Clausius Book
Clausius Book
Clausius Book
.:
//cisyf<^//
The
tine
original of
tliis
book
is in
restrictions in
text.
http://www.archive.org/details/cu31924101120883
THE
MECHANICAL THEORY
OP
HEAT.
THE
MECHANICAL THEOKY
OP
HEAT.
BY
R CLAUSIUS.
TRANSLATED BY
WALTER
E.
BROWNE,
COLLEOS,
M.A.,
OAMBKIDGB.
Hontjon
MACMILLAN AND
1879
CO.
PEINTED BY
TRANSLATOE'S PREFACE.
The following translation was undertaken at the of Dr T. Archer Hirst, F.R.S., the translator of
instance
the
first
Heat.
written
so completely re-
Dr
I trust
it
may
and connected
treatise
on
Thermodynamics,
use
in
students
for
more complete
this purpose, I
still
seemed of
are, (1)
interest, at
any rate
pro-
These
The Thermo-elastic
vi
translatoe's preface.
;
perties of Solids
(2)
The
;
application of Thermo-dynamical
(3)
principles to Capillarity
The Continuity
of the Liquid
and Gaseous
states of Matter.
F.E..S.,
My
Dr John Hopkinson,
investigation from
three points, and also for the original and very elegant
first
principles,
contained in the
first
My
through
the
first
suggestions
made
in connection with
it.
WALTER
E.
BROWNE.
10, ViCTOKiA
Chambees, Wesimiksteb,
November, 1879.
AUTHOR'S PREFACE.
Many
made
numerous papers "On the Mechanical Theory of Heat," which he had published at different
different quarters that the
many who, from the widespread interest now felt in this theory, were anxious to study them, he undertook some years back to publish a complete collection of his papers relating to the
times during a series of years, were inaccessible to
subject.
As a fresh edition of this book has now become necessary, he has determined to give it an entirely new form. The Mechanical Theory of Heat, in its present development, forms already an extensive and independent branch of science. But it is not easy 'to study such a subject from a series of separate papers, which, having been published at different times, are unconnected in their form, although they agree Notes and additions, however freely used in their .contents. to explain and supplement the papers, do not wholly overcome the difficulty. The author, therefore, thought it best so to re-model the papers that they might form a connected whole, and enable the work to become a text-book of the He felt himself the more bound to do this because science.
Vlll
AUTHOES PREFACE.
on the. Mechanical Theory
Heat at a Polytechnic School and at several Universities had taught him how the subject-matter should be arranged and represented, so as to render the new view and the new method of calculation adopted in this somewhat difficult theory the more readily intelligible. This plan also enabled him to make use of the investigations of other writers, and by that means to give the subject greater completeness and
of
.
These authorities of course have been in every case During the ten years which have elapsed since the first volume of papers appeared, many fresh
finish.
part
new
Theory of Heat,
owes
its
it
still,
as
contains so
that is fresh, he
to call it a
much new
work.
OLAUSIUS.,
TABLE OF CONTENTS.
MATHEMATICAL INTRODUCTION.
ON MECHANICAL WOKK, ON ENERGY, AND ON THE TREATMENT OP
NON-INTEGEABLE DIFFERENTIAL EQUATIONS.
PAGE
Defiuition
and Measurement
Integration of the Differential Equations for Work done Geometrical interpretation of the foregoing results, and obaerTations on Partial Differential Coefficients Extension of the foregoing to three Dimensions On the Brgal General Extension of the foregoing Eelation between Work and Vis Viva
.
On Energy
9 11 13 17 19
CHAPTER
FIRST MAIN PRINCIPLE OP
I.
1.
2.
3.
Starting Point of the theory Positive and Negative Values of Mechanical Expression for the First Main Principle
4.
5. 6.
7. 8. 9.
Numerical Eelation between Heat and Work of Heat Development of the First Main Principle Different Conditions of the quantities J, W, and
....
K
. . . .
...
21 22 23
24 25
27
.27
30 32 33 35
of the Body Cyclical Processes for Finite Changes of Condition ^Latent and Specific Heat 10. Total Heat . 11. Expression for the External Work in a particular case
Energy
Equations
CONTENTS.
CHAPTER
II.
ON PERFECT GASES.
1. 2. 3. 4. 5. 6. 7. 8. 9.
by Gases
ciple assumes in the case of Perfect Gases Deduction as to the two Specific Heats and transformation of the
foregoing equations Eelation between the two Specific Heats, and its appUcatiou to calculate the Mechanical Equivalent of Heat Various Formulse relating to the Specific Heats of Gases Numerical Calculation of the Specific Heat at Constant Volume Integration of the Differential Equations which express the First Main Principle in the case of Gases Determination of the External Work done during the change of Volume of a Gas
. . .
.... ....
first
...
PAOE 38 42
43 46 48 52 56 60 64
Main Prin-
CHAPTER
III.
H!eAT.
7.
8. 9.
Description of a special form of Cyclical Process Eesult of the CycHoal Process Cyclical Process in the case of a body composed partly of Liquid and partly of Vapour Camot's view as to the work performed during the Cyclical Process New Fundamental Principle concerning Heat Proof that the Eelation between the Heat carried over, and that converted into work, is independent of the matter which forms the medium of the change Determination of the Function (Tj Tj) Cyclical Processes of a more complicated character Cyclical Processes in which taking in of Heat and change of Temperature take place simultaneously
. . .
'
....
69 71 73 76 78
......
...
79 81 84 87
CHAPTER
IV.
THE SECOND MAIN PEINCIPLE UNDER ANOTHER FORM; OR PRINCIPLE OF THE EQUIVALENCE OF TEANSFORJIATIONS.
1. 2.
3.
On
91 92 86
CONTENTS.
XI
4. 5.
6. 7. 8.
PAOB Equivalence Values of the Transformations 97 Combined value of all the Transformations wliieh take place in a single CycHeal Process 101 . . . . , Proof that in a reversible Cyclical Process the total value of all the Transformations must be equal to nothing . . 102 On the Temperatures of the various quantities of Heat and the Entropy of the body 106 On the Function of Temperature, t . . . .107
'
CHAPTER
V.
5. 6.
7.
Discussion of the Variables which determine the Condition of the body , Elimination of the quantities U and S from the two fundamental equations Case in which the Temperature is taken as one of the Independent Variables Particular Assumptions as to the External Forces Frequently occurring forms of the differential equations Equations in the case of a body which undergoes a partial change in its condition of aggregation Clapeyron's Equation and Carnot's Function
. . ; . . . .
.
110
112 115 116 118
119 121
....
CHAPTER
APPLICATIOjST
VI.
1.
2. 3. i.
5. g 6.
I 7.
Fundamental equations for Saturated Vapour Specific Heat of Saturated Steam Numerical Value of h for Steam
....
. .
Numerical Value of h for other vapours Specific Heat of Saturated Steam as proved by experiment Specific Volume of Saturated Vapour Departure &om the law of Mariotte and Gay-Lussac in the case
of Saturated Steam
........
5 8.
Differential-coefficients of
9.
1
determine the Specific volume of Saturated Steam, and its comparison with Experiment 10. Determination of the Mechanical Equivalent of Heat from the behaviour of Saturated Steam
A Formula to
150
155
Xll
CONTENTS.
PAGE
Complete Differential Ec^aation for Q in the case of a body composed both of Liquid and Vapour Change of the Gaseous portion of the mass Belation between Volume and Temperature Determination of the Work as a function of Temperature
.... ....
.
CHAPTER
Fundamental Equations
VII.
1. 2.
3. 4.
5.
6.
for the process of Fusion Eelation between Pressure and Temperature of Fusion Experimental Verification of the Foregoing Eesult Experiments on Substances which expand during Fusion Eelation between the Heat contained in Fusion and the Temperature of Fusion Passage from the solid to the gaseous condition
.
.
. .
163*
....
CHAPTEE
VIII.
ON HOMOGENEOUS BODIES.
1. 2. 8.
Improved Denotation
Eelations
4.
5. 6. 7.
8.
between Pressxire, Volume, and Temperature Complete differential equations for Q Specifie Heat at constant volume and constant pressure Specific Heats under other oiroumstanees laentropic Variations of a body Special form of the fundamental equations for an Expanded Eod
.
175 176
Alteration of temperature during the extension of 9. 10. Further deductions from the equations
the rod
CHAPTEE
IX.
1.
General equations
Differential equations for the case in which only reversible changes take place, and in which the condition of the Body is
195
2.
....
197
CONTENTS.
Introdnction of the Temperature as one of the Independent
Xlll
PAGE
3.
4. 5.
6.
YatiaWes
Special case of the differential equations on the assumption that the only External Force is a uniform surface pressure . . Application of the foregoing equations to Homogeneous Bodies and in particular to perfect Gases Application of the equations to a body composed of matter in two States of Aggregation Kelation of the expressions Dxy and Axy
200 203
20S 207 209
7.
CHAPTER
X.
ON NON-BBVEESIBLE PKOCESSES.
1. 2. 3.
for the
Second
.
4.
5.
6.
7.
8.
Principle Magnitude of the Uncompensated Transformation . . Expansion of a Gas unaccompanied by Work Expansion of a Gas doing partial work Methods of experiment used by Thomson and Joule . . . Deyelopment of the equations relating to the above method . Besults of the experiments, and the equations of Elasticity for the gases, as deduced therefrom On the Behaviour of Vapour during expansion and under various
Main
....
224
228
circumstances
CHAPTER XL
APPLICATION OF THE THEOKT OF HEAT TO THE STEAM-ENGINE.
1.
Necessity of a
new
Bngiue
j2.
I
3.
4.
5.
6. 7.
Action of the Steam-Engine Assumptions for the purpose of SimpMoation Determination of the Work done during a single stroke . . Special forms of the expression fovmd in the last section Imperfections in the construction of the Steam-Engine Pambour's Eormulse for the relation between Volume and Pres-
On the
....
. .
.
sure
J
8.
single
.
)
i
9. 10.
Pambour's Value for the Work done per unit- weight of steam Changes in the Steam during its passage from the Boiler into
the Cylinder
244 247
248
XIV
CONTENTS,
Divergence of the Eesnlts obtained from Pambour's Assumption Determination of the work done during one stroke, taking iato . . consideration the imperfections abeady noticed Pressure of Steam in the Cylinder during the different Stages of the Process and corresponding Simplifications of the Equations Substitution of the Vohune for the corresponding Temperature in certain cases
.
PAOE 251
252
,
255
14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34.
Work per
....
. . .
.
Introduction of other measures of Pressure and Heat Determination of the temperatures T^ and Ta . . Determination of c and r Special Form of Equation (32) for an Engine working without expansion Numerical values of the constants The least possible value of V and the corresponding amount of
. . .
'
Work
Calculation of the work for other values of F . . . . Work done for a given value of by an engine with expansion . Summary of various cases relating to the working of the engine Work done per unit of heat delivered from the Source of Heat . Friction . General investigation of the action of Thermo-Dynamio Engines and of its relation to a Cyclical Process Equations for the work done during any cychoal process . . AppHoation of the above equations to the limiting case in which Cyclical Process in Steam-Engine the is reversible . . a Another form of the last expression Influence of the temperature of the Source of Heat . . . Example of the application of the Method of Subtraction . .
274
275
CHAPTEE
XII.
ON THE CONCENTRATION OP EATS OF LIGHT AND HEAT, AND ON THE LIMITS OF ITS ACTION.
1.
g 2.
3. 4.
Object of the investigation I. Beasons why the ordinary method of determining the mutual radiation of two surfaces does not extend to the present case Limitation of the treatment to perfectly black bodies and to homogeneous and unpolarized rays of heat Kirchhoff's Formula for the mutual radiation of two Elements of surface Indetermiuateness of the Formula in the case of the Concentration of rays
293
....
296
296
297 300
CONTENTS.
Determination of corresponding points and corresponding Elements of Surface in three planes cut by the rays . Equations between the co-ordinates of the points in which a ray cuts three given planes The relation of corresponding elements of surface Various fractions formed out of six quantities to express the Eolations between Corresponding Elements III. Determination of the mutual radiation when there is no
II.
.
.
XV
PAGE
301 301 304
.SC9
5.
6.
7.
concentration of rays
8. 9.
plane in a particular position Expressions for the quantities of heat which da, and ds, radiate to each other . 10. Eadiation as dependent on the surrounding medium IV. Determination of the mutual radiation of two Elements of Surface In the case when one is the Optical Image of the
. .
....... .......
....
310
on a
310
812 314
other
11. Relations between B,D,F,a.ndiE and to determine the relation 12. Application of
316 316
318
319
between the A C elements of surface 18. Eelation between the quantities of Heat which ds, and ds^ radiate to each other V. Eelation between the Increment of Area and the Eatio of . the two solid angles of an Elementaiy Pencil of Bays 14. Statement of the proportions for this case VI. General determination of the mutual radiation of two surfaces in the case where any concentration whatever may
take place
15. 16.
321 321
General view of the concentration of rays Mutual radiation of an element of surface and of a through an element of an intermediate plane
.
324 324
finite surface
. .
3r 5
3'.
17. Mutual radiation of entire surfaces 18. Consideration of various collateral circumstances 19. Summary of results
329 330
CHAPTER
DISCUSSIONS ON
XIII.
1.
2. 3.
4. 5.
Different views of the relation between Heat and Work . Papers on the subject by Thomson and the author On Eankine's Paper and Thomson's Second Paper
.
332 3S3
3i?4
Holtzmaan's Objections
Decher's Objections
337 339
of the second
.
6.
7,
8.
XVI
CONTENTS.
9. 10. 11.
... ...
Appenhix
I.
....
. . .
363
II.
CapiUarity
369 373
III.
MATHEMATICAL INTRODUCTION.
ON MECHANICAL WORK, ON ENEHGT, AND ON THE TREATMENT OF NON-INTEGRABLE DIFFERENTIAL
EQUATIONS.
1. JOefinition and Measurement of Mechanical Work. Every force tends to give fiiotion to the body on which it acts; but it may be prevented from doing so by other opposing forces, so that equilibrium results, and the body reniains at rest. In this case "the force performs no work. But as soon as the body moves under the influence of the
force,
Work
is
performed.
In order to investigate the subject of Work under the simplest possible conditions, we may assume that instead of an extended body the force acts upon a single material point. If this point, which we may call p, travels in the same straight line in which the force tends to move it, then the product of the force and the distance moved through is the mechanical work which the force performs during the motion." If on the other hand the motion of the:point is in any other direction than the line of action of the force, then ibhe work performed is represented by the product of the distance moved through, and the component of the force resolved in the direction of motion, This component of force in the line of motion may be
piopitive or
c.
it
tends in th
1
'
same direction in which the motion actually takes place, or in the opposite. The work likewise will be positive in To express the the first case, negative in the second. difference in words, which is for many reasons convenient, recourse may be had to a terminology proposed by the writer in a former treatise, and the force may be said to do or perform work in the former case, and to destroy work
in the latter.
From the
tities of
foregoing
it is
work numerically, we should take as unit that quantity of work which is performed by an unit of force acting through an unit of distance. In order to obtain a scale of measurement easy of application, we must choose, as our normal or standard force, some force' which is The force thoroughly known and easy of measurement.
usually chosen for this purpose
is that of gravity. Gravity acts on a given body as a force always tending downwards, and which for places not too far apart may be If now we wish to lift a taken as absolutely constant. weight upwards by means of any force at our disposal, we must in doing so overcojjae the force of-gravity; and gravity thus gives a measure of the force which we must Hence we take as our exert for any slow lifting action. unit of work that which must be performed in order to lift a unit of weight through a unit of length. The units of weight and length to be. chosen are of course matter of indifference ; in applied mechanics they are generally the kilogram and the metre respectively, and then the unit of work is called a kilogram metre. Thus to raise a weight of a kilograms through a height of b metres ab kilogrammetres of work are required and other quantities of work, in cases where gi;avity does not come directly into play, can also be expressed in kilogrammetres, by comparing the forces employed with the standard force of gravity.
;
2.
In the foregoing explanation it has been tacitly assumed that the active component of force has a constant value; throughout the whole of the distance traversed. In reality' this is not usually true for a distance of finite length. On
MATHEMATICAL INTRODUCTION.
the one hand the force need not itself he the same at different points of space ; and on the other, although the force may remain constant throughout, yet, if the path be not straight but curved, the component of force in the direction of motion will still vary. For this reason it is allowable to express work done by a simple product, only when the distance traversed is indefinitely small, i.e. for an element of space. Let ds be an element of space, and S the component in the direction of ds of the force acting on the point p. We have then the following equation to obtain dW, the work done during the movement through the indefinitely small space ds
dW=Sds
If
(1).
resultant force acting on the point p, and ^ the angle which the direction of this resultant makes with the direction of motion at the point under consideration,
P be the total
then
S = P cos
whence we
have,
<f),
by
(1),
dW = P cos ^ds
(2).
It is convenient for calculation to employ a system of rectangular co-ordinates, and to consider the projections of the element of space upon the axes of co-ordinates, and the components of force as resolved parallel to those axes. For the sake of simplicity we will assume that the motion takes place in a plane in which both the initial direction of will employ motion and the line of force are situated. rectangular axes of co-ordinates lying in this plane, and will call X and y the co-ordinates of the moving point ^ at a given time. If the point moves from this position in the plane of co-ordinates through an indefinitely small space ds, the projections of this motion on the axes will be called cfe and dy, and will be positive or negative, according as the co-ordinates x and y are increased or diminished by the motion. The components of the force F, resolved in the
We
and Y. Then, if a directions of the axes, will be called and b are the cosines of the angles which the line of force makes with the axes of a; and y respectively, we have
x'=aP;
Y=bP.
12
Again, if a and y3 are the cosines of the angles wliich the element of space ds makes with the axes, we have
dx = ads
dy = ^ds.
rom
these equations
we
obtain
= cos ^,
line of
force,
and the
by equation
(2),
dW=Xdx+
nitely small motion,
Ydy
(3).
This being the equation for the work done during an indefiwe must integrate it to determine the work done during a motion of finite extent.
done.
3.
In the integration of a differential equation of the form given in equation (3), in which and are functions of x and y, and which may therefore be written in the form
dW=<f>(!^)dx + y{r(a^\dy
(3a),
a distinction has to be drawn, which is of great importance, not only for this particular case, but also for the equations which occur later on in the Mechanical Theory of Heat j and
which will therefore be examined here at .some length, so that in future it will be sufficient simply to refer back to the ; present passage. According to the nature of the functions ^ (my) and \fr (o^), differential equations of the form (3) fall-iSto two classes, which differ widely both as to the treatment which they require, and the results to which they lead. To the
ilATHEMATICAI/ INTRODUCllOU.
first class
fulfil
X and Y
""^
''
dy
dx
'""
which
The second
is
this condition
not
fulfilled.
If the condition (4) is fulfilled, the expression on the right-hand side of equation (3) or (3a) becomes immediately integrable ; for it is the complete differential of some function of X and y, in which these may be treated as independent variables, and which is formed from the equations
dx
Thus we obtain
at once
'
dy
w = F {xy) + const
;
If condition (4) is not fulfilled, the right-hand side of the cannot be equation is not integrable and it follows that expressed as a function of x and y, considered as independent For, if we could put variables. Fiacy), we should have
W=
^^dW^ dF{xy)
dib
dx
dy
'
Y^dW^ dFjxy)
dy
'
whence
it
follows that
dX ^ d'Fjxy)
dy
dxdy
'
dY^ d^Fjxy)
dx
But
dxdy
dydx
whence
tion.
follows that-y-
=-5
'e-
condition
(4) is fulfilled
X and Y;
which
is
then the integration is impossible, so long as If however as independent variables. relation to hold between these two quantities, so that one may be expressed as a function of the For if we other, the integration again becomes possible. put
In
this case
/(^2/)
(6),
in which expresses any function whatever, then by means of this equation we can eliminate one of the variables and its differential from the differential equation. (The general form in which equation (6) is given of course comprises the special case in which one of the variables is taken as constant ; its differential then becomes zero, and the variable itSBlf only appears as part of the constant coefficient). Supposing y to be the variable eliminated, the equation (.3) takes the form dW=<^{x) dx, which is a simple differential equation, and gives on integration an equation of the form
w = F {x) \- comi
The two equations
(6)
(7).
ing together a solution of the differential equation. As the form of the function fixy) may be anything whatever, it is clear that the number of solutions thus to be obtained is
infinite.
The form of equation (7) may of course be modified. Thus if we had expressed x in terms of y by means of equation (6); and then eliminated x and dx from the differential equation, this latter would then have taken the form
dW = <^,{y)dy,
and on integrating we should have had an equation
''^=^1(2')+ const
(7a3.
(7)
by
substituting y for x in that equation by means of equation (6). Or, instead of completely eliminating x from (7), we may
MATHEMATICAL INTRODUCTION.
'
prefer a partial elimination. (cc) conFor if the function tains X several times over in different teriiis, (and if this does not hold in the original form of the equation, it can be easily introduced into it by writing instead of x an expression such
as (l
a)x + ax, ^,
X
in
&c.)
then
it is
possible to substitute
to let
for
some of these
expressions,
and
x remain
in
others.
+ const
(76),
which
is
It is of course understood that the three equa^ tions (7), (7a), (76), each of which has no meaning except when combined with equation (6), are not different solutions, but different expressions for one and the same solution of
special cases.
W=F{x,y) + const
(A).
When this condition is not fulfilled, we must first assume some relation between the variables, in order to make inte^ation possible; and we shall thereby obtain a system of two equations of the following form
/(*,2/)
= 0,
W=F{a;.,y)
+ const. \
,gv ^^^'
depends not only on in which the form of the function that of the original differential equation, but also on that of the function/, which may be assumed at pleasure.
8
i.
and
difference between the results in the two mentioned above is rendered more clear by treating them geometrically. In so doing we shall for the sake of {x, y) in equation (A) simplicity assume that the function is such that it has only a single value for any one point in the plane of co-ordinates. We shall also assume that in the movement of the point p its original and final positions are known, and given by the co-ordinates x^, y^, and a;,, y^ respectively. Then in the first case we can find an expression for the work done by the effective force during the motion,: without needing to know the actual path traversed. For it is clear, that this work will be expressed, according to concases
The important
dition {A)i by the difference F(x^p^ F(xy). Thus, while the moving point may pass from one position to the other by very different paths, the amount of work done by the. force is wholly independent of these, and is completely known aS soon as the original and final positions are given. In the second case it is otherwise. In the system of equations (B), which belongs to this case, the first equation must be treated as the equation to a curve ; and (since the form of the second depends upon it) the relation between them may be geometrically expressed by saying that the work done by the effective force during the motion of the point p can only be determinedj when the whole of the curve, on which the point moves, is known. If the original and final positions are given, the first equation must indeed be so chosen, that the curve which corresponds to it may pass through those two points ; but the number of such possible curves is infinite, and accordingly, in spite of their coincidence at theii: extremities, they will give an infinite number of possible quantities of "work done during the motion. If we assume that the point p describes a closed curve, so that the final and initial positions coincide, and thus the coordinates x^, y^ have the same value as x, y^, then in the first case the total work done is equal to zero: in the second case, on the Other hand, it need not equal zero, but may have any value positive or negative. The case here examined also illustrates the fact that a
'
MATHEMATICAL INTBOBUCTION.
9^.
quantity, which cannot be expressed as a function of x and y (so long as these are taken as independent variables),, may
yet have partial differential coefficients according to x and y, which are expressed by determinate- functions of those variables. For it is manifest that, in the strict sense of the words, the components and must be termed the partial differential coefficients of the work according to x and y smce, when x increases by dx, y remaining constant, the,
X
;
work increases by Xdx and when y increases by dy, x remaining constant, the work increases by Ydy. Now whether TF be a quantity generally expressible as a function of x and y, or one which can only be determined on knowing the path described by the inoving point, we may always employ the ordinary notation for the partial differential coefficients of W, and write
.(8).
this notation we may also write the condition (4), the fulfilment or non-fulfilment of which causes the distinction between the two modes of treating the differential equation,
Using
,g.
Thus we may say that the distinction which has to be' depends on whether drawn in reference to the quantity
W
is
-
the difference
t-(-tdy\dx J
-r- (7 dx\dy J
,
finite value.
5.
to three dimensions.
If the point p be not restricted in its movement to one plane, but left free in space, we then obtain for the element of work an expression very similar to that given in equation Let a, b, c be the cosines of the angles which the direc(3). tion of the force P> acting on ..the, point, makes with three
10
rectangular axes of co-ordinates ; then the three components X, Y, of this force will be given by the equations
X=aF,
Y=hP, Z=.cP.
Again, let a, 7 be the cosines of the angles, which the element of space ds makes with the axes ; then the three, projections dx, dy, dz of this element on those axes are given by the equations
dx = ads,
Hence we have
dy = ^ds,
ds
= 'yds.
{pa.
P and ds^
hence
&;8
Pds.
Comparing
this
with equation
(2),
we
obtain
(10).
dW = Xdx+Ydy + Zdz
This is the differential equation for determining the work may be any functions whatdone. The quantities X, Y, ever of the co-ordinates oc,y,z; since whatever may be the values of these three components at different points in space, may always be derived from them. a resultant force
we must
first
consider the
fol-
dX^dY
dy
satisfy
dY^dZ
^
dZ^dX
'
dx
dz
dy
dx
dz
'^'
on the right-hand
If these three conditions are satisfied, then the expression side of (11) is the complete differential of a function of x, y, z, in which these may all be treated as independent variables. The integration may therefore be at once effected, and we obtain an equation of the form
TF = P(a!^s)
+ const
(12).
MATHEMATICAL INTEOBUCTION.
If
11
j> to move from a given z^ to a given final position (ajj, y^, z^ the work done by the force during the motion will be repre-
we now
sented by
If then we suppose F{x, y, z) to be such that single value for any one point in space, the
it
work
completely determined by the original and final positions; and it follows that the work done by the force is always the same, whatever path may have been followed by the point in passing from one position to the other. If the three conditions (1) are not satisfied, the integration cannot be effected in the same general manner. If, however, the path be known in which the motion takes place, the integration becomes thereby possible. If in this case two points are given as the original and final positions, and various curves are conceived as drawn between these points, along any of which the point p may move, then for each of these paths we may obtain a determinate value for the work dpne; but the values corresponding to these different paths need not be equal, as in the first case, but on the contrary
are in general, different.
6.
OniheErgal.
In those cases in which equation (12) holds, or the work done can be simply expressed as a function of the co-ordirtates, this function plays a very important part in our calculations. Hamilton gave to it the special name of "force function"; a
applicable also to the more general case where, instead of a single moving point, any number of such points are considered, and where the condition is fulfilled that the work done depends only on the position of the points. In the later and more extended investigations with regard to the quantities which are expressed by this function, it has become needful to introduce a special name for the negative value of the function, or in other words for that quantity, the svhtraction of which gives the work performed; and Rankine proposed for this the term 'potential energy.' This name sets forth very clearly the character of the quantity ; but it
name
1'2
somewliat long, and the author has ventured to propose * its place the term " Ergal." Among the cases in which the force acting on a point has an Ergal, the most prominent is that in which the force originates in attractions or repulsions^ exerted on the moving point from fixed points, and the value of which depends only on the distance in other words the case ia which the force may be classed as a central force. Let us take as centre of force a fixed, point it, with co-ordinates ^, vj, and let p be its distance from the moving point p, so that
is
in
",
p=J{^-'cr+{v-yr+{^-^r-
as)-
Let us express the force which tt exerts on p by <^' {p), in which a positive value of the function expresses attraction, and a negative value repulsion we then have for the com;
^=fW^;
r
F=.^'(p)^.
r
:
^=^'(p)i:rf.
r
But by
that
(13)
^ = - ^-^^
hence
If
X
^ (p)
<i>'
(p)%, and
simi-
two axes.
be a function such
<l>(p)=j<}>'{p)dp....
.(14).
we may
j-_
and similarly
dj>{p)
dp
(i4>{p)
dp
dx
dx
''
F=
-^, ^=-^
(15a).
Hence we have
ItitHEMATlOAt INTRODUCTION,'
1'3
may
we
(16).
of
(p)
work is thus given by the negative differenwhence it follows that 6 (p) is in this case the
number
^'(P)>
Again, instead of a single fixed point, we may have any of fixed points tTj, tt^, w &c., the distances of which
p^,
it forces as in equation (14), we assume^j(jo), <^^{p), 4)^(p), &c, to be the integrals of the above Junctions, we obtain, exactly as in equation (15),
from p^ are
p^,
p^. Sec,
^'W. 0'(Pj).
&c.
Then
x^.
<^'^i(pi)
dx
^M
doa
#.,(p)
dsB
^ ^S^W
Similarly
(17).
r=-|2^(p), Z=-^^t^{p)
Xdx + Ydy + Zdz = - d% <^
{p)
(I7a),
whence
(18).
Hitherto
7.
S ^{p)
is
we have only considered a single moving point now extend the method to comprise a system composed of any number of moving points, which are in part
we
will
If this whole system makes an indefinitely small movement, the forces acting on any One point, which forces we may conceive as combined into a single resultant, will perform a quantity of work which may be represented by the
expression
{Xdx
+ Ydy + Zdz),
of all the
14
work done by all the forces acting in the system represented by an expression of the form
may be
This
peculiarity that it is the complete differential of some function of the co-ordinates of all the moving points ; in which case we call this function, taken negatively, the Ergal of the whole system. It follows from this that in a finite movement of the system the total work done is simply equal to
the difference between the initial and final values of the Ergal; and therefore (assuming that the function which, represents the Ergal is such as to have only one value for one position of the points) the work done is completely determined by the initial and final positions of the points, without its being needful to know the paths, by which these have moved from one position to the other.
This state of things, which, it is obvious, simplifies greatly the determination of the work done, occurs when all the forces acting in the system are central forces, which either act upon the moving points from fixed points, or are actions between the moving points themselves. First, as regards central forces acting from fixed points, we have already discussed their effect for a single moving point ; and this discussion will extend also to the motion of the whole system of points, since the quantity of work done in the motion of a number of points is simply equal to the sum of the quantities of work done in the motion of each several point. We can therefore express the part of the Ergal relating to the action of the fixed points, as before, by 2 (^ (p), if we only give such an extension to the summation, that it shall comprise not only as many terms as there are fixed points, but as many terms as there are combinations of one fixed and one moving point. Next as regards the forces acting between the moving points themselves, we will for the present consider only two points p and p\ w^Hh co-ordinates co, y, z, and x, y', z\
MATHEMATICAL INTKODXJCTION,
respectively.
15
If r be
tlie
we
have
r
= J{x'-a=y + (y'-yr+[,'-zy
(19).
We may denote
the force which the points exert on each other by/'(r), a positive value being used for attraction, and a negative for repulsion. Then the components of the force which the point p
exerts in this
/W^, /'^,
and the components of the opposite
force exerted
by p' are
/'W^', /'W^'.
But by
(19), differentiating
/'(r)i^'.
dr
x'
X
r
dr
dx'
x^
r
'
dx
'
x may
and
if
f(r)=jf'{r)dr
the foregoing
(20),
may
also
be written
,
-dfjr) dx
-df(r)
'
dd
may be
dy
'
dy'
'
-df{r)^ -df(r)
16
That part
of the total work done in the indefinitely small motion of the two points, which is due to the two opposite foTcea arising from their mutual ^,ction, may therefore be
expressed as follows
dz
J
y, z, quantities
oc',
But as r depends only on the six quantities x^ ,y, and f{r) can therefore be a function of these six
z,
and
the work done, as far as concerns the mutual action between the two points, may be simply expressed by tl)e function
In the same way may be expressed the work due to the" mutual action of every other pair of points and the total work done by all the forces which the points exert among themselves is expressed by the algebraical suni
;
-rf/(r)-J/(/)-d/(0-...;
or as
it
may be
otherwise written,
or
-<^[/W+/(^')+/(0 + -]
in which the summation there are combinations of
-dS/W;
must comprise as many terms as moving points, two and two. This
sum S/(r) is then the part of the Ergal relating to the mutual and opposite actions of all the moving points. If we now finally add the two kinds of forces together, we obtain, for the .to.tal work done in the indefinitely small motion of the system of points, the equation
2 {Xdx + Ydy + Zdz) =-dt(]>(p)^ dtfir) = ^d[tcl>{p)+-ZAr)] whence it foUows that the quantity %<f>{p) + '^f{r)
(21),
is the Ergal of the whole of the forces. acting together in ^he systeni,. The assumption lying at the root of the foregoing analysis, viz. that central forces are the only ones acting, is of course only one among all .the assumptions mathematically
MATHEMATICAL INTEODUCTION.
17
possible as to the forces; but it forms a case of peculiar importance, inasmuch as all the forces which occur in nature may apparently be classed as central forces.
8.
Relation between
Viva.
Hitherto we have only considered the forces which act on the points, and the change in position of the points themselves ; their masses and their velocities have been left out of account. wiU now take these also into consideration. The equations of motion for a freely moving point of
We
mass
m are
well
known
to
be as follows
If
we
dx
7,
dy
,^
dz
The
into
may be
dt.
transformed
md
2dt
or, if
m <%'<%'
^i(p.dt=%ldt=d(^v^ dt 2 dt V2 y
and the equation becomes
<'(f')=(^S+4!+4)*
n
'^')-
If, instead of a single freely moving point, a whole system of freely moving points is considered, we shall have for every
18
point a similar equation to, the above; we shall obtain the following:-
<ffi5^.2(x|+7|+43i.
Now the quantity 2
of points.
If
n-t)'
<^-
we take a simple
T=-Z''^v'
and put
(26),
..r=2(z| + r|+zg)c^*
for the
(27).
