurements of distance. Through our
intercommunication system we hear and approve as our friends on the ground
interpret distance shortening as relative acceleration — and relative acceleration as
Curvature alone accounts for
“gravitation.” But then they get into weighty discussions. They start speaking of relative acceleration
“gravitation” as action at a distance. We smile. W hat is at issue— we know — is not
action at a distance at all, but the geometry of curved space. All this talk about the
identity of “gravitational mass” and “inertial mass” completely obscures the ttuth.
Curvature and nothing more is all that is required to describe the increasing rate at
which A and B approach each other.
9.6 GRAVITATION AS CURVATURE OF

SPACETIME
spacetime curvature accounts for
tidal accelerations of objects
Einstein smiles, too, as he hears gravitation described as action at a distance. Curvature
Spacetim e curvature
of spacetime and nothing more, he tells us, is all that is required to describe the
dem onstrated by change in
separation of two originally
millimeter or two change in separation in 8 seconds of two ball bearings, originally 20
parallel worldlines meters apart in space above Earth, and endowed at the start with zero relative velocity.
Moreover, this curvature completely accounts for gravitation.
“W hat a preposterous claim!” is one’s first reaction. “ How can such minor— and
slow — changes in the distance between one tiny ball and another offer any kind of
understanding of the enormous velocity with which a falling mass hits Earth?” The
answer is simple: Many local reference frames, fitted together, make up the global
structure of spacetime. Each local Lorentz frame can be regarded as having one of the
Acceleration tow ard Earth:
ball bearings at its center. The ball bearings all simultaneously approach their neigh
Totalized effect of relative
accelerations, each particle
bors (curvature). Then the large-scale structure of spacetime bends and pulls nearer to
tow ard its neighbor, in a chain of Earth (Figure 9-6). In this way many local manifestations of curvature add up to give
test particles that girdles globe the appearance of long-range gravitation originating from Earth as a whole.
In brief, the geometry used to describe motion in any local free-float frame is the
flat-spacetime geometry of Lorentz (special relativity). Relative to such a local free-
float frame, every nearby electrically neutral test particle moves in a straight line with
constant velocity. Slightly more remote particles are detected as slowly changing their
velocities, or the directions of their worldlines in spacetime. These changes are de
scribed as tidal effects of gravitation. They are understood as originating in the local
curvature of spacetime.
From the point of view of the student of local physics, gravitation shows itself not at
all in the motion of one test particle but only in the change of separation of two or more
nearby rest particles. “Rather than have one global frame with gravitational forces we
have many local frames without gravitational forces. ’’ However, these local dimension
changes add up to an effect on the global spacetime structure that one interprets as
“gravitation” in its everyday manifestations.
In contrast, Newton supposed the existence of one ideal overall reference frame. For
him, “Absolute space, in its own nature, without relation to anything external,
remains always similar and immovable.” The ball bearing or spaceship is regarded by
Newton as actually accelerated with respect to this ideal frame. The “gravitational
force” that accelerates it acts mysteriously across space and is produced by distant
objects. That the man in the spaceship finds no evidence either of the acceleration or
the force is an accidenr of nature, according ro the Newtonian view. Pundits used to
interpret this accident of nature as the fortuitous equality of “gravitational mass” and
“inertial mass” or in other “ learned” ways.
In conversations with one of the authors of this book at various times over rhe years,
Einstein emphasized his great respect for Newton and, in particular, his admiration for
Newton’s courage and judgment. He stressed that Newton was even better aware than
his seventeenth-century critics of the difficulties with the ideas of absolute space and
time. To postulate those ideas was nevertheless the only practical way to get on with
the task of describing motion in Newton’s century. In effect, Newton chopped the
problem of motion into two parts: (1) space and time and their meaning: ideas that
were puzzling but usable and that were destined to be clarified only 230 years later and
(2) the laws of acceleration with respect to that idealized spacetime: laws that Newton
gave the world.
9.6 GRAVITATION AS CURVATURE OF SPACETIME 285
FIGURE 9 -6 . L o c a l c u r v a t u r e a d d i n g u p to t h e a p p e a r a n c e o f lo n g -r a n g e g r a v i ta t i o n . The
shortening o f d ista n ce between a n y one p a ir , A a n d B, o f h a ll hearings is sm a ll w hen the d istance its e lf is
sm all. H owever, sm a ll separation between each h a ll hearing a n d its p a rtn e r dem ands m any p a ir s to
encompass E arth. T he to ta lize d shortening o f the circumference in a n y g iv e n tim e — the shortening o f one
separation tim es the num ber o fse p a ra tio n s— is in dependent o f the fineness o f the subdiv ision. T h a t to ta lize d
p u llin g in o f the circumference carries the whole necklace o f masses in w a rd . T h is is fre e fa l l , th is is g ra v ity ,
th is is a large scale m otion interpreted a s a consequence o f local curvature. Exam ple:
O rig in a l separation between A a n d B — a n d every other p a ir: 2 0 meters

T im e o f observation: 8 seconds
Shortening o f separation in th a t tim e: 1 m illim eter
F ractio n a l shortening: 1 m illim e te rf 2 0 meters = 1 1 2 0 ,0 0 0
Circumference o f E a rth {length o f a iry necklace o f h a ll hearings): 4 .0 0 3 0 X 10^ meters
Shrinkage o f th is circumference in 8 seconds: 1 /2 0 ,0 0 0 X 4 .0 0 3 0 X 10^ meters = 2 0 0 1 .5 meters
Decrease in the d ista n ce fro m the center o f E a rth (drops by the sam e fa c to r 1 /2 0 ,0 0 0 ) :
1 /2 0 ,0 0 0 X 6 .3 7 1 X 10^ meters = 3 1 5 meters.
T h is app a ren tly large-scale effect is ca u sed — in E in s te in ’s p ic tu r e — by the a d d itio n o f a m u ltitu d e o f

sm all-scale effects: the changes in the local dim ensions associated w ith the curvature o f geometry (fa ilu r e of&
to rem ain a t rest a s observed in the fre e-flo a t fra m e associated w ith A).
W hat is the source of the curvature of spacetime? Momenergy is the source. In

Chaprer 8 we saw the primacy of momenergy in governing interactions between
particles. Crash of mass on mass, no matter how elastic or how destructive, leaves the
total momenergy of the system quite unaltered. By what miracle does this come about?
Education of momenergy from birth onward to good behavior? Goodness of heart?
HOW SPACETIME CURVATURE CARRIES INFLUENCE FROM

ONE MASS TO ANOTHER
The necklace of ball bearings (Figure 9-6) as they ap great, essentially uniform, essentially isolated sphere
proach Earth, examined more closely, reveals a re of mass. The curvature in its character is totally “ tide-
markable feature of spacetime curvature outside a producing,” totally “ noncontractile.”
A n arra y o f test masses covering the surface o f a hollow sphere

fre ely flo a tin g above the E a r th ’s surface w i l l sh rin k in tw o
dim ensions a n d lengthen in one. T he volum e rem ains constant;
only the shape changes. T h is change is evidence o f the noncon
tra c tile, tid e -d r iv in g spacetim e curvature outside E arth.
What do these descriptive terms mean, and how do yes; but a totally noncontractile curvature. Einstein’s
we verify that they apply? W e look at a cluster of ball famous equation, stated in simple terms, tells us how
bearings dotted here and there over the surface of an spacetime curvature responds to mass:
imaginary small sphere, all momentarily at rest relative
to each other and relative to Earth. That shape, how ap p ro p riate m easure of
ever, as the seconds tick by, changes from sphere to spacetim e contractile curvature
ellipsoid. How come? First let’s look at the two dimen at an y p lace, an y time,
sions of the sphere that lie perpendicular to each other in any Lorentz fram e
but parallel to Earth's surface. Both these dimensions of
density of e n e rg y \
the sphere shrink os the boll bearings converge toward
Earth’s center. The up-down dimension of the pattern,
( at that lo cale j
however, lengthens, and twice as much. Why? Newton perceived in that I
says because of the greater gravitational acceleration Lorentz fram e /
of the one nearer Earth. Einstein says because two-per
cent stretch in that dimension compensates one-percent Outside, no mass, no energy, a spacetime curvature
shrinkage in the other two dimensions and keeps the that is totally noncontractile. Inside Earth, however,
volume of the pattern unchanged. Spacetime curvature. there is mass, therefore there is energy — or in a mov-
Obedience to the eyes of a corps of bookkeepers? No, Einstein taught us. The
enforcing agency does not lie far away. It’s close at hand. It’s the geometry of
Spacetim e controls momenergy
spacetime, right where the crash takes place. N ot only does spacetime grip isolated
mass, telling it how to move. In addition, in a crash it sees to it that the participants
neither gain nor lose momenergy. But there is more! Spacetime, in so acting, cannot
9.6 GRAVITATION AS CURVATURE OF SPACETIME 287
ing frame, energy plus energy flow — and therefore Likewise spacetime does not tear. Its fabric just above
spacetime curvature there has a contractile character. Earth’s surface experiences the same lateral contractil
The ball bearings — when shafts are drilled for them so ity as it does just below the surface. Not so with the
that not one of them encounters any obstacle to free- curvature in the two-dimensional domain defined by
float motion — start to converge vertically as well as time and by direction perpendicular to Earth’s surface.
horizontally. The volume shrinks. That, overlooking de In that one plane, curvature within Earth is contractile
tails, is what we mean when we say that “ mass grips but suddenly jumps just above Earth’s surface to the
spacetime, telling it how to curve.” opposite character. Hence the tide-producing charac
ter of spacetime curvature outside Earth. A point twice
There is no Earth mass out at Moon's orbit. How then as far from Earth’s center lies on an imaginary Earth-
does Einstein’s spacetime geometry account for centered sphere that encompasses eight times the vol
Moon’s motion? Answer: Earth’s mass imposes on ume. There the tide-producing curvature experiences
spacetime a contractile curvature throughout Earth’s in eight times the dilution and has one eighth the strength.
terior, as a jumper’s feet impose a contractile curvature Despite this rapid dilution of tide-producing power with
on a trampoline. That contractile curvature, where the distance, it has strength enough at Moon, 60 Earth radii
feet push, forces on the surrounding nontear fabric a away from Earth’s center, to deform Moon from sphere
corresponding lateral stretch. That effect transmits itself to ellipsoid, 1738.35 kilometers in radius along the
in ever more dilute measure to the ever more remote Earth-M oon direction, 1738.15 kilometers in radius
regions of the trampoline. for each of the other two perpendicular directions.
Easy as it is to regard Earth as running the whole show.

Moon too has its part. Like an infant standing on the
trampoline some distance from its mother, it imposes its
own small curvature on top of the curvature evoked by
Earth. That additional curvature, contractile in Moon’s
interior, has tide-driving character outside. W ere the
Earth an ideal sphere covered by an ideal ocean of
uniform depth, then Moon would draw that ocean’s
surface 35.6 centimeters higher than the average in two
domains, one directly facing Moon, one directly oppo
site to it — simultaneously lowering those waters 17.8
centimeters below the average on the circle of points
midway between the two. (These low figures show how
important are funneling and resonant sloshing in deter
mining heights of actual ocean tides on Earth.)
The local contractile curvature of spacetime at Moon’s

location added up along Moon’s path yields the ap
pearance of long-range gravitation, similar to that il
T he deform ation o f the nontear tram poline fa b r ic u n d er the ju m p e r ’s fe e t a n d lustrated in Figure 9-6. Box 2-1 tells a little of the many
elsewhere is analogous to the nontear cu rva tu re o f spacetim e geometry inside influences that have to be taken into account in any
E a rth a n d elsewhere. fuller treatment of the tides.
maintain the perfection assumed in textbooks of old. To every action there is a

corresponding reaction. Spacetime acts on momenergy, telling it how to move; momenergy
M om energy tells spacetim e
reacts hack on spacetime, telling it how to curve. This “handshake” between momen how to curve
ergy and spacetime is the origin of momenergy conservation — and the source of
spacetime curvature that leads to gravitation (Box 9-1).
9.7 GRAVITY WAVES

gravitational energy moving at light speed
In the depths of an ill-fated, collapsing star, billions upon billions of tons of mass cave
Gravity waves from in and crash together. The crashing mass generates a wave in the geometry of
collapsing matter
space— a wave that rolls across a hundred thousand light-years of space to “jiggle”
the distance between two mirrors in our Earthbound gravity-wave laboratory.
A cork floating all alone on the Pacific Ocean may not reveal the passage of a wave.
But when a second cork is floating near it, then the passing of the wave is revealed by
the fluctuating separation between the two corks. So too for the separation of the two
mirrors. There is, however, this great difference. The cork-to-cork distance reveals a
momentary change in the two-dimensional geometry of the surface of the ocean. The
B O X 9- 2
COMPACT STELLAR OBJECTS
Three kinds of astronomical objects exist comparable in mass to Sun but very
much smaller. Two of these have been observed; the third seems an inevita
ble result of Einstein’s theory.
A white dwarf star is a star of about one solar mass, with radius about 5000
kilometers. (The radius of Earth is 6371 kilometers.)Thisgives the white dwarf
a density of approximately 10’ kilograms/meter^ (or one metric ton per cubic
centimeter). As of 1990, approximately 1500 white dwarfs have been iden
tified.
White dwarfs were observed and studied astronomically long before they
were understood theoretically. Today we have come to recognize that a
white dwarf is a star that quietly used up its fuel and settled gently into this
compact state. The electrons and nuclei that make up the body of a white
dwarf are not separated into atoms. Instead, the electrons form a gas in
which the nuclei swim. The pressure of this “ cold” electron gas keeps the
white dwarf from collapsing further.
S. Chandrasekhar calculated in 1930 that no white dwarf can be more mas

sive than approximately 1.4 solar masses (“ Chandrasekhar limit” ) without
collapsing under its own gravitational attraction. His analysis assumed the
mix of electrons and nuclei to be unaltered under compression by a load so
heavy, an assumption that had to be modified in later years. Today we
recognize that enormous compressions squeeze electrons into combining
with protons to make neutrons. At compressions near the Chandrasekhar
limit, the electron gas transforms into a neutron gas, the interior of the star
becomes a giant nucleus, and the whole nature of the compact object
changes to that of a neutron star.
A neutron star has roughly the same density as an atomic nucleus, of the
order of 10'^ kilograms/meter^, or one Earth mass per cube of edge length
400 meters. The radius of a neutron star is approximately 10 kilometers.
9.7 GRAVITY WAVES 289
mirror-to-mirror distance reveals a momentary change in the three-dimensional ge

How to detect gravity w aves
ometry of space itself.
The idea of extracting energy from ocean waves is old. After all, the ability of a
water wave to change a distance lets itself be translated into the ability to do work. The
same reasoning applies ro a gravity wave. Because it can change distance, it can do
work. Ir carries energy. Energy once resident as mass in the interior of a star has
radiated out to us and to all the universe.
O f all the workings of rhe grip of gravity, none is more fascinating or opens up for
exploration a wider realm of ideas rhan a gravity wave. None pushes to a higher pirch
the art of detecting a small effect, and none gives more promise of providing an
unsurpassable window on cataclysmic events deep inside troubled stars. Nevertheless,
no other great prediction of Einstein’s geometric theory of graviry srands today so far
from triumphant exploitation. As of this writing, not one of the nine ingenious
How often is a neutron star formed? Towards answering this still open ques
tion we have one important lead: In our own galaxy we see one supernova
explosion on average about every 300 years [most recent supernova in the
Large Magellanic Cloud, a satellite structure near our galaxy, on February
23, 1987; one seen by Kepler, October 13, 1604; one seen by Tycho Brahe,
November 6, 1572; earlier ones: 1181 a . d . ; July 4, 1054 a . d . ; 1006 a . d . (the
brightest); 185 a . d . ; and two possibles in 386 a . d . and 393 a . d .] . In such an
event a star teetering on the edge of instability finally collapses. The Niagara
Falls of infalling mass in some cases go too far and overcompress the inner
region of the star. That region thereupon acts like a spring, or explosive
charge, and drives off the outer portions of the star. This explains the spec
tacular luminosity that is such a prominent feature of a supernova. The core
that remains becomes a neutron star in some events, it is believed, in others a
black hole.
Neutron stars were predicted in 1934 but not observed until 1968. Many
neutron stars spin rapidly — with a period as short as a few milliseconds. A
neutron star typically has an immense magnetic field. When that field is
aligned at an angle relative to the axis of spin of the star (as in Earth, for
example), it sweeps around like a giant whisk brush through the plasma in the
space around the star. The periodic shock to the electrons of the plasma from
the periodic arrival of this field excites those electrons to radiate periodic
pulses of radio waves and visible light — both observed on Earth. Because of
this behavior, such neutron stars are called pulsars. As of 1990, nearly 500
pulsars have been identified.
A black hole is an object created when a star collapses to a size so small that
strong spacetime curvature prevents it from communicating outward with the
external universe. Even light cannot escape from a black hole, whence its
name. No one who accepts general relativity has found any way to escape
the prediction that black holes must exist in our galaxy. Strong evidence for
the existence of black holes has been found, but it is not yet convincing to all
astrophysicists. A black hole can have a mass as small as a few times the mass
of our Sun. A black hole of three solar masses would have a “ radius” of
about 9 kilometers. There is no theoretical upper limit to its mass.
detectors built to this day has proved sensitive enough to secure any generally agreed
detection of an arriving gravity wave.
Does any truly simple line of reasoning assure us that gravity will inescapably carry
energy away from two masses that undergo rapid change in relative position? Yes is the
conclusion of a little story that savors of mythology. The Atlas of our day, zooming
through space in free float, insists as much as ever on maintaining physical fitness. He
pumps iron, not by raising iron against the pull of Earth’s gravity, but by throwing
apart two identical great iron spheres. Alpha and Beta. He floats between those minor
moons and plays catch with them. Each time they fall together under the influence of
their mutual gtavity, he catches them, absorbs their energy of infall in his springlike
muscles, and flings them apart so that they always travel the same distance before
returning. It’s an enchanting game, but Atlas finds that it’s a losing game. When the
masses fly back together, they never yield up to him as much energy as he must supply
to throw them apart again. Why not?
Say the central point in two words: time delay. Like any force that makes itself felt
through the emptiness of space, the force of gravity cannot propagate faster than the
Gravity waves result from
speed of light. This limitation imposes a delay on the attraction between the two iron
time delay
spheres. Alpha, on each little stretch of its outbound path, feels a pull that originated
from Beta when the two were a tiny bit closer than they are now. The actual force that’s
slowing Alpha is therefore a tiny bit bigger than we would judge from thinking of
them as stationary at their momentary separation. On its return trip inbound along the
same little stretch of path. Alpha experiences a helping pull that originated from Beta
when the two had a separation slightly greater than its present value. The actual force
that’s speeding Alpha inward is therefore a tiny bit less than we would judge from
thinking of them as stationary at theit momentary separation. In each stretch of their
outbound trip, the two masses have to do more work against the pull of gravity than
they get back — in the form of work done on them by gravity— on the same stretch of
path inbound. A calculable amount of energy disappears from the local scene on each
out-in cycle of Atlas’s exercise. Yet the total energy must somehow be conserved.
Therefore the very gravity that steals energy from Atlas and his iron, or from any two
Year
FIGURE 9 -7 . T w o w h irlin g neutron sta rs fu r n is h a g ia n t do ck, whose tim e-keep in g h a n d is the line,
ever-tu rn in g, th a t separates the centers o f those tw o stars. T h a t h a n d does not today keep the “slo w ” schedule
(stra ig h t h o rizo n ta l line) one m ig h t have expected fro m its tim in g a s m easured in 1 9 7 4 . T he d o w n w a rd
sloping curve show s g ra v ity -w a v e theory's p rediction o f the shortening in the tim e required to accum ulate a ny
specified num ber o f revolutions. T he dots show the a c tu a l observed shortening in th a t tim e.
9.7 GRAVITY WAVES 291
masses that rapidly change their relative position, must somehow all the time be
transporting the stolen energy to the far-away. That inescapable theft of energy is in its
quality, its directional distribution, and its magnitude none other than what Einstein
had treated long before under the head of gravity radiation and what we now call
gravity waves.
Atlas couldn’t “see” those gravity waves. Neither have we today yet succeeded in
detecting directly the gravity waves we feel sure must be radiating from sources dotted
here and there in the galaxy and in the universe. However, we have an exciring indirecr
confirmation that gravity waves exist— not through their action on any receptor, bur
G ravity w aves steal en ergy from
through the energy they carry away from a whirling pair of neutron srars. That
orbiting neutron stars
particular “binary pulsar” first revealed itself to Joseph H. Taylor, Jr., and Russell A.
Hulse by periodic pulses of radio waves picked up on the huge disklike antenna at
Arecibo in Puerto Rico. As one of these neutron stars spins on its axis, its magnetic field
spins with it, giving timing comparable in accuracy to the best atomic clock ever built
(Box 9-2). Thanks to this happy circumstance, Taylor and his colleagues have been
able to follow the ever-shortening separation of the two stars and the ever-higher speed
they attain as they slowly spiral in toward an ultimate catastrophe some 400 million
years from now. The timing of the orbits gives us a measure of energy lost as the stars
spiral in. No reasonable way has ever been found to account for the thus observed loss
of energy except gtavitational radiation. As of September 1989, 14 years after first
observation, this loss of enetgy agrees with the rate predicted by theory to bettet than
one percent (Figure 9-7).
Gravity waves and pulses of gravity radiation are sweeping over us all the time from
sources of many kinds out in space. Detecting them, however, we are no better than
the primitive jungle dweller unable to detect and even totally unaware of the tadio
waves that carry past her every minure of the day music, words, and messages.
However, experimentalists are working out ingenious technology and building detec
tor instrumentation of evet-growing sensitivity (Figures 9-8 and 9-9). Few among
them have any doubt of their ability to detect pulses of gravity radiation from one or
another star catastrophe by sometime in the first decade of the twenty-first century.
FIGURE 9 -8 . The proposed M I T - Caltech g ra v ity -w a v e detector w ill ( I ) use the beam fro m a laser (left),
(2 ) sp lit i t by a device (center) analogous to a h a lf-silvere d mirror, (3 ) send one h a lf-stren g th beam to one
fa r a w a y m irror (top) a n d the other to the other fa r a w a y m irror (right), (5 ) allow these beams to undergo
m a n y m a n y reflections (not show n), a n d (6 ) recombine them a t the detector (bottom ). A g r a v ity -w a v e
in c id e n t on E a rth w ill slig h tly shorten the 4 -kilo m eter d ista n ce to the one m irror a n d s lig h tly lengthen the
4-kilo m eter d ista n ce to the other mirror. T h is rela tive a ltera tio n in the p a th length o f the laser beams, i f big
enough, a m p lified enough, a n d p ic k e d u p by detectors sensitive enough, w ill reveal the passage o f the g r a v ity
29 2 CHAPTER 9 GRAVITY: CURVED SPACETIME IN ACTION
FIGURE 9-9. P r o to ty p e g r a v i t y - w a v e d e
te c to r , C a l i f o r n ia I n s t i t u t e o f T e c h n o lo g y ,
P a s a d e n a . The laser beam is ta ilo red (lower
right} f o r entry in to the beam sp litte r (located
where the tw o long lig h t p ip es meet, ju s t to the
le ft o f center in the photograph). The m irrors a t
the ends o f these tw o eva cu a ted lig h t pipes lie
outside the boundary o f the photograph.
Astronomy uses signals of many kinds — light, radio waves, and X-rays among
them — to reveal the secrets of the stars. O f all signals from a star, none comes out
from deeper in the interior than a gravity wave. Among all violent events to be probed
deeply by a gravity wave, none is more fascinating than the dance of death of two
compact stars as they whirl around each other and undergo total collapse into . . . a
black hole!
9.8 BLACK HOLE

over the edge with a scream of radiation
A black hole is a domain whose mass is so tightly compacted that nothing can escape
from it, not even light. Everything that falls in is caught without hope of escape
(Figure 9-10).
‘Escape velocity c” implies To fire a missile from Moon’s surface so that it escapes that satellite’s attraction
black hole demands a speed of 2.3 8 kilometers per second or greater. The critical speed for escape
from Earth — in the absence of drag from the atmosphere— is 11.2 kilometers per
second. When the object does not rotate and is so compact that even light cannot
escape, the “effective radius’’ or so-called “horizon radius’’ is
'circumference of region'
out of which
^ light cannot escape ,
(effective radius)
27T
mass of black hole '
= 2 X (1.47 kilometers) X expressed in
, number of Sun masses j
When a star or cloud of matter collapses to a black hole it disappears from view as
totally as the Cheshire cat did in Alice in Wonderland. The cat, however, left its grin
Black hole still exerts
behind; and the black hole — via the effect of spacetime curvature that we call
“ pull” of gravity
gravity — exerts as much “pull” as ever on normal stars in orbit around it. They are
like participants in a formal dance with lights turned low. Only the white dress of the
girl is visible as she whirls around in the arms of her black-suited companion. From the
9.8 BLACK HOLE 293
p articles a s yet
undetected known
angular
particles gravitational and
momentum
electrom agnetic
mass
ch arg e
an gu lar momentum
FIGURE 9 -1 0 . W h a tever objects f a l l in to a black hole, they possess a t the e n d — a s seen fro m o u tsid e — only
mass, a n g u la r m om entum , a n d electric charge. N o t one other characteristic o f a n y in -fa llin g object rem ains to
betray its p a s t — not a h a ir. T h is leads to the saying, “A black hole has no h a ir .”
speed of the girl and the size of the circle in which she swirls, we know something of the
mass of the invisible companion. By such reasoning it was possible to conclude by
1972 that the optically invisible companion of one long-known star has a mass of
the order of 9.5 solar masses.
This remarkable object came first to attention because in December 1971 the
Uhuru orbiting X-ray observatory detected X-ray pulsations with time scales from one
Cygnus X -1 : A black hole?
tenth to tens of seconds from an object located in the Cygnus region close to the known
star. Why does it give off X-rays? And why does the intensity of the X-rays vary
rapidly from instant to instant? The gas wind from the visible companion varies from
instant to instant like the smoke from a factory chimney. This gas, falling on a compact
object, gets squeezed. To picture the how and why of this squeeze, look from a
low-flying plane at the streams of automobiles converging from many directions on a
football stadium for a Saturday afternoon game. The particles and the gas are pushed
together as surely as the cars in the traffic. The compression of the traffic raises the
temper of the driver, and the compression of the gas raises its temperature as air is
heated when pumped in a bicycle pump. However, because the gas falls from an object
of millions of kilometers in size to one a few kilometers across, the compression is so
stupendous that the temperature rises far above any normal star temperature, and
X-rays come off.
The time scale of the fluctuations in X-ray intensity depends on the size of the object
that is picking up the star smoke, a size less by a fantastic factor than that of any normal
star. Could the object be a white dwarf (Box 9-2)? No, because such a star would be
294 CHAPTER 9 GRAVITY; CURVED SPACETIME IN ACTION
-C ^ ^ B L E 92 T ]> -
BLACK HOLES FOR WHICH THERE WAS