But the right-hand side of this equation is the expression work done during the time dt. Integrate the equation from an initial time i to a time t, and call T^ the vis
viva at time
t^:
is
''-^-/>(^S+4!+^S)*
the meaning of which
(^''
may be
expressed as follows:
The Wo^h done during any time hy the forces acting upon a system is equal to the increase of the Yis Viva of the system during the same time.
In this expression a diminution of Vis Viva is of course treated as a negative increase. It was assumed at the commencement that all the points were moving freely. It may, however, happen that the poihts are subjected to certain constraints in reference to their motion. They may be so connected with each other that the motion of one point shall in part determine the .motion or there may be external constraints, as for inof others stance, if one of the points is compelled to move in a given fixed plane, or on a given fixed curve, whence it will natur;
* Translator's Note. The vis viva of a particle is here defined as half the mass multiplied by the square of the velocity, and not the whole mass, as was formerly the custom.
MATHEMATICAL INTEODUCTION.
with
19
ally follow that all those points, which are in any connection it, will also be to some extent constrained in their
motion.
If these conditions of constraint can be expressed by equations which contain only the co-ordinates of the points, it may be proved, by methods which we will not here consider more closely, that the reactions, which are implicitly comprised in these conditions, perform no work whatever during the motion of the points and therefore the principle given above, as expressing the relation between Vis Viva and Work done, is true for constrained, as well as for free motion. It is called the Principle of the Equivalence of Work and
;
Vis Viva.
is
9.
On
Energy.
t^
to t
I'Min which t is considered as the only independent variable, and the co-ordinates of the points and the components of the forces are taken as functions of time only. If these functions are known (for which it is-requisite that we should know the whole course of the motion of all the points), then the integration is always possible, and the work done can also be determined as a function of the time.
Cases however occur, as we have seen above, in which it not necessary to express all the quantities as functions of one variable, but where the integration may still be effected, by writing the differential in the form S (Xdx + l^y -t- Zdz), and considering the co-ordinates therein as independent variFor this it is necessary that this expression should ables. be a perfect differential of some function of the co-ordinates, or in other words the forces acting on the system must have an Ergal. This Efgal, which is the negative value of the The letter above, function, we will denote by a single letter. is generally chosen for this purpose in works on Mechanics: but in the Mechanical Theory of Heat that letter is needed to express another quantity, which will enter as
is
22
20
iHEAT.
we
will
Ergal by
J.
Hence we put:
%(X.Ax^Ydy-^Ziz) = -dJ......
whence
if J^ '0
(29),
fi
we have:
(30),
^{Xdx^Ydy^Zdz)=J,-J.
/,
is
equal to
substitute J^
J for
whence we have the following principle: The sum of the Vis Viva arid of the Ergal remains constant during the motion. This sum, which we will denote by the letter TJ, so that
U=T + J.
is
(32),
;
Energy of the system so that the above principle may be. more shortly expressed by saying: The Energy remains constant during the motion. This principle, which in recent times has received a much more extended application than formerly, and now forms one of the chief foundations of the whole structure of physical philosophy, is known by the
called tbe
name
(21)
CHAPTER
I.
THEORY OF
AND WORK.
1,
Nature of Heat.
it M'as the generally accepted view that special substance, which was present in bodies in
Until recently
Heat was a
greater or less quantity, and which produced thereby their higher or lower temperature; which was also sent forth from bodies, and in that case passed with immense speed through empty space and through such cavities as ponderable bodies contain, in the form of what is called radiant heat. In later days has arisen the other view that Heat is in reality a mode of motion. According to this view, the heat found in bodies and determining their temperature is treated as being a motion of their ponderable atoms, in which motion the ether existing within the bodies may also participate ; and radiant heat is looked upon as an undulatory motion propagated in that ether. It is not proposed here to set forth the facts, experiments, and inferences, through which men have been brought to this altered view on the subject ; this would entail a reference here to much which may be better described in its own place during the course of the book. The conformity with experience of the results deduced from this new theory will probably serve better than anything else to establish the
foundations of the theory itself will therefore start with the assumption that Heat consists in a motion of the ultimate particles of bodies and of ether, and that the quantity of heat is a measure of the Vis Viva of this motion. The nature of this motion we
We
22
shall not
ox
t:he
attempt to determine, but shall merely apply to principle of the equivalence of Vis Viva and Work, which applies to motion of every kind ; and thus establish a principle which may be called the first main Principle of
Heat the
total work.
of view.
also be looked at from another point If a material point has once been set in motion, it can continue this movement, on account of its momentum, even if the force acting on it tends in a direction opposite to
that of the motion; though its velocity, and therewith its Vis Viva, will of course be diminishing all- the time. material point acted on by gravity for example, if it has received an upward impulse, can continue to move against the force of gravity, although the latter is continually diminishing the velocity given by the impulse. In such a case the work, if considered as work done by the force, is Conversely however , we may reckon work as negative. positive in cases where a force is overcome by the momen* tum of a previously acquired motion, as negative in cases where the point follows the direction of the force. Applying the form of expression introduced in 1 of the Introduction, in which the distinction between the two opposite directions of the component of force is indicated by different words, we may express the foregoing more simply as follows we may determine that not the work done, but the work destroyed,
:
by a
On
principle of
23
the equivalence of Vis Viva and Work takes the following The decrease in ike Vis Viva is equal to the increase in the Work done, or The sum of the Vis Viva and Work done is constant. This latter form will be found. very convenient in
:
what follows. In the case of such forces as have an Ergal, the meaning of that quantity was defined (in 6 of the Introdi^oiiqg) in such a manner that we must say, The Work d^e^is 'equal to the decrease in the Ergal.' If we use the method of denoting work JMt ^described, we must say on the contrary, 'The work dop?^is equal to the increase in the Ergal;'- and if the
'
constant occurring as one term of the Ergal be determined in a particular way, we may then regard the Ergal as simply an expression for. the work done.
3.
Having fixed as above what is to be the positive sign for work done, we may now state as follows the first main
Principle of the Mechanical Theory of Heat.
In
all cases
where work
is
heat is consumed proportional to the work done; and inversely, by the expenditure of the same amount of work tfte same quantity of heat m,ay be produced. This follows, on the mechanical conception of heat, from
the equivalence of Vis Viva and Work, and is named The Principle of the Equivalence of Heat and Work. If heat is consumed, and work thereby produced, we may say that heat has transformed itself into work ; and conversely, if work is expended and heat thereby produced, we may say that work has transformed itself into heat. Using this mode of expression, the foregoing principle takes the following form Work may transform itself into heat, and heat conversely into work, the quantity of the one bearing always a fixed proportion to that of the other. This principle is established by means of many pheno-' mena which have been long recognized, and of late years has been confirmed by so many experiments of different kinds, that we may accept it, apart from the circumstance of its forming a special case of the general mechanical principle of the Conservation of Energy, as being a principle directly derived from experience and, observation.
:
24
'4.
While the mechanicd,! principle asserts that the changes in the Vis Viva and in the corresponding Work done are actually equal to each other, the principle which expresses the relation between Heat and Work is one of Proportion The reason is that heat and work are not measured only.
on the same scale. Work is measured by the mechanical imit of the kilogramme tre, whilst the unit of heat, chosen for convenience of measurement, is That amount of heat which is required to raise one kilogram of water from 0 to 1" (Centigrade). Hence the relation existing between heat and work can be one of proportion only, and the numerical value must be specially determined. If this numerical value is so chosen as to give the work corresponding to an unit of heat, it is called the Mechanical Equivalent of Heat ; if on the contrary it gives the heat corresponding, to an unit of work, it is called the Thermal shall denote the former by E, and Equivalent of Work.
,
We
the latter by -^
The determination
different ways.
is
effected in
sometimes been deduced from already existing data, as was first done on correct principles by Mayer (whose method will be further explained hereafter), although, from the imperfection of the then existing data, his At result must be admitted Hot to have been very exact. other times it has been sought to determine the number by experiments specially made with that view. To the distinguished English physicist Joule must be assigned the credit cf having established this value with the greatest circumspection and care. Some of his experiments, as well as
It has
determinations carried out at a later date by others, will more properly find their place after the development of the theory and we will here confine ourselves to stating those of Joule's experiments which are the most readily understood, and at the same time the most certain as to their results. Joule measured, under various circumstances, the heat
;
generated by friction, and compared it with the work consumed in producing the friction, for which purpose he employed descending weights. As accounts of these experi-
25
ments are given in many books, they need not here be described and it will suffice to state the results as given in
;
his paper, published in the Phil. Trans., for 1850. In the first series of experiments, a very extensive one,
water was agitated in a vessel by means of a revolving paddle wheel, which was so arranged that the whole quantity of water could not be brought into an equal state of rotation throughout, but the water, after being set in motion, was continually checked by striking against fixed blades, which occasioned numerous eddies, and so produced a large amount The result, expressed in English measures^ is of friction. that in order to produce an amount of heat which will raise 1 pound of water through 1 degree Fahrenheit, an amount of work equal to 772'695 foot-pounds must be consumed. In two other series of experiments quicksilver was agitated in the same way, and gave a result of 774'083 foot-pounds. Lastly, in two series of experiments pieces of cast iron were rubbed against each other under quicksilver, by which the heat given out was absorbed. The result was 774'987 foot-* pounds. Of all his results Joule considered those given by water as the most accurate; and as he thought that even this figure should be slightly reduced, to allow for the sound produced by the motion, he finally gave 772. foot-pounds as the most probable value for the number sought. Transforming this to French measures we obtain the result that. To produce the quantity of heat required to raise 1 kiiogramme of water through 1 degree Centigrade, work must be consumed to the amount o/'4!23'55 kilogrammetres. This appears to be the most trustworthy value among those hitherto determined, and accordingly we shall henceforward use it as the mechanical equivalent of heat, and write
J5=423-55
In most of our calculations to use the even number 424.
5.
it
(1).
will
be
sufficiently accurate
Having established the principle of the equivalence of Heat and Work, in consequence of which these two may be
26
opposed to each other in the same expression, we are often in the position of having to sum up quantities, in which heat and work enter as terms to be added together. As, however, heat and work are measured in different ways, we cannot in such a case say simply that the quantity is the sum of the work and the heat, but either that it is the sum of the heat and of the heat-equivalent of the work, or the sum of the work and of the work^equivalent of the heat. On account of this inconvenience Rankine proposed to employ a different unit for heat, viz. that amount of heat which is equivalent to an unit of work. This unit may be called simply the Mechanical Unit of Heat. There is an obstacle to its general introduction in the circumstance that the unit of heat hitherto used is a quantity which is closely connected with the ordinary calorimetric methods (which mainly depend on the heating of water), so that the reductions required are slight, and rest on measurements of the most while the mechanical unit, besides needreliable character ing the same reductions, also requires the mechanical equivalent of heat to be known, a requirement as yet only approximately fulfilled. At the same time, in the theoretical development of the Mechanical Theory of Heat, in which the relation between heat and work often occurs, the method of expressing heat in mechanical units effects siich important simplifications, that the author has felt himself bound to drop his former objections to this method, on the occasion of the present more connected exposition of that theory. Thus
;
in what follows, unless the contrary is expressly stated, it will be always understood that heat is expressed in mechanical
units.
first
this system of measurement the above mentioned main Principle of the Mechanical Theory of Heat takes a yet more precise form, since we may say that heat and its corresponding work 'are not merely proportional, but equal to
On
each other.
If later on it is desired to convert a quantity of heat expressed in mechanical units back again to ordinary heat units, all that wUl be necessary is to divide the number given in mechanical units by H, the mechanical equivalent of
heat.
27
6. Develc^ment of the Jtrst main Principle. Let any body whatever be given, and let its, condition as to temperature, volume, &c. be assumed to be known. If an indefinitely small quantity of heat dQ is imparted to this body, the question arises what becomes of it, and what effect it
produces. It may in part serve to increase the amount of heat actually existing in the body in part also, if in consequence of the imparting of this heat the body changes its condition, and that change includes the overcoming of some
;
may be absorbed in the work done thereby. If we denote the total heat existing in the body, or more briefly the Quantity of Heat of the body, by H, and the indefinitely small increment of this quantity by dR, and if we put dL for the indefinitely small quantity of work done, then we can
force, it
write
dQ = dH-\-dL
(I).
The forces against which the work is done may be divided into two classes: (1) those which the molecules of the body exert among themselves, and which are therefore dependent on the nature of the body itself, and (2) those which arise from external influences, to which the body is subjected. According to these two classes of forces, which have to be overcome, the work done is divided into internal and external work. If we denote these two quantities by
dJ and d W, we may
put
dL = dJ+dW
and then the foregoing equation becomes
(2),
dQ = dH+dJ+dW
7.
Different conditions of the Quantities J,
(II).
W, and H.
external work obey widely different internal work it is easy to see that if a body, starting from any initial condition whatever, goes through a cycle of changes, and finally returns to its original condition again, then the internal work done in the whole process must cancel itself exactly. For if any definite amount, positive or negative, of internal work remained over at the end, there must have been produced thereby either an
28
equivalent quantity of external work or a change in the body's quantity of heat ; and as the same process might be repeated any number of times it would in the positive case be possible to create work or heat out of nothing, and in the negative case to get rid of work or heat without obtaining any equivalent for it both of which results will be at once admitted to be impossible. If then at every return of the body to its original condition the internal work done becomes zero, it follows further that in any alteration whatever of the body's condition the internal work done can be determined from its initial and final conditions, without needing to know the way in which it has passed from one to the other. For if we suppose the body to be brought successively from the first condition to the second in several difierent ways, but always to be brought back to its first condition in exactly the same way, then the various quantities of internal- work done in difierent ways in the first set of changes must all be equivalent to one and the same quantity of internal work done in the second or return set of changes, which cannot be true unless they are all equal to each other. must therefore assume that the internal forces have an Ergal, which is a quantity fully determined by the existing condition of the body at any time, without its being requisite for us to know how it arrived at that condition. Thus the internal work done is ascertained by the increment of the Ergal, which we will call J; and for an indefinitely small change of the body the differential dJ of the Ergal forms the expression for the internal work, which agrees with the notation employed in equations 2 and II. If we now turn to the external work, we find the state of things wholly difi^erent. Even when the initial and final conditions of the body are given, the external work can take
;
We
very different forms. To show this by an example, let us choose for our body a Gas, whose condition is determined by its temperature t and volume v, and let us denote the initial' values of these by t^,v^, and the final values by %,v^; let us also assume that t^>t^, and v^> v^. Now if the change is carried out in the following way, viz. that the gas is first expanded, at the temperature \, from v^ to v^, and then is heated, at the volume v^, from <j to t^, then the external work will consist in overcoming the external pressure which
29
corresponds to the temperature f, On the other hand, if the change is can-ied out in the following way, viz. that the gas is first heated, at the volume Vj, from t^ to t^, and is then expanded, at the temperature t^, from v^ to v^, then the external work will consist in overcoming the external pressure which corresponds to the temperature t^. Since the latter pressure is greater than the former, the external work is greater in the second case than in the first; Lastly, if we suppose expansion and heating to succeed each other in stages of any kind, or to take place together according to any law, we continually obtain fresh pressures, and therewith an endless variety in the quantities of work done with the same
initial
is as follows. Let us take a given quantity of a liquid at temperature t, and transform it into saturated vapour of the higher temperature t^. This change can be carried out either by heating the liquid, as a liquid, to t^, and then vaporizing it; or by vaporizing it at i, and then heating the vapour to t^, compressing it at the same time sufficiently to keep it saturated at temperature t^; or finally by allowing the vaporization to take place at any intermediate temperature. The external work, which again shows itself in overcoming the external pressure during the alteration of volume, has different values in all these different
cases.
The difference in the mode of alteration which, by way of example, has been thus described for two special classes of bodies, may be generally expressed by saying that the body can pass 6y different paths from one condition to the
other.
Another difference, besides this, may come into play. If a body in changing its condition overcomes an external resistance, the latter may either be so great that the full force of the body is required to overcome it, or it may be less than this amount. Let us again take as example a given quantity of a gas, which at a given temperature and volume possesses a certain expansive force. If this gas expands, the external resistance which it has to overcome in so doing must clearly be smaller than the expansive force, or it would not overcome it but the difference between them may be as small as we please, and as a limiting case we may assume them to be
;
80
equal.
These may also however be cases in which this difference is a finite quantity more or less considerable. If e.g. the vessel, in which the gas is at first confined with a given force of expansion, is suddenly put in communication with some space in which a smaller pressure exists, or with a vessel which is entirely empty, then the gas in its expansion overcomes a less external resistance than it has the power of overcomipg, or in the second case no external resistance at and it performs in so doing a smaller amount of external all work than it might perform, or in the second case no external work whatsoever. In the original case, where pressure and reaction are at each instant equal, the gas may be compressed back again by exactly the same force which it has overcome in expanding. If however the resistance overcome is less than the force of expansion, the gas cannot be compressed back again by the same amount of force. The distinction may be expressed by saying that the expansion is reversible in the first case, and not reversible in the second. This mode of expression may be employed in other cases, where changes of condition take place in the overcoming of aiiy kind of resistance, and the distinction just mentioned in relation to the external work may be generally described as follows: with a given change of condition the external work may differ in amount, according as the change takes place in
;
reversible or
a non-reversible manner.
Energy of the Body. In addition to the two differentials dJ and dW, which depend on the work done, we have on the right-hand side of equation (II) a third, which is the differential of H, the total
8.
heat actually existing in the body, or its quantity of Heat. This quantity has clearly the property, also mentioned as belonging to J, that it is known as soon as the condition of the body is given, without needing to know the way in which the body has arrived at that condition. Since the heat existing in the body and the internal work are on the same footing as regards the above most important property, and since further, on account of our ignorance as to the internal work, we generally do not know the several amounts of these two quantities but only their
31
sum, the author, in his first Paper on Heat, published in 1850, combined the two under one designation. Following the same system, we will put
U=H+J.
which changes equation
(II) into
(3),
dQ = dU+I)W
The
function U,
first
(III).
mentioned paper, has been since adopted by other writers on Heat, and as the definition given by him that starting from any given initial condition it expresses the sum of the increment of the heat actually existing and of the heat consumed in internal work is somewhat long, various attempts have been made at. a shorter designation. Thomson, in his paper of 1851*, called it the mechanical energy of a body in a given state: Kirchoff has given it the name Function of
'
Activity' (Wirkungsfunction)-!": lastly Zeuner, in his 'Grundzuge des mechanischen Warmetheorie,' published 1860, has called the quantity U, when multiplied by the heat-equivalent of work, the Interior Heat of the body. This last name (as remarked in the author's former work of 1864) does not seem quite to correspond with the meaning of U; since only one part of this quaptity stands for heat actually existing in the body, i.e. for vis viva of its molecular motion, while the other part consists of heat which has been consumed in doing internal work, and therefore exists as heat no longer. In his second edition, published 1866, Zeuner has made the alteration of calling tl the Internal Work of the body. The author however is unable to accept this name any more than the other, inasmuch as it appears to be just Of the other as objectionably limited on the other side. two names, that of Energy, employed by Thomson, appears very appropriate, since the quantity under consideration corresponds exactly with that which is denoted by the same word in Mechanics. In what follows the quantity tT will therefore be called the Energy of the body.
' '
There
still
and of
Tram. Royal
t Pogg.
Amk
32
the Energy which comprises the Ergal. Since the Ergal expresses the work which the internal forces must have performed, while the body was passing from any initial condition, taken as the starting point, to its condition at the moment under consideration, we can only determine completely, the value of the Ergal for the present condition of the body, when we have previously ascertained once for all its initial condition. If this has not been done, we must conceive the function which expresses the Ergal as still containing an indeterminate constant, which depends on the initial condition. It will be obvious that it is not always necessary actually to write down this constant, but that we may conceive it as included in the function, so long as this latter is designated by a general symbol. Similarly we must conceive another such indeterminate constant as included in the other symbol which expresses the Energy of the body.
9.
Cyclical
processes.
conceive the equation (III), which relates to an change of condition, to be integrated for any given finite change, or for a series of successive finite changes, the integral of one term can be determined, at once. Eor the energy U, as mentioned above, depends only on the condition of the body at the moment, and not on the way in which it has arrived at that condition. If then we put U^ and CTjj for the initial and final values of U, we may write
If
indefinitely small
we
/
written
dU=U-U,.
(III)
may be
(4);
jdQ=
or if
U^- U^+jdTV.
integrals
j
dQ
and
idW
which occur in
this equation, and which represent respects ively the total heat imparted tp the body during, the change, or series of changes, and the external work done, then the
equation will be
Q= U,- U^+W.
(4a).
33
As a special case, we may assume that the body undergoes a series of changes such that it is finally brought round to its initial condition. To such a series the author gave the name of cyclical process. As in this case the initial and final conditions of the body are the same, U^ becomes equal to Z7j, and their difference to zero. Hence for a cyclical process equations (4) and (4a) become:
jdQ = dW.
f
(5),
Q=^W
(5a).
Thus in a cyclical process the total heat imparted to the body (i.e. the algebraical sum of all the several quantities of heat imparted in the course of the cyx;le, which quantities may be partly positive, partly negative) is simply equal to the total amount of external work performed.
10.
stance,
Total Heat
Latent and
Specific Heat.
In former times, when heat was considered to be a suband when it was assumed that this substance might exist in two different forms, which were distinguished by the .terms free and latent, a conception was introduced which was often made use of in calculations, and which was called the total heat of the body. By this was understood that quantity of heat which a body must have taken up in order to pass from a given initial condition into its present condition, and which is now contained in it, partly as free, partly as latent heat. It was supposed that this quantity of heat, if the initial condition of the body was known, could be completely determined from its present condition, without taking into account the way in which that condition had been reached. Since, however, we have obtained in equation (4a) an expression for the quantity of heat received by the body in passing from its initial to its final condition, which expression
contains the external work W, we must conclude that this quantity of heat, like the external work, depends not only on the initial and final conditionSj but also on the way in which the body has passed from the one to the other. The conception of the total heat as a quantity depending only on the present
c.
34
condition of the body is therefore, under the new theory, no longer allowable. The disappearance of heat during certain special changes of condition, e.g. fusion and vaporization, was formerly explained, as indicated above, by supposing this heat to pass into a special form, in which it was no longer sensible to our touch or to the thermometer, and in which it was therefore called Latent Heat. This mode of explanation has also been opposed by the author, who has laid down the principle that all heat existing in a body is appreciable by the touch and by the thermometer that the heat which disappears under the above changes of condition exists no longer as heat, but and that the heat which has been converted into work makes its appearance under the opposite changes (e,g. solidification and condensation) does not come from any concealed source, but is newly produced by work done on the body. Accordingly he. has proposed the term Work-heat as a substitute for Latent heat in general cases. This work, into which the heat is converted, and which in the opposite class of changes produces heat, may be of two kinds, internal or external. If e.g. a liquid is vaporized, the cohesion of its molecules must be overcome, and, since the vapour occupies a larger space than the liquid, the external pressure must be overcome also. In accordance with these two divisions of the work we also may divide the total workheat, and call the divisions the internal and external work; ;
heat respectively. That quantity of heat which must be imparted to a body in order to heat it simply, without making any change in its density, was formerly known under the general name of free heat, or more properly, of heat actually existing in the body a great part of this, however, falls into the same category as that which was formerly called latent heat, and for which the term work-Jieat has been proposed. For the heating of a body involves as a general rule a change in the arrangement of its molecules, which change produces in general an externally ^perceptible alteration of volume, but still may take place apart from such alteration. This change of arrangement requires a certain amount of work, which may be partly internal, partly external; and in doing this, work-heat is again consumed. The heat applied to the body thus serves
35
in part only to increase the heat actually existing, tHe other part serving as work-heat. On these principles the author attempted to explain (by way of example) the unusually great specific heat of water, which is much beyond that either of ice or of steam*: the assujnption being that of the quantity of heat, which each receives from without in the process of heating, a larger portion is consumed in the case of water in diminishing the cohesion of the particles, and thus serves as work-heat.
From the foregoing it is seen to be necessary that, in addition' to the various specific heats, which shew how much heat must be imparted to one unit-weight of a body in order
it through one degree under different circumstances the specific heat of a solid or liquid body under ordinary atmospheric pressure, and the specific heat of a gas at constant volume or at constant pressure), we must also take into consideration another quantity which shews hy how much the heat actually existing in one unit-weight of a substance (i.e. the vis viva of the motion of its ultimate pojrticles) is increased when the substance is heated through one degree of temperature. This quantity we will name the body's true heat-capacity. It would be. advantageous to confine this term 'heat-capacity (even if the word true be not prefixed) strictly to the heat actually existing in the body ; whereas for the total heat which must be imparted for the purpose of heating it under any given circumstances, and of which work-heat forms a part, the expression 'specific heat' might be always employed. As however the term 'heat-capacity' has hitherto been usually taken to have the same signification as specific heat' it is still necessary, in order to affix to it the above simplified meaning, to add the epithet ' true,'
to
warm
(e.g.
'
'
'
'
11,
case.
'
'
$W. No
(III) the external work is denoted generally by special assumption is thereby made as to the nature of the external forces which act on the body, and on which the external work depends. It is, however, wprth while to
In equation
p. 23.
32
36
OIT
and which leads to a very simple determination of the external work, viz. the case where the only external force acting on the body, or at least the only force which needs to be referred to in the determination of the work, is a pressure acting on the exterior surface of the body and in which this pressure (as is always the case with liquid and gaseous bodies, provided no other forces are acting, and "which may be the case even with solid bodies) is the same at all points of the surface, and everywhere normal to it. In this case there is no need,
;
in order to determine the external work, that we should consider the body's alterations in form and its expansion in particular directions, but only its total alteration in volume. As an illustrative case, let us take a cylinder, as shewn in Fig. 1, closed by an easily moving piston P, and containing
some expansible substance, e.g. a gas, under a pressure per unit-area represented by p. The section of the cylinder, or the area of the piston, we may call a. Then the total pressure which acts on the piston, and which must be overcome in raising it, is pa. Now if the piston stands originally at a height h above the bottom of the cylinder, and is then lifted through an indefinitely small distance dh, the external work performed in the lifting will be expressed by the equation
But
V
Kg. 1. V be the volume of the gas we have ah, and therefore dv = adh ; whence the above equation
if
dW=padh.
becomes
dW=pdv
This same simple form
is
(6),
assumed by the
differential of the
external work for any form of the body, and any kind of expansion whatever, as
may be easily shewn as follows. Let the full line in Fig. 2 represent the surface of the body in its original condition, and the dotted line
its
surface
after
an
indefi-
Fig. 2.
S7
and volume. Let us consider element dca of the original surface at the point A. Let a normal drawn to this element of surface cut the second surface at a distance du from the first, where du is taken as positive if "the position of the second, surface
.
is
surface,
and negative
us suppose an indefinite number of such normals to be drawn through every point in the perimeter of the surface-element dm to the second surface ; there will then be marked out an indefinitely small prismatic space, which, has deo as its base, and du as its height, and whose volume is therefore expressed by dadu. This indefinitely small volume forms the part of the increase of volume of the body corresponding to the element of surface da. If then we integrate the expression dcodu all over the surface of the body, we shall obtain the whole increase in volume, dv, of the body, and if we agree to express integration over the surface by an integral sign with
inside.
let
Now
suffix
CO,
we may
write
dv-
dudm
ta
(7).
Now
surface
if,
as before,
we denote the
dtu will be pdm. Therefore the part of the external work, which corresponds to the element dw, and is described by saying that the element under the action of the external force pdm is pushed outwards at right angles through the distance du, will be expressed by the i^rodact pdm du. Integrating this over the whole surface, we obtain for the total external work,
dW =
As p
is
pdudm.
ia'
may be
written
dW=p\
J
ta
dudm,
or,
by equation
is
(7),
dW=pdv,
which
the same as equation
(6)
given above.
38
this equation, we may give to equation (III) the case in which the only external force is a uniform pressure normal to the surface, the following form
for
dQ = dU+pdv
This
last equation,
(IV).
sion most in use for the first main principle of the Mechanical Theory of Heat, we will in the next place apply to a class of bodies, which are distinguished for the simplicity of their
liarly
for which the equation takes accordingly a pecusimple form, so that the required calculations can be easily performed.
laws,
and
(39)
CHAPTER
ON PEBFECT
1.
II.
GASES.
Among the laws which characterize bodies in the gaseous condition the foremost place must be given to those of Mariotte and Gay Lussac, which may be expressed together in a single equation as follows. Given a unit- weight of a gas, which at freezing temperature, and under any standard pressure p^ {e.g. that of the atmosphere) has the volume j then if p and v be its pressure and volume at any temperature t (in Ceritigrade measure) the following equation will
hold:
pv=-p^v^{\
a.t)
(1),
wherein the quantity a, which is usually termed the coefficient of expansion, although it really relates to the change of pressure as well as the change of volume, has one and the same value for every kind of gas. Regnault has indeed recently proved by careful experiment that these laws are not strictly accurate; but the deviations are for permanent gases very small, and become of importance only for gases which are capable of condensation. It seems to follow that the laws are the more nearly exact, the further a gas is removed, as to pressure and temperature,^ from its point of condensation. Since for permanent gases under ordinary conditions the exactness of the law is already, so great, that for most purposes of research it may be taken
>
as perfect, we may imagine for every gas an ultimate condition, in which the exactness is really perfect; and in what follows we will assume this, ideal condition to be actually
40
reached, calling for brevity's sake all gases, in which this is assumed to hold, Perfect Gases. As however the quantity a, according to Eegnault's determinations, is not absolutely the same for all the gases which have been examined, and has also somewhat different values for one and the same gas under different conditions, the question arises, what value we are to assign to a in the case of perfect gases, in which such differences can no longer appear. Here we must refer to the values of a which have been found to be correct for various permanent gases. By experiments made on the system of increasing the pressure while keeping the volume constant, Regnault found the following numbers to be correct for various permanent gases
Atmospheric Air
Hydrogen
Nitrogen Carbonic Oxide
'
differences here are so small, that it portance what choice we make ; but as it
The
spheric air that Regnault made the greatest number of experiments, and as Magnus was led in his researches to a precisely similar result, it appears most fitting to select the
number 0003665.
Regnault, however, by experiments made on the other system of keeping the pressure constant and increasing the volume, has obtaitied a somewhat different value for a in the He has further case of atmospheric, air, viz. 0'003670. observed that rarefied air gives a somewhat smaller, and compressed air a somewhat larger, coefficient of expansion than This latter circumstance has led air of ordinary density.
some physicists to the conclusion that, as rarefied air is nearer to the perfect gaseous condition than air of ordinary density, we ought to assume for perfect gases a smaller value than 0003665. Against this it may be urged, that Regnault observed no such dependence of the coefficient of expansion on the density in the case of hydrogen, but after increasing the density threefold obtained exactly the same value as before ; and that he also foud that hydrogen, in its deviation from the laws of Mariotte and Gay Lussac, acts altogether differently, and ior the, most 'part in exactly the opposite
ON PERFECT
GASES.
'
41
way, from atmospheric air. In. these circumstances the author considers that additional weight is given to the result taken above from the figure for atmospheric air; since it will hardly be questioned that hydrogen is at least as near as atmospheric air to the condition of a perfect gas, and therefore in drawing conclusions relative to that condition the behaviour of the one is as much to be noted as that of the other. It appears therefore to be the best course (so long as fresh observations have not furnished>a more satisfactory starting point for wider conclusions) to adhere to the figure which, under the pressure of one atmosphere, has been found to agree almost exactly for atmospheric air and for hydrogen and thus to write
a = 0003665 =3j^
If
(2).
we denote
the he reciprocal - by a
:
we may
equation thus
pv = Pf{a^-t)
(3).
And
if for
brevity
we put
i?=^5!!o
(4)
T=a+t
we then obtain the equation
in the form
.........(5),
pv = RT
.B is here
(6).
a constant which depends on the nature of the gas inversely proportional to its specific gravity*. represents the temperature, provided this is measured not from the freezing point, but from a zero point lying a degrees
and
is
lower.. The temperature thus measured from term the Absolute Temperature, a name which
a
will
we
shall
be more
* For S is proportional to the volume of a unit of weight of the gas at standard pressure and temperature and is therefore inyersely proportional to the wgight of a unit of -volume, i.e, to the specific gravity. [Translator^)
;
42
in equation (2)
we obtain
=l= T=273 + t]
<^
M
to
(7).
2.
Approadmate Principle as
In an experiment of Gay Lussac's, a vessel filled with air was put in communication with an exhausted receiver of equal size, so that half the air from the one passed over into the other. On measuring the temperature of each half, and comparing it with the original temperature, he found that the air which had passed over had become heated, and the air which remained behind had become cooled, to exactly the same degree; so that the mean temperature was the same
after the expansion as before.
He thus proved that in this kind of expansion, in which no external work was done, no loss of heat took place. Joule, and after him Regnault, carried out similar experiments with greater care, and both were led to the same result. The principle here involved may also be deduced, without reference to special experiments, from certain properties of gases otherwise ascertained, and its accuracy may thus be checked. Gases shew so marked a regularity in their behaviour (especially in the relation between volume, pressure, and temperature, expressed by the law of Mariotte and Gay
Lussac), that we are thereby led to the supposition that the mutual action between the molecules, which goes on in, the interior of solid and liquid bodies, is absent in the case of gases ; so that heat, which in the former cases has to overcome the internal resistances, as well as the external pressure, in order to produce expansion, in the case of gases has to do with external pressure alone. If this be so, then, if a gas expands at constant temperature, only so much heat can thereby be absorbed as is required for doing the external work. Again, we cannot suppose that the total amount of heat actually existing in the body is greater after it has expanded at constant temperature than before. On these assumptions we obtain the following principle a permanent gas, if it expands at a constant temperature, absorbs only
;
ON PEEFECT GASES.