SUBSTANTIAL EVIDENCE AS OF SEPTEMBER 1989
(Uncertainties in masses are of the order of 20 to 50 percent.)
Astronomical designation Mass

of black hole (in solar masses)
Cygnus X-1 9.5

LMC X-1 2.6
AO 620-00 3.2
LMC X-3 7.0
SS433 4.3
Black hole at center of our galaxy 3.5 X 10*
visible. A neutron star? No, because even matter compressed so tightly that it is
transformed to neutrons cannot support itself against gravity if it has a mass much over
two solar masses. No escape has been found from concluding that Cygnus X-1 is a
black hole. This great discovery transformed black holes from pencil-and-paper
objects into a lively and ever-growing part of modern astrophysics (Table 9-1).
Much attention went in the 1980s to a presumptive black hole with a mass of about
Black hole at center of three and a half million times the solar mass and a horizon radius of about ten million
our g a la x y? kilometers. It floats at the center of our galaxy, rhe Milky Way. Around it buzz visible
stars of the everyday kind, most of them fated to fall eventually into that black hole
and increase its mass and size. That stars close to the center of our galaxy go around as
fast as they do is one of the best indicators we have for the presence, and one of the best
measures we have for the mass, of the central black hole, which is itself invisible.
In contrast to dead solitary black holes, the most powerful source of energy we know
or conceive or see in all rhe universe is a black hole of many millions of solar masses,
gulping down enormous amounts of matter swirling around it. Maarten Schmidt,
working at the Mount Palomar Observatory in 1956, was the first to uncover evidence
for these quasistellar objects, or quasars, starlike sources of light located not billions of
Q u a sa r energy output from matter kilometers but billions of light-years away. Despite being far smaller than any galaxy,
swirling into black hole? the typical quasar manages to put out more than a hundred times as much energy as
our own Milky Way, with its hundred billion stars. Quasars, unsurpassed in brilliance
and remoteness, we call lighthouses of the heavens.
Observation and theory have come together to explain in broad outline how a
quasar operates. A black hole of some hundreds of millions of solar masses, itself built
by accretion, accretes more mass from its surroundings. The incoming gas, and
stars-converted-to-gas, does not fall in directly, any more than the water rushes
directly down the bathtub drain when the plug is pulled. Which way the gas swirls is a
matter of chance or past history or both, but it does swirl. This gas, as it goes round and
round, slowly makes its way inward to regions of ever-stronger gravity. Thus com
pressed, and by this compression heated, the gas breaks up into electrons — that is
negative ions — and positive ions, linked by magnetic fields of force into a gigantic
High-efflciency conversion of
accretion disk. Matter little by little makes its way to the inner boundary of this
gravitational en ergy to radiation
accretion disk and then, in a great swoop, falls into the black hole, on its way crossing
the horizon, the surface of no return. During that last swoop, hold on the particle is
relinquished. Therefore, the chance is lost to extract as energy the full 100 percent of
the mass of each infalling bit of matter. However, magnetic fields do hold onto the
ions effectively enough for long enough to extract, as energy, several percent of the
9.8 BLACK HOLE 295
I
Co
3O
u
o
w
c
A L B E R T E IN S T E IN
Ulm, G erm any, M a rch 14, 1 8 7 9 — Princeton, N e w Jersey, A p r il 18, 1 9 5 5
“Newton himself was better aware of the weaknesses inherent in his intellectual
edifice than the generations which followed him. This fact has always roused my
admiration.’’
* ★ ★
“Only the genius of Riemann, solitary and uncomprehended, had already won its
way by the middle of the last century to a new conception of space, in which space
was deprived of its rigidity, and in which its power to take part in physical events
was recognized as possible.’’
★ ★ ★
“ All of these endeavors are based on the belief that existence should have a
completely harmonious structure. Today we have less ground than ever before for
allowing ourselves to be forced away from this wonderful belief.’’
mass. In contrast, neither nuclear fission nor nuclear fusion is able to obtain a
conversion efficiency of more than a fraction of a percent. O f all methods to convert
bulk matter into energy, no one has ever seen evidence for a more effective process than
accretion into a black hole, and no one has even been able to come up with a more
feasible scheme for one.
O f all the features of black hole physics in action, none is more spectacular than a
quasar. And no lighthouse of the skies gives more dramatic evidence of the scale of the
universe.
9.9 THE COSMOS

a final crunch?
The more distant quasars and galaxies are, the greater the speed with which they are
Expanding universe; Evidence for
big bang beginning observed to be receding from us. This expansion argues that somewhere between ten
and twenty billion years ago the universe began with a big bang, a time before which
there was no time.
We see around us relics of the big bang, not only today’s rapidly receding galaxies
but also today’s abundance of the chemical elements — some among them still
radioactive, the “still warm ashes of creation” (V. F. Weisskopf) — and today’s
greatly cooled but still all-pervasive “primordial cosmic fireball radiation.” We now
believe that in the first instants of its life, the entire universe filled an infinitesimally
small space of enormous density and temperature where matter and energy fused in a
homogeneous soup. Immediately the universe began expanding. After about 10“ ^
seconds it had cooled enough that subatomic particles condensed from the matter-
energy soup. In the first three minutes after the big bang, neutrons and protons
“ O p e n ” universe expanding
forever? combined to make heavier elements. Eons later stars and galaxies formed. Never since
O r “ clo sed ” universe that has the universe paused in its continual spread outward.
recontracts to crunch? Will the universe continue expanding forever? Or will its expansion slow, halt, and
An open question! turn to contraction and crunch (Table 9-2), a crunch similar in character but on a far
larger scale than what happens in the formation of a black hole? Great question! No
one who cares deeply about this question can fail to celebrate each week that week’s
astrophysical advances: instruments, observations, conclusions.
We have come to the end of our journey. We have seen gravity turned to float, space
and time meld into spacetime, and spacetime transformed from stage to actor. We
have examined how spacetime grips mass, telling it how to move, and how mass grips
spacetime, telling it how to curve. O f all the indications that existence at bottom has a
simplicity beyond anything we imagine today, there is none more inspiring than the
unsutpassed simplicity of gravity as we now see it.
REFERENCES
Extended portions of this chapter were copied (and sometimes modified) from
John Archibald W h e e l e r , Into Gravity and Spacetime (Scientific Ameri
can Library, a division of HPHLP, New York, 1990).
For details of Galileo’s views on motion, see Galileo Galilei, Dialogues Concerning
Two New Sciences, originally published March 1638; one modern translation is
by Henry Crew and Alfonso de Salvio (Northwestern University Press, Evanston,
111., 1950).
How Newton came only in stages to the solution of the problem of fall is told
nowhere with such care for the fascinating documentation as in Alexander Koyre,
“A Documentary History of the Problem of Fall from Kepler to Newton,”
Transactions of the American Philosophical Society, Volume 45, Part 4 (1955).
Keynes quotation under Newton portrait: Reprinted by permission of the pub
lisher, Horizon Press, from Essays in Biography by John Maynard Keynes,
copyright 1951.
THE COSMOS 2 9 7
A CLOSED-MODEL UNIVERSE
COMPATIBLE WITH OBSERVATION
Radius at phase of maximum 18.9 X 10^ light-years or
expansion 1.79 X 10^^ meters
Time from start to maximum size 29.8 X 10^ years or

2.82 X 10^^ meters
Radius today 13.2 X 10^ light-years
Time from start to today’s size 10.0 X 109 years
Time it would have taken from start 20.0 X 109 years

to today’s size if the entire expansion
had occurred at today’s slowed rate of
expansion
Ptesent expansion rate An extra increment of recession

velocity of 15.0 kilometers/second for
every extra million light-years of
remoteness of the galactic cluster
113.2 degrees
Fraction of the way around the 3- ---------- ----- = 6 2 .9 %
sphere universe from which we can in 180 degrees
principle receive light today
Fraction of the matter in the 3-sphere 74.4%

universe that has been able to
communicate with us so far
Number of new galaxies that come One!

into view on average every three days
Average mass density today 14.8 X 10“ ^^ kilogram/meter^
Average mass density at phase of 5.0 X 10“ ^^ kilogram/meter^

maximum expansion
Rate of increase of volume today 1.82 X 10®^ meters^/second
Amount of mass Alconv “ 5.68 X 10’^ kilograms

In geometric units:
M= = 4.21 X 1026 njeters
Equivalent number of suns like ours 2.86 X 1023
Equivalent number of galaxies like ours 1.6 X 1012
Equivalent number of baryons 3.39 X 10«9

(neutrons and protons)
Total time, big bang to big crunch 59.52 X 109 years
/
/ /
Figure 9-2: Figure and data from Journal of Spacecraft, Volume 11 (September
1974), pages 6 3 7 -6 4 4 , published by the American Institure of Aeronautics and
Astronautics. Data also from D. B. De Bra, APL Technical Digest, Volume 12:
pages 1 4 -2 6 .
Figure 9-3 from Philosophiae Naturalis Principia Mathematica (Joseph Streater,

London, July 5, 1686); Morre translation into English revised and edited by
Florian Cajori and published in two paperback volumes (University of California
Press, Berkeley, 1962). This is also the source of rhe quote in Section 9.6:
“Absolute space, in its own namre, without relation to anything external, remains
always similar and immobile.”
Three quotations under the Einstein picture come from Albert Einstein, Essays in
Science (Philosophical Libraty, New York, 1934).
Quotation in Section 9.6: “Rather than have one global frame with gravitational
forces we have many local frames without gravitational forces.” Steven Schutz, in
January 1966 final examination in course in relativity, Princeton University.
For an exciting and readable overview of the experimental proofs of Einstein’s

general relarivity theory, see Clifford M. Will, Was Einstein Right? Putting
General Relativity to the Test (Basic Books, New York, 1986). In particular
(Chapter 10, pages 1 8 1 -2 0 6 ), he describes at some length the emission of
gravity waves by the binary pulsar system studied by Joseph H. Taylor, Jr., and
Russell A. Hulse.
ANSW ERS TO O D D -N U M B ER ED EXERCISES
chapter 1
1 - 1 a 10.2 meters b 270 meters c 10^ meters d 10"* kilometers ~ 2 times
Boston-San Francisco distance l-3 a 2.6 X 10*^ meters b 5.3 X 10“ ^ sec
ond c 1.85 X 10“ *° hours d 52 weeks e 5.4 X 10° furlongs l-5 a 4 years
b 4 /5 the speed of light = 2.4 X 10® meters/second l-7 a 4 meters b V7
meters = 2.65 meters c V l5 meters = 3.87 meters d 2 meters e 4 meters (same
as part a) l-9 a 2 X 10’ years h v — 0.995 c 6.33 X 10"* years, v =
0.9995 d r /= 1 - 5 X 10“ ** = 0.99999999995 1-1 la 2 X 10“ "* second
b 133 half-lives; (1/2)*’’ ~ lO” "*** c 3 half-lives d zero space separation (creation
and decay occur at the same place in rocket frame) e 3 half-lives = 4.5 X 10“ °
second
chapter 2
2 - la hit the ceiling b same answer c Rider cannot tell when elevator reaches
top. 2-3 Set clock to 10 seconds, start when reference flash arrives.
2 -5 a Experiment in progress for 1/0.96 = 1.04 meters of time. In this time, test
particle falls 6 X 10“ *^ meters, about 10“ ^ diameter of a nucleus, b 3 X 10“ "*
second, 10’ meters 2-7 3.6 millimeters; 19.7 seconds. Spacetime region: 20
meters X 20 meters X 20 meters of space X 59 X 10® meters of time 2 -9a de
crease (think of each ball bearing in an elliptical orbit around the center of Earth)
b apart c No, you cannot distinguish rising from falling. At the top you notice
nothing inside the coach. 2-11 = 0.735 thespeedoflight. 2 - 13a Effec
tive time of fall: 4.67 seconds. Net velocity of fall: 1284 meters/second. b Angle of
deflection: 4.3 X 10“ ° radian = 2.5 X 10“ "* degree = 0.88 second of arc
chapter 3
3 - la 6 0 seconds b 4 5 seconds against the current, 2 2 .5 seconds with the current,
6 7 .5 seconds round trip c No 3-3 If different kinds of clocks ran at different rates
in a free-float rocket frame, then this difference could be used to detect the relative
velocity of the laboratory from inside the rocket, which violates the Principle of
Relativity. This does not mean that the common rate of rocket clocks will be the same
as measured in rocket and laboratory frames. 3 -5 a 11.5 light-years b 9 .4 3
years c v = 0 .6 d 8 years = the interval between the two events. 3-7 The
bullet misses. Coincidence of A and A ' (event 1) and firing of the bullet at the other
end of spaceship 0 (event 2) cannot be simultaneous in both rocket reference frames.
The right panel of the figure is wrong. Consistent with the Train Paradox (Section
3 .4 ), spaceship O' (standing in for the train frame) will observe the bullet to be fired
before coincidence of A and A ', thus accounting for the fact that bullet misses.
3-9a sin yy = (in meters/meter) b siny/ ~ \j/ ~ 10“ "*radian = 21 seconds
of arc c sin y / and tan y / are both approximately equal to y / for small y/. Therefore the
difference between the two predictions cannot be used to distinguish between relativ
istic and nonrelativistic ptedictions. d in a direction 0 .5 2 4 radians = 3 0 degrees
ahead of transverse 3 -llg (l) = 10“ ^, r'buUet “ 2 X 10“ °. Their product is
2 X 10“ *’, very small compared with 1; thetefore we expect t'buiiet to be the sum of
t'buUct ^tid the form verified in everyday experience at low speeds.
(2) Vb,ji„ = 2 4 / 2 5 = 0 .9 6 (3) I'buiiet = t'ught = + 1 ("<) t'buUet = t'light = “ 1•
Yes, expected from the Principle of Relativity. 3 -1 3 a 0.32 meters = 1 .1 nanosec
ond b 6.0 X 10’ periods c No shift would imply the speed of light is the same for
the frame of Earth going one way around Sun as compared with frame of Earth going
the opposite direction around Sun. d dc = — ( 2 /n^){/^l/T)dn and dc/c = 2
299
300 ANSWERS TO ODD-NUMBERED EXERCISES
dn/n For d n = 'iY . 10“ ^ and n = 6.0 X 10’, we have the maximum value of
d c /c = 1 X 10“ ®(sign not important). Hence dc ^ 3 meters/second is the maxi
mum change in the speed of light that could have escaped detection in this very
sensitive experiment. 3-15a visual distance apart = time lapse between
images = (1 — v)Ac, visual speed of approach = = v /{ l ~ v)- r '^ p ^ =
4 when = 4 /5 ; r'approach “ 99 when v — 0.99 b visual distance apart = vAt-,
time lapse between images = (1 + v)At\ visual speed of recession = ~
(1 + for t'approach = 4 when v = 4 /5 , then = 4 /9 = 0.44; for t'.p p ^ =
99 when V = 0.99, then J'rccede ” 0.497 3-17a Light leaves E one meter of time
earlier than light from Gin order to enter the eye at the same time. In this time the cube
moves V meter of distance, equal to x in the top right figure, b The angle <f) is given
by the expression sin (f) = v. For v ~ * 0, this visual angle of rotation goes to zero, as
we experience in everyday life. F o r > 1, this visual angle of rotation goes to 90
degrees, and the cube shows us its back side as it passes overhead, c The word
“really” is not appropriate; each mode of observation is valid; some will be more
useful than others for different applications. (Requested speech to each observer not
included here.) d The “cube” will look sheared, with top EF pulled backward a
distance x with respect to bottom GH in the left panel.
special topic
L -la = 3/1 7 = 0.176 for speed of Super 6 times speed of light h v ^ = 1 /
3 = 0.333 for infinite speed of Super L-3b 128days e (l) 0.1 meter of time; too
small for either wristwatch or electronic clock (2) about 1O'*meters of time; too small
for wristwatch but easily detected by electronic clock (3) distance is 10*^ meters, or
about 6.7 times the Earth-Sun distance. L-5d = 0.944 L-11 The man
hole is tilted, so it passes over the meter stick without collision. L-13a At the
beginning and the end of their trip (and all during their trip), Dick and Jane are
separated by 12 light-years as measured in the Earth frame. Final velocity: v = 3 /4 .
Aging of each astronaut = proper time along either worldline = sum of the space-
time intervals along each segment of either worldline = Vl5 + V l24-V 7 years =
10 years, b Yes. Yes. c (l) Jane stops accelerating 13.6 years earlier than Dick.
(2) 30 years (3) 30 years (4) 43.6 years (5) Dick: 50 years old. Jane: 63.6 years
old. (6) 18.1 lightyears, which is just / = 1.51 times their 12-light-year separation
measured in the Earth frame by Mom and Dad. (d)(1) Yes (2) Yes Yes
(3) Jane’s (4) Yes. No. (5) It’s the old story: relativity of simultaneity, in this case
the fact that Dick and Jane stop accelerating simultaneously only in the Earth frame,
e Then, by symmetry, Dick will be older than Jane in their final rest frame. All the
numbers will otherwise be the same, f Then they will start and stop simultaneously
in Earth frame and also in the final rocket frame; they will be the same age at these
stopping events in both frames. L-15c For the extreme relativistic case when
^ 1, then > 1 also.
chapter 4
4 -la 11.6years b 18.6years c30.2years d 7 .6 7 years e l4 .6 7 y e a rs f 22.34
years g 5.75 light-years h 7.67 years, 5.07 years i 14.67 years, 30.2 — 5.1 =
25.1 years 4-3a The engineer is wrong, b Frequency of oscillation increases by
V2 when voltage doubles, c frequency in cycles/second ~ f ~ (.qVJSmL^Y^^,
where m and q are mass and charge of the electron, I/, is the voltage applied, and L is
the width of the box = 1 meter, d Minimum round-trip time across box at the
speed of light is 2 L /c s o f ^ = c/2L. e For the Newtonian K g i o n ,f / f ^ = {q V J
(2wc^)]^/^. For the extreme relativistic region, f / f ^ = 1. The quantity qV„ is a
measure of electron potential energy at the wall or electron kinetic energy at the screen.
ANSWERS TO ODD-NUMBERED EXERCISES 301
W e expect the Newtonian analysis to be correct when this energy of motion is very
much less than the rest energy mc^. The extreme relativistic analysis will be correct
when qVo is very much greater than mc'^. The crossover occurs (the two dashed curves
intersect) where qV„ ^ 2mc^ or ~ 10^ volts, f For low speeds, the ratio
will follow the Newtonian curve. At extreme relativistic speeds, the proper time
for one period —* 0 and the proper frequency infinity.
chapter 5
5 - l a ( l ) 1 year (2 ) 1.94years (3) 0.87 year (4 ) 3.81 years b 5 .2 0 yearsc solid
line traveler will be younger 5-3a event A is at {x, t) = (0,0); event B is at (0,1);
event C is a t(1.5, 3 .5 )o r(— 1.5, 3 .5 );e v e n tD isa t(3 ,6 )o r(—3 ,6 ) b eventD isat
(x, t) = (0, 0); event C is at (0, —2); event B is at (0, —4); event A is at (—0.75,
— 5.25) or (4-0.75, — 5.25) c = + 0.6 d Yes 5-5d 3136 cycles/sec-
ond e 31.4 cycles/second 5-7 Hint: As with most paradoxes in relativity, the
solution has to do with the relativity of simultaneity.
chapter 6
6 - la Events 1 a n d 2: (1) Proper time: 4 meters (2) Yes (3) Yes (4) No
Events 1 a n d 3: (1) Proper distance: 4 meters (2 ) No (3 ) No (4) Yes E v en ts2
a n d 3: (1) zero (2 ) Yes (3) No (4) No b vî = 3 /5 in 4- x-direction for
both 6 -3 a Set t ' = 0 in the first inverse Lorentz transformation equation (L-11)
and solve for . b Set x ^ = 0 in the second equation (L-11) and solve fot .
(Why does the result look so funny?) 6-5a Yes, explosion. (Sorry!) b No change
in prediction. (The impact at A and activation of the detonation switch are spacelike
events; the laser pulse cannot connect them.)
chapter 7
7 - la [5w, V24»?, 0, 0} b [w, 0, 0, 0] c [Vl0»?, 0, 0, 3»«] d [ 5 m ,0 ,— 'l24m ,0]
e [10 m, 2.66 m, 5.32 7.98 »?}. 7 -3a 0.05 milligram b O . l milligram
7 -7a wristwatch time: 32 seconds; Earth time: 1000 centuries b E/m = 10^^. 1.7
million metric tons. 7-9a Eg = 9 units b = V32 units = 5.66 units
c mg = 7 units d greater: m^ — 15 units > 4- Wg = 9 units 7-1 la proton:
938 MeV; electron: 0.511 MeV b P i ^ ^ 0.12. Proton kinetic energy at limit ~ 6
MeV. Electron kinetic energy at limit ~ 3.4 X 10~^ MeV = 3.4 keV. Yes, designer
of color TV tubes (electron kinetic energy — 25 keV) must use special relativity.
chapter 8
8 -la approximately 35 X 10“ ® kilograms = 35 micrograms b approximately
600 kilograms. More, c approximately 6 X 10^^ seconds or about 2 million years!
Chemical burning in Eric’s body produces large quantities of waste products. Elimi
nation of these products carries away mass enormously faster than mass is carried away
as energy. 8-3a Force is approximately 3 X 10“ ^ newtons, or the weight of
3 X 10“ kilograms. You should not be able to feel it. b pressure on a perfectly
absotbing satellite == 5 X 10“ ^ newton/meter^; on a perfectly reflecting satellite ==
9 X 10“ ^ newton/meter^; somewhere in between for a partially absorbing surface.
Total enetgy absorbed/meter^, not color of the incident light, determines pres
sure. c acceleration approximately 10“ ^ g d particle radius approximately 10~®
meter, independent of the distance from Sun 8-7 density approximately 5 X
10^® kilograms/meter^ = 5 X 1 0 ’ grams/centimeter^, or 50 million times the den-
302 ANSWERS TO ODD-NUMBERED EXERCISES
sity of water! 8-9 m^)/{2m) 8-1 la From conservation equa

tions, show that cos (/) > 1, which is impossible, b If the total momentum is zero
after the collision, it must be zero before the collision. But the alleged single photon
before the collision cannot have zero momentum. Therefore the reaction is
impossible. 8-13 2Ec = E^-\- {Ey^ — and 2Ep = E ^ ~ ( E ^ —
If the particle is at rest, then Ej^ = m and E ^ = Ep = m /2. 8-15a = w(£ -b
m)l\E + m - (E^ - cos 0c] 8 -17a 1.8 TeV b £ « 1700 TeV
8-19e No 8-21 When the bulb is seen way ahead, its light is very intense and
radically blue-shifted. While still seen ahead, there is an angle of observation (de
pending on the speed) at which the light is red, but dim. As the bulb is seen to pass the
observer, its light is infrared and very dim. As the bulb is seen to retreat into the
distance, its light is extremely dim and radically red-shifted. 8-23a v = 0.38
b 13 X 10^ years c Allowance for gravitational slowing will decrease the estimated
time back to the start of the expansion. 8-25 l ^ f / f ~ The
observed frequency will increase for molecules approaching the observer and decrease
for molecules receding from the observer. The overall effect— at temperatures for
which Newtonian expressions are valid — is to produce a spread of frequencies
approximated by the expression above (“ Doppler line broadening”). 8-
21 E ' = m /2, E = m, (f) = 30 degrees. 8-35a The incident gamma ray (with
excitation energy £ ) imparts a small kinetic energy K to the iron atom, for which
Newtonian expression is valid: K = p'^/2m = E'^/2m, since p = E {os the gamma
ray. Then (energy of recoil)/(energy for excitation) = K /£ ~ £/(2w ) = 1.4 X
10“ ^. But fractional resonance width (6 X 10“ *^) is smaller than this by a factor of
almost a million, so the iron nucleus cannot accept the gamma ray and conserve
energy, b One gram is about 10^^ atoms. If the m in the above equation increases by
the factor 10^^, then the energy of recoil is a factor 10^^ smaller, and the nucleus will
not notice the residual mismatch in energy. 8-37 A / / / = — gz/eY, v = 0.7 X
10“ ® meter/second towards emitter 8-39 A //( /,A T ) = (3/2)/^/(wc^) ~
1.2 X 10“ '^ per degree.
INDEX
Abbott, James, 135 (Exercise 4-1) cannonball, human, 45 (Exercise 2-1)