43
SO much heat as is required for the eootemal work which it performs in so doing. We cannot of course give to this principle any greater validity than that of the principles from which it springs, but must rather suppose that for any given gas it is true to the same extent only in which the law of Mariotte and Gay Lussac is true. It is only for perfect gases that its absolute accuracy may be assumed. It is on this understanding that the author brought this principle into application, combined it as an approximate assumption with the two main principles of the Mechanical Theory of Heat, and used it for establishing more extended conclusions. More recently Mr, now Sir William Thomson, who at first did not agree with one of the conclusions so deduced, undertook in conjunction with Joule to test experimentally the accuracy of the principle*; and for this purpose instituted with great care a series of skilfully conceived experiments, whichj on account of their importance, will be more fully and exactly discussed further on. These have completely confirmed the truth not only of the general principle, but also of the remark added by the author as to its degree of exactness. In the permanent gases on which they experimented, viz. aianospheric air and hydrogen, the principle was found so nearly exact that the deviations might for the purpose of most calculations be neglected; while in the non-permanent gas selected for experiment (Carbonic Acid) somewhat greater deviations were observed, exactly as might have been expected from the behaviour of that gas in other respects. After this we may with the less scruple apply the principle, as being exact for actually existing gases in the same degree as the law .of Mariotte and Gay Lussac, and absolutely
first
3.
On
the
Form
We now
dQ = dU+pdv,
in order to apply it to the case of a perfect gas, of assume as before one unit of weight to be given.
which we
44
The condition of the gas is completely determined, when temperature and volume are known ; or it may be determined by its temperature and pressure, or by its volume and We will at present choose the first-named quanpressure. tities, temperature and volume, to determine the condition, and accordingly treat T and v as the independent variables, on which all other quantities relating to the condition of the of the gas as gas depend. If then we regard the energy being also a function of these two variables, we may write
its
dr
v^hence equation (IV) becomes
dv
dU
,^. fdU
^^-^'^-^if^^y^
for a gas,
()
This equation, which in the above form holds not only but for any body whose condition is determined' by its temperature and volume, may be considerably simplified for gaseous bodies, on account of their peculiar properties.
The quantity of heat, which a gas must absorb in expanding at constant temperature through a volume dv, is
generally denoted
by
-r^ dv.
tion of the last Section this heat is equal to the work done in the expansion, which is expressed by pdv, we have the
equation
dQ,
But from equation
(8)
dQ
dQ^dV^
dv dv
last
f-
()
ON PERFECT
GASES.
45
Hence we conclude that in a perfect gas the energy is independent of the volume, and can only be a function of
the temperature.
77r
If in equation (8)
j-~,
we put
-5-
= 0,
and substitute
for
it
becomes
(10).
dQ= C,dT+pdv
From the form of this equation we see that C denotes the Specific Heat of the Oas at constant volwme, since GjiT expresses the quantity of heat which must be imparted to the gas in order to heat it from 2' to T+dT, when du is
equal to zero.
As
this Specific
Heat
-tj,, i.e. is
the
differ-
ential coefl&cient with respect to temperature of a function of the temperature only, it can itself also be only a function of temperature. In equation (10) all the three quantities T, v, and p are found but since by equation (6) pv Rt, it is easy to eliminate one of them ; and by eliminating each in succession we obtain three different forms of the equation.
;
Eliminating p
we
obtain,
dQ'=C,dT + -^d'B
Again, to eliminate v
(11).
we put v =^
7? 7'
;
whence we have
dv =
dT^T-
ip.
If we substitute this value of dv in equation (10), and then combine the two terms of the equation which contain dT, we obtain
dQ =
differentiation,
{C,
+ B)dT-?^dp
we
obtain from equation
(12).
Lastly, to eliminate T,
(6),
by
-,_vdp+pdii
4G
dQ=^vdp + ^^pdv
(13).
4. Deductions as to the two Specific Heats, and transformation of the foregdi ig equations. ^^"VtrfiAKfju^
In the same way as We see from eqiwiion (10) that the quantity C,, which appears as factoi/of dT, denotes the specific heat at constant '^feempgratuje, we may see from equation (12) that the factorof d!Z in that equation, viz. If C, + R, expresses the Specific Heat at constant pressiire. therefore we denote this Specific Heat by G^ we may put
(7^=a,
Heats. Since
+ i2....
'....;
(14),
When the author first drew in this manner from the Mechanical Theory of Heat the conclusion that the two Specific Heats of a permanent gas must be independent of its density, or in other words of the pressute to which it is subjected, and could depend only on its temperature and when he added the further remark that they were thus in all probability constant be found himself in opposition to the then prevailing views on the subject. At that time it was considered to be established from the experiments of Suermann, and from those of de la Roche and B^rard, that the specific heat of a gas depended on the pressure and the
; ; ;
circumstance that the new theory led to an opposite conclusion produced mistrust of the theory itself, and was used by A. von Holtzmann as a weapon of attack against it. Some years later, however, followed the first publication of the splendid experiments of Eegnault on the specific heat of gases*, in which the influence of pressure and temperature on the specific heat was made a subject of special
* Comptes Vol. II.
Bendm,
Vol. xxxn.,
ON PERFECT
GASES.
47
investigation. Regnault tested atmospheric air at pressures from 1 to 12 atmospheres, and hydrogen at from 1 to 9 atmospheres, but could detect no difference in their specific
heats.
'
He
tested
them
between - 30 and + 10, between 0 and 100, and between 0 and 200 ; and here also he found the specific heatt always the same *. The result of his experiments may thus be expressed by saying that, within the limits of pressure and temperature to which his observations extended, the specific heat of permanent gases was found to be constant. It is true that these direct explanatory researches were
confined to the specific heat at constant pressure; but there be little scruple raised as to assuming the same to be correct for the other specific heat, which by equation (14) differs from the former only by the constant B. Accordingly in what follows we shall treat the two, specific heats, at least for perfect gases, as being constant quantities. By help of equation (14) we may transform the three equations (11), (12) and (13), which express the first main principle of the Mechanical Theory of Heat as applied to gases, in such a way that they rnay contain, instead of the Specific Heat at constant volume, the Specific Heat at constant density which may perhaps appear more suitable, since the latter, as being determined by direct observation, ought The resulting to be used more frequently than the former. equations are
will
;
.(15).
we may
* The numbers obtained for atmospheric air (Be/, des Exp. Vol. are as follows in ordiuary heat units 0'23771, between - 30 and + 10 00 0-23741, I'OO" 208 0-23751, O" which may be taken as practically the same.
,
p. 108)
48
to^ and
v,
as follows
.(16).
G
In the above equations the speciiic heats are expressed If we wish to express them in ordinaryheat units, we have only to divide these values by the Mechanical Equivalent of Heat. Thus if we denote the specific heats, as expressed in ordinary heat units, by c, and Cj,, we may put
in mechanical units.
.=
^=^
to equation (14),
(17)-
and dividing
= c, + |,
(18).
5. Relation between the two Specific Heats, and its application to calculate the Mechanical Equivalent of Heat.
If a system of Sound-waves spreads itself through any gas, atmospheric air, the gas becomes in turn condensed and rarefied; and the velocity with which the sound spreads depends, as was seen by Newton, on the nature of the changes For very of pressure produced by these changes of density. small changes of density and pressure the relation between the two is expressed by the differential coefficient of the pressure with respect to the density, or (if the density, i.e. the weight of a unit of volume, is denoted by p) by the differential
e.g.
coefficient -y-.
Applying
this principle
we
we
in which
= V;
g represents
<'^''
ON PERFECT
GASES.
49'
Now
coefficient
to
other.
He
therefore put
= constant,
whence by
differentiation
pdp-pdp
.-
and therefore
t-l
whence
(19)
m:
.(21).
becomes
-\/:-
The velocity calculated by this formula did not however agree with experiment, and the reason of this divergence, after it had been long sought for in vain, was at last discovered by
Laplace.
The law of Mariotte in fact holds only if the change of density takes place at constant temperature. But in sound vibrations this is not the case, since in every condensation a heating of the air takes place, and in every rarefaction g, cooling. Accordingly at each condensation the pressure is increased, and at each rarefaction diminished, to a greater extent than accords with Mariotte's law. The question now
arises
jof -y-
dp be determined.
Since the condensations and rarefactions follow each other with great rapidity, the exchange of heat that can take place during each short period between the condensed and rarefied parts of the gas must be very small. Neglecting this, we have to do with a change of density, in which the quantity of gas under consideration receives no heat and gives forth none; and we may thus, in applying to this case the differential equa*
Tkis law
c.
is
commonly known
in (Translator.)
England as
'
So
dQ = 0,
Hence,
e.g.
from the
last
Qf equations
(16),,
we
olstain:
or
G^vdp
since the
+ Oj,pdv = 0.
18
Now,
the re-
we may put
= -,
r
and therefore
,
dv =.
j^
p
or
p
(22).
# = p2
differs
dp
from that
deduced from Mariotte's law, and given in (20), by containing as factor the ratio of the two Specific Heats. If for simplicity we put
^=
the last equation becomes
(23),
= hP ^ dp p
Substituting this value of
of (21)
(24),
-^
in equation (19),
we
get instead
= \/%^
(25)v
From
if is
sound u can be calculated known or, on the other hand, if the velocity of sound known by experiment, we can apply the equation to calcuthis equation the' velocity of
;
is
late k,
changing
it firs':
h^"^^
gp
(26). J ^
ON PEBFECT
GASES.
51
The velocity of sound iu air has been several times determined with great care by various physicists, -whose results
agree with each other very closely. According to the experiments of Bravais and Martens* the velocity at freezing temperature is 332"4 m. per second (10906 feet). We will inSiert this value in equation (26). We may also give g its recognized value 9'809 m. (32-2 feet). To determine the
quotient -
we may give
the pressure
any value we
please,
but we must then assign to the density p the value corresponding to that pressure. We will assume p to be the pressure of 1 atmosphere. This must be expressed in the fonnula by the amount of weight supported per unit of surface. As' this weight is equal to that of a column of quicksilver, whose base is 1 sq. m. and height 760 mm., and which therefore has a volume of 760 cubic decimetres, and as,
according to Regnault, the Specific Weight of quicksilver at 0, as compared with water at 4, is 13"696, we obtain
p = l atmosphere = 760 x
Lastly,
13"596
= 10333
the weight of a cubic metre of air under the 1 atmosphere and at temperature 0, Substituting which, according to Regnault, is 1'2932 kg. these values in equation (26) we obtain
is
assumed pressure of
*~
(332-4)-
1-2 932,
9-809x10333
~*^"-
air,
i.e.
Ha-ving thus determined the quantity h for atmospheric we can now use equation (18) to calculate the quantity E, the Mechanical Equivalent of Heat, as was first done by
Mayer.
(18)
and,
if
is
the
Ann. de Chim.
iii.,
13, 5;
and Fogg.
p. 351.
42
52
C
-^ for
we have
''^-
^-of%-:
Here we may substitute
and
for for c^ its value as given
To do
this, let
us
agam
take Pn as the pressure of 1 atmosphere, which, as seen above, is equal to 10333, and we then have for v the volume in cubic metres of 1 kg. of air under the, above pressure of 1 atmosphere and at temperature 0", which according to Regnault is 0'7733. Lastly we have already assumed the for atmospheric air value of a to be 273. The value of will therefore be given by the equation
P _ 10333 X 07733
273
__
B
we
in equation (27) Substituting these values for k, c^, and obtain 1-410 X 29-27 _ ^ ~ 0-410 X 0-2375 ~*'^'^^-
This figure agrees very closely with that determined by In fact it Joule from the friction of water, viz. 423-55. must be admitted that the agreement is more close than, considering the degree of uncertainty as to the data used in the calculation, we could have had any right to expect; so that chance must have assisted in some degree to produce it. In any case, however, the agreement forms a striking confirmation of the equations deduced for permanent gases.
6.
to the
Specific
Heats of
we
^as
known, we may apply that equation to calculate the specific heat at constant volume from that at constant pressure, which is known from experiment. This application is of special importance, because the method of deducing the ratio of the
ON PERFECT
two
GASES.
53
specific heats from the velocity of sound is only practicable in the case of the very few gases for which that velocity has been experimentally determined. For all others,
equation (18) offers the only means as yet discovered of calculating the specific heat at constant volume from that at constant pressure. It must here be observed that equation (18) is exactly true only for perfect gases, although it gives at least approximate results for other gases. The circumstance has also to be considered, that the determination of the specific heat of a gas at constant pressure is the more difficult, and therefore the value determined the less reliable, in proportion as the gas is less permanent in its character, and thus diverges more widely in its behaviour from the laws of a perfect gas therefore, as there is no need to seek in our calculations a greater accuracy than the experimental values themselves can possibly possess, we may treat the mode of calculation employed as sufficiently complete for our purpose. Accordingly we begin by putting equation (18) in the form
o.
= c,-
(28).
S is determined
=
a
where pv are the pressure and volume at the temperature of freezing. Should it be difficult to make observations on the
gas at this temperature we may also, by equation
(as is
(6),
i?=5
where
p, v,
(29),
T are any three corresponding values of and absolute temperature. This quantity B, as already observed, is only so far dependent on the nature of the gas, that it is inversely proportional to its specific gravity. For if we denote by v' the volume of. a unit of weight of air at temperature T and pressure p.
and
pressure, volume,
54
O'S
we have
Combining
this
with equation
(29),
B = R'^,
V
But
-, is
and is therefore the reciprocal of the ratio of the weights of equal volumes, which ratio is called the Specific Gravity of the gas, as compared with common air. If we call this specific gravity d, the last equation becomes
gases,
^=f
Substituting this value of
(^)-
B in
(28)
we
obtain
(^1)-
'^-'^'-M
The quantity here denoted by R', i.e. the Value of for atmospheric air, has been already determined in 5 to be equal to 29-27. Hence further,
whence the equation, which serves to determine the Heat at constant volume, takes this simple form
:
Specific
''
= ".
0-0691
(32).
we apply
which
d=l, and
this equation to the case of air, for the sake of distinction denote by
air,
we
get the
=c;- 00691
(33),
ON PERFECT
GASES.
55
and substituting for c'^ its value according to Kegnault, which is 02375, we obtain the result
c'.
(34.)*
given in the
^^^^-0:0691
^3.^^
which, as will be seen later, is specially convenient for the application of the values given by Eegnault for specific heats at constant pressure. The specific heats denoted by c^ and c relate to a unit of weight of the gas, and have for unit the ordinary unit of heat, i.e. the quantity of heat required to raise a unit of weight of water from the temperature 0 to 1". may thus say that the gas, in relation to the heat which it requires to raise its temperature either at constant pressure or constant volume, is referred as regards weight to the standard of water. With gases however it is desirable to refer to the standard of air as regards volume ; i.e. so to determine the specific heat, as to compare the quantity of heat, which the gas Requires to raise its temperature through 1, with the quantity of heat which an equal volume of air; taken at the same temperature and pressure, requires to raise its temperature to the same extent. may use this kind of comparison in the case of both the specific heats, inasmuch as we assume in the one case that both the gas under consideration and the atmospheric air are heated at constant pressure, and in the other that they are both heated at constant volume. The specific heats thus determined may
We
We
be denoted by 7^ and 7^. As we denote by v the volume which a unit weight of gas assumes at a given pressure and temperature, the quantity of heat, which a unit-volume of the gas absorbs at constant
pressure in being heated through
It will be Been that cj and
perfect gases;
1,
will
be expressed by
,
-^
e,' fulfil
= 1-410.
{Translator.)
56
by -f
The
specific
heat
latter, or,
^^
= ^x-^;=%x^ = ^cZ
(36).
Similarly
% = ^d c.
we may
;
(37).
becomes
^=
we have
0^
(^)-
In the second we may put for c', according to (34)^ the value 0"1684<, and for c, the expression given in (35); whence
cd- 00691
,Q>
7.
Numerical Calculation of
the Specific
Heat
at con-
stant Volume.
The formulae developed in the last section have been applied by the author to calculate from the values which Eegnault has determined by his researches for the Specific Heat at constant Pressure of a large number of gases and vapours, the corresponding values of the Specific Heat at constant Volume. In so doing he has in some sort recalculated one of the two series of numbers given by Regnault himself; who has expressed the Specific Heat at constant Pressure in two different ways, and has brought together the resulting numbers in two series, one of which is superscribed ' en poids,' and the other ' en volume.' The first series contains the values which result, if the gases in question are compared weight by weight with water, in relation to the quantity .of heat. required to warm them through 1; in other words, the values of the quantities denoted above by Cp. The numbers in the second series are simply obtained from those
ON PERFECT GASES.
57
in the first by multiplying them by the corresponding specific gravity, i. e. they are the values of the product Cpd. These latter numbers were no doubt those most easily calculated from the observed values of c^,; but their signification is somewhat complicated. With them the quantity of heat has for its unit the ordinary unit of heat, whilst the volume to which they refer is that which a unit-weight of atmospheric air assumes, when under the same temperature and pressure as the gas under consideration. The tediousness of the verbal description thus required makes the numbers troublesome to understand and to apply ; moreover this mode of expressing the Specific Heat of gases has been used, so far as the author knows, by no previous writer. In considering gases with reference to volume, it has in all other cases been custoinary to compare the quantity of heat, which a given gas requires to raise its temperature through 1, with the quantity of heat which an equal volume of atmospheric air requires under the same conditions for the same purpose, or, as briefly expressed above, by comparing the gas, volume for volv/me, with air. The numbers thus obtained are remarkable for their simplicity, and allow the laws which hold as to the specific heats of the gas to be treated with special clearness. It will therefore, the author believes, be found an advantage that he has calculated, from the values given by
Eegnault under the heading 'en volume' for the product All that Gj,d, the values of the quantity 7^,, defined above. was required for this, by (38), was to divide the values of c^d
by
0-2375.
tions
has further calculated the values of c and 7,; calculawhich by equations (35) and (39) could be very simply performed, by taking fi:om the values of the product c^d the number 0'0691, and dividing the remainder by d, or by 01684, respectively.
He
The numerical values thus calculated are brought together in the annexed table, in which the different columns have the following signification
Column
I.
gives the
name
of the gas.
Column II. gives the Chemical composition, and this expressed in such a way that the diminution of volume pro-
S8
duced by the combination can be immediately observed. For in each case those volumes of the simple gas are given, which must combine in order to give Two Volumes of the compound gas. Thus we assume for Carbon, as a gas, such an hypothetical volume as we must assume, in order to say that one
volume of Carbon unites with one volume of Oxygen to make Carbonic Oxide, or, with two volumes to make Carbonic acid. Again, when, e.g. Alcohol is denoted in the Table by CjjHgO, this means that two volumes of the hypothetical carbon gas, six volumes of Hydrogen, and one volume of Oxygen, make up together two volumes of Alcoholic vapour. For sulphur-gas the specific gravity used to determine its volume is that found by Sainte-Claire Deville and Troost for very
high temperatures, viz. 2"23. In the five last combinations in the Table, which contain Silicon, Phosphorus, Arsenic, Titanium, and Tin, these simple elements are denoted by their ordinary chemical signs, without reference to their volumes in the gaseous condition, because the gaseous, volumes of these elements are partly still unknown, partly hampered with certain irregularities not yet thoroughly cleared
up.
Column
III. gives
given by Kegnault.
Column IV. gives the Specific Heat at constant Pressure as compared, weight for weight, with water, or in other words referred to a unit-weight of the gas and expressed in ordinary
These are the numbers given by Regnault under the heading en poids.'
units of heat.
'
Column V. gives the Specific Heat at constant Pressure compared, volume for volume, with air, calculated by dividing by 0'2375 the numbers given by Regnault under the heading ' en volume.' Column VI. gives the Specific Heat at constant Volume compared, weight for weight, with water, calculated by equation (35).
Column VII.
(39).
Heat
air,
ON PERFECT
GASES.
59
n.
III.
IV.
Specific heat at con-
VI.
VII.
Name
of the Gaa.
Chemical Composition.
stant Pressure
stant
Volume
Density.
compared compared weight for compared weight for volume for weight volume for weight volume volume with with with with Air, Air. Water. "Water.
compared
Atmosi)]ierio Air
Oxygen
Nitrogen
Hydrogen
Chlorine
&.
Br,
Bromine
Nitric Oxide
NO
CO
HCl
CO,
CarbonicOxide
Hydrochloric Acid.... Carbonic Acid Nitric Acid
Steam
"
Sulphuric Acid SOa Hydro-sulphuric Acid. H^S OSj Carbonic di-sulphide Carburetted Hydrogen OH4
1-1056 0-9713 0-0692 2-4S02 5-4772 1-0384 0-9673 1-2596 1-5201 1-5241 0-6219 2-2113
.
1-1747,
Chloroform
Olefiant
Gas
CHCI3 CjH^
Ammonia
Benzine
Oil of Turpentine. Wood Spirit
NH3
6
fl
CH^O
C^HeO
C^HioO
C^Hju S
Alcohol
0,H,C1 CaHsBr
CaH^CI,
CjHjO
C.HsOa
S1CI3
2-6258 0-5527 4-1244 0-9672 0-5894 2'6942 4-6978 1-1055 1-5890 2-5573 3-1101 2-2269 3-7058 3-4174 2-0036 3-0400 5-8833
0-2375 0-21751 0-24380 3-40900 0-12099 0-05552 0-2317 0-2450 0-1852 0-2169 0-2262 0-4805 0-1544 0-2432 0-1569 0-5939 0-1567 0-4040 0-5084 0-3754 0-5061 0-4380 0-4534 0-4797 0-4008 0-2738 0-1896 0-2293 0-4125 0-4008 0-1322
0-1347
1-013 0-997 0-993 1-248 1-280 1-013 0-998 0-982 1-39 1-45 1-26 1-44 1-20 1-74 1'38 2-72
1-75 1-26
4-26 10-01 2-18 3-03 5-16 5-25 2-57 2-96 3-30 3-48 5-13 3-27
2-69
0-1684 0-1551 0-1727 2-411 0-0928 0-0429 0-1652 0-1736 0-1304 0-172 0-181 0-370 0-123 0-184 0-131 0-468 0-140 0-359 0-391 0-350 0-491 0-395 0-410 0-453 0-379 0-243 0-171 0-209 0-378 0-378 0-120 0-120
0-101
1 1-018
0-996 0-990 1-350 1-395 1-018 0-997 0-975 1-55 1-64 1-36 1-62 1-29 2-04 1-54 3-43 2-06 1-37 5-60 13-71 2-60 3-87 6-87 6-99 3-21 3-76 4-24 4-50 6-82 4-21
phorus
PCI3
4-7464
6-2667 6-6402 8-9654
3-39
3-77
0-1122
0-1290
3-61
tanium
TiCI^
0-119
0-086
4-67
4-59'
Tetra-chloride of Tin.
SnCl^
0-0939
3-54
60
8. Integration of the Differential Equations which express the first main Principle in the case of Oases.
The differential equations deduced in sections 3 and 4, which in various forms express the first main principle of the Mechanical Theory of Heat in the case of gases, are not immediately integrable, as can be seen by inspection; and must therefore
be treated after the method developed in 3 of the Introduction. In other words, the integration becomes possible as soon as we subject the variables occurring in the equation to some one condition, thus determining the path of the change of condition of the body. We shall here give only two very simple examples of the process, the results of which are important for our further investigations.
Pressure,
Example 1. The gas changes its volume at Constant and the quantity of heat required for such change is
known. In this case we select from the above equations one which contains p and v as independent variables, e.g. the last of Equations (15), which is
dQ = ^^vdp + ^piv.
As the
dp
dQ = -^p^dv,
which gives on integration
oi v)
<3
(if
we
call v^
= Fi(-i)
(40).
Example 2. The gas changes its volume at Constant Temperature, and the quantity of heat required for such change is known. In this case we select an equation which contains y and v as independent variables, e.g. Equation (11), which is
dQ = C,dT +
dv.
ON PERFECT GASES.
61
As Tis to be we have
constant,
we put
T = T^,
1
and
dT =
whence
Integrating,
.Q = RT>g^
Hence
(41).
is_ derived the Principle that if a Gas changes its volume without change of temperature, the quantities of heat
series,
while the
Again,
if
we put
for
R its
value
^^
we have
Q=p,v,log^
_
(42).
If we suppose this equation.to refer, not directly to a unit weight of the gas, but to a quantity of it such that at pressure p it assumes a volume v^, and then suppose that this volume changes under constant temperature to v, then the equation contains nothing which depends on the special nature of the gas. Therefore the quantity of heat absorbed is independent of the nature of the gas. Further, it does not depend on the temperature, but only on the pressure, being proportional to the original pressure.
Another application of the differential equations deduced in sections 3 and 4 consists in making some assumption as to the heat to be imparted to the gas during its change of condition, and then enquiring what course the change of The simplest condition will take under such circumstances. and at the same time most important assumption of this kind is that no heat whatever is imparted to or taken from the gas
during its change of condition. For this purpose we may imagine the gas confined in a vessel impermeable to heat, or that the change is so rapid that no appreciable heat can pass to or ffom the gas in the time.
62
On
becomes
this
dQ = 0.
Let us do
of these
Then the
first
C,dT+{G^-G:)^dv = 0.
Dividing by
Tx
(7
h,
we have
f+(*-i)?=o.
Integrating,
log
or
T+(k-l)logv = Const.
Tv^^
= Const.
v,
we may
eliminate the
.(43).
If this equation be applied for example to atmospheric air, then, writing k= 1"410, we can easily calculate the change of temperature which corresponds to any given change of volume. If e.g. we assume a certain quantity of air to be taken at freezing temperature and at any pressure whatever, and to be compressed, either in a vessel impermeable to heat, or with great rapidity, to half its volume, then Tj = 273
(absolute temperature)
and
=2
comes
^'=2^' =
whence
or
if t
1-329,
T = 273 x
1-329
= 363,
point,
T- 273 = 90.
If a similar calculation is made for the compression of the gas to J and -^ of its original volume, results are obtained,
ON PEEFECT GASES.
^3
which, combined with the former, are presented in the following Table.
Value of
64
9.
the change
There is one quantity connected witli the expansion of a gas which still requires to be specially considered, viz. the External Work done in the process. The element of this work, as determined in Equation (6), Ch. I, is
d
This work
W =pdv.
very clearly set forth by a graphic adopt a rectangular system of co-ordinates, in which the abscissa represents the volume v, and the ordiiiate the pressure p. If we now suppose p to be expressed as a function of v, say p =f (v), then this equation is the equation to a curve, whose ordinates express the values of pr corresponding to the different values of v, and which for brevity we will call the Pressure-curve. In Fig. 3 let rs be
representation.
may be
We
will
sent the volume v existing at a certain instant, the ordinate ef drawn at e will represent the pressure at the same instant.
If further eg represent
an indefinitely small element of volume dv, and the ordinate gh is drawn at g, then we shall have an indefinitely small paralellogram efhg, whose area represents the external work done in an indefinitely small Fig. 3. expansion of the body and which differs from the product pdv only by an indefinitely small quantity of the second order, which may be neglected. The same holds for any other indefinitely smaU expansion; and hence in the case of a finite expansion (say from the volume Vj, represented by the abscissa oa, to that otv^, represented by the abscissa oc) the external work, for which we have the equation
;
W=
pdv
.(46),
ON PERFECT
IS
GASES.
Go
represented by the quadrangular figure abdc, which is bounded by the difference of abscissae ac, the ordinates ab and cd, and the portion of the pressure-curve bd. In order actually to perform the integration in equation (46) we must know the function of v which expresses the
we
will select as
us assume that the Pressure p is constant. of pressure is a straight line parallel to- the and abdc is a rectangle (see Fig. 4) whose area is
Fig. 4.
equal to the product of ac and ab. In this case then obtain from (46), denoting the constant pressure hy p^,
we
W=p,{v^-v,)
(47),
Secondly, let us assume that the Temperature remains constant during the expansion of the gas. Then the. law of Mariotte holds for the relation between pressure and volume, and is expressed by the equation
pv= const.
From the form of this equation we see that the curve of pressure is an equilateral hyperbola (Fig. 5) having the axes pressure- curve of this kind, of co-ordinates as asymptotes. which involves the special condition that the temperature is constant, is usually called an Isothermal Curve. To effect the integration in this case we may write for p
c.
66
the value
^'
where
'p^^ is
;
(46)
^=PA''7=i'i^i^og^^
(*)-
We
much
in equation (42) for Q ; the reason for this being that the gas, while expanding at constant temperature, absorbs only so
heat as is required for the external work. Joule has employed the equation (48) in one of his determinations of the Mechanical Equivalent of Heat. For this purpose he forced atmospheric air into a strong receiver, up to ten or twenty times its normal density. The receiver and pump were meantime kept under water, so that all the heat which was developed in pumping could be measured in the is water. The apparatus is represented in Fig. 6, in which the receiver, and G the pump. The vessel G, as will be easily understood, was used for the drying of the air, and the vessel with the spiral tube served to give to the air, before its entrance into the pump, an exactly known temperature. From the total quantity of heat given in the calorimeter Joule subtracted the part due to the friction of the pump, the amount of which he determined by working the pump for exactly the same length of time, and under the same mean pressure, but without allowing the entrance of air, and then observing the heat produced. The remainder, after this was subtracted, he took as being the quantity of heat developed by the compression of the air ; and this he compared with the work required for the compression as given by equation (48). By this means he obtained as the mean of two series of experiments the value of 444 kilogrammetres as the Mechanical Equivalent of Heat. This value, it must be admitted, does not agree very well with the value 424 obtained by the friction of water the reason of which is probably to be found in the far larger sources of erroi: attending experiments on air. Nevertheless at that time, when the fact that the work required for developing a given quantity of heat was equal under all circumstances was not yet placed on a firm basis, the agreement
ON PERFECT
of-
GASES.
67
close
the values found by such wholly different methods was enough to aid considerably in the establishment of the
principle.
As a
we may
cy;^
impermeable to heat or, which comes to the same thing, that the change of volume takes place too rapidly to allow of the passing of any appreciable quantity of heat to or from the body during the time.
;
52
Pi
\vj
"The curve of pressure corresponding to this equation more steeply than that delineated in Fig. 5. ilankine has given to this special class of pressurecurves, which correspond to the case of expansion within an envelope impermeable to heat, the name of Adiabatic curves (from Sia/3alveiv, to pass through). On the other hand Gibbs {Trans. Connecticut Academy, vol. II. p. 309) has proposed to
(Fig. 7) falls
I>=FiK
whence
(46)
becomes
(49).
(69
CHAPTEE
III.
1,
In order to prove and to make intelligible the second Principle of the Mechanical Theory of Heat, we shall commence by following out in all its parts, and graphically representing in the manner already described, one special form of cyclical process. For the latter purpose we will assume that the condition of the variable body is determined by its volume V and its pressure p, and will employ, as before, a rectangular system of co-ordinates, in which the abscissse represent'volumes, and the ordinates pressures. Any point on the plane of co-ordinates will then correspond to a certain condition of the body, in which its volume and pressure have the same volumes as the abscissa and ordinate of the point. Further, every variation of the body's condition will be represented by a line, whose extreme points determine the initial and final condition of the body, and whose form shews the way in which the pressure and volume have simultaneously varied. In Fig. 8 let the initial condition of the body, at which the cyclical process commences, be given by the point a, so that the abscissa oe = Vi and the ordinate ea=^p^ represent the initial volume and pressure respectively. By means of these two quantities the initial temperature, which we wUl
call Tj, is also fixed.
70
Now let the body in the first place expand, while retaining the same temperature T,. If no heat were imparted to it
e
Fig. 8.
it would necessarily become cooler: we assume that it is put in communication with a body K, acting as a reservoir of heat, which body has the same temperature T,, and does not appreciably vary from this during the cyclical process. From this body the variable body is supposed to draw during the expansion just sufficient heat to keep itself also at the temperature T^. The curve, which duritig this expansion expresses the change of pressure, is part of an isothermal curve. In order
during expansion,
will therefore
we may give definite forms to the graphic representations of this curve, and of others yet to be described, we will, without limiting the investigation itself to any particular bodies, draw the figure as it would appear in the case of a perfect gas. Then the isothermal curve, as explained above, will be an equilateral hyperbola; and, if the expansion take place from the volume oe = v^ to the volume o/= Fj, we shall obtain the part ah of such an equilateral hyperbola. When the volume V^ has been reached, let us suppose the
tha,t
and let the variable body be left by itself, without any heat hieing imparted to it. The temperature must then fall, and we obtain as curve of pressure an iseniropio curve, which descends more steeply than the isothermal curve. Let this expansion continue till the volume V^ is reached, giving us the portion of an isentropic curve be. The lower temperature thus
to be withdrawn, to continue its expansion
body K^
attained
we may
call
T.^.
7l
From henceforward let the body be compressed, so as to bring it back, to its original volume. Let the compression first take place at the constant temperature T^, for which purpose we may suppose that the body is connected with a body K^ at temperature T^, acting as a reservoir of heat, and that it gives up to K^ just so much heat as suffices to keep itself also at temperature T^. The pressure-curve corresponding to this compression is again an isothermal curve, and in the special case of a perfect gas is another equilateral hyperbola, of which we obtain the portion cd during the reduction
of
volume to oh=v^.