aberration of starlight, 81 (Exercise 3-9) Canopus, trip to, 121-134 (Chapter 4)
absolute elsewhere, 181 cat, Cheshire, 292
“absolute” space and time (Newton), 160, 284 causaliry, light speed limit on, 171 (Section 6.1), 180-183
abuse of the concept of mass, 244-251 (Section 8.8) center of momentum frame, 246-251
acceleration, relative, as witness to gravity, 30-36 (Sec Cerenkov radiation, 80-81 (Exercise 3-8)
tions 2.3, 2.4), 280-287 (Sections 9.4, 9.5, 9.6) Chandrasekhar, S., 288
acceleration-proof clocks, 152 Chandrasekhar limir, 288-289 (Box 9-2)
active future, 182 chemistry, relativistic, 254 (Exercise 8-2)
addition of velocities, 8 2 -84 (Exercise 3-11), 103-110 Civil War, American, 25
(Secrion L.7) Cleoparra, 228
Aging, Principle of Maximal, 150 clock
Akihito, Emperor of Japan, 138 acceleration-proof, 152
American Civil War, 25 atomic, test of twin effect, 131
Andromeda galaxy bad, 112-113 (Exercise L-2)
Enterprise in, 106-107 (Box L-2) construction of, 78 (Exercise 3-3)
trip to by rocket, 2 2-23 (Exercise 1-9) light-flash, 12
trip to by Transporter, 23 (Exercise 1-10) reference, 37
angles, transformation of, 114-115 (Exercise L-6) clock paradox, see Twin Paradox
annihilation, positron-electron, 237-238, 242-243 clocks
(Sample Problem 8-4), 260 (Exercises 8-14, 8-15) latricework of, 37-39 (Secrion 2.6), 4 5 -4 6 (Exer
appearance, visual, of relarivistic objects, 64, 9 2-93 (Ex cises 2-3, 2-4)
ercise 3-17) plane of agreement of, 120 (Exercise L-15)
Arecibo radio antenna (Puerto Rico), 291 run at different rates in gravitational field, 118 (Ex
arrow of momenergy, 191-195 (Section 7.2) ercise L-13)
autobiography of a phoron, 184-185 (Exercise 6-4) “run slow?”, 76-77 (Box 3-4)
available interaction energy, 261 (Exercise 8-17) collapse, gravitational, 288, 292-295 (Section 9.8)
colliders, 261-262 (Exercise 8-17)
backyard zoo of particles, 235 (Box 8-1)
collision, 221-252 (Chapter 8)
bad clock, 112-113 (Exercise L-2)
analyzing, 239 (Box 8-2)
barn and pole paradox, 166 (Exercise 5-4)
elastic, 222, 240-241 (Sample Problem 8-3)
Barrlett, Sreven, 19
inelastic, 222-223
Bay of Fundy, tides in, 32-33 (Box 2-1)
solving problems, 239 (Box 8-2)
Berman, Eric, 254 (Exercise 8-1)
comet, 35
beta (Greek p), symbol for speed, 41, 253
communication, time delay in, 39-40
Betrayal, Great, 108-109 (Box L-1)
communications storm, 48 (Exercise 2-11)
black hole, 289 (Box 9-2), 292-295 (Section 9.8)
compact stellar objects, 288-289 (Box 9-2)
as source of neutrinos, 80 (Exercise 3-8)
components of momenergy, 195- 199 (Section 7.3)
bomb
energy, 201-206 (Section 7.5)
fission, 249
momentum, 199-200 (Section 7.4)
hydrogen (fusion), 248-249
Compton, Arthur Holly, 229
Super, 108-109 (Box L-1)
Compton scattering, 229, 231, 267-268 (Exercise 8-29)
bounce, free-float, 45 (Exercise 2-2)
examples of, 268 (Exercise 8-30)
Braginsky, Vladimir, 36, 223
inverse, 269-270 (Exercise 8-32)
broadening of spectral lines, Doppler, 264 (Exercise 8-25)
computer size, 22 (Exercise 1-8)
bulb
cone, light, partition in spacetime, 177- 183 (Section 6.3)
flickering, paradox of, 186-187 (Exercise 6-7)
conscience-guided satellite, 277-279
speeding, 264 (Exercise 8-21)
conservation laws, see energy; momentum; momenergy
c (speed of lighr), see light speed conserved, defined, 208-209 (Box 7-3)
Caesar, Julius, 106-107 (Sample Problem L-2) constant, defined, 208-209 (Box 7-3)
303
304 INDEX
contracting train paradox, 187-188 (Exercise 6-8) E = m?, 203, 206, 250
contraction, Lorentz, 63-65 (Seaion 3.5), 126-127 from Doppler shift, 264-265 (Exercise 8-26)
(Section 4.7) Earth
for cosmic rays, 215-216 (Exercise 7-7) curved, 281-283 (Section 9.5)
described by stretch factor, 157 mass in units of meters, 258
how it occurs, 119-120 (Exercise L-14) surface of as a free-float frame, 46 (Exercise 2-5)
or rotation?, 9 2-93 (Exercise 3-17) Eigenzeit, 11; ree also proper time
conversion factors Einstein, Albert
for energy, 203, 250 admiration for Newton, 284, 295
miles to meters, 2, 16, 58-59 (Box 3-2) curvature equation, 286
for momentum, 200 eliminate gravity, 28
seconds to meters, 6, 12, 16, 58-59 (Box 3-2) epigram, iii
conversion of mass to energy, 237-244 (Section 8.7), equivalence of energy and mass, 250, 254-258
254 (Exercise 8.1) (Exercise 8-5)
cosmic rays, 160, 215-216 (Exercise 7-7) and Galileo and Newton, 275-276 (Section 9.2)
cosmos, 296-297 (Section 9.9) and gravity, 275-298 (Chapter 9)
creation of proton-antiproton pair by an electron, 261 happiest thought of life, 25, 44
(Exercise 8-16) picture and quotes, 295
curvature special relativity, 5
of Earth, 281-283 (Section 9.5) Train Paradox, 62-63
equation, Einstein’s, 286 Einstein puzzler, 78 (Exercise 3-2)
of spacetime, 280-287 (Sections 9.4, 9.5, 9.6) elastic collision, 222, 240-241 (Sample Problem 8-3)
electrodynamics, quantum, 185 (Exercise 6-4)
Daytime surveyor, 1 -4 (Section 1.1), 16-17 (Box 1-1) electron, 235 (Box 8-1)
decay creation of proton - antiproton pair by, 261 (Exercise
mu meson, 2 3-24 (Exercise 1-11) 8-16)
pi-naught meson, 267 (Exercise 8-28) electron-positron annihilation, 237-238, 242-243
pi-plus meson, 24 (Exercise 1-12) (Sample Problem 8-4), 260 (Exercises 8-14, 8-15)
positronium, 260 (Exercise 8-13) electron-positron pair production, see photon
defleaion of starlight by Sun, 50-51 (Exercise 2-13) electrons, fast, 215 (Exercise 7-6)
density of companion of Sirius, 258-259 (Exercise 8-7) elsewhere, absolute, 181
detonator paradox, 185-186 (Exercise 6-5) Emperor Akihito, 138
deuterium, combined with helium, 237 Emperor Hirohito, 137
Dicke experiment, 36, 4 8 -5 0 (Exercise 2-12) emptiness of spacetime, 56-57 (Box 3-1)
dimension, transverse, invariance of, 65-67 (Section 3.6) encounter, particle, 239 (Box 8-2)
distance energy, 196, 213 (Table 7.1)
invariance of, 4, 17 conserved in a collision, 1 8 9 -1 9 0 (Section 7.1),
proper, 174 207, 2 2 2 -2 2 3 (Section 8.2), 239 (Box 8.2)
dog and passenger paradox, 25-26 conversion of mass to, 2 3 7 -2 4 4 , (Section 8.7), 254
Dog Star (Sirius), 135 (Exercise 4-1), 258-259 (Exercise (Exercise 8-1)
8-7) interaction, 261 (Exercise 8-17)
Doppler shift kinetic, 201, 203, 206
along x-direction, 114 (Exercise L-5), 263 (Exercise Newtonian, low-velocity limit, 190, 203, 205 (Box
8-18) 7- 2)
at limb of Sun, 264 (Exercise 8-22) and mass, 201, 203, 206, 250-251, 254-258
E = mi?- from, 264-265 (Exercise 8-26) (Exercise 8-5)
equations, 263 (Exercise 8-19) production of, in Sun, 242-245 (Sample Problem
line broadening, 264 (Exercise 8-25) 8- 5)
measurement of by resonant scattering, 271-272 quantities related to, 213 (Table 7-1)
(Exercise 8-36) rest, 201, 203, 250
Twin Paradox using, 264 (Exercise 8-24) shift of, due to recoil of emitter, 270 (Exercise 8-33)
down with relativity!, 79 (Exercise 3-6) threshold, 236, 259 (Exercise 8-12), 261 (Exercise
DUMAND experiment, 80 (Exercise 3-8) 8-16)
dwarf, white, 258-259 (Exercise 8-7), 288 (Box 9-2) as “time” part of momenergy, 201-206 (Section 7.5)
INDEX 305
transformation of, 215 (Exercise 7-5) Lorentz, see free-float frame

in unit of mass, 190, 203 reference, 5; see also free-float frame
without mass (photon), 228-233 (Section 8.4), rocket, 41-43 (Section 2.9)
273-274 (Exercise 8-40) super-rocket, 69, 71, 140-142
energy of light, 230 free float, 25-45 (Chapter 2)
energy of photon and frequency of light, 268-269 (Exer free-float bounce, 45 (Exercise 2-2)
cise 8-31) free-float (inertial) frame, 26-29 (Section 2.2)
Engelsberg, Stanley, 4 5-46 (Exercise 2-4) defined, 31
Enterprise, Starship, 106-107 Earth surface as, 46 (Exercise 2-5)
Eotvos, Baron Roland von, 36 extent of, near Earth, 30-34 (Section 2.3), 46 (Ex
equivalence of energy and mass, 250, 254-258 (Exercise ercise 2-6), 47 (Exercise 2-8), 285
8-5) extent of, near Moon, 4 6-47 (Exercise 2-7)
ether theory of light propagation, 84, 88 local, 30-34 (Section 2.3), 284
Euclidean 3-vector, 192 (Box 7-1) rocket, 41-43 (Section 2.9)
Euclidean geometry, 8, 11, 126, 143, 151, 172, 177, stripped down, 121-122 (Section 4.2)
192 (Box 7-1), 198, 279 super-rocket, 69, 71, 140-142
event, 10, 16 and test of twin effect, 133
and interval, 9 -1 1 (Section 1.3) touring spacetime without, 160-162 (Section 5.9)
locating, with latticework of clocks, 37-39 (Section verifying, 41, 279
2 . 6) what is same in different, 60-62 (Section 3.3)
not owned by any frame, 43 what is not same in different, 56-60 (Section 3.2)
reference, 38 frequency of light and energy of a photon, 268-269
events (Exercise 8-31)
relation between, 11, 172 —177 (Section 6.2) Fundy, Bay of, 32 —33 (Box 2-1)
time of, 38, 137-139 (Section 5-1) fusion, 237-238
evidence, experimental, for Twin Paradox 131-134 (Sec fusion bomb, 248-249
tion 4.10) future, active, 182
expanding universe, 82 (Exercise 3-10), 264 (Exercise
8-23), 296-297 (Section 9.9) Galilean principle of relativity, 53-55
experimental evidence for Twin Paradox, 131-134 (Sec Galilean transformation, 113 (Exercise L-3)
tion 4.10), 272-273 (Exercise 8-39) Galilei, Galileo
and gravitational acceleration, 36
fast electrons, 215 (Exercise 7-6) and Newton and Einstein, 275-276 (Seaion 9.2)
fast protons, 214-215 (Exercise 7-4) picture and quotes, 54
faster than light?, see light, faster than? and Leaning Tower of Pisa, 36
Federation, 108-109 (Box L-1) and Principle of Relativity, 53-55
Feynman, Richard, 1 and tides, 32
firing meson, 110 (Sample Problem L-3) gamma (Greek y), stretch factor, 99, 155-160 (Section
fission, 237-238 5.8)
bomb, 249 gamma rays, 237; see also photon
Fizeau experiment, 120 (Exercise L-16) General Conference on Weights and Measures, 12, 58
flash, reference, 38 general relativity, 275-298 (Chapter 9)
flickering bulb paradox, 186- 187 (Exercise 6-7) needed for Twin Paradox?, 132 (Box 4-1)
floating to Moon, 2 5-26 (Section 2.1) when required, 34, 35, 133, 276, 281
force of gravity, eliminate, 26-29 (Section 2.2) geometry
four times light speed, 89 (Exercise 3-15) Euclidean, 8, 11, 126, 143, 151, 172, 177, 192
four-vector, momenergy as, 191- 195 (Section 7.2) (Box 7-1), 198, 279
frame curved space, 280-281 (Section 9.4),
center of momentum, 246-251 curved spacetime 284-287 (Section 9.6)
Earth, 46 (Exercise 2-5) Lorentz, 8, 11, 126, 143, 151, 172, 177, 192
free-float, see free-floar frame (Box 7-1), 198, 284
inertial, see free-float frame gigaflop, 22 (Exercise 1-8)
laboratory, 5, 41 gravitation
local, see free-float frame effect of on clocks, 118 (Exercise L-13)
306 INDEX
as curvature of spacetime, 284-287 (Section 9.6) Kennedy-Thorndike, 86-88 (Exercise 3-13)

tutorial in Newtonian, 258 (Exercise 8-6) Michelson-Morley, 8 4-86 (Exercise 3-12)
gravitational attraction of system containing photons, 257 verifying free-float frame using, 46 (Exercise 2-5)
gravitational collapse, 288, 292-295 (Section 9.8) interstellar travel, 274 (Exercise 8-41)
gravitational radiation, 288-292 (Section 9.7) interval, 6
gravitational red shift, 258 (Exercise 8-6) and event, 9-11 (Section 1.3)
test of, 272 (Exercises 8-37, 8-38) invariance of, see invariance of interval
graviton, 153, 176 as lightlike relation between events, 175-177
gravity as spacelike relation between events, 11, 173-174
as curved spacetime, 284-287 (Section 9.6) as timelike relation between events, 11, 172-173
in brief, 275 (Section 9.1) invariance of distance, 4,17
eliminate, 28-29 (Section 2.2) invariance of interval, 6 -7 , 17, 18
radiation, 288-292 (Section 9.7) for all free-float frames, 71-73 (Section 3.8)
relative acceleration as witness to, 30-36 (Sections preserves cause and effect, 180-183
2.3, 2.4), 280-287 (Sections 9.4, 9.5, 9.6) proved, 6 7-70 (Section 3.7)
waves, 288-292 (Section 9.7) and spacetime hyperbola, 143 (Section 5.3), 173, 174
Great Betrayal, 108-109 (Box L-1) and spacetime map, 143 (Section 5.3)
Great Pyramid, 209 used in derivation of the Lorentz transformation, 102
grid, paradox of skateboard and, 116-117 (Exercise L-12) invariance of mass, 197, 246
invariance of momenergy, 194, 198, 210
h, Planck’s constant, 265, 268-269 (Exercise 8-31)
invariance of speed of light, 60; 86-88 (Exercise 3-13)
handle showing invariant magnitude of momenergy vec
invariance of transverse dimension, 6 5-67 (Section 3.6)
tor, 198
invariant, defined, 208-209 (Box 7-3)
headlight effect, 115 (Exercise L-9)
invariant hyperbola, 143 (Section 5.3), 173, 174
heat
inverse Compton scattering, 269-270 (Exercise 8-32)
as system property, 224
inverse Lorentz transformation, 102-103 (Section L.6)
weighing, 223
helium in Sun, 242-245 (Sample Problem 8-5) Japan, 27, 96-97, l6 l
Himalaya Mountains, 48-49 Japan Microgravity Center (JAMIC), 27 (Figure 2-3)
Hirohito, Emperor of Japan, 137 Julius Caesar, 106- 107 (Sample Problem L-2)
hole, black, 289 (Box 9-2), 292-295 (Section 9.8)
as source of neutrinos, 80 (Exercise 3-8) K"''-meson, 72 (Sample Problem 3-2)
Horwitz, Paul, 186 (Exercise 6-6) Kamisunagawa, 27
Hubble, Edwin, 264 Kennedy-Thorndike experiment, 86-88 (Exercise 3-13)
Hubble constant, 264 Kepler, Johannes, 32
Hubble time, 264 kinetic energy, 201, 203, 206
Hull, Penny, 19, 264, 272 kinked worldline, 152- 155 (Section 5.7)
Hulse, Russell A., 291 Klingons, 108-109 (Box L-1)
human cannonball, 45 (Exercise 2-1) Krotkov, Robert V., 36
hydrogen bomb, 248-249
laboratory frame, 5, 41
hydrogen burning in Sun, 242 - 245 (Sample Problem 8-5)
lattice clocks, synchronizing, 37-38, 45-46, (Exercises
hydrogen molecule ion, 233
2-3, 2-4)
hyperbola
latticework of clocks, 37-39 (Section 2.6)
invariant, 143 (Section 5.3), 173-174
momenergy, 198 Law of Addition of Velocities, 8 2-84 (Exercise 3-11),
103-110 (Section L.7)
identically accelerated twins paradox, 117-118 (Exercise laws, conservation, see energy, momentum; momenergy
L-13) Laws, Kenneth L., 77
index of refraction and speed of light, 185 (Exercise 6-4) Leaning Tower of Pisa, 36
inelastic collision, 222-223 length
inertia, 31, 189 mass in units of, 258 (Exercise 8-6)
inettial frame, see free-float frame time in units of, 11-13 (Section 1.4)
integrity of photon, 259 (Exercise 8-11) length along a path, 147- 148 (Section 5.5)
interaction energy, available, 261 (Exercise 8-17) length contraction, see Lorentz contraction
interferometer less is more, 154-155 (Sample Problem 5-1), 163-164
Fizeau, 120 (Exercise L-16) (Exercise 5-1)
INDEX 307
light Lorentz-FitzGerald contraction hypothesis, 88

deflection of by Sun, 50-51 (Exercise 2-13) Lorentz geometry, 8, 11, 126, 143, 151, 172, 177, 192
frequency of and energy of a photon, 268-269 (Box 7-1), 198, 284
(Exercise 8-31) Lorentz interval, 6; see also interval; invariance of interval
gravitarional red shift of, 258-259 (Exercises 8-6, Lorentz transformation, 95-111 (Special Topic)
8-7) equations, 102
pressure of, 254 (Exercise 8-3), 255 form of, 100 (Section L.4)
rocket propelled by, 274 (Exercise 8-41) inverse equations, 102-103 (Section L.6)
speed of, see lighr speed for momenergy componenrs, 215 (Exercise 7-5)
See also photon usefulness of, 95 (Secrion L.l) •
light, faster rhan?, 74-75 (Box 3-3), 9 6-99 (Section
L.2), 108-109 (Box L-1), 122-123 (Section 4.3) manhole, paradox of rising, 116 (Exercise L-11)
four times the speed of light?, 8 9 -9 0 (Exercise 3-15) map, spacetime, see spacetime map
superluminal expansion of quasar 3C273?, 90-92 mapmaking
(Exercise 3-16) in space, 10, 21-22 (Exercise 1-6)
things that move faster than light, 8 8-89 (Exercise in spacetime, 164-166 (Exercise 5-3)
3-14) mass
light bulb abuse of the concept of, 244-251 (Section 8.8)
flickering, 186-187 (Exercise 6-7) change in nuclear, 237-238
speeding, 264 (Exercise 8-21) conversion of to enetgy, 237-244 (Section 8.7),
light cone as partition in spacetime, 177- 183 (Section 6.3) 254 (Exercise 8-1)
light-flash clock, 12 created by material parricle, 234-236 (Section 8.6)
lightlike relation between events, 172-177 (Section 6.2) created by photon, 233-234 (Section 8.5)
light propagation, ether theory of, 84, 88 and energy, 201, 203, 206, 250-251, 254-258
light speed (Exercise 8-5)
as conversion factor, 6, 12, 16, 58-59 (Box 3-2), energy in unit of, 190, 203
200, 203, 250 energy without (photon), 228-233 (Section 8.4)
index of refraction and, 185 (Exercise 6-4) invariance of, 197, 246
invarianr magnitude of, 60 (Kennedy-Thorndike local moving orders for, 277-280 (Section 9-3)
experiment), 8 6-88 (Exercise 3-13) loss by Sun of, 242-245 (Sample Problem 8-5)
isotropic (Michelson-Morley experiment), 84-86 as magnitude of momenergy 4-vector, 195, 197
(Exercise 3-12) momentum in unit of, 190, 200
as limit on causality, 171 (Section 6.1), 180-183 momentum without?, 273-274 (Exercise 8-40)
as limit on observation, 39-40 photon used to create, 233-234 (Section 8.5)
See also light, faster than? proof, 277, 279
light-second, 11-13 (Section 1.4) “relativistic,” 250-251
light-year, 12 “rest,” 251
limb of Sun, Doppler shift at, 264 (Exercise 8-22) as unit of length, 258 (Exercise 8-6)
limits of Newtonian mechanics, 34, 113-114 (Exercise use and abuse of the concept of, 244-251 (Section
L-4), 217 (Exercise 7-11) 8 .8)
line, world, see worldline mass of photon, 230
line broadening, Doppler, 264 (Exercise 8-25) mass of system of particles, 214 (Exercise 7-2), 224-228
linear acceleraror, Stanford, 215 (Exercise 7-6) (Section 8.3), 247
local inertial frame, see free-float frame Maximal Aging, Principle of, 150
local moving orders for mass, 277-280 (Section 9.3) maximum speed of walking, 186 (Exercise 6-6)
local time, see proper rime; interval mechanics
locating events with latticework, 37-39 (Section 2.6) Newtonian, 113-114 (Exercise L-4), 192 (Box
Lorentz, Hendrik, 5 7-1), 217 (Exercise 7-11)
Lorentz contraction, 63-65 (Section 3.5), 126- 127 relarivistic, 192 (Box 7-1)
(Section 4.7) megaflop, 22 (Exercise 1-8)
for cosmic rays, 216 (Exercise 7-7) meson
described by strerch factor, 157 decay of pi-naught, 267 (Exercise 8-28)
how ir occurs, 119-120 (Exercise L-14) firing, 110 (Sample Problem L-3)
or rotarion, 9 2-93 (Exercise 3-17) time stretching with, 23-24 (Exercise 1-11), 24
Lorentz frame, see free-float frame (Exercise 1-12), 72-73 (Sample Problem 3-2)
308 INDEX
meter nanosecond, 5
defined, 58-59 (Box 3-2) Neptune, images from, 20 (Exercise 1-2)
of rime, 11-13 (Section 1.4) neutral or unreachable region, 182
as unit of mass, 258 (Exercise 8-6) neutrino
meter stick, tilted, 115-116 (Exercise L-10) described, 235 (Box 8-1)
Michelson - Morley experiment, 8 4 -8 6 (Exercise 3-12) detection of, 80 (Exercise 3-8)
microgravity, 27 (Figure 2-3), 277 (Figure 9-2) neutron, described, 235 (Box 8-1)
Minkowski, Hermann, 15 neutron star, 288-289 (Box 9-2)
mile and gravity waves, 290-291
defined, 58-59 (Box 3-2) Newton, Isaac, 275-280
as sacred unit, 1 -4 absolute space and time, 160, 284
minus sign, 6 -8 , 26, 190, 197 Einstein’s admiration for, 284, 295
minute, unit of distance and time, 11-13 (Section 1.4) First Law of Motion, 31
momenergy and Galileo and Einstein, 275-276 (Section 9.2)
as 4-vector, 191, 192 (Box 7-1) picture and quotes, 278
analogy of to tree, 210 Newtonian mechanics, 192 (Box 7-1)
arrow, 191-195 (Section 7.2) First Law of Motion, 31
components of, 195-199 (Section 7.3), 204 gravitational theory, tutorial, 258 (Exercise 8-6)
(Sample Problem 7-3) limits of, 34, 113-114 (Exercise L-4), 217 (Exer
conservation of, 189-190 (Seaion 7.1), 207-210 cise 7-11)
(Section 7.6), 247 Nighttime surveyor, 1 -4 (Section 1.1), 16-17 (Box 1-1)
defined, 191-195 (Section 7.2) nuclear excitation, 259 (Exercise 8-8)
energy as “time” part of, 201-206 (Section 7.5)
observer, 39-40 (Section 2.7)
handle showing invariant magnitude, 198
oozing!, 12
invariance of, 194, 198, 210
oscillator, relativistic, 135-136 (Exercise 4-3)
magnitude of is mass, 195, 197
oscilloscope writing speed, 89 (Exercise 3-14)
momentum as “space” part of, 199-200 (Section
7.4) pair production by photon(s), 233-234 (Section 8.5),
quantities related to, 213 (Table 7-1) 259 (Exercises 8-11, 8-12)
tree, analogy of, 210 Parable of the Surveyors, 1 -4 (Section 1.1), 16-17
transformation of components of, 215 (Exercise 7-5) (Box 1-1)
units of, 194, 195, 200, 203 Parable of the Two Travelers, 281-283 (Seaion 9.5)
momentum, 196, 213 (Table 7.1) paradoxes
components of, 196 contracting train, 187-188 (Exercise 6-8)
conserved in a collision, 189- 190 (Section 7.1), detonator, 185-186 (Exercise 6-5)
207, 222-223 (Section 8.2), 239 (Box 8.2) Einstein’s train, 62-63
derived from conservation law, 217-219 (Exercise flickering bulb, 186-187 (Exercise 6-7)
7-12) four times light speed, 89 (Exercise 3-15)
of light, 230 identically accelerated twins, 117-118 (Exercise
Newtonian expression for, 190, 200 L-13)
as “space” part of momenergy, 199-200 (Section passenger and dog, 25-26
7.4) pole and barn, 166 (Exercise 5-4)
transformation of, 215 (Exercise 7-5) rising manhole, 116 (Exercise L-11)
in unit of mass, 190, 200 mnner on the train, 168 (Exercise 5-7)
without mass?, 273-274 (Exercise 8-40) scissors, 88 (Exercise 3-14)
momentum-energy 4-vector, see momenergy skateboard and grid, 116-117 (Exercise L-12)
Moon, 2 5 -2 6 (Section 2.1), 32-33 (Box 2-1) space war, 79-80 (Exercise 3-7)
Moral Principle, Wheeler’s First, 20 tilted meter stick, 115-116 (Exercise L-10)
Mossbauer effect, 270 See also Twin Paradox
Minkowski, Hermann, 15 particle, test, 36 (Section 2.5), 4 7-48 (Exercise 2-10)
more is less, 154-155 (Sample Problem 5-1), 163-164 particles
(Exercise 5-1) backyard zoo of, 235 (Box 8-1)
moving orders for mass, local, 277-280 (Seaion 9.3) creation of, 234-236 (Section 8.6), 261-262 (Ex
muons, time stretching with, 23 (Exercise 1-11) ercises 8-16, 8-17)
INDEX 309
creation of by photons, 233-234 (Section 8.5), positron-electron annihilation, 237-238, 242-243