Finally, let the last compression, which brings the variabody back to its initial volume, take place without
ble
the presence of the body K^, so that the temperature rises, and the pressure follows the line of an isentropic curve. wUl assume that the volume oh = v^, up to which the compression went on according to the first mode, is so chosen, that the compression which begins from this volume and continues to volume oe = v^ is just sufficient to raise the temperature again from T^ to T^. If then the initial temperature is thus regained at the same time as the initial volume, the pressure must also return to its initial value, and the last curve of pressure must therefore exactly hit the point a. When the body is thus brought back again to the original condition, expressed by the point a, the cyclical process is complete.
We
Result of tJie Cyclical Process. During the two expansions which take place in the course <jf the cyclical process the external pressure must be over2.
' ^'
the expansion ab
is
come, and therefore external work must be performed whereas conversely during the compressions external work is absorbed to perform them. These quantities of work are given directly by the figure, which is here reproduced. The work performed during represented by the quadrangle eabf, and
72
that perfortned during the expansion be by the quadrangle fbcg. Again, the work absorbed for the compression cd is represented by the quadrangle cfcdh, and that absorbed for the compression da by the quadrangle hdae. The two latter quantities, on account of the lower temperature which obtains during the compression, are smaller than the two former; and, if we subtract them from these, there remains an overplus of external work performed, which is represented by the
quadrangle abed, and which we will call W. To the external work thus gained must correspond, according to equation (5a) of Chapter I., a quantity Q, equal to Now the it in value, which is required for its production. variable body, during the first expansion, expressed by ab, which took place in connection with the body if,, received from this latter a certain quantity of heat, which we may call Q, and again during the first compression, expressed by cd, which took place in connection with the body K^, it imparted to this latter a certain quantity of heat, which may be During the second expansion 6c and the second called Qj. compression da the body neither imparted nor received heat. Now, since in the course of the whole cyclical process a certain quantity of heat Q is. absorbed in work, it follows that the quantity of heat Q,, received by the variable body, is larger than the quantity of heat Q^ which it gives out, so that the difference Q^ Q^ is equal to Q. We may accordingly put
;
Q.
= Q,+ Q
(1),
and can tben distinguish in the quantity of heat Q,, which the variable body has drawn from the body iT,, two parts, of which one Q is converted into work, whilst the other Q, is given back as heat into the body K^. Since in all the other relations of the body the original condition is restored at the end of the cyclical process, and accordingly every variation which takes place at one part of the process is counterbalanced by an equal and opposite variation which takes place at some other part of the process, we may finally describe the result of the cyclical process in the following terms The one quantity of heat Q, derived from the bodyK, is transformed into work, and the other quantity Q^ has passed over
from
the hotter
73'
The whole of the cyclical process just described may also be supposed to take place in t^e reverse order. If we again begin with the conditions represented by the point a, in which the variable body has the volume , and the temperature Tj, we may suppose that it first expands, without any heat being imparted to it, to the volume v^, thus describing the curve ad, in which its temperature sinks from T^ to T^; that it is then connected with the body K^, and expands at constant temperature T^ from v^ to Fj, describing the curve dc, during which it draws heat from the body K^; that it then, without parting with its heat, is compressed from V^ to F,, describing the curve ch, during which its temperature rises from Tjj to T^ finally that it is connected with the body K^, at the constant temperature T^, and whilst imparting its heat to K^ is again compressed from Fj to the initial volume v^,
;
describing the ctirve ha. In this reversed process the quantities of work represented by the quadrangles eadh and hdcg are work performed or positive, those represented by gchf and fhae are work absorbed or negative. The latter amount is larger than the former, and the remainder, as represented by the quadrangle ahcd, is in this case work absorbed. In addition the variable body has drawn the quantity of heat Qj from the body K^, and has given out to the body K^ the quantity of heat Q^=Q^+ Q- Of the two parts of which Q^ consists, the one Q corresponds to the work absorbed, and is generated from it, whilst the other Q^ has passed over as heat from the body K^ to the body K^. Hence the result of the cyclical process may here be described as follows: the
quantity of heat Q is generated out of work, and is given off body K^, and the quantity of heat Q^ has passed over from the colder body K^ to the hotter body K^.
to the
Cyclical process in the case of a body composed partly of liquid, partly of vapour.
sections,
3.
although in describing the limiting the nature of the variable body, yet the graphic representation of the process was made to correspond to the case of a perfect gas.' It is perhaps as well therefore to examine the cyclical process.
In the foregoing
cyclical process
we made no assumptions
7i
over again in the case of a body of a different kind, in order to how its appearance may vary with the nature of the body will select for this examination a body operated on. which has not all its molecules in one and the same state in all its parts, but consists partly of liquid, partly of vapour at the maximum density. Let us suppose a liquid contained in an expansible envelope, but only filling a part of it, and leaving the remainder free for vapour having the maximum density corresponding to the existing temperature T^. The combined volumes of liquid and vapour are represented in Fig. 10 by the abscissa oe, and the pressure of the vapoiir by the ordinate ea. a Now suppose the envelope to yield to the pressure and enlarge, while at the same time the liquid and vapour are connected with a body K^ of constant temperature T^. As the volume increases, more liquid becomes vaporised, but the heat consumed in the
see
We
7o
volume oe and let this compression take place-, partly in connection with the body K^, of constant temperature T^, to which all the heat produced by condensation of vapour passes over, so that the temperature T^ remains unaltered partly apart from this body, so that the temperature rises. Let it also be arranged, that the first compression shall extend only so far (to oh) as that the decrease of volume he then remaining may be just sufficient to raise the temperature again from T^ to T^. During this first compression the pressure remains unaltered, at the value gc the external work thus absorbed is therefore represented by the rectangle gcdh. During the last compression the pressure increases, and is represented by the isentropic curve da, which must end exactly at the point a, since with the original temperature T^ we must also have the original pressure ea. The external work absorbed in this last operation is represented by hdae. At the end of the operation the liquid and vapour are again in their original condition, and the cyclical process is complete. The surplus of the positive ^bove the negative external work, or the external work which has been gained on the whole in the course of the process, is represented as before by the quadrangle abed. To this work must correspond the absorption of an equivalent quantity of heat Q and if we denote by Q^ the heat imparted during the expansion, and by Qj the heat given out during the contraction, we may put Qj = Q+^j, and the final result of the cyclical process is again expressed by saying, that the quantity of heat Q is converted into work, and the quantity Q.^ has passed over fi;om the hotter body K^ to the colder ^j. This cyclical process may also be carried out in the reverse direction, and then the quantity of heat Q will be generated out of work, and given off to the body K^, while the quantity Q^ will pass over from the colder body /C, to the hotter K^. In a similar manner cyclical processes of this kind may be carried out with other variable bodies, and graphically represented by two isothermal and two isentropic lines ; in which cases, while the form of the curves depends on the nature of the body, the result of the process is always of the same kind, viz. that one quantity of heat is converted into work, or generated out of work, and that another quantity passes over from a hotter to a colder body, or vice versa.
: ;
76
question now arises. Whether the quantity of heat converted into work, or generated out of work, stands in a generally constant proportion to the quantity which passes over from the hotter to the colder body, or vice versd; or whether the proportion existing between them varies according to the
is
the
medium of
the
4.
Cyclical Process.
Camot, who was the first to remark that in the production of mechanical work heat passes from' a hotter into a colder body, and that conversely in the consumption of mechanical work heat can be brought from a colder into a hotter body, and who also conceived the simple cyclical process above described (which was first represented graphically by
Clapeyron), took a special view of his own as to the fundamental connection of these processes *In his time the doctrine was still generally prevalent that heat was a special kind of matter, which might exist within a body in greater or lesser quantity, and thereby occasion differences of temperature. In accordance with this doctrine it was supposed that heat might change the character of its distribution, in passing from one body into another, and further that it could exist in different conditions, which were denominated respectively 'free' and 'latent'; but that the whole quantity of heat existing in the universe could neither be increased nor diminished, inasmuch as matter can neither be created nor destroyed. Camot shared these views, and accordingly treated it as self-evident that the quantities of heat, which the variable body in the course of the cyclical process receives from and gives out to the surrounding space, are equal to each other, and consequently cancel each other. He lays this down very " we shall distinctly in 27 of his work, where he says assume that the quantities of heat absorbed and emitted in these different transformations compensate each other exactly. This fact has never been held in doubt admitted at first without reflection, it has since been verified in many instances by
: ;
du feu.
Paris, 1824.
"
"'
77
experiments with the calorimeter. To deny it would be to subVert the whole theory of heat, which rests on it as its basis." Now since on this assumption the quantity of heat existing in the body was the same after the cyclical process as before it, and yet a certain amount of work had been achieved, Carnot sought to explain this latter fact from the circumstance of the heat falling from a higher to a lower temperature. He drew a comparison between this descending passage of heat (which is especially striking in the steam-engine, where the fire gives off heat to the boiler, and conversely the cold water of the condenser absorbs heat) and the falling of water from a higher to a lower level, by means of which a machine can be set in motion, and work done. Accordingly in 28, after making use of the expression fall of water,' he applies the corresponding expression fall of caloric' to the sinking of heat from a higher to a lower temperature. Starting from these premises, he laid down the principle that the quantity of work done must bear a certain constant relation to the 'passage of heat,' i.e. the quantity of heat passing over at the time, and to the temperature of the bodies between which it passes ; and that this relation is independent of the nature of the substance which serves as a medium for the performance of work and passage of heat. His proof of the necessary existence of this constant relation rests on the principle " That it is impossible to create moving
' '
That perpetual motion is an impossibility." This mode of dealing with the question does not accord with our present views, inasmuch as we rather assume that in the production of work a corresponding quantity of heat is consumed, and that in consequence the quantity of heat given out to the surrounding space during the cyclical process Now if for the production is less than that received from it. of work heat is consumed, then, whether at the same time with this consumption of heat there takes place the passage of another quantity of heat from a hotter to a colder body, or not, at least there is no ground whatever for saying that
the work is created out of nothing. Accordingly not only must the principle enunciated by Carnot receive some modification, but a different basis of proof from that used by him
must be
discovered.
78
5. New Fundamental Principle concerning Heat. Various considerations as to the conditions and nature of heat had led the author to the conviction that the tendency of heat to pass from a warmer to a colder body, and thereby equalize existing differences of temperature (as prominently shewn in the phenomena of conduction and ordinary radiation), was so intimately Tsound up with its whole constitution that it must have a predominant influence under all conceivable circumstances. He thereupon propounded the following as a fundamental principle " Heat cannot, of itself, pass from a colder to a hotter body." The words of itself,' here used for the sake of brevity, require, in order to be completely understood, a further explanation, as given in various parts of the author's papers. In the first place they express the fact that heat can never, through conduction or radiation, accumulate itself in the warmer body at the cost of the colder. This, which was already known as respects direct radiation, must thus be further extended to cases in which by refraction or reflection the course of the ray is diverted and a concentration of rays thereby produced. In the second place the principle must be applicable to processes which are a combination of several different steps, such as e.g. cyclical processes of the kind described above. It is true that by such a process (as we have seen by going through the original cycle in the reverse direction) heat may be carried over from a colder into a hotter body our principle however declares that simultaneously with this passage of heat from a colder to a hotter body there must either take place an opposite passage of heat from a hotter to a colder body, or else some change or other which has the special property that it is not reversible, except under the condition that it occasions, whether directly or indirectly, such an opposite passage of heat. This simultaneous passage of heat ia the opposite direction, or this special change entaUing an. opposite passage of heat, is then to be treated as a compensation for the passage of heat from the colder to the warmer body and if we apply this conception we may replace' ^ihe words " of itself" by " without compensation," and then enunciate the principle as follows " passage of heat from a colder to a hotter body cannot take place without compensation."
:
'
79
This proposition, laid down as a Fundamental Principle by the author, has met with much opposition but, having repeatedly had occasion to defend it, he has always been able to shew that the objections raised were due to the fact that the phenomena, in which it was believed that an uncompensated passage of heat from a colder to a hotter body was to be found, had not been correctly understood. To state these objections and their answers at this place would interrupt too In the discusseriously the course of the present treatise. sions which follow, the principle, which, as the author believes, is acknowledged at present by most physicists as being correct,
;
but the be simply used as a fundamental principle author proposes to retixm to it further on, and then to consider more closely the points of discussion which have been raised
will
;
upon
it.
6. Proof that the relation between the quantity of heat carried over, and that converted into work, is independent of the nature of the matter which forms the medium of the change.
Assuming the foregoing principle to be correct, it may be proved that between the quantity of heat Q, which in a cyclical process of the kind described above is transformed into work (or, where the process is in the reverse order, generated by work), and the quantity of heat Q^, which is transferred at the same time from a hotter to a colder body (or vice versS,), there exists a relation independent of the nature of the variable body which acts as the medium of the transformation and transfer and thus that, if several cyclical processes are performed, with the same reservoirs of heat K^ and K^^, but with
;
for
If we suppose the processes so arranged, according to all. their magnitude, that the quantity of heat Q, which is transformed into work, has in all of them a constant value, then we have only to consider the nmgnitude of the quantity of heat Qj which is transferred, and the principle which is to be " If ^here two different proved takes the following form
:
variable bodies are used, the quantity of heat Q transformed into work is the same, then the quantity of heat Q^, which
is
be the same."
80
(e.g. the Let there, if possible, be two bodies G and perfect gas and the combined mass of liquid and vapour, described above) for which the values of Q are equal, but
those of the transferred quantities of heat are different, and he called Q^ and Q\ respectively Now let us in the first (y^ being the greater of the two. place subject the body (7 to a cyclical process, such that the quantity of heat Q is transformed into work, and the quantity to a Next let us subject Qjj is transferred from K^ to K^. cyclical process of the reverse description, so that the quantity of heat Q is generated out of work, and the quantity Q'^ is transferred from K^ to K^. Then the above two changes, from heat into work, and work into heat, will cancel each other ; since we may suppose that when in the first process the heat Q has been taken from the body K^ and transformed into work, this same work is expended in the second process in producing the heat Q, which is then returned to the same body K^. In all other respects also the bodies will have returned, at the end of the two operations, to their original condition, with one exception only. The quantity of heat Q\, transferred from K^ to K^, has been assumed to be greater than the quantity Q^ transferred from K^ to K^. Hence these two do not cancel each other, but there remains at the end a quantity of heat, represented by the difference Q\ Q^, which has passed over from K^ to K^. Hence a passage of heat will have taken place from a colder to a warmer body without any other compensating change. But this contradicts the fundamental principle. Hence the assumption that Q\ is greater
let these different values
than Qj must be false. Again, if we make the opposite assumption, that Q^ is less than Q^, we may suppose the body to undergo the cyclical process in the first, and G in the reverse direction. We then arrive similarly at the result that a quantity of heat Q^ Q\ has passed from the colder body K, to the hotter K^, which is again contrary to the principle. Since then Q^ can be neither greater nor less than Q^ it must be equal to Q^; which was to be proved. We will now give to the result thus obtained the mathematical form most convenient for bur subsequent reasoning.
is
81
variable body, it can only depend on the temperature of the two bodies JT^ and ^j, which act as heat reservoirs. The same will of course be true of the sum
This last
;
ratio,
which
is
and the heat transferred, we shall select for further tion and shall express the result obtained in this
follows
:
Tj and T^"
I^'^c^'.t;)
(2),
in which <^ {T^T^ is some function of the two temperatures, which is independent of the nature of the variable body.
Determination of the Function <f> (T^T^. that the function given in equation (2) is independent of the nature of the variable body, offers a ready means of determining this function, since as soon as we
7.
The circumstance
have found
its
form
foi;
it
is
known
for all
bodies whatsoever.
Of
all classes of
for such a deterniination, since their laws are the most accuwill therefore consider the case of a perrately known.
We
fect gas subjected to a cyclical process, similar to that graphically expressed in Fig. 8, 1 ; which figure may be here repro-
duced
(Fig. 11),
inasmuch as a perfect gas was there taken as an example of the variable body. In this process the
gas takes up a quantity of heat Q during its expansion ah, and gives out a quantity of heat Qj during its compression cd. These quantities
l
we
shall calculate,
and
each
'
then
other.
compare
with
Fig. 11.
first
For this purpose we must turn our attention to the volumes represented by the
c.
82
that
aBscissse oe, oh, of, off, and denoted we may ascesrtain the relation
volumes v^v^ (represented by oe, change of volume to which the isentropic curve ad refers, and which may be considered at pleasure as an expansion or a compression. Such a change of volume, during which the gas neither takes in nor gives out any heat, has been treated of in 8 of the last chapter, in which we arrived at the following equation (43), p. 62
'
2;
\v)
and v are the temperature and volume at any point Substituting for these in the present case the in the curve. final values T and v,, "2' we have
where
S.ft\
!=
(=^
In exactly the same way we obtain for the change of volume represented by the isentropic curve be (of which the initial and final temperatures are also T^T^
;
fW
T^
/F\*^
(*
= -,
or
'=
(o).
We must now turn to the change of volume represented by the isothermal curve ab, which takes place at the constant temperature T^, and between the limits of volume v^ and Fj. The quantity of heat received or given off during such a change of volume has been determined in 8 of the last chapter, and by the equation (41) there given, p. 61, we
may
put in the present case
Q,
= ET,log'^K
(6).
83
Similarly for the change of volume represented by the isothermal curve do, which takes place at temperature T^ between the limits of volume v^ and V^, we have
Q, = i?T,logJ'
(7).
From
we
obtain by division
^=^
smcebyfS)
'
(8)
V = *. V
The function occurring in equation (2) is now determined, since to bring this equation into unison with the last equation
(8)
we must have
<^(2;rj
= -JJ
'a
(9).
We can now use in place of equation (2) the more determinate equation (8), which may also be written as follows
:
&-^^ =
(10).
The form of this equation may be yet further changed, by , affixing positive and negative signs to Q^, Q^. Hitherto these
have been treated as absolute quantities, and the distinction that the one represents heat taken in, the other heat given out, has been always expressed in words. Let us now for convenience agre6 to speak of heat taken in only, and to treat heat given out as a negative quantity of heat taken in. If accordingly we say that the variable body has taken in during the cyclical process the quantities of heat Qj and Q,, we must here conceive Q^ as a negative quantity, i.e. the same quantity which has hitherto been expressed by Q^. On this supposition equation (10) becomes
% + %' =
(11>,
84
8,
more complicated
character.
Hitherto we have confined ourselves to cyclical processes in which the taking in of quantities of heat, positive or Such pronegative, takes place at two temperatures only. cesses we shall in future call for brevity's sake Simple Cyclical Processes. But it is now time to treat of cyclical processes, in which the taking in of positive and negative quantities of heat takes place at more than two temperatures.
may first consider a cyclical process with heat taken in at three temperatures. This is represented graphically by the figure abcdefa (Fig. 12), which, as in the former cases, These consists of isentropic and isothermal curves only. curves are again drawn, by way of example, in the form which they would take in the case of a perfect gas, but this.
We
^-^e
Fig. 12.
is
not essential. The curve ab represents an expansion at. constant temperature T^; be an expansion without taking in heat, during which the temperature falls from T, to T^; cd de an expansion an expansion at constant temperature T^ without taking in heat, during which the temperature falls from jTj to Tg ef a, compression at constant temperature T^ and lastly fa a compression withoiit taking in heat, during which the temperature rises from T^ to T^, and which brings back the variable body to its exact original volume. In thp
; ;
85
expansions ab and cd the variable body takes in positive quantities of beat Q, and Q,, and in the compression ef the negative quantity of heat Q^. It now remains to find a relation between these three quantities.
For this purpose let us suppose the isentropic curve 6c produced in the dotted line eg. The whole process is thereby divided into two Simple Processes abgfa and cdegc. In the first the body starts from the condition a and returns to the same again. In the second we may suppose a body of the same nature to start from the condition e, and to r6tum to the same again. The negative quantity of heat Q^, which is taken in during the compression ef, we may suppose divided into two parts q^ and g,', of which the first is taken in during the compression gf, and the second during the compression We can now form the two equations, corresponding to eg. equation (11), which will hold for the two simple processes. These equations are, for the process abgfa,
^+53 ~ fp = 'p
and
for the process cdego
'
-tj
'3
-'i
'i
'a
or,
since
93+9a=Qa'
.(12).
In exactly the same way we may treat a process in which is taken in at four temperatures, as represented by the annexed figure abcdefgha. Fig. 13, which again consists solely of isentropic and isothermal lines. The expansions ab and cd, and the compressions ef and gh, take place at temperatures Tj, Tjj, Tj, T^, and during these times the quantities of heat Qi>Qi> Qi> Qi ^^ taken inrespectively the two former being
heat
;
86
positive,
and the two latter negative. Ptoduce the isentropic curves 6c and fg in the dotted lines ci and gk respectively. Then the whole process is subdivided into three Simple Processes akghja,
hUfk, and
cdeic,
1^
^e
Fig. 13.
may supcarried out with three exactly similar bodies. pose the quantity of heat Q^, taken in during the expansion ah, to be divided into two parts q^ and q^, corresponding to expansions ak and kh ; and the negative quantity Q^, taken ia during the compression ef, to be likewise divided into q^ and Then we can q^, corresponding to compressions if and ci. form the followiag equations for the three simple processes
:
We
First, for
akgha
'a
'a
87
we
obtaan
or
^1
rp
rp
-^H
rr
-^S
"^
m>
-^4
"
.(13).
In exactly the same way any other cyclical process, which can be represented by a figure consisting solely of isentropic and isothermal lines, and which has any given number of temperatures at which heat is taken in, may be made to yield
viz.
^l
-^2
-^3
-^4
or generally
4-0.
.(11).
of Heat and 9. Cyclical Processes, in which taking change of Temperatv/re take place simultameously.
have lastly to consider such cyclical processes as are represented by figures not consisting solely of isentropic and isothermal lines, but altogether general in form. The mode of treatment is as foUows. Let point a in
We
Fig. 14.
Fig.
let
pq be an
14 represent any given condition of the variable body ; arc of the isothermal curve which passes through
88
a, rs
an arc of the isentropic curve wliicli passes througt thfe Now let the body un4ergo a variation which is point. expressed by a pressure-curve not coinciding with either of the above, but taking some other course such as be or de. Then we may consider. such a variation as made up of a very grSat number dfvery small variations, in which we have alternately change of temperature without taking in of heat, This and taking in of heat without chahge of temperature. series of successive variations wiU be represented by a dis-
same
continuous
line,
made up
Fig. 15.
The smaller the elements of which the disbe and de. continuous curve is made up, the more closely will it coincide with the continuous line, and if these are indefinitely small the coincidence wiU be indefinitely close. In this case it can only make an indefinitely small difference, in relation to the quantities of heat taken in and their temperatures, if we substitute for the variation reprpsented by the continuous line the indefinitely large number of alternating variations, which are represented by the discontinuous line. are now in a position to consider a complete cyclical process, in which the taking in of heat is simultaneous with changes of temperature, and which may be represented graphically by curves of any form whatever, or merely by a
We
'single
continuoTls
Fig. 16.
The area
89
work consumed. Let it be divided into indefinitely thin strips by means of adjacent isentropical curves, as shewn by the dotted lines in Fig, 16. Let us suppose these curves joined at the top and bottom by indefinitely small elements
Fig. 16.
which cut the given curve, so that we have a broken line, which is everywhere in indefinitely close coincidence with it. By the above reasoning we may substitute for the process represented by the continuous line the other process represented by the broken line, without producing any perceptible alteration in the quantities of heat taken in, or in their temperatures. Further, we may again substitute for the process represented by the broken fine an indefinitely great number of Simple Processes, which will be represented by the indefinitely small quadrangular strips, made up each of two adjacent isentropic curves, and two indefinitely smaU elements of isothermal curves. If then for each one of these last processes we form an equation similar to (11), in which the two quantities of heat are indefinitely small, and can therefore be denoted by differentials of Q and if all these equations be finally added together we shall then obtain an equation of the same form as (14), but in which the sign of summation is replaced by the
of isothermal lines,
its
throughout
length
I'-
-f
^^^^-
90
93, p. 500), forms a very convenient expression for the second main Principle of the Mechanical Theory of Heat, as far as it relates to reversible processes. This Principle may be expressed in words as follows If in a
vol.
:
reversible
Cyclical Process every element of heat taken in {positive or negative) he divided by the absolute temperatwre at which it is taken in, and the differential so formed be inte-
grated for the whole cov/rse of the process, the integral so obtained is equal to zero.
If the integral l-m-j corresponding to any given succession
of variations of a body, be always equal to zero provided the body returns finally to its original condition, whatever the intervening conditions may be, then it follows that the ex-
-^ must
,
be the perfect
which depends only on the present condition of the body, and is altogether independent of the way in which it has been brought into that condition. If we denote this quantity by 8, we may put
or
dQ = Td8...,
(VI),
.
an equation which forms another expression, very convenient in the case of certain investigations, for the second main principle of the Mechanical Theory of Heat.
(91)
CHAPTER
IV.
THE SECOND MAIN PRINCIPLE UNDER ANOTHER FORM, OR PRINCIPLE OF THE EQUIVALENCE OF TRANSFORMATIONS.
1.
On
the
In the last chapter it was shewn that in a Simple Cyclical Process two variations in respect to heat take place, viz. that a certain quantity of heat is converted into work (or generated out of work), and another quantity of heat passes from a hotter into a colder body (or vice versS,). It was found further that between the quantity of heat transformed into work (or generated out of work) and the quantity of heat transferred, there must be a definite relation, which is independent of the nature of the variable body, and therefore can only depend on the temperatures of the two bodies which
serve as reservoirs of heat. For the former of these two variations we have already employed the word "transformation," inasmuch as we said, when work was expended and heat thereby produced, or conversely when heat was expended and work thereby produced, that the one had been " transformed " into the other. may use the word " transformation " to express the second variation also (which consists in the passage of heat from one body into another, which may be colder or hotter than the first), inasmuch as we may say that heat of one temperature " transforms " itself into heat of another tempera-
We
ture.
describe the result of a simple terms Two transformations are produced, a transformation from heat into work (or vice versS,) and a transformation from heat of a higher temperaprinciple
cyclical process in the following
:
On this
we may
92
ture to heat of a lower (or vice versS.). The relation between these two transformations is therefore that which is to be expressed by the second Main Principle. Now, in the first place, as concerns the transformation of heat at one temperature to heat at another, it is evident at once that the two temperatures, between which the transformation takes place, must come under consideration. But the further question now arises, whether in the transformation from work into heat, or from heat into work, the temperature of the particular quantity of heat concerned plays an essential part, or whether in this transformation the particular temperature is matter of indifference. If we seek to deduce the answer to this question from the consideration of a Simple Cyclical Process, as described above, we find that it is too limited for our purpose. For since in this process there are only two bodies which act as heat, reservoirs, it is tacitly assumed that the heat which is transformed into work is derived from (or conversely the heat generated out of work is taken in by) one or other of these same two bodies, between which the transference of heat also takes place. Hence a definite assumption is made from the beginning as to the temperature of the heat transformed into work (or conversely generated out of work), viz. that it coincides with one of the two temperatures at which the transference of heat takes place and this limitation prevents us from learning what influence it would have on the relation between the two transformations if the first-mentioned temperature were to alter, while the two latter remained un;
altered.
To
ascertain
this
influence,
we may
revert to those
more complicated
which have also been described in the last chapter, 8, and to the equations derived from them. But in order to give a clearer and simpler view of the question it is better to consider a single process specially chosen for this investigation, and by its help to bring out the second Main Principle anew in an altered
cyclical processes,
form.
2.
On a
.
Cyclical Process
of special form.
Let us again take a variable body, whose condition is completely determined by its volume and pressure, so that
93
we can
represent its variations graphically in the manner already described. will once more by way of example construct the figure in the form it assumes for a perfect gas,
We
but without making in the investigation itself any limiting assumption whatever as to the nature of the body. Let the body be first taken in the condition defined by the point a in Fig. 17, its volume being given by the abscissa
Fig. 17.
oh, and its pressure by the 'ordinate ha. Let be the temperature corresponding to these two quantities, and determined by them. will now subject the body to the following successive variations The temperature T of the gas is change.d to T^ (1) which we will suppose less than T. This may be done by enclosing the gas within a non-conducting envelope, so that it can neither take in nor give out heat, and then allowing it to expand. The decrease of pressure caused by the simultaneous increase of volume and faU of temperature will be represented by the isentropic curve ah; so that, when the temperature of the gas has reached T,, its volume and pressure have become ovand ih respectively. The variable body is placed in communication with (2) a body iTj of temperature T,, and then allowed to expand still further, but so that all the heat lost in expansion is restored by K^. With respect to the latter it is assumed that, on account of its magnitude or from some other pause, its temperature is hot
We
94
perceptibly altered by this giving out of beat, and may tbereHence the variable body will also fore be taken as constant. preserve during its expansion the same constant temperature r,, and its diminution of pressure will be represented by an isothermal curve be. Let the quantity of heat thus given off
by Z, be
(3)
body is disconnected from K^ and allowed to expand still further, without being able either to take in or give out heat, until its temperature has faUen from Let this diminution of pressure be represented by T^ to fj. the isentropic curve cd. The variable body is placed in communication with (4) a body K^, of constant temperature T^, and is then compressed, parting with all the heat generated by the compression to jKj. This compression goes on until K^ has received the same quantity of heat Q,, as was formerly abstracted from In this case the pressure increases according to the isothermal curve de. The variable body is disconnected from K^, and (5) compressed, without being able to take in or give out heat, until its temperature has risen from T^ to its original value T, the pressure increasing according to the isentropic curveThe volume on, to which the body is brought by this ef. process, is less than the original volume oh, since the pressure to be overcome, and consequently the external work to be transformed into heat, is less during the compression de than during the expansion be; so that, in order to restore the same quantity of heat Q,, the compression must be continued further than would have been necessary merely to annul the expansion. The variable body is placed in communication with a (6) body-^ of constant temperature T, and allowed to expand to its original volume oh, the heat lost in expansion being restored from K. Let Q be the quantity of heat thus required.- If the body attains the original volume oh at the original temperature T, then the pressure must also revert to its original value, and the isothermal curve, which represents this last expansion, will therefore terminate exactly in the point a.
The
The above
cess, since
six variations
make up
is finally
95
Of the three bodies, K, K^, K^, which in the whole process only come under consideration in so far as they serve as sources or reservoirs of heat, the two first have at the end lost the quantities of heat Q, Q^ respectively, whilst the last has gained the quantity of heat Q^ ; this may be expressed by saying that Q^ has passed from K^ to K^, while Q has disappeared altogether. This last quantity of heat must, by the first fundamental principle, have been transformed into external work. This gain of external work is due to the fact that in this cyclical process the pressure during expansion is greater than during compression, and therefore the positive work greater than the negative its amount is represented, as is easily seen, by the area of the closed curve abode/a. If we call this work W, we have by equation (5a) of Chapter I.
;
Q=W.
It is easily seen that the above Cyclical Process embraces as a special case the process treated of at the commencement of Chapter III., and represented in Fig. 8. For if we make the special assumption that the temperature T of the body
body K^, we may then altogether, and use K^ instead. The result dp away witL of the process will then be that one part of the heat given out by the body iT^ has been transformed into work, and the other part has been transferred to the body R^, just as was the case in the process above mentioned. The whole of this cyclical process may also be carried out in the reverse order. The first step will then be to connect the variable body with K, and to produce, instead of the final expansion fa of the former case, an initial compression af: and similarly the expansions fe and ed, and the compressions dc, cb, and ba will be produced one after another, under exactly the same circumstances as the
is
2",
of the
converse variations in the former case. It is obvious that the quantities of heat Q and Qj^ will now be taken in and K^ respectively, and the quantity by the bodies of heat Qi will be given ovi by the body K^. At the same time the negative work is now greater than the positive, so that the area of the closed figure now represents a loss of work. The result of the reversed process is therefore that
96
the quantity of heat Q^ has been transferred from K^ to K^^, And that the quantity of heat Q has been generated out of work and given to the body K.
3.
On Uquipalent Transformations.
For
all
such
it
last chapter (p. 78) that the quantity of heat Q,, transferred from ifJ to K^, must stand in a constant relation to the
quantity Q transformed into work. For let us suppose that there were two such processes, in which, while Q was the
same in both, Q was different then we might successively execute that in which Q^ was the smaller in the direct order, and the other in the reverse. In this case the quantity of heat Q, which in the first process would have been trans:
formed into work, would in the second process be transformed again into heat and given back to the body K; and in other respects also everything would at the conclusion be restored to its original condition, with this single exception that the quantity of heat transferred from K^ to K^ in the second process, would be greater than the quantity transferred from K^ to K^ in the first process. Thus on the whole we have a transfer of heat from the colder body K^ to the hotter
-,
97
with nothing to compensate for it. As this contradicts the fundamental principle, it follows that the above supposition cannot be true ; in other words Q must always stand in the
same
ratio to Q^.
transformations in a reversible process such as the above, either can replace the other, provided this latter be taken in the reverse direction : in other words, if a trans-
Of the two
formation of the one kind has taken place, this can be again and a transformation of the other kind substituted in its place, without the occurrence of any other permanent change. For example, let a quantity of heat Q be in any way generated out of work, and taken in by the body K; then by the cyclical process above described it can be again withdrawn from the body K, and transformed back into work, but in so doing a quantity of heat Q^ will be transferred from the body JST, to the body K^. Again, if the quantity of heat Q^ has previously passed from K^^ to K^, it can by performing the above process in the reverse order be transferred back again to K^, whilst at the same time the quantity of heat Q, at the temperature of the body K, wiU be generated out of work. It is thus seen that these two kinds of transformation may be treated as processes of the same nature j and two such transformations, which may mutually replace each other in the way indicated, wUl be henceforth called " Equivalent Transformations."
reversed,
4.
We have now to find the law according to which the above transformations must be expressed mathematically, so that the equivalence of the two may appear from the equality The mathematical value of a transformation of their values. may be termed, thus determined, its " Equivalence-Value." We must u:st settle the order in which each transformation is to be taken as positive this may be chosen arbitrarily for one of the two classes, but it will then be fixed for the other, since clearly we must regard a transformation in the latter class as positive, if it is equivalent to a positive transformation in the former. In all that follows we shall consider the transformation of Work into Heat, and therefore the passage of heat from a higher to a lower temperature, as being positive quantities. It will be seen later why this
:
C.
'
98
choice as to the positi-ve and negative sign is preferable to the opposite. With regard to the magnitude of the equivalence-value, it is at once seen that the value of a change from work into heat must he proportional to the quantity of heat generated, and that beyond this it can only depend on its temperature. We may therefore express generally the equivalence-value of the generation out of work of the quantity of heat Q, of temperature T, by the formula Q xf(T), where f{T) is a function of temperature which is the same for all eases. If Q is negative in this formula, what is expressed is that the quantity of heat Q 'has been transformed, not out of work into Tieat, but out of heat into work. Similarly the value of the passage of a quantity of heat Q from ihe temperature Tj to the temperature 2", must be proportional to the quantity of heat which passes, and beyond this can only depend on the two temperatures. We may therefore {T^, T^, in which express it generally by Dhe formula Q x [T^, T^ is a function of the two temperatures, also constant for all eases, and which we cannot at present determine more closely; but of which it is clear from the commencement that, if the two temperatures are interchanged, it must change its sign, without changing its numerical value. may
We
therefore write,
FiTT,) = -F{T,,T,)
we have
(1).
kind given above the two transformations that take place must be equal in magnitude but of opposite sign, so that
their algebraical
sum
is zero.
Thus
if
we
ment the
particular process fully described above ( 2), the quantity of heat Q, at temperature T, is then transformed into work, giving as its equivalence-value Qx-fiT); and the quantity of heat Q^ passes from temperature T^ to T^, thus giving as its equivalence-value (?, x F{T^, T^. Therefore the following equation must hold
-Qxf{T) + Q,xF{TT,)=0
(2).
Let us now suppose a similar process performed in the reverse order, and under the conditions that the bodies -K"
99
and K^, amd the quantity of heat Q^ which passes between them, remain the same, but that for the body of temperature T is substituted another body K' of temperature T': and let us call the heat generated out of work in this case Q'. Then, corresponding to the formier equation, we have the
following
Q' ^/{T')
+ Q,x F{T
Q'
T,)
=
we
(3).
Adding
(2)
and
(3)
(1)
obtain,
(4).
- Q xf{T) +
Now let us consider,
as
xf{T')
allowable, that these two successive processes make up together a single process ; then in this latter the two transferences of heat between K^ and K^ will cancel each other and disappear from the result ; we
is clearly
left the transformation into work of the quantity of heat Q, given off by K, and the generation out of work of the quantity of heat Q' taken in by K'. These two transformations, which are of the same kind, can however be so broken up and re-arranged as to appear in the light of transformations of different kinds. For if we simply hold fast to the main fact, that the one body has lost the quantity of heat Q, and the other K' gained the quantity Q", then the heat equivalent to the smaller of these two quantities may be considered as having been transferred directly from to K', and it is only the difference between the two which remains to be considered as a transformation of work into heat or vice versS,. For example, let the temperature be higher than T; then the transference of heat on the above view is from a hotter to a colder body, and is therefore .positive. Accordingly the other transformation must be negative, i.e. a transformation from heat into work whence it follows that the quantity of heat Q given off by is greater than the quantity Q' taken in by K'. Thus if we divide Q into its two component parts and Q Q', then the first of these will have passed over from to K', and the second is the quantity of heat transformed into work. On this view the two processes appear as combined into a single process of the same kindj for the circumstance that the heat transformed into work is not derived from a third body, but from one or other of the same two bodies, between
72
100
whicli the transference of heat takes place, makes no The temperature of the essential difference in the result. heat transformed into work is optional, and can therefore have the same value as the temperature of one of the two bodies in which case the third
body
is
no longer required
Accord-
ingly for the two quantities of heat Q' and be an equation of the same form as (2), namely
Q Q
there
must
{T,
T) = 0.
and then striking
(5).
Q by means
of equation (4)
we
obtain
F{TT')=f{T)-f{T)
As the temperatures
of
T' are any whatever, the function which holds for the second kind of transformation, is thus shewn to agree, with the function of one temperature/ (2'), which holds for the first kind. For the latter function we will for brevity use a simpler symbol. For this it is convenient, for a reason which will be apparent later on, to express by the new symbol not the function itself, but its reciprocal. We wiU therefore put
T and
two temperatures
jP'(T'2"),
^=7^0^/(20 = ^
so,
(6),
that T is now ihe unknown function of temperature which enters into the Equivalence-value. If special values of this function have to be written down, corresponding to temperatures Tj, T^, etc.', or T', T' , etc., then this can be done by simply using the indices or accents for t itself. Thus equation (5) wiE become
i^(r, 2")
= -,--.
T
Principle of the Mechanical Theory of Heat, which in this form may perhaps be called the principle of the Equivalence of Transformations, can be expressed in the following terms "K we call two transformations which may cancel each other without requiring any other permanent change to take place, Equivalent Transformations, then the generation out of work
101
value
and the
<-^>
in which t is a function of temperature independent of the kind of process by which the transformation is accomplished."
5. Combined value of all the transformations which take place in a single Cyclical Process.
If
we
in the form
we
of heat Q from temperature T to T^ has the same equivalence-value as a double transformation of the first kind, viz. the transformation of the quantity Q from heat of temperature T, into work, and again out of work into heat of temperature T^. The examination of the question how far this external agreement has its actual foundation in the nature of the process would here be out of place but in any case we may, in the mathematical determination of the Equivalence-Value, treat every transference of heat, in whatever way it may have taken place, as a combination of two opposite transformations of the first kind. By this rule it is easy for any Cyclical Process however complicated, in which any number of transformations of both kinds take place, to deduce the mathematical expression which represents the combined value of all these transFor this purpose, when a quantity of heat is formations.
;
given off by a heat reservoir, we have no need first to enquire what portion of it is transformed into work, and what becomes of the remainder ; but may instead reckon every quantity of heat given off by the hea,t reservoirs which occur in the cyclical process as being wholly transformed into work, and every quantity of heat taken in as being generated out of work. Thus if we assume that the bodies K-^, K^, K^, etc. of temperatures T^, T^, T,, etc. occur as heat reservoirs, and if Q^, Q^, Q^, etc. are the quantities of heat given off during
102
the Cyclical Process (in wliicli we will now consider quantities of heat taken in as negative quantities of heat given out*), then the combined value of all the transformations, which we may call N, will be represented as follows
j^=_i^_Q._^_etc.,
Ta
^2
T3
iV
2^
T
(7).
It is here supposed that the temperatures of the bodies K^, K^, K^, etc. are constant,' or at least so nearly so that If however the temperatheir variations may be neglected. ture of any one of the bodies varies so much, either through the taking in of the quantity of heat Q itself,. or through any other cause, that this variation must be taken into account, then we must, for every element of heat dQ which is taken in, use the temperature which the body has at the moment of its being taken in. This naturally leads to an integration. If for thje sake of generality we assume this to hold for all the bodies, th.en the foregoing equation takes the following form
"It
f''^
(8),
in which the integral is to be taken for heat given off by the different bodies.
all
the quantities of
6. Proof that in a reversible Cyclical Process the total value of all the transformations must he equal to nothing.
If th-e Oyclical Process under consideration is reversible, then, however complicated it may be, it can be proved that the transformations which occur in it must cancel each other, so that their algebraical sum is equal to nothing.
* This choice of positive and negative signs for the quantities of heat agrees with that which we made in the last chapter, where we considered a quantity of heat taken in by the variable body as positive, and a quantity given out by it as negative for a quantity given out by a heat reservoir is taken in by the variable body, and vice versa.
;
103
not the case, i.e. that this' then let us imagine the following process applied. Let all the transformations which take place be divided into two parts, of which the first has its algebraical sum equal to nothing, and the second is made up of transformations all having the same sign. Let the transformations of the first division be separated out into pairs, each composed of two transformations of equal magnitude but opposite signs. If all the heat reservoirs are of constant temperature, so that in the Cyclical Process there is only a finite number of definite temperatures, then the number of pairs which have to be formed will be also finite; but should the temperatures of the heat reservoirs vary continuously, so that the number of temperatures is indefinitely great, and therefore the quantities of heat given off and taken in must be distributed in indefinitely small elements, then the number of pairs which have to be formed will be indefinitely large. This however, by our principle,, makes no difference. The two transformations of each pair are now capable of being done backwards by one or two Cyclical Processes of the form
tliat this is
us suppose
algebraical
other value
described in 2. Thus in the first place let the two given transformations be of different kinds, e.g. let the quantity of heat Q of temperature T be transformed into work, and the quantity of heat Q^ be transferred from a body K^ of temperature T, to a body K^ of temperature T^. The symbols Q and Q^ are here supposed to represent the absolute values of the quantities. Let it be also assumed that the magnitudes of the two quantities stand in such relation to each other that the following equation, corresponding to equation (2), will hold, viz.
Then let us suppose the Cyclical Process to be performed in the reverse order, whereby the quantity of heat Q, of temperature T, is generated out of work, and another quantity Thi of heat is tra,nsferred from the body K^ to the body K^. lattier quantity must then be exactly equal to the quantity Q^, given in the above equation, and the given transformations have thus been performed backwards. Again let there be one transformation from work into
104
lON
heat and one from lieat into work, e.g. let the quantity of heat Q of temperature T be generated out of work, and of temperature 2" be transformed the quantity of heat into work, and let fbese two stand in such relation to each other that we may put
us suppose in the first place that the same process as has been performed, whereby the quantity of heat Q of teinperature T has been transformed into work, and another quantity Q^ has been transferred from a body K^ to another body K^. Next let us suppose a second process performed in the reverse direction, in which the last-named quantity Q^ is transferred back again from K^ to K^, and a quantity of heat of temperature T' is at the same time generated out of work. This transformation from work into heat must, independently of sign, be equivalent to the former trajisformation from heat into work, since they are both equivalent to one and the same transference of heat. The quantity of heat of temperature T', generated out of work, must therefore be exactly as great as the quantity Q' found in the above equation, and the given transformations have thus been made backwards.
let
Then
last described
Finally, let there be two transferences of heat, e.g. the quantity of heat Q transferred from a body K^ of temperature Tj to a body K^^ of temperature T^, and the quantity Q\, from a body K\ of temperature T'j to a body K\ of temperature T\, and let these be so related that we may put
.(i-i).e'.(^-ij=o.
Then let us suppose two Cyclical Processes performed, in one of which the quantity Q,^ is transferred from K^ to K^, and the quantity Q of temperature thereby generated out of work, whilst in the second the same quantity Q is again transformed into work, and thereby another quantity of heat transferred from to K\. This second quantity must then be exactly equal to the given quantity Q\, and the two given transferences of heat have thus been done backwards.
K\
105
When by operations of this kind all the transformations of the first division have been done backwards, there then rernain the transformations, all of the like sign, of the second division, and no others whatever. Now first, if these transformations are negative then they can only be transformations from heat into work and transferences from a lower to a bigher temperature; and of these the transformations of the *first kind may be replaced by transformations of the second kind. For if a quantity of heat Q of temperature is transformed into_ work, then we have only to perform in reverse order the cyclical process described in 2, in which the quantity of heat Q of temperature is generated out of work, and at the same time another quantity $j is transferred from a body K^ of temperature T^ to another body K^ of the higher temperature Tj^. Thereby the given transfonnation from heat into work is done backwards, and replaced by the transference of heat from K^ to Ky By the application of this method, we shall at last have nothing left except transferences of heat from a lower to a higher temperature which are not compensated in any way. As this contradicts our fundamental principle, the supposition that tlie transformations of the second division are negative must be incorrect.
Secondly,
if
since the cyclical process under consideration is reversible, the whole process might be performed in reverse order; in which case all the transformations which occur in it
would take the opposite sign, and every transformation of the second division would become negative. We are thus brought back to the case already considered, which has been found to contradict the fundamental principle.
As then the transformations of the second division can neither be positive nor negative they cannot exist at all ; and the first division, whose algebraical sum is zero, must embrace all the transformations which occur in the cyclical process. may therefore write iV= in equation (8), and thereby we obtain as the analytical expression of the Second Main Principle of the Mechanical Theory of Heat for reversiblg processes the equation
.
We
'"'^^
/'
0...,,
,(VII).
106
7.
On
the
In the development of Equation VII. the temperatures of the quantities of heat treated of were determined by those of the heat reservoirs from which they came, or into which they passed. But let us now consider a cyclical process, which is such that a body passes through a series of changes of condition and at last returns to its original state. This variable body, if placed in connection with the heat reservoir to receive or give off heat, must have the same temperature as the reservoir ; for it is only in this case that the heat can pass as readily from the reservoir to the body as in the reverse direction, and if the process is reversible it is requisite that this should be the case. This condition cannot indeed be exactly fulfilled, since between equal temperatures there can in general be no passage of heat whatever but we may at least assume it to be so nearly fulfilled that the small remaining dififerences of temperature may be neglected. In this case it is obviously the same thing whether we consider the temperature of a quantity of heat which is being transferred as being equal to that of the reservoir or of the variable body, since these are practically the same. If however we choose the latter and suppose that in forming Equation VII. every element of heat Q is taken of that temperature which the variable body possesses at the moment it is taken in, then we can now ascribe to the heat reservoirs any other temperatures we please, without thereby making
;
any
With
this assumption
as to the temperatures we may consider Equation VII. as holding, without troubling ourselves as to whence the heat comes which the variable body takes in, or where that goes which it gives off, provided the process is on the whole a reversible one.
The
given,
expression
if it
is the differential of a quantity which depends on the condition of the body, and at the same time is fully determined as soon as the condition of the body at the moment is known, without our needing to. know the path by which
107
the body has arrived at that condition for it is only in this case that the integral will always become equal to zero as often as the body after any given variations returns to its original condition. In another paper*, after introducing a further development. of the equivalence of transformations, the author proposed to call this quantity, after the Greek word Tpoirrj, Transformation, the Entropy of the body. The complete explanation of this name and the proof that it correctly expresses the conditions of the quantity under consideration can indeed only "be given at a later period, after the development just mentioned has been treated of; but for the sake of convenience we shall use the name henceforward.
If
we
;Si
we may put
T
or otherwise
dQ = rdS.,
8.
(VIII).
t.
On
the
Temperature Fimction
We
QrT,
^^^-
On the other hand, if we apply Equation VII. to this simple cyclical process, whilst at the same time we treat the
* Fogg. Ana^ Vol. cxxv.
p. 390.
108
giving out of the quantity of heat Q as equivalent to the taking in of the negative quantity Q, we have the following equation
^^"^^.
From
equations (9) and (10)
we obtain
or
^^^T
we now take
(U).
If as
T as being any temperature whatever and 2\ some given temperature, we may write the last equation
thus
T=Tx Const
and the temperature function t
factor.
is
(12),
What
value
rent, since it
we ascribe to the constant factor is indiffemay be struck out of Equation VII. and thus
has no influence on any calculations performed by means of the equation. We will therefore choose the simplest value, viz. unity, and write the foregoing equation
r=T
The temperature
,....(13).
function is now nothing more than the absolute temperature itself. Since the foregoing determination of the function t rests on the equations deduced for the case of gases, one of the foundations on which this determination rests will be the approximate assumption made in the treatment of gases, viz. that a perfect gas, if it expand at constant temperature, absorbs only just so much heat as is required for the external work thereby performed. Should anyone on this account have any hesitation in regarding this determination as perfectly satisfactory, he may in Equations VII. and VIII. regard t as the symbol for the tempierature function as yet
109
undetennined, and use the equations in that form. Any such hesitation would not, in the author's opinion, be justifiable, and in what follows T will always be used in the place of t. Equations VII. and VIII. wUl then be written in the following forms, which have already been given under Equations V. and VI. of the last chapter, viz.
dQ = TdS.
CHAPTER
V.
In the general treatment of the subject hitherto adopted succeeded in expressing the two main principles of the Mechanical Theory of Heat by two very simple equations numbered III. and VI. (pp. 31 and 90).
we have
dQ = dU+dW dQ = TdS
..(Ill),
(VI).
We will now throw these equations into altered forms which make them more convenient for our further calculations.
Both equations relate to an indefinitely small alteration of condition in the body, and in the latter it is further assumed that this alteration is affected in such a way as to be reversible. For the truth of the first equation this assumption is not necessary we will howevpr make it, and in the following calculation will assume, as hitherto, that we have
:
only to do with reversible variations. suppose the condition of the body under consideration to be determined by the values of certain magnitudes, and for the present we will assume that two such magnitudes are sufficient. The cases which occur most frequently are those in which the condition of the body is determined by its temperature and volume, or by its temperature and pressure, or lastly by its volume and pressure. will not however tie ourselves to any particular magnitudes, but will at first
We
We
Ill
assume that the condition of the body is determined by any two magnitudes which may be called x and y ; and these magnitudes
independent variables of our In special cases we are of course always free to take one or both of these yariables as representing either one or two of the above-named magnitudes. Temperature, Volume
calculations.
we
and Pressure. If the magnitudes x and y determine the condition of the body, we can in the above equations treat the Energy IT and the Entropy S as being functions of the variables. In the same way the temperature T, whenever it does not itself
form one of these variables, may be considered as a function of the two variables. The magnitudes and Q on the contrary, as remarked above, cannot be determined so simply, but must be treated in another fashion.
The
magnitudes we shall
denote as follows
dW
dW
(^)-
f = ^= f = ^
and
These differential coefficients are definite functions of x For suppose the variable x is changed into x+dx y. while y remains constant, and that this alteration of condition in the body is such as to be reversible, then we are dealing with a completely determinate process, and the external work done in that process must therefore be also
determinate, whence
it
dW -^
must
equally have a determinate value. The same will hold if we suppose y to change to 1/ + dy while x remains constant. If then the differential coefficients of the external work are determinate functions of x and y it follows from Equation III. that the differential coefficients of the quantity of heat Q taken in by the body are also determinate functions of x
and
y.
and dQ their expressions as Let us now write for functions of dx and dy, neglecting those terms which are of a
dW
112
and dy.
We then have,
(3),
"
dW = mdx + ndi/
dQ = Mdx + Ndi/
(4),
and we thus obtain two complete differential equations, which cannot be integrated so long as the variables a; and y are independent of each other, since the magnitudes m, n and M, do not fulfil the conditions of integrability, viz.
dm _dn dy ~ dx
dM _dN
'
dy
dx
The magnitudes and Q thus belong to that class which was described in the mathematical introduction, of which the
peculiarity is that, although their differential coefficients are determinate functions of the two independent variables, yet they themselves cannot be expressed as such functions, and can only be determined when a further relation between the variables is given, and thereby the way in which the variations took place is known.
U and
S from
the
Let us now return to Equation III., and substitute in it and dQ expressions (3) and (4) then, collecting together the terms in dx and dy, the equation becomes,
for
dW
Mdx + Ndy =
dx and
dy,
iV=-5-H-W.
113
equation according to
y,
and the
we
dy
obtain,
dM^ffU_
dxdy
dm
dy
dN^
dx
cP
dydx
dn dx
may apply to JJ the principle which holds for every function of two independent variables, viz. that if they are differentiated according to both variables, the order of differentiation
is
We
a matter of indifference.
Hence we have
d'u ^ dru_
dxdy
dydx
Subtracting one of the two above equations from the other we 'obtain,
dM dN_ dm_dn
dy
dx
dy
dx
We may
Putting for comes,
now
dQ
treat Equation VII. [in the same manner. and d8 their complete expressions, it be-
M^
or
N.
itself,
dS,
dS,
M^dS
T T
dx'
F_d8
dy'
Differentiating the first of these according to y, and the second according to x, we obtain,
114'
01?
HEAT.
T^
_dxdy
dx
:
dx
T^
d'S dydx'
But
as before,
d'S
d'S
dxdy
dydx
we
obtain
dy
dy
dx
dx
_^
,.
^^^-
(5)
and
(6)
To
-^ and
right
by those
differenwill
themselves.
,
Similarly for
dW dW write -i and t dx dy
(5)
and n we
Then the
hand
side of equation,
may be
wiitten
d^
/dW\ _ d^(dW\
J
dy \dx
dx\dy J
'
Thus the magnitude represented by this expression is a function of X and y, which may generally.be considered as known, inasmuch as the external forces acting on the body are open to direct observation, and from these the external work can be determined. The above difference, which will occur veryfrequently from henceforward,
we
shall
call
"The Work
115
a special
..
^^^
(5)
and
(6)
are trans-
dy\dx) dy\dx)
rfi\
^>-'
^(dQ\_fdQ\^lfdT^dQ_dT^dQ\
T\dy
dx dx
dyj"
''
These two equations form the analytical expressions of the two fundamental principles for reversible variations, in the case in which the condition of the body is determined by two given variables. From these equations follows a third, which is so far simpler as it contains only differential coefficients of Q of the first degree, viz.
^Jx|f-Sx|=-..-
(-)
3. Case in which the Temperature is taken as one of the Independent Variables. The above three equations are very much simplified if the temperature of the body is selected as one of the independent Let us for this purpose put y = T, so that the two variables. independent variables are the temperature T, and the still undetermined quantity x. Then we have first,
^=1 dy
Again, referring to the differential coefficient
dT
;,,
it is
its formation that, while x is changed into x + dx, the other variable, hitherto called y, remains constant. But as T is now itself the other variable, it follows that T must remain constant in the above differential coefficient, or in other words
assumed in
dx
82
116
we form
T, this
wUl run
as follows
^e^m\_d_idw\
^"^
dT\dx)
(8), (9), (10)
dxKdTj
'
Hence equations
^-s(S)-^
(dQ\_d_(d^\^ldQ
dT\dx)
dxKdT)
('^)-
,
^
Tdx
''
S=^^-
(1^)'
d fdQ\
dD,^
4.
Particular Assumptions as
to the
External Forces.
We have hitherto made no particular assumptions as to the external forces to which the body is subjected, and to which the external work done during its alterations of condiWe will now consider more closely a special tion is related. case, which occurs frequently in practice, namely' that in which the only external force which exists, or at least the only one which is of sufficient importance to be taken into consideration, is a pressure acting on the surface of the body, everywhere normal to that surface, and of uniform intensity. In this case external work can only be done during changes in the volume of the body. Let p be the pressure per unit of surface, v the volume then the external work done, where this volume is increased to ?; + dv, is given by the equation
;
dW=pdv
(16).
117
Let US now suppose tlie condition of the body to be given Then the of two given variables x and y. pressure p and volume v must be considered as functions of X and y. We may thus write the last equation as follows
by the values
"^'^(S-^+l*)-
118
=-^,
dj/
""
(21),
(22).
D^ = l
is
Lastly in the cases in which either the volume or pressure taken as one of the independent variables, we may choose the temperature for* the other. We must then put T for y in equations (20) and (21), which then become
D.
dp 'It-
(23),
(24).
Forms of
the
Differential
In the circumstances described above, where the only external force is a uniform pressure normal to the surface, it is usual to choose as the independent variables, which are to determine the condition of the body, the quantities last mentioned in the foregoing section, viz. volume and temperature, pressure, and temperature, or lastly, volume and pressure. The systems of differential equations which hold for these three, cases may be easily deduced from the more general systems given above but on account of their frequent use it may be well to collect them together in this place. The first system is the one which the author has usually employed in the treatment of special cases.
,
If V and
T are taken
d^(dQ\
_d_(^\_dp
d^
dT
d_
/^\ _
fdQ
dT\dv)
dQ_
dv
dv[dT, dp
IdQ T dv
..(2.5).
dT'
d^(dQ
dv
Q-
dr
119
and
T are taken
.(26).
dp dp [dTj
If V and
dT'
dT'
are taken
dp \dvj dp \dv
dv \dp
'
6.
dQ\
dp
'
(27).
dp
dv
dp
a Partial Change in
case whict permits a still further simplification peculiar to itself, and which is of special interest from its frequent occurrence, is that in which the changes of condition in the
partial change in its Condition of Aggregation. Suppo.se a body to be given, of which one part is in one particular state of aggregation, and the remainder in another. As an example we may conceive one part of the body to be in the condition of liquid and the remainder in the condition of vapour, and vapour of the particular density which it assumes when in contact with liquid the equations deduced will however hold equally if one part of the body is in the solid condition and the other in the liquid, or one part in the solid and the other in the vaporous shall, for the sake of generality, not attempt condition. to define more nearly the two conditions of aggregation which we are treating of, but merely call them the first and the second conditions.
:
We
120
Let a certain quantity of this, substance be inclosed in a vessel of given volume, and let one part have the first, and the other the second condition of aggregation. If the specific volumes (or volumes per unit of weight), which the substance assumes at a given temperature in the two different conditions of aggregation, are different, then in a given space the two parts which are in the different conditions of aggregation cannot be any we please, but can only have determinate values. For if the part which exists in the condition of greater specific volume increases in size, then the pressure is thereby increased which the inclosed substance exerts on the inclosing walls, and consequently the reaction which those walls exert upon it, and finally a point will be reached, where this pressure is so great that it prevents any further passage of the substance
into this condition of aggregation. When this point is reached, then, so long as tHe temperature of the mass and its volume, i.e. the content of the vessel, remain constant, the magnitude of the parts which are in the two conditions of aggregation can undergo no further change. If, however,
whilst the temperature remains constant, the content of the vessel increases, then the part which is in the conditionj of aggregation of greater specific volume can again increase at the cost of the other, but only until the same pressure as before is attained and any further passage from one condition to the other thereby prevented. Hence arises the peculiarity, which distinguishes this case from all others. For suppose we choose the temperature and the volume of the mass as the two independent variables which are to determine its condition; then the pressure is not a function of both these variables, but of the temperature alone. The same holds, if instead of the volume we take as the second independent variable some other quantity which can vary independently of the temperature, and which in conjunction with the temperature determines fuUy the condition of the body. The pressure must then be inde-
pendent of this
latter variable.
The two
quantities, tem-
perature and pressure, cannot in this case be chosen as the two variables which are to serve for the determination of the body's condition. will now take, in addition to the temperature T, any
We
121
other magnitude x, as yet left undetermined, for the second variable which is to determine the body's condition. Let us then consider the expression given in equation (19) for the work-diflference referred to xT, viz.
T\
_dp
dv
dp
dv
put -^
= 0,
^.-J4:
The three equations
'g: into the following:
(^)-
thereby changed
d_
fdQ\
d^
fdQ\
_dp
dr' iT\dool
dx\dT)~ df
dv dx
''
dQ
dx
^
''
S = ^*x|
7.
(^'>-
Clapeyron's Equation
of the Fundamental Equawhich have formed the subject of the present chapter, we may consider the equation which Clapeyron* deduced as
tions
a fundamentajl equation from the theory of Carnot, in order what relation it stands to the equations we have' here developed. As however Clapeyron's equation contains an unknown function of temperature, usually designated as Carnot's function, it will be advisable beforehand to throw our equations, so far as they will come under consideration, into the form which they take, if the temperature function t, introduced in the last chapter, is treated as still indeterminate, and is not, in accordance with the value
to see in
* Journal de VEcole Foly technique, Vol. xiy. (1834).
122
there determined, put equal to the absolute temperature T. shall thereby obtain the advantage of fixing the relation between our temperature-function t and Carnot's Function. If instead of equation
We
dQ=Td8,
we use the
less
dQ = rdS,
and eliminate
as in 2,
we
8 from this eque-tion, in the same manner obtain instead of equation (9) the following
:
(dQ\_d^(dQ\l^dj^^dQ_dT^dQ\
dy\dx)
Combining
tion (16),
dx\dy)
this
T\dy
(8)
dx
dx
dyj"
''
with
we
^^S-S"!"^-
(^>
If we now assume that the only external force is a uniform and normal pressure on the surface, we may use for D^y the expression given in equation (18), and the above equation becomes
dr dO dr dQ X X-^=T{-f x^ dy dx dx dy \dy dx
-7^
{dip
dv
-,
dp dv f X i) dx dy.
w-
If further
putting x
= v,
dp
But
as T
is
^ _ !^
dv
dv
'^^
r^ "I
^
dp
''
only a function of T, we
may put
dr^dr^
dv
dT
dv
dT^
123
^
dp
we
dv
dv
dp
is
dT
^"'
''
dT
It is here
cal units.
we
must divide the expression on the right-hand mechanical equivalent of heat, which gives
by the
dT^^dQ
dp
dv
dT^JQ
dv
r
zp
dp
dr
dl'
''
this, since
it
is
fx^_fx^=(7 dp dv dp dv
(38), ^
where C is an undetermined function of temperature, and the same as Carnot's Function already mentioned.
If we equate the right-hand expressions of (.37) and we obtain the relation between C and t, as follows
is
(38),
^==-fd^
If,
= Zd^7[
^^^^'
^dT ^df~
adhering to the determination of t given by the author, it to be nothing more than the absolute temperature T, then C takes the simpler form,
we assume
c=|
As equation (33) is formed by the combination
tions,
m.
of two equaand second Main Principles, it
lix. p. 574.
first
Ann. Vol.
124
must be considered not as an expression of the second Main Principle in the form assumed by the author, but as the expression of a principle, which may be deduced from the combination of the first and second As concerns the manner in which Clapeyron has principles. treated his differential equation, this differs widely from the method adopted by the author. Like Carnot, he started from the assumption that the quantity of heat which must be imparted to a body during its passage from one condition to another, may be fully ascertained from its initial and final conditions, without its being necessary to know in what way and by what path the passage has taken place. Accordingly he considered ^ as a function of p and v, and deduced by integrating his differential equation the following expression
follows that Clapeyron's equation
for
Q'.
Q=F{T)-C,t>{p,v)
(41),
in which F{T) is any function whatever of the temperature and ^(p,v) is a, function of p and v which satisfies the following more simple differential equation
:
f:.#_f+^/=i av dp dp av
To
temperature
gas,
(42). ^
integrate this last equation we must be able to express the for the body in question as a function of p and V. If we assume that the body in question is a perfect
we have
2^=5
whence
(43).
dT_p dv B
Hence equation
(42)
dT_v_
dp
B'
becomes
^^-^f = ^
Integrating,
(**)
we have
125
of the product pv. For (43) substitute any function whatever of the temperature, so that the equation becomes,
we may by equation
(^{pv)
If
= Blogp + ^fr{T)
<p
(45).
we
{p, v) in (41),
and put
where
perature,
we
obtain,
Q = B{B-Clogp)
This
is
(46).
case of gases.
CHAPTER
VI,
1.
Among the equations of the last chapter, those deduced in 6, which refer to a partial change in the body's state of aggregation, may conveniently be treated first; inasmuch as the circumstance there mentioned, viz. that the pressure is only a function of the temperature, greatly facilitates the treatment of the subject. will in the first place consider the passage from the liquid to the vaporous condition.
.
We
of any given substance be inclosed in an Let a weight expansible envelope: of this let the partm be in the- condition of vapour, and that vapour (as necessarily follows from its contact with the liquid) at its maximum density; and let the be liquid. If the temperature T of the remainder mass is given, the condition of the vaporous part, and at the same time that of the liquid part, is thereby determined. If be also given and thereby the magnitudes of both parts known, then we know the condition of the whole mass. We and as the independent variables, will accordingly choose and will substitute for x in equations (29), (30), (.31) of the last chapter. Then these equations become
M m
dT\dm)
dT\dm)
dm
[dTJ
~ dT^ dm
W'
^'^^'
d^fdQ\ _ d_ fdQ\
_l^dQ
T
dv^
dm\dTj
dm
dQ^j^dp dm dT
dm
^"^J-
SATUEATED VAPOUR.
127
(i.e.
We may
now denote
the volume
of a unit of weight) of the saturated vapour by s, and the specific volume of the liquid by a: Both these magnitudes bear some relation to the temperature and its corresponding pressure, and are therefore, like the pressure, functions of
volume
If
we
further denote
by
v the total
v='nis+
We
cr),
a simpler ex(4),
u = s-awhence
it
follows that
V = mu
+ Ma-
(5)
whence
;^'="
^^
"(6)'
The quantity of heat which must be applied to the mass, if a unit of weight of the substance, at temperature T and under the corresponding pressure, is to pass from the liquid into the vaporous condition, and which may be shortly called the vaporizing heat, may be denoted by p then we have
;
dQ dm = P
,
(')
will further introduce into the equations the specific heat of the substance in the liquid and vaporous condition. The specific heat here treated of is not however that at constant volume, nor yet that at constant pressure, but belongs to the case in which the pressure increases with the temperature in the same manner as the maximum expansive power of the. saturated vapour. This increase of pressure has very little influence on the specific heat of the liquid, since liquids are but slightly compressible by such pressures as are herein conshall hereafter explain how this influence may sidered. be calculated, in our researches on the different kinds of For specific heat, and a single example will suffice here. water at boiling-point the difference between the specific
We
We
128
heat
at
constant
which may be
neglected. Accordingly, we may for the purposes of calculation take the specific heat of the liquid here considered as being equal to the specific heat at constant pressure, although their meaning is different. will call this specific
We
heat 0.