259-260 (Exercises 8-11 and 8-12) (Sample Problem 8-4), 260 (Exercises 8-14, 8-15)
encounter, 239 (Box 8-2) positron-electron pair production, 233-234 (Section
measuring speed of, 40-41 (Section 2.8) 8.5), 259 (Exetcises 8-11, 8-12)
system of, 214 (Exercise 7-2), 221 (Section 8.1), positronium, decay of, 260 (Exercise 8-13)
224-228 (Section 8.3), 244-251 (Section 8-8) practical synchronization of clocks, 4 5 -4 6 (Exercises 2-3,
timelike wotldline of, 172 2-4)
used to create mass, 234-236 (Section 8.6) pressure of light, 254 (Exercise 8-3), 255
virmal, 56-57 principle of invariance of distance, 4, 17
wotldline of, 143- 147 (Section 5.4) Principle of Maximal Aging, 150
partition in spacetime, light cone as, 177- 183 (Section 6.3) Principle of Relativity, 53-60 (Sections 3.1, 3.2, 3.3)
passenger and dog paradox, 25-26 examples of, 61-62 (Sample Problem 3-1), 78
passive past, 182 (Exercise 3-4)
path, length along, 147—148 (Section 5.5) Galilean, 5 3 -5 5
Peace Treaty of Shalimar, 108- 109 (Box L-1) used in proof of invariance of interval, 73
Philoponus, John, of Alexandria, 36 proof mass (conscience), 277, 279
photon, 228-233 (Section 8.4), 246 proper clock, 10
from annihilation, 237-238 (Section 8.7) proper distance, 174, 184 (Exercise 6-3)
autobiography of, 184-185 (Exercise 6-4) proper time, 10, 184 (Exercise 6-3)
braking, 259 (Exercise 8-9) along a worldline, 148- 152 (Section 5.6)
Compton scattering of, 229, 231, 267-270 (Exer tau as symbol of, 155
cises 8-29, 8-30, 8-32) proton, described, 235 (Box 8-1)
creation of particle-antiparticle pair using, 233- proton - antiproton pair, creation of, 236
234 (Section 8.5) protons, fast, 214-215 (Exercise 7-4)
energy of, 228-233 (Section 8.4), 268-269 (Exer pulsar, 289
cise 8-31) puppy, 224
energy measurement of, 254 (Exercise 8-4) puzzler, Einstein, 78 (Exercise 3-2)
energy shift of due to recoil of emitter, 270 (Exercise Pyramid, Great, 209
8-33) Pythagorean theorem, 2, 7
gravitational red shift of, 258-259 (Exercises 8-6
quantum electrodynamics, 185 (Exercise 6-4)
and 8-7)
quasar, 90-92 (Exercise 3-16), 114 (Exercise L-5),
integrity of, 259 (Exercise 8-11)
294-295
mass of, 228-231 (Section 8.4)
momentum of, 230 radar speed trap, 166- 168 (Exercise 5-5)
pair production by, 233-234 (Section 8.5), 259- radiation, Cerenkov, 80-81 (Exercise 3-8)
260 (Exercises 8-11, 8-12) radiation, gravitational, 288-292 (Section 9.7)
resonant scattering of, 271-272 (Exercises 8-35, radius of a black hole, 292
8-36) railway coach
rocket and interstellar travel, 274 (Exercise 8-41) rising, 47 (Exercise 2-9)
used to create mass, 233-234 (Section 8.5) and tidal accelerations, 30-34 (Section 2.3), 281
physicist and the traffic light, 263-264 (Exercise 8-20) ray, gamma, see photon
pi-naught meson, decay of, 267 (Exercise 8-28) ray, X-, see photon
pipes, speeding (thought experiment), 66 rays, cosmic, 160, 215-216 (Exercise 7-7)
pi-plus mesons, time stretching with, 24 (Exercise 1-12) recoilless processes, 270-271 (Exercise 8-34)
Pisa, Leaning Tower of, 36 recoil of emitter, energy shift due to, 270 (Exercise 8-33)
place, fundamental to surveying, 9, 16 red shift, gravitational, 258 (Exercise 8-6), 272 (Exercises
plane of agreement of clocks, 120 (Exercise L-15) 8-37, 8-38)
Planck, Max, 229 reference clock, 37
Planck’s constant, 265, 268-269 (Exercise 8-31) reference event, 38
plumb bob, deflection of by Himalaya Mountains, reference flash, 38
4 8-49 reference frame, 5; see also free-float frame
Poincare, Henri, 5 -6 refraction, index of, and speed of light, 185 (Exercise 6-4)
pole and barn paradox, 166 (Exercise 5-4) regions of spacetime, 34-36 (Section 2.4), 171-183
polyelectron, 233 (Chapter 6)
positron, 233-235 relations between events, 172-177 (Section 6.2)
310 INDEX
relative acceleration as witness to gravity, 30-36 (Sec Sheldon, Eric, 19

tions 2.3, 2.4), 280-287 (Sections 9.4, 9.5, 9.6) shift, see Doppler shift; red shift
relative synchronization of clocks, 130 Shurdiff, William A., 19, 77, 198, 213
relativistic chemistry, 254 (Exercise 8-2) simultaneity,
“relativistic” mass, 250-251 relativity of, 62-63 (Section 3-4), 64, 128-131
relativistic mechanics, 192 (Box 7-1) (Section 4.9)
relativistic momentum, 217-219 (Exercise 7-12) and transverse plane, 66-67
relativistic oscillator, 135-136 (Exercise 4-3) See also paradoxes
relativity Sirius, density of companion of, 258-259 (Exercise 8-7)
general, 34, 35, 132 (Box 4-1), 133, 276, 281 skateboard and grid paradox, 116-117 (Exercise L-12)
principle of, 53-62 (Sections 3.1, 3.2, 3.3), 78 Smith, Richard C., 19
special, 5, 18, 73, 78 (Exercise 3-1), 79 (Exercise Sommerfeld, Arthur, 53
3-6), 131-134 (Section 4.10), 270-273 (Exer solar constant, 242, 254 (Exercise 8-3)
cises 8-33 to 8-39) solar wind, 245
relativity of simultaneity, 62-63 (Section 3.4), 128-131 space
(Section 4.9) “absolute” (Newton), 284
and contraction of length, 64 as different from time, 18
See also paradoxes is ours!, 123-124 (Section 4.4)
resonant scattering, 271 (Exercise 8-35) spacelike relation between events, 11, 172-177 (Section
measurement of Doppler shift by, 271-272 (Exer 6 .2)
cise 8-36) space travel, practical, 135 (Exercise 4-1)
rest energy, 201, 203, 250 space war, 79-81 (Exercise 3-7)
“rest mass,” 251 spacetime
Riemann, G. F. B., 295 as absolute elsewhere, 181
“rigid body” not an invariant concept, 116-117 (Exer active future of, 182
cise L-12), 119-120 (Exercise L-14) emptiness of, 56-57 (Box 3-1)
rising manhole paradox, 116 (Exercise L-11) exploded view of regions of, 182 (Figure 6-5)
rising railway coach, 47 (Exercise 2-9) “Et m . . . ?”, 106-107 (Sample Problem L-2)
rocket frame, 4 1-43 (Section 2.9) light cone as partition of, 177- 183 (Seaion 6.3)
rocket, photon, and interstellar travel, 274 (Exercise 8-41) Lorentz geometry of, 8, 192 (Box 7-1)
rods, latticework of, 37-39 (Section 2.6) mapmaking in, 164-166 (Exercise 5-3)
Roll, Peter G., 36 neutral region of, 182
rotation or contraction?, 9 2-93 (Exercise 3-17) overview of, 1-19 (Chapter 1)
Rumford, Count (Benjamin Thompson), 223 passive past of, 182
Ruml, Frances Towne, 29 regions of, 34-36 (Section 2.4), 171- 183 (Chap
runner on the train paradox, 168 (Exercise 5-7) ter 6)
surveying, 5 -8 (Section 1.2)
sacred unit touring without tefetence ftame, 1 6 0 -1 6 2 (Section
mile, 1 -4 5.9)
second, 5 -7 trekking through, 137-163 (Chapter 5)
Satellite (dog), 26 units of, 20-21 (Exercises 1-2 and 1-3)
satellite unity of, 7, 15-18 (Section 1.5)
conscience-guided, 277-279 unreachable region of, 182
pressure of light on, 254 (Exercise 8-3) spacetime curvamre, 280-287 (Sections 9.4, 9.5, 9.6)
scattering contractile, 286-287 (Box 9-1)
Compton, 229, 231, 267-270 (Exercises 8-29, equation (Einstein), 286
8-30, 8-32) gravitation as, 284-287 (Section 9.6)
resonant, 271-272 (Exercises 8-35, 8-36) noncontractile, 286-287 (Box 9-1)
scissors paradox, 88 (Exercise 3-14) spacetime diagram, see spacetime map
Schmidt, Maarten, 294 spacetime displacement as 4-vector, 191-194
second spacetime geometry, see spacetime; spacetime curvature
defined, 58-59 (Box 3-2) spacetime interval, see interval; invariance of interval
as sacred unit, 5 -7 spacetime map, 22 (Exercise 1-7), 137-139 (Section 5.1)
as unit of distance and time, 11-13 (Section 1.4) constructing, 164-166 (Exercise 5-3)
Shalimar, Peace Treaty of 108-109 (Box L-1) special relativity, 5, 18
INDEX 311
down with, 79 (Exercise 3-6) not isolated, 228