With vapour it is otherwise. The specific heat here considered refers, as shewn above, to that quantity of heat which saturated vapour requires to heat it through 1", if it is at the same time so powerfully compressed that even at the higher temperature it again returns to the saturated condition. As this compression is very considerable, this kind of specific heat is very different from all which we have hitherto treated of shall call it the Specific Heat of Saturated Steam,
We
shall denote it by H. Bringing in the two symbols G and H, we may now at once "write down the quantity of heat which is necessary to give the increase of temperature dT to the quantity of vapour m, and the quantity of liquid m. The result will be as
and
follows
mHdT+(M-m)CdT,
whence
'^ = mH+{M-m)C,
or otherwise
^ = m{H-G)+MG
From
equations
(7)
(8).
and
(8)
we have
dp
d^(dQ\
U^h^-'
tions
(^>-
(1, 2, 3)
SATUBATED VAPOUR.
129
(U).
%^C-H=u%
^+a-F=|
P
(12),
= Tu^T
^^^)-
These are the fundamental equations of the Mechanical Theory of Heat as relates to the generation of vapour. Equation (11) is a deduction from the first fundamental principle, (12) from the second, and (13) from both together. If it is desired to use the ordinary and not the mechanical measures for the quantities of heat, we need only divide all
the members of the foregoing equations by the mechanical equivalent of heat. In this case we will denote the two specific heats and the hjeat of vaporization by new symbols, putting
c
= ^; ^ = 2;!
dr
''
= A'
(^*)-
u fdp\
,, .,
dT^'-^^EKdT)
^^'^'
_+C-A.= y
(16),
2.
Specific
As the foregoing equations (15), (16) and (17), of which however only two are independentjhave thus been obtained by means of the Mechanical Theory of Heat, we may make use.of them in order to determine more closely two magnitudes, of which one was previously quite unknown and the otber only known imperfectly; viz. the magnitude h and the magnitude
s contained in m. If we first apply ourselves to the magnitude h, or the Specific Heat of Saturated Steam, it may be worth while in
c.
ISO
the first place to give some account of the views formerly promulgated concerning this magnitude. The magnitude h is of special importance in the theory of the steam engine, and in fact the first who published, any distinct views upon it was the celebrated improver of the steam engine, James Watt. In his treatment of the subject he naturally started from those views which were based on the
older theory of heat.
To
that the so-called total Jheat, i.e. the total quantity of heat taken in by a body during its passage from a given initial condition to its present condition, depends only on the present condition and not on the way in which the body has been brought into it ; and that it accordingly may be expressed as a function of those variables on which the condition of the body depends. According to this view we must in our case, in which the condition of the body composed of liquid and vapour is determined by the quantities T and m, consider this quantity of heat ^ as a function of T and accordingly we have the equation
I., viz.
mentioned in Chapter
(i^ = A (^ _ A dT dm [dTj
\<hnj
If
n "'
we here substitute for the two second differentials their values given in equations (9) and (10), we have
^T +
or dividing
dp
C-B=0,
0,
by JE
+c-h =
whence we have, to determine
h,
the equation,
c
^=
^+
(18).
This was in fact the equation which was formerly used to determine h, though not quite in the same form. To calculate this equation we must know the differential cofi from
efficient
change of temperature,
SATURATED VAPOUR.
131
Watt had made experiments on the vaporizing heat of water at different temperatures, and was thereby led to a result, which may be expressed by a very simple law, commonly called Watt's law. This in its shortest form is as follows " The sum of the free and ktent heat is always constant." By this is meant that the sum of the two quantities of heat, which must be imparted to a unit of weight of water, in order to raise it from freezing point to temperature T, and then at that temperature convert it into steam, is independent of the temperature T itself. The quantity of heat required for heating the water is expressed by the
:
integral
in which a is the absolute temperature of freezing point. The heat of vaporization is represented by r. Watt's law therefore leads to the equation
r-l-f
Differentiating,
CZ!Z'
= Constant
(19).
J+c =
,
(20),
combining
this
=
:
(21).
If steam of maximum density the' following principle changes its volume within a vessel impermeable to heat, it always preserves its maximum density. More recently Eegnault made very careful experiments on the changes in the heat of vaporization with changes in the temperature* and thereby discovered that Watt's law, according to which the sum of the free and latent heat is always constant, does not agree with the facts, but that this sum hasan increasing value as the temperature rises. The result of his experiments is expressed in the following equation, in
by
* Relations de Experiences,
t. i. ;
Mem.
de VAcad.,
t.
ixi., 1847.
9-2
132
which instead of the absolute temperature T we introduce the temperature t reckoned from the freezing point
+ [ccif= 606-5
'
+ 0-305 i
t,
(22).
for -j-
^+c = 0-305
Combining
this equation with (18)
(2.3).
we have
(24).
^=-305
This 'was the value of h, which it was supposed, after the publication of Eegnault's experiments, must be substituted for the value zero and applied to the theory of the steam engine. Hence arose the idea that if saturated steam be compressed, and thereby heated, in such a way that it always has exactly the temperature for which the density is a maximum, it must take in heat from without and conversely in expanding, in order to cool itself in a manner corresponding to the expansion, it must give out heat from itself Hence we must further conclude that if saturated steam be compressed in a vessel impermeable to heat, a fall of temperature must take place whilst in expanding the steam will not' remain at its maximum density, inasmuch as its temperature
; ;
cannot
in
fall so
Having thus explained the conclusions previously drawn relation to h, we will now see what may be concluded
from the equations here developed. The quantity h occurs in the two equations (15) and (16); but the first of these also contains the quantity u, which cannot at present be considered as accurately known it is therefore less convenient for determining h than the latter, which in addition to A contains only such quantities as the experiments of Regnault have determined with great accuracy in the case of water, and of many other fluids. This equation may be written , dr r
;
^^dT'^''~T
f^25)-
SATURATED VAPOUR.
133
have thus obtained by the Mechanical Theory of Heat a equation for determining h, which differs from the equation (18), previously assumed to be correct, by the negative T quantity^, the value of which quantity is thus of great
We
new
importance.
3.
following Regnault, give to the sum of the first two symbols on the right-hand side the value 0'305. To determine the last symbol we must know the value of r, as a function of the temperature. Equation (22) gives us
= 606-5 -t-0-305-fcd{
Jo
:
(26).
The specific heat of water c is determined according Regnault by the following formulae
c
to
=1
-f-
0-00004!!
+ 00000009''
(27).
Applying
this,
by
i A
= no0-30o
606-5
-0-695-000002f- 0-0000003*'
2^3-j-^
,,
...(29).
The expression for r given in (28) is inconvenient from its length, and the experiments on the heat of vaporization at different temperatures, valuable as they are, can scarcely possess such a degree of accuracy as to make so long a formula necessary to represent them. Accordingly in his paper on the theory of the steam engine the author preferred to use the following formula
r
= 607-0-708<
(30).
The manner in which the two constants in the formula, are determined will be more closely examined further on,
134
Here "vve will only give a comparison of some values determined by both formulse, in order to shew that the difference between them is so small,that one may be substituted for the other without danger
in describing the steam engine.
*
SATURATED VAPOUR,
135
only that it remains exactly in the condition of saturated steam. Accordingly if this last condition is to hold heat must be imparted during the expansion. Should the original saturated steam be contained in a vessel impermeable to heat, it will be superheated during compression and will be in part condensed during expansion.
The conclusion that the specific heat of saturated steam negative was drawn by Rankine and by the author independently, and about the same time*. Eankine however developed only the first of the two equations (15) and (16), which contain A, and this in a somewhat different form. The second it was impossible for him to develope, since he was without the second fundamental principle, which was necessary thereto. Since in the first equation there occurs together with h the specific volume of the saturated steam, which is contained in u, Rankine in order to determine this applied to saturated steam the law of Mariotte and Gay-Lussac, which, as we shall see further on, is inaccurate. More exact determinations of h could only be accomplished by means of equation (16), which was first established by the author.
is
4.
When equation (25) was first published, Eegnault's determinations of the specific heat and heat of vaporization as functions of temperature had been performed only for the case of waterf; and therefore the numerical value of h could be given for water only. Regnault has since extended his measurements to other fluids J, and it is now possible to apply the equation to obtain the numerical value of h for these fluids. We thus obtain the following results.
Bisulphide of Carbon : have
[ cdt
CSjj.
According to Regnault we
= 0-23523 + O'OOOOSlSi',
Jo
* Eankine's paper was read tefore the Eoyal Society of Edinburgli in February, 1850, and printed in the Transactions, Vol. xx. p. 147. The author's paper was read before the academy of Berlin ill February, 1850, and printed in Poggendorf's Annalen, Vol. lxxix. pp. 368 and 500.
i.,
ii.,
136
+
Jo
90
+ OUeOlt - 00004123^
whence we have
9000-0-08922-00004938<''
273+T
'
SATURATED VAPOUR.
137
In the case of Ether therefore the specific heat of saturated steam has, at least at ordinary temperatures, a positive
value.
Chloroform
CHCI3.
cd<
f Jo'
= 0-23235 + 0-0000507a',
= 67 + 0-1375
r+l
whence we have
c =-
cdt
67
- 0-09485* - 0-00005072i(''
'
273T"^
deduced
and from
138
and from
deduced
SATURATED VAPOUR.
139
In all the formulae for h given above we see that its value increases as the temperature rises. In the only case, that of ether, in which it is positive at ordinary temperatures, its absolute value increases as the temperature rises. In the
other cases, in which it is negative, its absolute value diminishes; it thus approaches to zero, and it would appear that
some higher temperature it would attain the value zero, and at stiU higher temperatures would become positive. To determine the temperature at which A = 0, we have by equaat
tion (25)
^+c-J=0
In
(32).
this equation we must, as above, express c and r as functions of t, and then solve it with regard to t. The empirical formulis of Regnault, by means of which we have expressed c and r as functions of t, should not of course be applied much beyond the limits of temperature within which Regnault carried out his experiments. Hence the determination of the temperature for which A. = is in many cases impossible, as for instance with water, where the equations obtained by putting A = in (29) and (31) would lead to a value for t of about 600", whereas the equations are only applicable up to somewhat over 200". But with other fluids the temperature for which h = Q, and above which h is positive, lies within the limits of application of the formulae. Thus Cazin* calculates this temperature for Chloroform at 123"48, and for Bi-chloride of Carbon at 128-9.
The result arrived at by theory, that the Specific Heat of saturated steam is negative, and that therefore saturated steam, if expanded in a non-conducting envelope, must partially condense, has since been experimentally proved by cylindrical vessel of metal was fitted at the two Hirn-)-. ends with parallel plates of glass, so that it could be seen through. This, when filled with steam at high pressure was
* Annates de Chimie et de Physique, Series iv. Vol. xiv. t Bulletin 133 de la Soci4U Industrielle de Mulhmise, p.
137.
140
perfectly cldar; but when a cock was suddenly opened, so that part of the steam escaped, and the remainder expanded, a thick cloud appeared in the interior of the cylinder, proving
Subsequently,
when
Renault's Relation des Eceperiences had appeared, containing the data, given above, to determine h for other fluids, and shewing that for ether h must be positive. Him proceeded to experiment with that vapour also. His description " To the neck of a strong crystal flask I is as follows * connected a pump, the capacity of which was nearly equal to that of the flask, and which was provided with a cock at the bottom. Some ether was poured into the flask, and it was immersed to the neck in water at about 50. The cock was then kept open, until all the air was assumed to be expelled. Then the cock was closed, and the pump plunged into the hot water with the flask: whereupon the ether vapour raised the piston to the top. The apparatus was now suddenly taken from the water, and the piston forced rapidly down. At this moment, but for a moment only, the flask became filled with a distinct cloud." It was thus shewn that ether vapour behaves conversely to steam, partially condensing, not during expansion, but during compression; a fact which is in accordance with the opposite sign of h in the two
II.
Volume
of
cases.
To check this conclusion Him made an exactly similar experiment with Bi-sulphide of Carbon. The result was that on forcing down the piston the flask remained perfectly transparent. This is again in accordance with the theory, since with Bi-sulphide of Carbon, as with water, h is negative, and compression of the vapour produces a rise of temperature, and not a fall. Some years later Cazin"f, aided by the Association scientifique, made with great care
and skill a similar series of experiments, in some respects more extended. He used as before a cylinder of metal, fitted with glass at the ends. This was placed in a bath of oil, so
as to give
it the exact temperature proper for the experiment. The first series of experiments embraced only the expansion of steam; the arrangement was such that, when the cylinder was filled with vapour, a cock could be opened, * Cosmos, 10 April, 1863.
+ Anwilcs de Chimie
et
de J^hjfsique, Series
it.
Vol, xit.
SATURATED VAPOUR.
141
through which a part of the vapour escaped either into the atmosphere or into an air vessel, the pressure in which could be kept at any given point below the pressure of the vapour. In a second series of experiments a pump was connected with the cylinder; this was placed in the same bath of oil, and the piston could be moved ra^dly backwards or forwards by specigj. mechanism, so as to increase or diminish the volume of the vapour. By these experiments the results obtained by Hirn for steam and ether were confirmed, and with the second apparatus a double proof was given in each case, viz. both by rarefaction and condensation. Steam formed a cloud during rarefaction, whilst it remained quite clear during condensation. Ether, on the contrary, formed a cloud during condensation, and remained clear during rarefaction. Some special experiments were further made with vapour of chloroform. As mentioned above, in the case of chloroform A, which is negative at lower temperatures, becomes zero at a temperature which Cazin has calculated at 123'48, and at This vapour must thus still higher temperatures is positive. partially condense during expansion at lower temperatures, and must partially condense during compression at higher temperatures beyond the point of transition. With the first apparatus, which only allowed of expansion, clouds were
observed during expansion at temperatures up to 123. At temperatures above 145 there was no formation of cloud. Between 123 and 145 the conditions were somewhat different according to the degree of expansion. With a small degree of expansion there was no cloud with a higher degree some formation of a cloud appeared towards the end of the proThe explanation of this is simple. The high expansion cess. had produced a considerable fall of temperature, and the vapour had thereby been reduced to the temperature at which expansion is accompanied by a fall of tempera^ The result is thus completely in accordance with the ture. theory. With the second apparatus the vapour of chloroform formed a cloud during expansion up to 130, whilst it remained perfectly transparent during compression. Above 136 a cloud was formed during compression, whilst it remained clear during expansion. The theory is hereby more fully established than by the first apparatus. The circumstance
;
142
that the temperature at which the law of the vajpour changes appeared in these experiments to lie between 130 and 136, whilst theory gives it at 123-48'', is not a matter of great importance. On the one hand these experiments are not adapted for an accurate determination of this temperature, because they always involve finite changes of volume of considerable magnitude, whereas the theory embraces indefinitely small changes only. On the other hand, Cazin hiriiself mentions that his chloroform was not chemically pure, and required for a given vapour-pressure a higher temperature than that found by Eegnault. Having regard to these circumstances, the theory must be considered as being fully confirmed by experiment.
6.
Specific
will
consider the second of the two quantities viz. s, or the specific volume of the saturated vapour.
We
now
It was formerly the custom to use the law of Mariotte and Gay-Lussac, in order to calculate the volume which a gas assumes under difierent conditions of temperature and pressure, and to take no account of whether the vapour was in the saturated or superheated condition. It is true that from many quarters doubts were expressed, whether vapours really followed this law up to the saturation point: but, as the experimental determination of the volumes offered great difficulties, and a theoretical determination was impossible from the want of well-established principles, it remained the custom to apply the above law in this case, so as at least to arrive at some sort of determination of the volume of saturated vapour. But the' equations obtained by the author, and given at the end of 1, now offer us a means of arriving at a strict theoretical calculation for the volume of saturated vapour, which, when the data are given, may be worked out in practice. For in these equations occurs the quantity u, which =s a; where a is the specific volume of the fluid. This, as a rule, is very small in comparison with s, and may be neglected in many calculations but it is still a known quantity, and may be taken account of without diffi;
culty.
SATURATED VAPOUK.
143
we obtain
T(s-a)
^
^~~ ^ dT
dp
^^^^'
or,
s,
s=-=f- + <r IT
pour
Er
{U).
specific
7.
We
If
will first
whether saturated steam follows the law of Mariotte and GayLussac, or how far or in what way it departs from it.
it
must hold
^=
or,
const.,
substituting a
+ f for
1
T,
and multiplying by ^,
a
7
TV Ps U^ a+t
, = const.
(33), substituting
a+<
for T,
we
obtain
y,p{s-a)---=
^
=-T-
(35).
'pat
As
the difference
[s ->
side of these
two equations
144
ascertain
otte
how saturated steam is related to the law of Mariand Gay-Lussac, we have only to enquire whether the
right-hand side of the last equation is constant, or varies with the temperature. To ascertain whether the successive values of an expression are equal to each other, or in what way they depart from each other, is a very simple matter and the form of equation (35) is very well adapted for this purThe author has calculated the values of this expression pose. for a series of temperatures from 0 to 200, applying the numbers given by Regnault to r and p. For r, the heat of vaporization, the equation (28) was used, viz. r
606-5
simple formula (30) might have been used without difference in the results. To obtain p, the author first applied the numbers which Regnault has published in his well-known large Tables, in which the pressure of steam 32 to 230 is given. He found howfor every degree from
The more
any great
ever some peculiar variations from the regular course of the numbers, which in certain ranges of temperature had quite a different character from what they had in others and he soon discovered that the source of these variations lay in the fact that Regnault had calculated his numbers by empirical formulae, and that for different ranges of temperature he
;
It then appeared desirable emancipate himself entirely from the influence of empirical formulae, and to confine himself to those numbers Which express simply the results of the observations, because these are specially adapted for comparison with theoretical results. Regnault, in order to obtain from his numerous observations the most probable values, used the aid of graphical methods. He constructed curves of which the abscissae represent the temperature, and the ordinates the pressure p, and which run from 33 to \- 230. From 100 to 230 he also constructed a curve in which the ordinates represented not p itself, but the logarithm oip. From this have been obtained the following values, which may be considered the most direct results of his observations, and from which were also taken the values which served for the
different formulae.
to the author to
had empjoyed
SATURATED VAPOUK.
t
145
in Centi-
146
By
and of the values of r given by the equation stated dt above, and lastly of the value 273 for a, the values were calculated which the expression on the right-hand side of equation (35) and therefore likewise the expression
-'^-
assumes for the temperatures 5, 15", 25, etc. These values are given in the second column of the table below. For temperatures over 100 the two series of numbers found above for p were both made use of, and the two results thus obtained are placed side by side. The meaning of the third and fourth columns will be more fully explained below.
1.
SATUEATED VAPOUK.
This table at once shews that
as
147
(s
<r)
- is not
constant,
it must be if the law of Mariotte and Gay-Lussac holds, but decreases decidedly as the temperature rises. Between 35 and 95 this decrease appears very regular. Below 35 the decrease is less regular, the simple explanation being that
p and
-^ are very
small, and therefore small errors in their determination, which are quite within the limits of errors of observation, may yet
relatively important. Above 100 the values of the expression are not so regular as between 25 and 95, but shew on the whole a similar rate of decrease and if we make a graphic representation of these values, it is found that the curve, which below 100 passes exactly through the points determined by the numbers contained in the table, can be readily continued up to 230 in such a way that these points are distributed equally on both sides of it. The course of this curve can be expressed with sufficient accuracy for the whole extent of the table by an equation of the form
:
become
^P(-'^)^i = "*-'**
where
e is the
(37),
base of Napierian logarithms, and m, n, k are latter from the values which the curve gives for 45, 125, and 205, we obtain
constants.
If
we determine the
(37a),
convenience
we use common
logarithms,
we
finally
obtain
Logr31-549--^.p(5-o-)
a-ht
The numbers contained in the third column of the table are calculated from this equation, and in the fourth column are given the differences between these and the nUmbers in
the second cplumn,
102
148
8.
whick we can ascertain more exactly the mode in which steam departs from the law of Mariotte and Gay-Lussac, Assuming, this law to hold, we shall be able to put
PI: a
+t
0".
The
differential
viz.
at \psgj
the
well-known
coefficient of
= 0003665,
s -in
Instead
(s
we use
a
place of
tr),
the equation
me''*
X
F^o
+t
,
(39).
whence
d
/ps_
=
dt [ps^
-X a
m - n [I + k (a + t)] mn
e''t
,.
(40).
Thus the differential coefficient is not a constant, but a function which decreases as the temperature increases. This function, if we substitute for m, n, and Ic the numbers given in (37a), has amongst others the following values, for different temperatures.
t.
SATURATED VAPOUR,
140
From the above table we see that it is only at low temperatures that the variations from the law of Mariotte and Gay-Lussac are small, and that at higher temperatures, e.g. above 100, they can by no means be neglected. glance at the tabl-e is sufficient to shew that the values
found for -n at
\psj
-,
whereas
it
is
the gases which vary considerably from the law of Mariotte and Gay-Lussac, such as carbonicacid and sulphuric
acid, the coefficient of expansion is not smaller, but greater than the above number. Hence this differential coefficient cannot be taken to correspond with the coefficient of expansion which relates to increase of volume by heating at constant pressure, nor yet with the figure obtained if we leave the volume constant during the heating, and then observe the Thus we have here a third increase of the expansive force.
>
viz.
that in which the pressure increases during the heating under the same conditions as in the case of steam, when this retains its maximum density; and this case must be considered for carbonic acid likewise, if we are to establish a comparison.
Steam has at about 108 an expansive force equal to 1 metre of mercury, and at 129J equal to 2 metres. We will examine what takes place with carbonic acid, if this is also heated by 21^, and thereby the pressure increased from 1 to 2 metres. Regnault* gives as the coefficient of expansion of carbonic acid at constant pressure 0'00371 if the pressure is 760 mm., and 0"003846 if the pressure is 2520 mm. For a pressure of 1500 mm., the mean between 1 and 2 metres, the coefficient of expansion, if assumed to increase in proportion to the '003767. If carbonic acid pressure, will have the value under this mean pressure is heated from 0 to Sl^'", the
quantitv
wiU
1
increase from 1 to
+ 0-003767
21-5
= 1-08099.
that, i|
1. 1.
Mem. Mem.
1.
6.
150
carbonic acid, which had a pressure of about 1 m. at a temperature of aiiout 0, is loaded with a pressure of 1 '98292 m., the quantity pv decreases in the ratio of 1 0'99146 or for an increase of pressure from 1 to 2 ms., it will decrease in the 0'99131. Now if both of these take place ratio of 1 together, viz. a rise of temperature from 0 to 21^, and
: ; :
must
nearly,
increase from 1 to 108099 x 0-99131 = 1-071596 very whence we obtain as the mean value of the differen-
d
tial coefficient
/
,
pv^
at \pi\.
^ ^Q^f^ 21-5
000333.
We thus see that for the case here considered we have a value for carbonic acid which is less than 0003665, and there is therefore less, ground for surprise at obtaining the
same
If
result for
steam at
its
maximum
density:
seek to determine on the other hand the actual coefficient of expansion for steam^ or the number which shews how far a quantity of steam expands, if it is taken at a given temperature and at its maximum density and then heated, apart from water, at constant pressure, we shall certainly obtain a value which is larger, and probably much larger, than 0-003665. 9. Formula to determine the Specifia Volume of Saturated Stea/m, and its comparison with experiment.
we
From
equation
(37),
relative values of s
a;
and equally from equation (34), the and therefore to a close approxima-
s, may be calculated for different temperatures, without needing to know the Mechanical Equivalent E. If however we wish to calculate from the equations the absolute value of s, we must either know E^ or must attempt to eliminate it by the help of some other known quantity. At the time when the author first began these researches, several values of had been given by Joule, taken from various methods of experiment: these differed widely from each other, and Joule had not announced which he considered the
tion those of
SATURATED VAPOUR.
151
most probable. In this uncertainty the author determined to attempt the determination of the absolute value of s from another starting point, and he believes that his method still possesses interest enough to merit description. The specific weight of gases and vapours is generally expressed by comparing the weight of a unit of volume of the gas or vapour with the weight of a unit of volume of atmospheric air at the same pressure and .temperature. Similarly the specific volume may be expressed by comparing the volume of a unit of weight of the gas or vapour with the volume of a unit of weight of atmospheric air at the same pressure and temperature. If we apply this latter method to saturated steam, for which we have denoted the volume of a unit of weight by s, and if we designate by v' the volume of a unit of weight of atmospheric air at the same pressure and temperature, then the quantity under consideration is
given by the fraction
For
we have the
<r
:
by neglecting
g^E(a+J)^
For
v'
we have by
the equation
V =Jti
,
jj,
+
.
P
These two equations give
-,
= -( e*')
for
(42).
If
t^,
we
obtain
[->)
p7-(TO-e*').
152
If
by the help of
this equation
we
factor T=rr-
Ka
from
(42),
we
obtain
t^(s_\
The question
t,
m-n^
for
^^g.
is
now whether,
its reciprocal
which expresses
the specific weight of the steam at temperatute t, can be determined with sufficient certainty. The ordinary values given for the specific weight of steam refer not to saturated but to highly superheated steam. They agree very well, as is known, with the theoretical values which may be deduced from the well-known law as to the relation between the volume of a compound gas, and those of the gases which compose it. Thus e.g. Gay-Lussac found for the specific weight of steam the experimental value 0"6235; whilst the theoretical value obtained by assuming two units of hydrogen and one unit of oxygen to form, by combining, 2 units of steam, is
2 X 006926 X 1-10563
^ = 0-622.
This value of the specific weight we cannot in general apply to saturated steam, since the table in the last section,
<
divergence from the law of Mariotte and Gay-Lussac. On the other hand the table shews that the divergences are smaller as the temperature is lower; hence, the error will be insignificant if we assume that at freezing temperature saturated steam follows exactly the law of Mariotte and GayLussac, and accordingly take 0-622 as the specific heat at that temperature. In strict accuracy we must go yet further and put the temperature, at which the specific weight of saturated steam has its theoretical value, still lower than freezing point. But, as it would be somewhat questionable to use equation (37), which is only empirical, at such low temperatures, we shaU content ourselves with the above
SATURATED VAPOUH.
assumption.
153
0,
Thus giving
to
t^
the value
and
at the
,
same
ob-
= 0-622
and therefore
\ /o
(-)\ \v Jo
= jr^ we
0622
.(44).
From
this
h
s,
calculated for each temperature. The foregoing equa,tion may be thrown into a more convenient form by putting.
may be
.(45),
and by giving to the constants M, iV, and a the following values, calculated from those of to, n, and k :
-3f= 1-663; iV=0-05527,- a = 1-007164.. .(45a).
To give some
we
give
re-
and of
*
its
ciprocal
which
by
154
judges. The author believes however that it is now generally accepted as correct. It has also received an experimental verification by the experiments of Fairbairn and Tate*, published in 1860. The results of their experiments are compared in the following table, on the^ one hand with the results previously obtained by assuming the specific weight to be 0-622 at all
temperatures, and on the other hand with the values calculated by equation (4-5)'.
Temperature in Degrees
Centigrade.
SATURATED VAPOUR.
155
perimental values are in general yet further removed from those previously obtained than are the values derived from the author's formula.
10.
Heat from
Since we have determined the absolute values of s, without assuming the mechanical equivalent of heat to be known, we may now applj' these values, by means of equation (17), to determine the mechanical equivalent itself. For this purpose we may give that equation the following form
=
The
coefficient of s
T^Cs-.')
().
for different
o- in this equation may be calculated temperatures by means of Regnault's tables. For example, to calculate its values for 100", we have given for
-ij
metres of mercury. To reduce this to the measure here employed, viz. kilogrammes per square metre, we must multiply by the weight of a column of mercury at temperature 0", 1 square metre in area and 1 millimetre in height, that is by the weight of 1 cubic decimetre of mercury at 0". As Regnault gives this weight at 13"o96 kilogrammes, the multiplication gives us the number 369 "8. The values of {a + t) and of r Hence we have at 100 are 373 and 536'5 respectively.
/ a. rt ^"' "^ '
;:
^
~dt
_ 373
X 369-8 _ _.^
536-5
"^''^'
and equation
(46)
becomes
^=257(s-o-)
(47).
We have
known,
to determine the quantity (s a), or, since o- is the quantity s for steam at 100. The method formerly
now
pursued, i.e. to use for saturated steam the same specific weight, which for superheated steam had been found by experiment or deduced theoretically from the condensation of
156
water, led to the result, that a kilogram of steam at 100 should have a volume of 1'696 cuhic metres. From the foregoing however it appears that this value must be considerably too large, and must therefore give too large a value for Taking the specific heat the mechanical equivalent of teat. as calculated by equation (45), which for 100 is 0"645, we obtain for s the value 1'638. Applying this value of s we get from equation (47)
^ = 421
....(48).
This method therefore gives for the mechanical equivalent of heat a value which agrees in a very satisfactory manner with the value found by Joule from the friction of water, and with that deduced in Chapter II. from the behaviour of
gases
;
ment may
both of which are about equal to 424. This agreeserve as a verification of the author's theory as to
dm
^'
^ = m{E-G) + Ma
Hence we may
first
dQ = pdm +
(49).
By
equation (12) we
may put
TT
whence equation
(49)
becomes
(50).
dQ=pdm+ m(^^-^ + Ma dT
satueated vapour.
Since p
is
157
a function of
only,
and therefore
we have
or
(51),
(52).
These equations are not integrable if the two quantities, whose dififerentials are on the right-haad side, are independent of each other, and the mode of the variations thus left undetermined. They become integrable as soon as this mode is determined in any way. We can then perform with them calculations exactly similar to those given for gases in Chapter II. We will for the sake of example take a case which' on the one hand has an importance of its own, and on the other derives an interest from the fact that it plays a prominent
part in the theory of the steam-engine. The assumption is that the mass consisting both of liquid and vapour changes its volume, without any heat being imparted to it or taken from it. In this case the temperature and magnitude of the gaseous portion also suffers a change, and some external work, positive or negative, must at the same time be perof the gaseous portion, its volume formed. The magnitude V, and the external work W, must now be determined as functions of the temperature.
12.
the
Mass.
As
we may put dQ =
0.
becomes:
Td (^-P^+MCdT=0
If
(53).
we divide this equation by E, the quantities p and C, which relate to the mechanical measure of heat, change into r and If we also c, which relate to the ordinary measure of heat. divide the equation by T, it becomes:
dl-jrj+Mc-^=0
(53a).
.,
158
The first member of this equation is a simple differential, and may at once be integrated the integration of the second is also always possible, since c varies only with the temperature T. If we merely indicate this integration, and denote the initial values of the various magnitudes by annexing
:
the figure
1,
we
or
mr
f^ dT "^^^-mQ^
m,r,
,,
(54).
Actually to perform the integration thus indicated, we the empirical formula for c given by Regnault, For water this formula, already given in (27), is as follows
may employ
= 1 + 000004 +
00000009*'.
thus seen to vary v6ry slightly with the temperawater assume c to be constant, which will not seriously affect the accuracy of the Since
c is
ture,
we
results.
Hence
(54)
becomes:
(oo),
~^=-^-Mclog-^
whence
m = f(^-^clogJj
(56).
If we here substitute for r the expression given in (28), or in a simpler form in (30), then will be determined as a func-
tion of tenaperature. To give a general idea of the values of this function, some values have been calculated for a special case, and collected in the following table. The assumption is that the vessel contains at first no water in a liquid condition, but
with steam at its maximum density, so that in equation (56) we may put m^ = M. Let there now be an expansion of the vessel. compression would not be admissible, because on the assumption of the absence of water at the commencement, the steam would not remain at its
is filled
SATURATED VAPOTJE.
159
maximum
density, but would be superheated by the heat developed in the compression. In expansion on the other hand the steam not only remains at its maximum density, but a part of it is precipitated as water ; and the diminution of thus produced is exhibited in the table. The initial
temperature
is
given for the moments when the temperature has sunk through expansion to 125", 100, etc. As before, the temperature is reckoned from freezing point, and is denoted by t, to distinguish it from the absolute temperature T.
160
HEAT."
The
calculated
by this equation
for
Under
if
the two
ordinary assumptions in the theory of the steam-engine were correct viz. (1) that steam in expansion remains at its maxidensity without any part of it condensing; (2) that steam follows the law of Mariotte and Gay-Lussac. On these assumptions we shall have
:
mum
.t
SATUEATED VAPOUK.
In
this equation
161
we may substitute
for
mu -
the expres-
sion given in equation (57), and may then perform the inteThe result however is obtained in a more convenient form as follows. By (13) we have:
gration.
and from
(53)
we
obtain
MCdT = -d[m(p-up)]-MGdT
we
obtain
(61).
W = m;(p^-u^;)-m(p-wp) + MC{T^-T)...(Q2).
If in this equation we substitute for p and (7, according to equation (14), the values JEr and JEc, and collect together the as a factor, we have terms which contain
(63).
we may
calculate
W,
since
mr and
W the values of
made
of the
the equations (56) and (58). This for the same case as before, and
-jr, i.e.
of weight
during expansion, are given in the following table ; the unit of weight is here a kilogram and the unit of work a kilois 42355, as found by grammetre. The value used for
Joule.
As a comparison with
c.
it
may be
11
162
again mentioned that the value obtained for the work done during the actual formation of steam, as this overcomes the external pressure, is 18700 kilogrammetres per kilogram of water, evaporated at temperature 150 and at the corresponding pressure.
CHAPTER
VII.
1.
Fundamental
of Fusion.