four ideas of, 7 3 system property, heat as, 224
and swimming, 78 (Exercise 3-1)
tests of, 131-134 (Section 4.10), 270-273 (Exer
cises 8-33 through 8-39) tangent vector to worldline, 194-195
spectral lines, Doppler broadening of, 264 (Exercise 8-25) tau (Greek T), symbol for proper time, 155
speed, measuring, 40-41 (Section 2.8) Taylot
speeding light bulb, 264 (Exercise 8-21) Bradley James, 179
speeding pipes thought experiment, 66 Katherine Rose, 311
speeding train thought experiment, 65-66 Joseph H., 291
speed of light, see light speed Meredith Christine, 171
speed of walking, maximum, 186 (Exercise 6-6) Samantha Marie, 23 (Exercise 1-10)
speed trap, radar, 166-168 (Exercise 5-5) teraflop, 22 (Exercise 1-8)
speeds, comparing, 20 (Exercise 1-1) test particle, 36 (Section 2.5), 4 7 -4 8 (Exercise 2-10)
Stanford linear accelerator, 215 (Exercise 7-6) tests of relativity, 131-134 (Seaion 4.10), 270-273
starlight (Exercises 8-33 through 8-39)
aberration of, 81 (Exercise 3-9) Thompson, Benjamin (Count Rumford), 223
deflection of by Sun, 50-51 (Exercise 2-13) thought experiments
Starship Enterprise, 106-107 speeding pipes, 66
stellar aberration, 81 (Exercise 3-9) speeding train, 6 5-66
stellar objects, compact, 288-289 (Box 9-2) three-vectors. Euclidean, 192 (Box 7-1)
storm, communicarions, 48 (Exercise 2-11) threshold energy, 236, 259 (Exercise 8-12), 261 (Exercise
stretch factor, 99, 155-160 (Section 5.8) 8-16)
and Lorentz contraction, 157 tidal effects of large frame, 3 0 - 3 4 (Section 2.3), 2 8 0 -
as measure of speed, 157 281 (Section 9.4)
stripped down free-float frame, 121-122 (Section 4.2) tide-driving power of Moon and Sun, 32-33 (Box 2-1)
Sun tides, 32-33 (Box 2-1), 281, 286-287 (Box 9-1)
conversion of mass to energy in, 242-245 (Sample tilted meter stick paradox 115-116 (Exercise L-10)
Problem 8-5) time
deflection of starlight by, 50-51 (Exercise 2-13) “absolute” (Newton), I 6 O
Doppler shift at limb of, 264 (Exercise 8-22) as different from space, 18
explosion of, 171 of an event, 38, 137-139 (Section 5.1)
gravitational red shift of light from, 258 (Exercise 8-6) Hubble, 264
helium in, 242-245 (Sample Problem 8-5) and length, 11-13 (Section 1.4)
mass of in units of meters, 258 and Lorentz transformation, 102
tide-driving power of, 32-33 (Box 2-1) meter of, 12
sunspot, 179-180 (Sample Problem 6-3) proper, 10, 148-152 (Section 5.6), 155, 184
Super (superluminal bomb), 108-109 (Box L-1) wrisrwatch 10, 148-152 (Section 5.6)
super cosmic rays, 215-216 (Exercise 7-7) time delay in communication, 39-40
superluminal expansion of quasar 3C273?, 90-92 (Exer timelike relation between events, 11, 172-177 (Section
cise 3-16) 6 . 2)
supernova, 177, 289 timelike worldline of a particle, 172
super-rocket frame, 69, 71, 140-142 time stretching
super-speed Super, 112 (Exercise L-1) experimental evidence of, 131-134 (Section 4.10),
surveying spacetime, 5 -8 (Section 1.2) 272-273 (Exercise 8-39)
Surveyors, Parable of, 1- 4 (Section 1.1), 16-17 (Box 1-1) with K+ mesons, 72-73 (Sample Problem 3-2)
swimming and relativity, 78 (Exercise 3-1) with mu-mesons, 2 3-24 (Exercise 1-11)
symmetric elastic collision, 240 - 241 (Sample Problem 8-3) with pi-plus mesons, 24 (Exercise (1-12)
synchronization of clocks, relative, 130 and spacetime interval, 21 (Exercise 1-4)
synchronizing lattice clocks, 37-38, 4 5 -4 6 (Exercises See also Twin Paradox
2-3, 2-4) time traveler, 127- 128 (Section 4.8)
system of particles, 221 (Section 8.1), 244-251 (Section touring spacetime without a reference frame, 160-162
8 - 8) (Section 5.9)
mass of, 214 (Exercise 7-2), 224-228 (Section traffic light, physicist and, 263-264 (Exercise 8-20)
8.3), 247-248 train, mass effeas of in collision, 214 (Exercise 7-3)
312 INDEX
train paradoxes, 62-63, 168 (Exercise 5-7), 187-188 use and abuse of the concept of mass, 244-251 (Section
(Exercise 6-8) 8 . 8)
train rhought experiment, 65-66
Van Dam, Hendrik, 79 (Exercise 3-6)
transformation
vector, defined, 192 (Box 7-1)
Galilean, 113 (Exercise L-3)
velocities, addition of, 82-84 (Exercise 3-11), 103-110
Lorentz, 95-111 (Special Topic)
(Section L.7)
transformation of angles, 114-115 (Exercise L-6)
velocity
transformation of velocity direction, 115 (Exercises L-7,
measuring, 40-41 (Section 2.8)
L-8)
velocity of recession from Doppler shift, 114 (Exercise
transforming worldlines, 164 (Exercise 5-2)
L-5), 264 (Exercise 8-23)
transverse dimension, invariance of, 6 5-67 (Section 3.6)
velocity of recession from period of light, 82 (Exercise 3-10)
travel, inrerstellar, 274 (Exercise 8-41)
velocity direction, transformation of, 115 (Exercises L-7,
traveler, time, 127-128 (Section 4,8)
L-8)
Travelers, Parable of the Two, 281-283 (Section 9.5)
Verne, Jules, 25-26
traveling clock, synchronization using, 4 5-46 (Exercise
virtual particles, 56-57 (Box 3-1)
2-4)
Treaty of Shalimar, 108-109 (Box L-1) visual appearance of relativistic objects, 64, 92-93 (Exer
cise 3-17)
tree analogy to momenergy, 210
von Jagow, Peter, 44
Twin Paradox, 125-126 (Section 4.6)
atomic clocks (“airliner”) test of, 131 walking, maximum speed of, 186 (Exercise 6-6)
circling airplane test of, 133 war
general relativity needed for?, 132 (Box 4-1) American Civil, 25
one-way, 135 (Exercise 4-2) space, 79-81 (Exercise 3-7)
oscillating iron atom test of, 134, 272-273 (Exer waves, gravity, 288-292 (Section 9.7)
cise 8-39) weighing heat, 223
put to rest, 169-170 (Exercise 5-8) Weights and Measures, General Conference on, 12, 58
radioactive particle test of, 133 Weisskopf, V. W., 296
using Doppler shift, 264 (Exercise 8-24) Weyl, Herman, quote, 189
twins, paradox of identically accelerated, 117-118 (Exer Wheeler's First Moral Principle, 20
cise L-13) white dwarf star, 258-259 (Exercise 8-7), 288 (Box 9-2)
Two Travelers, Parable of, 281-283 (Section 9.5) wind, solar 245
worldline, 143-147 (Section 5.4)
unit, same for space and time, 11-13 (Section 1.4)
kinked, 152-155 (Section 5.7)
units, 213 (Table 7-1)
timelike, of a particle, 172
units of energy, 203
ttansforming, 164 (Exercise 5-2)
units of momenergy, 194
unit tangent vector to, 194-195
units of momentum, 200
wristwatch (proper) time along, 148-152 (Section
units of spacetime, 11-13 (Section 1.4), 20-21 (Exer
5.6)
cises 1-2, 1-3)
wristwatch time, 10-11
unit tangent vector to worldline, 194-195
along a worldline, 148-152 (Section 5.6)
unity of spacetime, 15-18 (Section 1.5)
universe X-ray, see photon
expanding, 82 (Exercise 3-10), 264 (Exercise 8-23),
y-velocity, transformation of, 115 (Exercise L-7)
297 (Table 9-2)
year as unit of distance and time, 11-13 (Section 1.4)
models of, 296-297 (Seaion 9.9)
unreachable region, 182 zero mass for photon, 230
uranium bomb, 249 zero-total-momentum frame, 246-251
uranium fission, 237 zoo of particles, backyard, 235 (Box 8-1)
THE AUTHORS
J ohn A rchibald W heeler (Ph.D., Johns Hopkins University) is one of the

world s foremost relativists. He is Joseph Henry Professor Emeritus at Princeton
University and, until his retirement in 1986, Blumberg Professor of Physics and
Director, Center for Theoretical Physics, at the University of Texas at Austin. A
past president of the American Physical Society, he is a recipient of the Enrico
Fermi Award (1968), the National Medal of Science (1971), and the Niels Bohr
International Gold Medal (1982).
Since the appearance of the First Edition of Spacetime Physics, John Wheeler has
published a graduate text in general relativity, GRAVITATION, with Kip S.
Thorne and Charles W Misner (W H. Freeman, 1970), and a popular treatment
of gravity, A Journey into Gravity and Spacetime (Scientific American Library, 1990;
distributed by W H. Freeman).
E dw in F. T aylor (Ph.D., Harvard University) taught physics for 26 years at the
Massachusetts Institute of Technology. He is currently Research Professor in the
Department of Physics at Boston University. He is the author of a textbook on in
troductory mechanics and An Introduction to Quantum Physics with A. P French
(W W Norton, 1978). He has served as Editor of the American Journal of Physics.
With MIT undergraduates, Edwin Taylor produced interactive computer pro
grams to help students visualize and solve problems in special relativity. These
won the 1988 EDUCOM/NCRIPTAL Higher Education Software Awards for Best
Physics Software and Best Tool Software.
THE BOOK
Collaboration on the First Edition of Spacetime Physics began in the mid-1960s
when Edwin Taylor took a junior faculty sabbatical at Princeton University where
John Wheeler was a professor. The resulting text emphasized the unity of space-
time and those quantities (such as proper time, proper distance, mass) that are in
variant, the same for all observers, rather than those quantities (such as space and
time separations) that are relative, different for different observers. The text has
become a standard for modern physics and relativity courses, as well as introduc
tory physics.
The Second Edition of Spacetime Physics embodies what the authors have learned
during an additional quarter century of teaching and research. They have updated
the text to reflect the immense strides in physics during the same perkxl and
modernized and increased the number of exercises, for which the First Edition was
famous. Enrichment boxes provide expanded coverage of intriguing topics. Sample
pioblems encourage students to exercise their newfound power. An enlarged final
chapter on general relativity includes new material on gravity waves, black holes,
and cosmology.
The Second Edition of Spacetime Physics provides a new generation of students with
a deep and simple overview of the principles of relativity.
ISBN-13: 978-0-7167-2327-1
ISBN-10: 0-7167-2327-1
W. H. Freeman and Company

4l Madison Avenue New York, NY 10010
20 Beaumont Street Oxford 0X1 2NQ

Spacetime Physics - Taylor, Wheeler

Uploaded by

Copyright:

Available Formats

Spacetime Physics - Taylor, Wheeler

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Spacetime Physics - Taylor, Wheeler

Uploaded by

Copyright:

Available Formats

What is the book about?

What is the book about?

What topics are covered in the first chapter?

What topics are covered in the first chapter?

V 'Mr

John Archibald Wheeler

W . H. Freeman and Com pany

© 1992 by Edwin F. Taylor and John Archibald Wheeler

No part o f this book may be reproduced by any mechanical, photographic, or electronic

Printed in the United States o f America

1.1 Parable of the Surveyors 1 1 .2 Surveying Spacetim e 5 1 .3 Events

Chapter 2 FLOATING FREE 25

Chapter 3 SAME LAWS FOR ALL 53

3.1 The Principle of Relativity 53 3 .2 W hat Is NOT the Sam e in Different

4.1 Invitation to C anopus 121 4 .2 Stripped-Down Free-Floot Frame 121

Chapter 5 TREKKING THROUGH SPACETIME 137

5.1 Time? No. Spacetim e M ap ? Y es. 137 5 .2 Sam e Events; Different

Chapter 6 REGIONS OF SPACETIME 171

6.1 Light Speed : Limit on C ausality 171 6 .2 Relation Between Events:

Chapter 7 MOMENERGY 189

C h apters COLLIDE. CREATE. ANNIHILATE. 221

8.1 The System 221 8 .2 Three M odest Experiments 2 2 2 8 .3 M ass of a

ANSWERS TO ODD-NUMBERED EXERCISES 299

•' ■•vV’j’- S ' ; '

■' . > '

1.1 PARABLE OF THE SURVEYORS

NIGHTTIME: NORTH-STAR NORTH

------------------------------ iC^^ ^ B L E 1 - 1 ^ -------------------------------

TWO DIFFERENT SETS OF RECORDS; SAME PLOT OF LAND

“ INVARIANT DISTANCE” FROM CENTER OF TOWN SQUARE TO STAKE C

Daytime measurements Nighttime measurements

Northward separation Northward separation

DAYTIME: MAGNETIC NORTH

FIGURE 1-2. The distance between stake A

He said he had discovered the p rin cip le o f invariance o f distance; he reckoned

1.2 SURVEYING SPACETIME

-------------------- C [ ^ B L E l - 3 ^ > --------------------

SPACE AND TIME LOCATIONS OF THE SAME

c = 299,792,458 meters/second = 2.99792458 X 10* meters/second

“ INVARIANT SPACETIME INTERVAL” FROM REFERENCE SPARK TO SPARK A

Laboratory measurements Rocket measurements

Time lapse Time lapse

The in v arian ce o f th e spacetim e in terv al — its independence of the state of

SAMPLE PROBLEM l-i;

16.6782048 nanoseconds X 0.299792458 meters/nanosecond

(interval)^ = (laboratory time)^ — (laboratory distance)^

(interval)^ = (rocket time)^ — (rocket distance)^

(4 /5 ) X 2.99792458 X 10® meters/second

1.3 EVENTS AND INTERVALS ALONE!

(interval)^ — (time separation)^ — (space separation)^

That returns us to a universal, frame-independent description of the physical world.

{interval^ = (time separation^ — (space separation)^

How can the square of the spacetime interval be negative?

(interval)^ = (space separation)^ — (time separation)^ [when spacelike]

This way, the squared interval is never negative.

1.4 SAME UNIT FOR SPACE AND TIME:

SOME LIGHT-TRAVEL TIMES

Telephone call one way:

W e are not accustomed to measuring time in meters. So as a reminder to ourselves

(proton time)^ — (proton distance)^ = (interval)^ = (lab time)^ — (lab distance)^

rest relative to them both — that has been synchro

a Verify the statement in Section 2.5 that a can

IJJJJJJJWJJJ clusion th a t n eith er p ip e can he m ea