Whilst in the case of vaporization the influence of the external pressure was eaxly observed, and was everywhere taken into account, it had hitherto been left out of account in the case of fusion, where it is much less easily noticed. little consideration however shews that, if the volume of a body
changes during fusion, the external pressure must have an influence on the process. For, if the volume increases, an increase of pressure will make the fusion more difficult, whence it may be concluded that a higher temperature is necessary for fusion at a high than at a low pressure. If on the other hand the volume decreases, an increase of pressure will facilitate the fusion, and the temperature required will be less, as the pressure is greater. To examine more exactly the connection between pressure and fusion-point, and the peculiar changes which are sometimes connected with a change of pressure, we must form the
equations which are supplied for the process of fusion by the two fundamental principles of the Mechanical Theory of Heat. For this purpose we pursue the same course as for vaporization. We conceive an expansible vessel containing a certain quantity iHf of a substance, which is partly in the Let the liquid part solid, and partly in the liquid condition. have the magnitude m, and therefore the solid part the magnitude Mm. The two together are su^osed to fill the
vessel completely, so that the content of tffi; vessel is equal to V, the volume of the body. are given, thie If this volume v and the temperature
magnitude
m is
thereby determined.
To prove
this, let
us
112
164
first
Let
it
be
is exactly the also in such a condition that the temp'ei-ature if in melting temperature at that particular pressure.
Now
condition the magnitude of the liquid part were to increase at the expense of the solid, the expansion which must then result would produce an increase of pressure against the walls of the vessel, and therefore an increased reaction of the walls against the body. This increased pressure would produce a rise in the fusion-point, and since the existing temperature would then be too low for fusion, a If on the consolidification of the liquid part must begin. trary the solid part were to increase at the expense of the liquid, the point of fusion would thereupon sink, and since the existing temperature would then be higher than the fusion -point, a fusion of the solid part must begin. Next, if we make the opposite assumption, viz. that the volume decreases during fusion, then if the solid part increase there must be a rise of pressure and in consequence a partial melting, and if the liquid part increase there must be a fall in pressure and in consequence a pai-tial solidification. Thus on either assumption we have the same result, viz. that only the original proportions of the liquid and solid parts (which proportions correspond to the pressure which gives a temperature of fusion equal to the given temperature) can be permanently maintained. Since then the magnitude is determined by the temperature and volume, this volume will conversely be determined by the temperature and the magnitude m; and we may choose and as the variables which serve to determine the condition of the body. It now remains to express ^ as a function of T. Here we may again apply -equations (1), (2), (3) of the last chapter, viz.
this
dT [dm]
.d_
dm \dTJ '^'dT^d^'
d_ fdQ\
fdQ\
_ jL^
dv_
dQ
dT\dm)
dm\dT)
^
T^dm'
iQ^rpdp dm dT
dm
If we denote by o-, as before, the specific volume (or volume of a unit of weight) for the liquid condition of the body,
165
and the
specific volume for the solid condition for the total volume v of the body,
by
t,
we have
V = maor
+ {M m) t, v=m{a--T) + MT
(1),
^^^e^e
If further
p',
^='^-'^
(2).
d^-p
To
express
(^^-
differential coefficient of Q,
we
must employ symbols for the specific heat of the body in the liquid and in the solid condition. Here, however, we must make the same remark as in the case of vaporization, viz.
that it is not the specific heat at constant pressure which is treated of, but the specific heat for the particular case in which the pressure alters with the temperature in such a way that the temperature shall always be the temperature of In the case of vaporizafusion for that particular pressure. tion, where the changes of pressure are generally small, it was possible to neglect the influence of the change of pressure on the specific heat of the liquid body, and to consider the specific heat of a liquid body, as found in the formula, to be In the equivalent to the specific heat at constant pressure. present case small changes of temperature produce such great changes of pressure, that the influence of these oq the will, therefore, under specific heat must not be neglected. the present circumstances, denote by 0" the specific heat of the liquid, which in the formula for vaporization we denoted by C. The specific heat of the solid body may be denoted in this case by K'. Applying these symbols we may
We
write
^=m{0'-K')+MK'
(4).
166
From
we have
'
dT\dm)~dT
,dm \<
Inserting these sralues,
U^)-^-""'
for
(^)-
-^
given in
(3),
in
%+,K'-0' = i.-r)^
(7),
%-^^'-^'4
/ = T(<r-r)^
^'\
(9).
In these equations the heat is supposed to be measured by mechanical units. If the heat is to be measured in ordinary
units,
following sj'mbols
The
+'--'-^'
g,+ A'-c' = J
<">
(12),
'=-^
T(a--T) fdp\
two.
(')
These are the equations required, of which the first corresponds to the first Fundamental Principle, and the second to the second, whilst the third is a combination of the other
2.
167
and
Temperature of
Ficsion.
The
pendent,
foregoing equations, only two of -which are indemaybe applied to determine two quantities hitherto
unknown.
way
in which the temperature of fusion depends on the pressure. The equation may be written
^T Tj.-r)
dp
Er'
^^*>-
This equation in the first place justifies the remark already made, that if a body expand during fusion the point of fusion rises as the pressure increases ; whereas if the body contracts the point of fusion falls. For according as cr is greater or less than t so is the difference a- t, and therefore also the
differential
coefficient
dT -^
..
positive or negative.
Again, by
this equation
we may
We will
lume
calculate this value for the case of water. The voin cubic metres, or the value of <r, for a kilogramme of water at 4 C. is O'OOl. At freezing point it is a little greater, but the difference is so small that it may be neglected. The volume in cubic metres, or the value of t, for a_ kilogram The heat of fusion for water, or the of ice is 0001087. value of r, is according to Person 79. At freezing point equals 773, and for we wiU take the value 424. Hence
we
obtain
dT _ "
dp
If the pressure is given not in mechanical units (kilograms per square metre), but in atmospheres, we must multiply the
This gives us
^=-000733,
d/p
i.e. an increase of pressure of one atmosphere point of fusion by G'00733 of a degree Cent.
168
3.
Experimental Verification of the Foregoing Result. The conclusioii that the melting point of ice is lowered by an increase of pressure, and the first calculation of the amount, are due to James Thomson, -who derived from Gamot's theory an equation -which differs from our equation
(14) only in
tliis,
it
contams an un-
known
function of temperature, whose particular value for the freezing point was determined from Regnault's data on the heat of vaporization and- pressure of steam. Sir William Thomson afterwards applied to this theoretical result a very accurate test by experiment*. In order to measure small differences of temperature, he prepared a thermometer filled with ether-sulphide, the bulb Of this 5^ in. of which was 3^ in. long and the tube 6 J in. were divided in 220 equal parts, and 212 of these parts comprised an interval of temperature of 3 Fahr., so that each This thermoof a degree Fahr. part was about equal to meter was hermetically enclosed in a larger glass tube, to protect it from the external pressure, and so enclosed was placed in an Oersted press, filled with water and lumps of clear ice, and containing an ordinary air gauge to measure the pressure. When the thermometer had become stationary at a point corresponding to the melting point of ice at atmospheric pressure, the pressure was increased by screwing do-wn the press. The thermometer was at once seen to fall, as the mass of water and ice assumed the lower melting temperaOn taking off, ture corresponding to the higher pressure. this pressure the thermometer returned to its original position. The table below gives the fall of temperature observed for two
Pressure.
169
dp
dT t which was
,
found in the
mospheric pressure. We see that the observed and calculated numbers agree very closely together, and thus another result of theory has been verified by experiment. More recently a very striking experiment was performed by Mousson*, who by the application of enormous pressures melted ice which was kept during the experiment at a temperature of 18 to 20. The pressure employed he calculates approximately at about 13000 atmospheres ; on which it may be remarked that it may be possible to produce the melting with a much smaller pressure, since with his arrangement all that could be known was that the ice had somehow melted during the experiment, and not the exact time at which the melting took place.
4.
Fusion.
Bunsent was the first to institute experiments on substances which expand during fusion, and of which the fusion point must therefore rise as the temperature increases. The substances he chose were spermaceti and paraffin. By an ingenious arrangement he obtained in an
extremely simple manner a very high and at the same time measurable increase of pressure, and was able to observe portions of the same substance side by side under ordinary atmospheric, and under the increased pressure. He took a tube of thick glass about the size of a straw and 1 foot in length, and drew it at one end into a capillary tube 15 to 20 inches long, and at the other end into a somewhat larger one only 1 J inches long. The latter, which was placed lowest in the apparatus, was bent round until it stood up parallel This short curved part to the lower part of the glass tube. was filled with the substance to be tested, and the larger glass tube with quicksilver, whilst the long capillary tube remained filled with air. Both capillary tubes were sealed
* Pogg. Ann., Vol. cv. p. 161. lxjcxi. p. 5G2.
170
at
the ends. On heating the apparatus the quicksilver expanded, rose in the longer capillary tube, and compressed the air within it. The reaction of this air compressed first the quicksilver and then the substance in the shorter tube, and the magnitude of the pressure, which was capable of rising to above 100 atmospheres, could be measured by the volume of air left in the upper tube. This apparatus was fixed on a board close to another arranged in the same way except that the upper air-tube was not sealed ; so that no compression of the air, and consequent rise of pressure, could take place. The two tubes were now plunged in water, the temperature of which was somewhat higher than the melting point of the substance to be tested. Thus when the lower tube filled with the substance was once completely under water, it was only needed to sink it still deeper in order to heat a larger part of the quicksilver, and so to obtain a higher pressure in the closed upper tube. Under these conditions Bunsen repeatedly melted the substance in both tubes, and then by cooling the water allowed it again to solidify, observing the temperature at which this took place. The result was that this solidification always took place at a higher temperature in the tube in which the pressure was increased than in^ the other. The following were the numerical results.
Spermaceti.
FUSION
ANB VAPOEIZATION OF
SOLID BODIES.
171
More recently Hopkins* experimented with spermaceti, wax, sulphur, and stearine, producing pressures by means of a weighted lever to 800 atmospheres and upwards. With all the above substances a rise of the melting point under an increase of pressure was observed. The particular temperatures observed with different pressures shewed however considerable irregularities. In the case of wax, with which the rise of temperature was most regular, an increase of pressure of 808 atmospheres produced a rise in the melting point of 15J Cent. The calculation of the rise in the melting point from the theoretical formula cannot well be performed for the substances tested by Bunsen and Hopkins, since the data required are not known with sufficient accuracy.
5.
BddUon
and
the temperature
of Fusion.
determine the
we
will follows
between the temperature of fusion and the pressure, now turn to equation (12), which may be written as
g=C'-A'+^y
(15).
This equation shews that, if the temperature of fusion is changed by a change of pressure, the quantity of heat r required for fusion also changes. The amount of this change can be determined from the equation. In this the symbols c' and y denote the specific heat of the substance in the liquid and in the solid condition, not however, as already observed, the specific heat at constant pressure, but the specific heat in the case in which the pressure changes with the temperature in the manner indicated by equation (13). The mode of determining this kind of specific heat will be described in the next chapter. Here we will merely by way of example give the numeyical values in the case
The specific heat at constant pressure, i.e. that heat which is simply measured at atmospheric pressure, has in the neighbourhood of 0 the value 1 for watgr, andy according to Personf, the value 0"48 for ice,
of water.
specific
* Report, Brit. Assoc, 1854, p, 57. Cojuptes Rendvs, Vol. zxx, p, S26.
172
specific heat for the case here considered has on the contrary for water and ice the values c'=0"945 and k'=OQM. thus obtain For / we may take Person's value 79.
The
We
=0-945 -0-631 +
^3
6'.
Passage from
Hitherto we have considered the passage from the liquid to the gaseous and from the solid to the liquid condition. It may however happen that a substance passes direct from the solid to the gaseous condition. In this case three equations will hold of the same form as equations (15), (16) and (17) of the last chapter, or (11), (12), (13) of this we must only remember to choose the specific heats and specific volumes relating to the different states of aggregation, and the quantities of heat consumed in the passage from one condition to the other, in the manner corresponding to the pre:
sent case.
The circumstance that the heat expended is greater in the passage from the solid condition to the gaseous than from the liquid, leads to a conclusion which has already been drawn by Kirchhoff *. For if we Consider a substance when just at its melting point, vapour may be developed at this temperature both from the liquid and from the solid. At
*
173
temperatures above the melting point we have only to do with vapour developed from a liquid, and at temperatures below with vapour developed from a solid, leaving out of account the special case mentioned in the last section, in which a liquid kept perfectly Still remains fluid in spite of having reached a lower temperature. If for these two cases, i. e. for temperatures above and below the melting point, we express the pressure of vapour p as a function of temperature, and construct for each case the curve which has the temperatures for abscissae and the pressures for ordinates, the question arises how the curves corresponding to the two cases are related to each other at the common limit, viz. the temperature of fusion. In the first place, so far as concerns the value of p itself, we may consider it as known by experience to be equal in the two cases ; and thus the two curves will meet in one point at the temperature of fusion. But with regard to the differential coeflScient
-, , the last of the above-named three equations shews that
;
it
has different values in the two cases and thus the tangents to the two curves at their point of intersection have different
directions.
Equation (17) of Chapter VI. which relates to the passage from the liquid to the gaseous condition, may be written as
follows
dip
Er
T{s-iT)
dT~
^^''^
To form the corresponding equation for the passage from the solid to the gaseous condition, we should set on the left hand the pressure of the vapour given off by the solid body, which for distinction we may ca,ll P. On the right hand we must put, instead of <t, which is the specific volume of"the liquid, the specific volume of the solid which we may call t ; the difference thus indicated is however very small, since these two specific volumes differ very slightly from each other, and in addition are small in comparison with s, the It is of more specific volume of. the substance as a gas. importance to substitute for r, which is the heat required to cause the passage from the liquid to the gaseous condition, the quantity of heat required for the passage from the solid
174
*
This latter equals r r', where r' to the gaseous condition. Thus in the present case is the heat required for melting.
the equation
is
.^ ''
difference
Combining this eqjiation with (16), and neglecting the small between <t and t, we have
dP dT
dp
Er'
^
dT~T{s-0-)
If we apply this equation to water, we must put T= 273, r'= 79, s = 205, a- = 0-001, and giving^ the known value 424 we have
dP dT
c?p_ 424x79
drr 273x205"
If we wish to express the pressure in millimetres of mercury, instead of kilogrammes per square metre, we must, as
remarked in Chapter VI. 10, divide the above result by 13'596; then putting for p and P the Greek letters tt and IT, we have
^-^ = dT dT^^^0-044
It
may be added
for the
diffe-
rential coefficient
has for
CHAPTER
VIII.
ON HOMOGENEOUS BODIES.
Changes of Condition without Change in of Aggregation.
1.
the
Con-
dition
We will now return to the general equations of Chapter V. and will apply them to cases, in which a body undergoes changes which do not extend so far as to alter its condition of aggregation, but in which all parts of the body are always in the same condition. We will suppose these changes to be produced by changes in the temperature and in the external pressure. In consequence of these, changes take place in the arrangement of the molecules of the body, which are indicated by changes in form and volume. With regard to the external force, the simplest case is that in which an uniform normal pressure alone acts on the body in this case no account need be taken of changes in the body's form, in determining the external work, but only of its alteration in volume. Here we may take the condition of the body as known, if of the three magnitudes, temperature, pressure and volume, which we will denote as before by T, p and v, any two are given. According as we choose for this purpose v and T, or p and T, or v and p, so we obtain one of the three systems of equations, which in Chapter V. are numbered (26), (26) and (27) these equations we will now use to determine the different specific heats and other quantities, related to changes in temperature, pressure, and volume.
;
:
176
cients.
2.
If the above-named equations of Chapter V. are referred to a unit of weight of the substance, the differential coefScient
-^
constant volume, and in equations (26) the specific heat at constant pressure.
Similarly
-p has
and
(27)
and -^ has
and
(27).
Such
indeterminate cases always occur where the -nature of the question occasions the magnitudes chosen as independent variables to be sometimes interchanged. If we have chosen any two magnitudes as independent variables, it follows that in differentiating according to one we must take the other as constant. But if, whilst keeping the first of these as one independent variable throughout, we then choose for the other different magnitudes in succession, we naturally arrive at a corresponding number of different significations for the differential coefficients taken according to the first
variable.
This fact induced the author, in his paper " On various convenient forms of the fundamental equations of the Mechanical Theory of Heat,"* to propose a system of denotation which so far as he knows had riot been in use before. This was to subjoin to the differential coefficient as an index the magnitude which was taken as constant in differentiating. For this purpose the differential coefficient was inclosed in brackets and the index written close to it, a line being drawn above' the latter, to distinguish it from other indices, which might appear at the same place. The two differential coefficients named above, which represent the specific heat at constant volume and at constant pressure, would thus be
written respectively
(-T^jand
\^ji)_-
ON HOMOGENEOUS
left
BODIES.
17T
out for the sake of conveniencei More recently* the author introduced a simplei' form of writing, which yet retained th? essential advantage of the method. This consisted in placing the index next to d, the sign of differentiation. The brackets were thus rendered needless and also the horizontal line, because no other index is in general placed
in this position.
cients
will
differential
;
coeffi-
would thus be
written -^ and -~
dT
dl
and
this
method
be adopted in what
follows.
Coefficients
of
If the condition of the body is determined by any two of the magnitudes. Temperature, Volume, and Pressure, we may consider each of these as a function of the two others, and thus form the following six differential coeffiicients
:
d,p
d^p
djo
djV
dJI
dv
'
dT
dp'
In these the suffixes, which shew which magnitude is to be taken as constant, may be omitted, provided we agree once for all that in any differential coefficient that one of the three magnitudes, T,p, v, which does not appear, is to be considered as constant for that occasion. shall however retain them for the sake of clearness, and because we shall meet with other differential coefficients between the same m.agnitudes, for which the constant magnitude is not the same as here. The investigations to be made by help of these six dif-
We
ferential coefficients will be facilitated, if the relations which In the first exist between t}iem are laid down beforehand.
place it is clear that amongst the six there are three pairs which are the reciprocals of each other. If we take v as constant, T and p will then be so connected that each may be treated as a simple function of the other. The same holds with T and v where p is constant, and with v and p when T
* "On the principle of the Mean Brgal and its application to the molecular motions of Gases." Proceedings of the Niederrhein. Ges. fur Naturund Seilkunde, 1874, p, 183.
c.
12
178
is
constant.
d^
dT'
<y
dp
pairs,
d^
dv
L =
^ dT'
_L = d^
dv
^ df
(^^ ^'
^
To examine further the relation between these three we will by way of example treat ^ as a function of T and V. Then the complete differential equation for p is
Itp
is
constant,
we must put
dp = 6,
in this equation,
dv
= ^dT; dT
d^
whence
it
becomes
= ^dT+^,dT: dT dv dT
whence
= -l ^x^x^ dT
dv dp
(2).
of this equation combined with equations (1), of the six differential coefficients by the product or the quotient of two other differential coefficients.
By means
4.
and given out by the body. If we denote the specific heat at constant volume by C^, and at constant pressure by Gp, and take the weight of the body as unity, we have
in
dT
.
"
dT~
"'
We have also the equations (25) and (26) of Chapter V., which with our present notation will be written as follows
djQ^rpd^
dv
dTQ__rpd
dp
dT'
dT'
ON HOMOGENEOUS
Hence we can
tial
BODIES,
179
write
equations
dQ = C,dT+T^,dv dQ = C^dT-T^^^dp
(3),
(4).
From these two we easily obtain a third differential equation for Q, which relates to t) and p as independent variables. For multiplying the first equation by G^ and the second by G subtracting, and dividing the result by (7^ 0 we have
These three equations correspond exactly to those obtained in Chapter II. for perfect gases, except that the latter are simplified by applying the law of Mariotte and Gay-Lussac. The equation expressing this law is
pv = RT,
whence we have
^ ^ dT
-B,
'
d^ _R dT p'
putting -^ for T,
we get
dQ=G,dT + ^dv,
dQ = G^dT-^dp,
C dQ = Q-_jQ pdv + Q
Zc
^'^^'
These equations are the same as (11), (15) and (16) of Chapter II. The equations (3), (4) and (5) are not immediately integrable, as has been already shewn with respect to the special equations holding for gases. For equations (.3) and (4) this
122
180
follows
from equations already given. If in the last equations of the systems (25) and (26) of Chapter :. we use the symbols G, and G^,, and also the method above explained of
writing the differential coefficients,
we have
dv
dT"
dp
dr
^"
fulfilled, if (3)
and
dflv^rpdlp
dv
d^, dT^'^dT'
d^G^^
dp
djv
dj)
dr
dT'
a similar but longer process we may shew that equais not integrable; as may at once be concluded from the fact that it is derived from equations (3) and (4), These three equations thus belong to that class of complete differential equations which are described in the Introduction, and which can only be integrated if a further relation between the variables is given, and the path of the variation thereby fixed.
By
tion (5)
5.
Specific
Seat
at Gonstant
Volwne and
at Gonstant
Pressure.
If in equation (4)
differential
we
dp the expression
special
case in which the body changes its temperature by dT, the volume remaining constant. If we divide hj dTwe have on
-^
which
is
the specific heat at constant volume and has been denoted by (7. Hence we obtain the following relation between
Cf.
andC,: C.=
C,-2^g,x^
(^)-
by
this equation,
Substituting in equation (5) the value of G^, G given we obtain the following simpler form
dQ^G/^dv + G/^dp
.(8).
ON HOMOGENEOUS BODIES.
If
181
by means
specific
sure, it
of equation (7) we proceed to determine the heat at constant volume from that at constant presis requisite first to make a slight change in the
equation.
The
differential coefficient
-^ aT
contained therein
ture,
expresses the expansion of the body upon a rise of temperaand may generally be taken as known, but the other
differential coefficient
-^
for solid
However
d^
d,p^_dT
dT
'
djv'
dp
In this fraction the numerator is tlie differential coefficient already discussed, and the denominator expresses, if taken with a negative sign, the diminution of volume by an increase of pressure, or the compressibility of the body ; and
this for a large number of liquids has been directly measured, whilst for solids it may be approximately calculated from the Equation (7) now becomes coefficients of elasticity.
(d^X
dp
If the specific heats are expressed not in mechanical but in ordinary units, we may denote them by c and c,, the equation then takes the form
;
T\dT
,,
''-'^^E-W
dp
iinust
^^^)-
In applying this equation to a numerical calculation we remember, :that in the differential coefficients the unit of volume must be the cube of the unit of length which has
182
and that the unit of pressure been used for determining must be the pressure which a unit of weight exerts on a unit
:
If, as is usually the case, the coefficients of expansion and compression refer to other units, they must be reduced to those above mentioned.
of surface.
coefficifent
~-
is
specific heat at constant volume must always be less than that at constant pressure. The other differential coefficient
jm
is
generally positive.
it is
zero at
the temperature of maximum density, and accordingly the two specific heats are equal at that temperature. At all other temperatures, both above and below, the specific heat at constant volume is less than that at constant pressure; for
which occurs in the formula, it has no influence on its value. As an example of the application of equation (7), we will calculate the difference between the two specific heats in the case of water at certain known temperatures. According to the observations of Kopp (see his tables in Lehrbuch der Phys. u. theor. Ghimie, p. 204), we have the following coefficients of expansion in the case of water, its volume at 4 being taken as unity
at
25
According to the observations of Grassi*, we have for the compressil^ility of water the following numbers, which express
the diminution of volume upon an increase of pressure of one atmosphere, in the form of a decimal of the volume at the original pressure at 0" 0-000050,
25 50 ,i
*
0-000046,
0-000044.
p. 129.
Ann. de Chim.
et
ON HOMOGENEOUS BODIES.
183
will now, as an example, perform the calculation for the temperature of 25. The unit of length may be the metre, the unit of weight the kilogram. We must then take the cubic metre as the unit of volume and, since a kilogram of water at 4 contains O'OOl cubic metres, we
'
We
must, in order to obtain -^, multiply the coefficient of expansion given above by O'OOl.
1)
Thus we have
^=
0-00000025
= 25 X
lO"*.
For compression we must, by what has been said, take as unit the volume which the water contained at the temperature in question and at the original pressure (which latter we may assume to be the ordinary pressure of one atmosphere). This volume at 25 = 0-001003 cubic metres. Further we have taken one atmosphere as unit of pressure, whereas we must take the pressure of 1 kilogramme on 1 square metre ; in which case a pressure of one atmosphere is expressed by 10,333. Accordingly, we must put
_ ^dj,v
0-000046 X 0-001003
Ipss
at 25,
^"""^^
;
, .
_i3
Further,
we have
7= 273 + 25 = 298
IT"
and
for
we
in equation (76),
we get
X -7^ X ^7^5 424 ^ 45 ^ 10"
25^
- c, = 298 T^
= 0-0098.
obtain. from the values given above expansion and compression at 0 and
50 the following
numbers
at
0,
50,
If we now give to Cp, or the specific heat at constant pressure, the experimental values found by Regnault, we obtain for the two specific heats the following pairs of values:
184
9995,
25
,.
50-
c,
6.
Specific
at constant
In the same way as we have determined the specific heat volume in the last section, we may determine also
the specific heat corresponding to various other circumstances, since we may by equation (4) fix its relation to the specific heat at constant pressure. Thus, if the circumstances are given under which the heating takes place, the two differentials dT and dp are no longer independent, but the one is determined by the other.
We
dl
dT, in which
known
dp
specific
circumstances,
we
obtain
'^="--^^4?
If the specific heat
is
(9a) ^^"^
to
c
c^c--^^ E'dfdT
"
We will employ this equation, by way of example, to determine the specific heats which came under consideration in the two last chapters, viz. (1) the specific heat of water, wh^n in contact with steam at maximum pressure ; (2) the specific
ON HOMOGENEOUS
BODIES.
185
heat of water and ice, when the pressure changes with the temperature in such a way that the temperature of melting corresponding to the pressure at any moment is always equal to the temperature which exists at that moment.
In the
first
case
we have simply
to give to
-^ the
value
corresponding to the intensity of pressure of the steam. For the temperature 100 this value is 370, taking as unit of pressure a kUbgram per square metre. With regard to
expansion of water at 100, taking the volume of water at 4 This number must be multiplied by Q-OOl, in order to
obtain the value of
-^
in the case
when a
cubic metre
is
taken as the unit of volume and a kilogram as the unit of weight: we thus obtain the number 0'0000008. Lastly we write for T the absolute temperature for 100, or 373, and for E, as usual, 424. Then equation (9a) becomes
c
= c^ -
ll^
X 0-0000008 X 370
= c^ - 0-00026.
-
take for the specific heat of water at constant pres 100, the values derived from the empirical formula of Regnault, we obtain for the two specific heats which we wish to compare, the following simultaneous values
If
sure,
we
and at
c,
= 1-013, = 1-01274.
It thus appears that these two quantities are so nearly equal, that it would have been useless to take account of the difference between them in the calculations as to saturated steam. The consideration of the influence of pressure on the freezing point of liquids shews that a great change in the pressure only produces a very slight alteration in the freezing
-^
must be very
large.
If
we
assume, according
,to.
186
increase of pressure of one atmosphere lowers the freezing point by O-OOTSS" C, we have
dp _
hnce equation
(9ffl)
dT ~
freezing point, viz. 273,
10 333 0'00733
273
'^
=*
+ 424
"*
000733 ^
E 10333_^_
dT~ * + ^"^""" dT
.908000^
Applying this equation first to water, we will tal^e Kopp's value for the coefficient of expansion of water at 0, viz. 0000061 ; then, using the kilogram as unit of weight, and the cubic metre as unit of volume, we have
^= - 0-000000061
whence, from the equation above,
c
= c^- 0-055.
unit,
As
Cj,
is
here
we have
finally
Next, to apply the equation to ice, we will take the linear expansion of ice at 0'000051, following the experiments of Schumacher, Pohrt, and Moritz whence the cubic coefficient will be 0-000153. In order to reduce this number to the required units, we must multiply it by 0-001087, the volume of a kilogram of ice in cubic metres:
coefficient of
;
whence we obtain
^=0-000000166. dl
Substituting this value, the equation becomes
According to Person*
finally
ON HOMOGENEOUS BODIES.
187
These values, 0'945 and 0'631, were those employed in Chapby which the relation between the heat expended in fusion and the temperature of fusion was determined.
ter VII. for the calculation
7. Isentropic Variations of a Body. Instead of determining the kind of variation of condition, which a body is to undergo, by means of an equation containing one or more of the quantities T, p, v, we will now lay down as a condition, that no heat is imparted to or withdrawn from the body during its variation. This is expressed mathematically by the equation
we have
further
body remains constant. We kind of variation the designation isentropic, already applied to the curves of pressure which
that
is,
the entropy
of the
correspond to
efficients
it and will characterize the differential coformed in discussing it by the index 8. If in equation (3) we put dQ = 0, we have
:
If
we
dT
an isentropic
d,^=.^^i
c'^dT
(4),
no)
''^"^
Similarly
we
d,T dp
Applying either equation
....(11).
equation
(7),
we have
0-c/dv^C.pp;
188
whence
d,v
-7
dp
= 7=r T~ Gp d^ dT
(2), this
C,
dT
equation becomes
,^
^
dp _0^ d dp'o^'^dp
If here
,> ^'
we
give to 0,
dv
its
we
obtain
,
.
^^^>-
If
we take the
reciprocal of (12),
^^ = g.x^
dv
C,
dv
This equation,
gives
if
way
as (12),
dv
dv
C,\dTj
''
^
These
differential coeflScients
entropy being constant, have been applied to calculate the velocity of propagation of sound in gases and liquids, as has been already described in Chapter II. for the case of perfect gases.
sure, for the case of the
*
8.
Special
Forms of
the
an Expanded Bod.
Hitherto we have always considered the external force to be a uniform surface pressure. We will now give an example of a different kind of force, and will take the case of an elastic rod or bar, which is extended lengthwise by a tensional strain, e.g. a hanging weight, whilst no forces act
Upon
Instead of a tensional we long as the rod is not thereby This we should simply treat in the formulae as a bent. negative tension. The condition that no transverse force acts on the rod would be exactly fulfilled only if the rod were
it
in a ti'ansVerse direction.
strain, so
ON HOMOGENEOUS BODIES.
189
placed in vacuo and thus freed from the atmospheric presBut, since the longitudinal strain, which acts on the cross section, is very large in comparison with the atmospheric pressure upon an equal area, the latter may be
sure.
neglected.
Let be the force, and I the length of the rod, when acted on by the force and at temperature T. The length, and in general the whole condition, of the rod is under these conditions determined by the quantities and T; and we may therefore choose these as independent variables. Let us now suppose that by an indefinitely small change in the force or temperature or both, the length I is increased by dl. The work Fdl will then have been done by the force P. Since however in our formulae we have taken as positive not the work done but the work destroyed by a force, the equation for determining the external work must be written
Taking
dW=-Pdl P and T, we
(16).
may
whence
_=-P_;
first
Differentiating the
and
T are
independent variables,
dP
-jfjj
= 0, we
have
dT\dP)
A(W]
p. dPdT'
d'l
dP\dT)~
If
dT
dTdP-
subtract the second of these from the first, and substitute for the difference on the left-hand side of the resulting equation the symbol already employed for the
we
same purpose,
viz.
Df^,
we have
I>rT
= ,--
(17).
190
This value of 2)pj, we will apply to equations (12), (13), (14), (15) of Chapter V., substituting P for x in this partiWe then obtain the fundamental cular case throughout. equations, in the following form
:
dT\dP)
dP\dT)
dT
^'
(dQ\_(dQ\_l_dQ
dT\dP)
dP\dT)~ T dT
''
g=^
dP [dTj~-^ dT'
the
9.
(^)'
'
Mod.
The form of equation. (20) indicates a special relation between two processes, viz. the alteration in temperature produced by an alteration in length, and the alteration Thus in length produced by an alteration in temperature.
if,
as
is
when heated
under a constant
and
-^
is
t=
is also
positive
whence
it
follows
that, if the rod is lengthened by an increase in the external must take in heat from without if it is to keep its temperature constant, or in other words, if no heat is imparted to it, it will cool during extension. On the other
force, it
hand
if,
as
may happen
when heated
tive,
at constant pressure,
is
nega-
negative.
In
heat,
when lengthened by
a constant temperait
an increase of
ture
;
strain, if it is to preserve
must grow
ON HOMOGENEOUS BODIES.
191
takes place
of the alteration of temperature which the force is varied, without any heat being imparted to or taken from the rod, is easily determined if we form the complete differential equation of the first order for Q, in the same way as we have already done in the case of bodies under a uniform surface pressure. The differenif
f?
The magnitude
(1
tial coefficient
-jn
is
determined by equation
the fuller form
(20), in
which
we
-ttt,
dl
-^
dl
-^
dT
form, we may denote the specific heat of the rod under constant strain by C>, and the weight of the rod by M. Then
we have
dT
and the complete
-^^^^'
:
dQ = MG^dT+T^
(22).
If we now assume that no heat is imparted to or taken from the rod, we must put dQ = 0, which gives
= MC^dT+T^dP.
If
we
dT
ex-
according to P, in presses that differential coefficient of the formation of which the entropy is taken as constant;
it
We
thus
T dj d,T_ 'dF~~Mcr,''dT
This equation was
different form,
^^^^
first developed, though in a slightly Sir William Thomson, and its correctness was experimentally verified by Joule*. The agreement of
by
192
observation and tteory was specially brought out by a pbenomenon occurring with India rubber, which had already been noticed by Gough, but was then observed also by Joule and verified by accurate measurements. So long as India rubber is not extended at all, or only by a very small force, it behaves, with regard to alterations in length produced by alterations in temperature, in the same way as other bodies i. e. it lengthens when heated and shortens when cooled. When however it is extended by a greater force its behaviour is the opposite; i.e. it shortens when heated and lengthens
when
cooled.
The
differential coefficient
-^ is
thu^ positive
in the first case and negative in the second. In accordance with this it exhibits the peculiarity that it is cooled by an increase of the strain, so long as the strain remains small, but is heated by an increase of the strain when the strain
is large.
10.
The complete differential equation (22) may also be so formed as to present T and I, or I and P, as the independent variables. For this purpose we must first state the relation which, holds between the differential coefficients of the This relation will be expressed by quantities T, I, and P. an equation of the same form as (2), viz,
'
U/^Jr
(tpt
a^ J.
W''df''~dP^~^
First, to
^^*''-
contains T and I as independent variables, we must consider and I, and accordingly write (22) in, the as a function of
form
dQ = MG,dT+ T
^ (g dT + ^
dl)
-[MO.^Tf/^)dT + T%'-dl
ON aoMOGENEOTD'S! BODIESi
Transforming the
last
193^
(24),
we have
dQ=^(MG, + T^x^^)dT-T^,dl
(25).
If we denote by 0, the specific heat at constant length, the coefficient of dT in this equation must be equal to MCi ;
whence
^'=^'+J4^f
(24),
(27)-
we have
UtI
^'-^'-M^'-dj; dp
form:
dQ = MO,dT-T^dl
(28).
Secondly, to form the complete differential equation which as independent variables, we must consider contains I and T as a function of I and P. Equation (22) then becomes
dQ = MC^{^+%dP) + T'^dP
= Mc/-^dl + (Mc/^^T'f^dP.
Transforming the
coefficient of dP,
we have
.
dQ = M-G/-^dl,-(MO^+T'^.^^)l^dP.
By
equation (26)
MOi
in brackets.
dQ = MG,^dl + MC,^dP
(29).
(28)
and
13
194
= 0. The
'-i-m^i?
aad the
secoead
dl
But by equation
(24)
we may
C^,
dji _ dp-o,''dP
ds^
^^^^'
Giving to
(7,'
its
we have
"
,
dP~dP MOMt)
The
is
dl_dl
T_ /dpZy
.
''
relation
expressed by the
is
here de-
termined,
velocity of sound in
that which has to be applied to calculate the an elastic rod, in place of the relation
is
commonly
is determined by the coefficient of elasticity. In the same way, to calculate the velocity of sound ii;i gaseous and liquid bodies, we must use the relation between
-~
by -J-
We may
of the propagation of sound, in cases where the force is not large, we may in equation (82), which serves to deter-
mine
-J-,
denoted by Cp, the specific heat at constant pressure, as measured in the ordinary way under the pressure of the atmosphere.
CHAPTER
IX.
1.
General Equations.
In former chapters we have repeatedly spoken of the Energy and Entropy of a body as being two magnitudes of great importance in the Science of Heat, which are determined by the condition of the body at the moment, without its being necessary to know the way in which the body has come into this condition. Knowing these magnitudes, we can easily make by their aid various calculations relating to the body's changes in condition, and the quantity of heat
thereby brought into action. One of these, the Energy, has already been made the subject of many valuable researches, especially by Kirchhoff*, and the method of determining it is therefore more accurately known. We will here treat of
Energy and Entropy simultaneously, and set forth side by side the equations which serve to determine them. In Chapters I. and III. the two following fundamental equations, denoted by (III.) and (VI.) were developed
dQ = dU+dW. dQ^TdS.
Here
(Ill),
(YI).
and S denote the Energy and Entropy of the and dS' the changes produced in them by an dQ is the indefinitely small change in the body's condition quantity of heat taken in by the body during its change;
body, and
dU
cm.
p. 177.
132
196
the external work performed; and T tlie absolute temThe first equaperature at which the change takes place. tion is applicable to any indefinitely small change of condition, in whatever way it takes place, but the latter can be applied only to such changes as are in their nature reversible. These two equations we will now write in
the form
dW
dU=dQ-dW
dS =
(1),
^
U
(2).
and 8. Their integration will then determine Here we must first notice a point which has already been mentioned with regard to energy in Chapter I., 8. It is not possible to determine the whole energy of a body, but only the increase which the energy has received, whilst the body was passing into its present condition from some other which we choose as its initial condition and the same is also true of the Entropy. Now to apply equation (1). Let us suppose that the bodyhas been brought into its present condition from the given initial condition, the energy of which we will denote by U, by any convenient path, and in any way reversible or not reversible and let us suppose dU to he integrated through the range of this change in condition. The value of this integral will be simply IT t/,. The integrals oi dQ and represent the whole quantity of heat which the body has taken in, and the whole external work which it has performed, during the change in condition. These we will denote by Q and W. Then we have the equation
; ;
dW
U=U,+ Q-W...
Hence
a given
,
(3).
it
initial
we can determine the heat taken in and the work performed, we thereby know also the energy of the body, except as
regards one constant depending on the initial condition. Next to apply equation (2). Let us suppose that the body has been brought into its present condition from the given initial condition, the entropy of which we denote by S, by any path whatever, but by a process which is reversible;
197
condition.
us suppose the equation integrated for this change in The integral of dS will have the value 8 8:
whence we have
S-S, +
j^
(4).
Hence it follows that if for any passage -of the body, by a reversible method but by any path whatever, from a given initial condition to its present condition, we can determine
/ -fp-
we
shall thereby
know
2. Differential Equations for ike Case in, which only Reversible Changes take place, a/nd. in which the condition of the Body is determined by two Independmi Variables.
If
equations
(III.)
indefinitely small
and
reversible
dQ
will
Hence we
(5).
have
^dS=dU+dW
We
will
now assume
is
determined by two variables, wMch, as in Chapter VI., we will generally denote by x and y, signifying by these certain magnitudes to be fixed later on, such as temperature, volume, pressure. If the condition is determined by x and y, then all magnitudes, the values of which are fixed by the condition of the body at the moment, without its being necessary to know the way in which the body has come into that condition, are capable of being represented by functions of these variables; in which functions the variables must be' considered as independent of each other. Accordingly the entropy 8 and the energy U must be looked upon as On the other hand, the external work functions of x and y. W, as we have repeatedly observed, holds a completely difIt is true that the differ-!, ferent position in this relation. ential coefficients of W, so far as concerns reversible changes, itself may be considered as known functions of x and y however caimot be represented by such a function, $,nd can
:
198
only be determined,
and
it
we have given not only the initial the body, but also the path by which has passed from one to the other. If in equation (5) we put
final conditions of
dU^-^da, + ~^dy,
,^^
that equation becomes
dW.
dW.
^dS
As
rpdS
dW\,
this equation
it
and dy,
it
must hold for any values whatever of dx must hold for the cases alnongst others in which
one or other of these differentials is equal to zero. divides into the two following equations
Hence
rpdS^md^
doe
dx
dy
dx
^dS^dJl
dy
dW.
dy
.(6).
From
S or U may
For
this
be eliminated by
a second differentiation.
We
purpose
we must
differentiate the first of equations (6) according to y, and the shall write the second differential second according to x.
We
coefficients of
8 and U in
ential coefficients of
dW dW -y- and -^ we
dx
dy
but the
differ-
This is with the same obiect as in ). dx \dy J. viz. to shew that they are not second differential Finally we may observe coefficients of a function of x and y. that Tf the absolute temperature of the body, which in this
-r-
(^-
land-^- f-i
''
199
we assume
dT^dS ^
dy
dfX
d'S
dxdy
^ <PU
dxdy
dT^dS
dx
dy
dxdy
dydx'
and
d^U dxdy
d^U
dydx'
we have
dT ^dS
dy dx
dT^^dS_d/dW\
dx
dy
d /dW\
dy\dxj
dx\dy
)'
The right-hand side of this equation we have named in Chapter V., "the work, difference referred to xy" and have denoted it by D^/, hence we may put
J. "'
_d
fdW\ ,dy\dxj
d /dW\ dx\~d^)
.^
'''''
dT
This
is
dS
dT
dS_
^
^
dy^dx
which serves
dx^dy~^^
(5),
(6),
we
write
them
d,8_i^dU
dW
dy- T
dy '^T dy-
200
Differentiating the first of these equations according to .and the second according to cs, we have
d'S
dxdy~ T
d'S
1
d?U dxdy
dT
dy
dU
dx
dJJ
'^
dW\ d dy \T dxl'
d^_
dT
dydx
T^ dydx
T^ dx ^1^^ dx\T
U
n^\
dyj
'
Subtracting the second of these equations from the first, putting all the terms containing in the resulting equation on the left-hand side, and multiplying the whole equation by
I",
we have
dy
X
^ - = r F
dx
dx
X dy
f[dy \T dx
^)~ ) -A(l dx \T dy
j )
We
will adopt
""
following relation:
[dy\T dx)
dx\T dy}\
^''
is
D^ and
A,^ there
the
S^
K
(10).
Using
dTdU_dTdU
dy dx
dx dy
(11).
This
(5),
is
The above equations take a specially simple form, if the temperature 2^ is chosen as one of the independent variables. If we put T=y,vfe have
= = dy dx
1
'
201
We
relation
between
A,,,
and
D^
A,,= TD,,Equations
(8)
dW
dx
.(12).
and
(11) also
become
dx
'
dU
.(13);
The differential coefficients of the two functions 8 and with regard to x are thus known. For their differential coefficients with respect to T we will take the expressions which follow directly from (2) and (1) on the assumption that the condition of the body is determined by T and x, viz.
dS
dT~
202
which
is
dx\dT)~dx\dT)~ dT
For by
(12)
^'''
last.
we have
dK,_
Now, remembering
'^~dT\dx)
we may
dx\dTJ'
dA^^
dT On
dD,,_d^(dW\ dT dx\dTj'
-~ in equation
(17),
we
are
brought back to the form of equation (16). We have now to determine 8 and U themselves, by integrating equations (15). Let us suppose that the body has been brought into its present condition, by any path we please, from an initial condition for which the quantities T, 8, X, TJ have the values T^, 8^, x, U respectively and let this particular change of condition give the range of the integration. As an example, let us suppose that the body is first heated from the temperature T to the temperature T, while the other variable keeps its initial value x^, and then that this other variable changes its value from to x, while the temperature remains constant. Then we have
:
(18).
203
side
T and x.
Let us now make the opposite assumption, viz. that the change of x first takes place at the initial value of T, and then the change of T at the final value of x. Then we
obtain
(19).
In both these equations the first integral on the right-hand side is a simple function of x, and the second of T and x. By what has been said above, we may choose any other path whatever, instead of that which we have taken as our example, in which path the changes of T and x may be transposed in any way, or may both take place at once according to any law. We should naturally in each special case choose that path, for which the data requisite to perform the calculation are most accurately known.
4. Bpedal case of the Differential equations on the assumption that the only external force is a Uniform Surface
Pressure.
If we assume as the only External Force a Uniform Pressure normal to the surface, we must put
dW=pdv.
Hence
The
dW
t^ dx
dv
^
Ptdx
^^^d
r- =Pj-. dy " dy
dW
dv
expressions for D^ and A^ then assume peculiar forms. Those for D^ have been already considered in Chapter V. We have first
_ d
^\
^^~d^V'&)
"*
^^ _^f di[Pd^)'
dx)
.
Uy
\T
dx\T^
dy)
204 In the
brevity
these equations
we
will
put
^ = f,
whereby
it
(20),
becomes
A " T^\ ^
Performing
the
^'''\
\jly\
dx)
^ dx
^'^\
dyl
equations,
differentiation
in
these
and
remembering that
dxay
dydx
we have
^.'%^i-t-%
If the temperature
w.
become
_dp
^''~ dT
d^
-
d^
^
'^''
dx
dx^ dT
^.'=^-4--S
or,
<^*)^
restoring to
ir its
value '^
^^'-^Uy
The
db
.jy
d^
dx
dT)
^dx
:
^'^^'''
t""
dT'^^^\dT''dx
dv\
dx'^dTr"'
dv
dir
^^''^'
fdQ
rmfdir
dv\
,_.
205
.(26a).
If we further choose for the second variable, as yet undetermined, the volume v, and thus put x = v, we have
,
dv
dv
.(27).
(4^i')^
If
the pressure
variable, so that
'^-VS'^^-pp'
(28).
^H'dT-^S)'H'TT-4)''
5. Application of the foregoing Equations geneous Bodies, and in particular to Perfect Gases.
to
Homo-
is
For Homogeneous Bodies, where the only external force a uniform pressure normal to the surface, it is usual, as at the end of the last section, to choose for independent
two of the quantities T,v, p; and -j^ then takes
variables
the simple significations which we have several times alluded Thus if T and -v are the independent, variables, and if to.
206
or, if specific heat at constant volume independent variables, the specific heat at constant pressure. Equations (27) and (28) become in these cases
.(29),
dU=C,dT+(T^-p)dv
d8 = ^dT- ^rpdp,
(30).
dv
dU={c^-p)dT-{T%.p'^)dp\
If
we wish to apply these equations to a may use the following well known equation pv=IiT.
if
perfect gas,
we
Hence,
T and
dT~
and equations
(29)
v'
then become
d8=G,f^E'^'
dU=C,dT
.(31).
As in this case G, must be regarded as a constant, these equations can at once be integrated, and give
8 = S,+ CJog^+Rlogl^
.(32).
If
we choose
and
dv
as independent variables,
,
we may
put
dT
=R P
and ^dp
dv
ET
P
207
become
'^'''.f-^fl
Whence
"we obtain
(S3,.
by integration
8 = S. + C^lo4-Blogl\
^3^^_
The integration of the general equations (29) and (30) can of course only be accomplished if, in (29), p and G, are known functions of T and v, or if, in (30), v and. C^ are known functions of T and p.
6. Application of the Equations to a Body composed of matter in two Different States of Aggregation.
As another special case we may select the state of things treated of in Chapters VI. and VII., viz, the case in which the body under consideration is partly in one state of aggregation and partly in another, and when the change, which the body may undergo at constant temperature, is such that the magnitudes of the parts in the two different states of aggregation are altered, with a corresponding change in the volume, but no change in the pressure. In this case the pressure jp depends only on the temperature and we may
;
therefore put
-^
= 0,
by which equations
(25)
and
(26) are
transformed as follows
(35).
in Chapters, VI. and VII., let us denote by ilf the weight the weight of the part in the of the whole mass, and by in place of x second state of aggregation ; and let us take
As
208
for
ON
I'HE
dm
for which,
'
by equation
(12),
Chapter
"VI.,
we may
substitute
dv
dm
J,
dp
dT
Then the ahove equations become
<^S
= -^^dT+jjdm,
>
(36).
''H-''S)''*f-^yTo integrate these equations we may take as a starting is in the first point the condition that the whole mass state of aggregation, that its temperature is T^, and that its pressure is the pressure corresponding to that tempera:ture. The passage from this to its present condition (in which of the whole the temperature is T, and in which the part mass is in the second, and the part Jf in the first state of aggregation) may be supposed to take place in the following way First let the mass, still remaining entirely in the first state of aggregation, be heated from T to T, and let the pressure change at the same time, in such a way that it is always the pressure corresponding to the temperature at the moment then let the part pass at temperature T from the first to the second state of aggregation. The integration has to be performed according to these' two successive stages. During the first stage dm = 0, and thus it is only the first term on the right-hand side which has to be integrated.
Here
-^
MG, where C
heat of the body in its first state of aggregation, and for the case in which the pressure changes during the heating in the way described above. This kind of specific heat has been already discussed several times, and the conclusions drawn in Chap. VIII., 6, shew that where the first state of aggre-
209
the solid or liquid, and the second the gaseous, be taken, for purposes of numerical calculation, as equal to the specific heat at constant pressure. It is only at very high temperatures, for which the vapour tension increases very rapidly with the temperature, that the difference between the specific heat and the specific heat at constant pressures is important enough to be taken into account. Further, during the first change the volume v has the value Ma, where a is the specific volume of the substance in the first state of aggregation. During the second stage dT= 0, and therefore it is only the second term on the right-hand This inteside of equation (3a) which has to be integrated. gration can be at once performed for both equations, since the coefficient of dm is a constant with regard to m. The resulting equations therefore are
may
safely
S = S, + Mf^^dT+^,
U=U + M /:(--!?) dT + mp
.(37).
= or = il/, we obtain If in these equations we put the entropy and energy for the two cases in which the mass is either entirely in the first or entirely in the second state of aggregation, under the temperature T, and under the pressure corresponding to that temperature. For example, if the first state is the liquid and the second the gaseous, then if we put = 0, the expressions Prelate to .the case of liquid under temperature T, and under a pressure equal to the maximum vapour tension at that temperature or if we put m = M, they relate to saturated vapour at temperature T.
delations of the Eocpressions T>^^ and A^^. this chapter it is worth while to refer again to the expressions D,^ and A, which by (7) and (9) have~the following meanings
7.
In concluding
_ d fdW\ '^~dy\dxl
A,
d (dW\ dx\dy)'
=r
d^ (l
d'W\
d_
/I
dM\
14
dy\Tdx)
dx\T^dy}_
210
These are both functions of x and y but if to determine the condition of the body we choose instead of x and y any two other variables which we may call ^ and ?;, we may form corresponding expressions D^^ and Aj, as follows
I>i.
'dvKd^J
mi
d^\dr,)'
Hn--
,(38).
These are of course functions of f and 17, as the former were of x and y. But if we compare one of them, e.g. that for D^^, with the corresponding expresssion for D^^, we find that these are not simply two expressions for one and the same magnitude referred to different variables, but are actually two different magnitudes. For this reason D^ has not been called simply the work difference, but the work difference referred to xy, so that it may be distinguished from Di,,, the work difference referred to ^t]. The same holds true of Aj,j, and A|,. The relation which exists between D^^,. and Z)j, may be found as follows. The differential coefficients which occur in the expression for i)^, in (38), may be derived by firfjt forming the differential coefficients according to x and y, and then treating each of these as a function of ^ and rj. Thus we have om^^dW dx d^ dx d^ d^ dy d^'
dW
dy
dW^dW
dr]
'
dx
drj
dW
dy
drj'
dx
Differentiating the
first
77
fdW\ dxdx
d fdWX dx dy
dx\dx) d^ dr]
d (dW\_ dvKd^J-
dy\dx J d^ dr]
dW d^
dx
d^d/r]
d /dW\ dx dy dx\dy)dr]d^
dW
dy
d'y
'"
d^dr]
211
(dW\ dxdx
\ djo J
d^
/dF\ dm dy
d/q
dx
d_
d^
drj
dy\dx J
d^
/dW\
dW
dx
d^x
d^dt)
d /dW\dydy dy\dy) d^
d/rj
dW
dy
d^y
d^d/ij
If
subtract the second of these equations from the first, the terms on the right-hand side disappear except four, which may be expressed as the product of two binomial terms in the following equation
all
we
dx
dy\
fdM\
dy\dx)
dx \dy ) \'
dv\d^J
d^[dvJ~\d^^dv'
'dv^'d^J
D^^ dx
drj
dx
dy
dr)
dy
i'.-Cd^
Similarly
)l>.
.(39).
d^,
we may
obtain
fdx
Hi'
d^
dr)
dx dy\ 'd^'^t^)^"'.
.(39a).
If we substitute one new variable only, e.g. if we keep the variable x, but replace y by 17, we must put a; = |^ in the two
dcs
last
equations,
whence
-tz.
=^
and 3-
dss
= 0-
The equations
then become
A,
--^D and A
-^A
.(40).
If we retain the original variables, but change their order of sequence, the expressions simply take the opposite sign, as Hence is seen, at once on inspection of (7) 'and (9).
D,^
= - J?^,
and A,^ =
- A^^,.
(41).
142
CHAPTER
X.
ON NON-REVEKSIBLE PE0CESSE3.
1. Completion of the Mathematical Expression for the second main Principle.
In the proof of the second main principle, and in the investigations connected therewith, it was throughout assumed must that all the variations are such as to be reversible.
We
now
consider how far the results are altered, when the investigations embrace non-reversible processes. Such processes occur in very different forms, although in their substance they are nearly related to each other. One case of this kind has already been mentioned in Chapter I., viz., that in which the force under which a body changes its condition, e.g. the force of expansion of a gas, does not meet with a resistance equal to itself, and therefore does not perform the whole amount of work which it might perform during the change in condition. Other cases of the kind are the generation of heat by friction and by the resistance of the air, and also the generation of heat by a galvanic current in overcoming the resistance of the wire. Lastly the direct passage of heat from a hot to a cold body, by conduction or radiation, falls into this class. will now return to the investigation by which it was proved in Chapter IV. that in a reversible process the sum of all the transformations must be equal to zero. For one kind of transformation, viz. the passage of heat between bodies of different temperatures, it was taken as a fundamental principle depending on the nature of heat, that the passage fropi a lower to a higher temperature, which represents negative transformation, cannot take place without compensation. On
.
We
ON NON-REVERSIBLE PROCESSES.
this rested the proof that the
213
sum of all the transformations in a cyclical process could not be negative, because, if any negative transformation remained over at the end, it could always be reduced to the case of a passage from a lower to a higher temperature. It was finally shewn that the sum of the transformations could not be positive, because it would then only be necessary to perform the process in a reverse order, in order to make the sum a negative quantity. Of this proof the first part, that which shews that the sum of the transformations cannot be negative, still holds without alteration in cases where non-reversible transformations occur in the process under consideration. But the argument which shews that the sum cannot be positive is obviously inapplicable if the process is a non-reversible one. In fact a direct consideration of the question shews that there may verywell be a balance left over of positive transformations since in many processes, e.g. the generation of heat by friction, and the passage of heat by conduction from a hot to a cold body, a positive transformation alone takes place, unaccompanied by any other change. Thus, instead of the former principle, that the sum of all the transformations must be zero, we must lay down our principle as follows, in order to include non-reversible variations The algebraic svmi of all the transformations which occur in a cyclical process must always he positive, or in the limit
;
:
equal to zero.
'
'
give the name of uncompensated transformations to such as at the end of a cyclical process remain over without anything to balance them ; and we may then express our principle more briefly as follows ; Uncompensated transformations must always he positive.
We
may
In order to obtain the mathematical expression for this extended principle we need only remember that the sum of all the transformations in a cyclical process is given
.,-/< 7^
'
Thus
we must
/f-io...
,.(ix).
214
Equation
Chapter
III.,
then becomes
(X.)
dQ<TdS
2.
Magnitude of the Uncompensated Transformation. In many cases the magnitude of the Uncompensated
Transformation is obtained directly from! the equivalence value of the transformations, as determined by the method If for example a quantity of heat Q is of Chapter IV. generated by any process such as friction, and this is finally imparted to a body of temperature T, the uncompensated
transformation thus produced has the value
j,
.
Again,
if
quantity of heat Q has passed by conduction from a body of temperature 2\ to another of temperature T^, then the
uncompensated transformation
is
Q(nr
-^-j.
If a body
has passed through a non-reversible cyclical process, and we wish to determine the resulting uncompensated transformation, which we may call N, we have, by the principles explained in Chapter IV., the equation
N-.
./f
(1)..
As however a cyclical process may be made up of several individual changes of condition in a given body, some of which may be reversible, others non-reversible, it is in many cases interesting to know how much any particular one of the latter has contributed towards making up the whole sum of uncompensated transformations. For this purpose we may suppose that after the change of condition which we wish to enquire into, the variable body is brought by any reversible process into its former condition. By this means we form a smallei: cyclical process, in which equation (1) may be applied just as well as in the whole process. Thus if we know the quantities of heat which the body has taken in during this process, and the temperatures which appertain
to
-yp-
gives the
uncom-
But
as
ON NON-EEVEESIBLE PROCESSES.
215
the return to the original condition, which has taken place in a reversible manner, cEln have contributed nothing to increase this sum, the expression above gives the uncompensated transformation which was sought, and which was caused by the given change in condition. If we examine in this way all the parts of the whole process which are non-reversible, and thereby find the values of N^, N.^ etc., which must all be individually positive, then the sum of these gives the magnitude relating to the whole cyclical process, without requiring ua to bring under review those parts of it which are known to be reversible.
3.
Work.
It may be worth while to examine more closely those changes of condition, mentioned in 1, which take place in a non-reversible manner because the resistances to be overcome are less than the forces at work our object being to determine the amount of heat taken in during the process. As however there are a great number of different changes ,of this kind, which are produced in a great number of ways, we must confine ourselves to a few cases, which are either especially noteworthy on account of their simplicity, or have some special interest on other
;
grounds.
The general equation for determining the quantity of heat which a body takes in, whilst it undergoes any given
change
follows
of
condition,
reversible
or
non-reversible,
is
as
Q=U,-U,+ W
.....(2);
in which U^ and 'U^ are the energy in the initial and final is the external work done during the conditions, and
variation.
To determine the energy we can employ the equations of Chapter IX. If the only external force is a uniform pressure, and if the condition of the body is determined by its tempei-ature and volume, then we may use equation (29), viz.
dU=0,dT+{T^P,-p\dv
(3).
2lQ
This must be integrated for a passage in some reversible manner from the initial to the final condition. If the temperature is equal in the two conditions, as we shall,
assume in the examples which follow, then the integration be performed at constant temperature, and the result will be, if we denote the initial and final volumes by t)j and v^,
may
^.-^."i:(4H*
whence equation
(2)
<*'i
becomes
^^>
^=;(^S'"^)'^^-^^
As the first and simplest case we may take that in which a gas expands without doing any external work. may suppose a quantity of the gas to be contained in a vessel and that this vessel is put in connection with another in which is a vacuum, so that part of the gas can pass from one to the other without meeting any external resistance. The quantity of heat which the gas must in this case take in, in order to keep its temperature unaltered, is determined by putting in the last equation ; thus we have
We
W=0
=;(^S-^)'^"
If
(^>-
we make the
gas
is
perfect one,
RT, we
have
dp
_R
v'
dT~
whence
dT~^ v~ R
whence
(6)
v~P'
(7).
becomes
<3
As mentioned, Gay-Lussac, Joule, and Eegnault have experimented on expansion apart from external work. Joule annexed to his experiment^, described in Chapter II., by which he determined the heat. generated in the compression
alreiady
ON NON-REVERSIBLE PROCESSES.
of
217
air, other experiments upon the expansion of air. The receiver R, shewn in Fig. 6, was filled with air condensed to 22 atmospheres, and was then connected, in the manner shewn in Fig. 18, with an empty receiver B', so that the communication between the two was only closed by the cock. The two receivers Pig. 18. were placed together in a water calorimeter, and the cock was then opened, whereupon the air passing over to the receiver B' expanded to about twice its former volume. The calorimeter 'shewed no loss of heat, and thus, so far as could be measured by this apparatus, no heat seemed to be required for the expansion of the air. The above result however holds only for the process as a whole, and not for its individual parts. In the first receiver, in which the expansion takes place and the motion originates, heat is required; in the second, on the contrary, in which the motion ceases, and the air which rushes in first is compressed by that which follows, heat is generated and so also in the places where friction has to be overcome during the passage.. Since however the heat generated and the heat required are equal, they cancel each other; and we may say, so far as the general result of the whole process is concerned, that no expenditure of heat takes place.
.
Fig. 19.
To observe specially the different parts of the process. Joule varied his experiment by placing the two receivers and
218
the pipe carrying the cock in three different calorimeters, as shewn in Fig. 19. Then the calorimeter in which was the receiver containing the air shewed a loss of heat, and the two other calorimeters a gain. The whole gain and the whole loss were so nearly equal that Joule considers the difference to be within the limits of error of the observation.
Eaypansiori of a
4.
If a gas in expanding has a resistance to overcome, but one which is less than its expansive force, then an amount of work will be performed less than the amount which the gas could perform during the expansion. An example of this is the case of a gas rushing into the atmosphere out of a vessel in which it has a pressure higher than atmospheric pressure. In this case also the process is a complicated one. We have not only to deal with the work necessary for the expansion and the corresponding consumption of heat, but in addition heat is consumed in producing the velocity with which the gas escapes and heat is again generated when this
;
velocity is subsequently checked. Similarly, heat is consumed in overcoming the resistance of friction, and is generated by the friction itself. To investigate accurately
these individual parts of the process would involve us in great difficulties. If however we only wish to determine the quantity of heat, which on the whole must be taken in from without in order to keep the temperature of the gas constant, the case is can then leave out of account those parts of the simple. process which balance each other, and need only consider the initial and final volume of the gas, and so much of the work done as is not transformed back again into heat. Then the internal work is the same as in any other case of the gas expanding at the same temperature and between the same initial and final volumes while the external work is simply represented by the product of the increase of volume and the
all
We
atmospheric pressure. To determine the required quantity of heat, we start again from equation (5), and there substitute for the expression for the external work performed in the present case, viz. p^ (v^ v^, where p^ is the atmospheric pressure.
ON NON-SEVERSIBLE PROCESSES.
The equation thus becomes
219
<^-f!^{''%-py^-^p^(^^-^:)-
.(8).
side, as
If the gas is a perfect one, the integral on the right-hand shewn in the last section, will = 0, and the equation
(9),
whence
we have
Q=^(v,-v,).
.(9a).
This kind of expansion has also been experimented on by Having as before compressed air to a high pressure Joule. in a receiver, he allowed it to escape under atmospheric pressure. In. order to bring the escaping air back to the original temperature, he caused it, after leaving the receiver, to pass through a long coil of pipe, as shewn in Eig. 20, which was placed together with the receiver
The cooling of the rimeter. calorimeter gave the quantity of heat given off to the air during
Fig. 20.
220
its-
heat, Joule
Applying equation (9a) to this quantity of was able to use this experiment as a means of The numbers calculating the mechanical equivalent of heat. obtained from three series of experiments gave a mean value of 438 (in English measures 798) a value which agrees closely with the value 444 found by the compression of air, and does not differ from the value 424, found by the friction of water, more widely than can be explained by the causes of error
;
5.
Joule.
The above-mentioned experiments of Joule, in which air contained in a receiver was expanded either by espaping into another receiver or into the atmosphere, shewed that the conclusions drawn under the assumption that air is a perfect gas are in close accordance with experience. If however we wished to know to what degree of approximation air or any other gas obeys the laws of perfect gases, and what are the laws of kny variations that may occur from, the conditions of a perfect gas, then the above mode of experiment is not sufficiently accurate ; since the mass of the gas is too small compared with that of the vessels and other bodies which take part in the variation of heat, and therefore the sources of error derived from these have too great an influence on the result. very ingenious method of making more accurate experiments was devised by W. Thomson, and the experiments were carried out by him and Joule with great care and skill. Let us imagine a pipe, through which is forced a continuous current of gas. At one place in this let a porous plug be inserted, which so impedes the passage of the gas, that even when there is a considerable difference between the pressure before and behind the plug, it is only a moderate amount of gas, suitable for the experiment, which can pass through in a unit of time. Thomson and Joule used as plug a quantity of cotton wool or waste silk, which, as shewn in Fig. 21, was compressed between two pierced plates,
.
AB
and GH, one before and one behind the plug, but at such a distance that the unequal motions which may occur in the neighbourhood of the plug
EF
ON NON-EEVERSIBLE PROCESSES.
are not discernible, and there is only a uniform current of gas to deal with. Then the whole process of expansion, corresponding to the difference of pressure before and behind the plug, takes place in the small space between these two sections.
221
kept uniform motion is produced, in which all the fixed parts of the apparatus keep their temperature unaltered, and neither take in nor give off heat. Then if, as was done by Thomson and Joule, we surround this space with a non-conducting substance, so that no heat can either pass into it from without Pig. 21. or vice versa, the gas can only give out or take in the quantity of heat expended or generated in the process; and thus, even where this quantity is very small, a difference of temperature may exist sufficient to be easily noticed and accurately measured.
is
6.
to the
above
method.
In order to determine theoretically the difference of temperature in the above case, we will first form the general equations determining the quantity of heat which the gas must have taken in, if the temperature at the second section is to have any required value. From this we can readily find the temperature at which the heat imparted will be
nothing.
The separate parts of the process in the present case are connected partly with consumption, partly with generation of heat. Heat will be consumed in overcoming the frictional resistance due to the passage through the porous plug; whilst by the friction itself the same amount of heat will be generated. At certain points in the passage heat is consumed in increasing the velocity; whilst at other points heat is generated as the velocity decreases. To determine the quantity of heat which on the whole must be imparted to the gas, we may leave out of account the parts of the process which balance each other ; since it is sufficient for our purpose to know what is the work which remains over as external work done or consumed, and at the same time the actual perngianent change in
222
the vis viva of the current. For this we need only consider the work done at the entrance of the gas into the space between the sections, i.e. at section EF, and also at the exit from that space, i.e. at section GH; and similarly the velocities of the current at those two sections. With regard to the velocities, the difference between their If however they are at vis viva can readily be calculated. each section so small as they were in Thomson and Joule's experiments, their vis viva may be altogether neglected. It then remains only to determine the work done at the two sections. The absolute values of these quantities of work may be obtained as follows. Let us denote the pressure at section ISF by Pj, and suppose the density of the gas at this section to be such that a unit-weight at this density has the volume v^. Then the work done during the passage through the section of a unit- weight of the gas equals p^v^. .Similarly the work done at section GJS will be p.jv^, where p^ is the pressure and v^ the specific volume at that section. These two quantities must however be affected with opposite signs. At section OH, where the gas is escaping from the given space, the external pressure has to be overcome, in which case the work done must be taken as positive while in section SF, where the gas is entering the space and thus moving in the same direction as the external pressure, the work must be conThus the net external work persidered as negative. formed on the whole will be represented by the difference
;
We have now further to determine the quantity of heat, which a unit-weight of the gas must take in while it passes through the distance between the two sections; supposing the gas to have at the first section, where the pressure is p^, the temperature T^, and at the second section, where the pressure is p^, the temperature T^. For this purpose we must use the equation which applies to the case in which a unit-weight of the gas passes from a condition determined by the magnitudes ^^ and T^ into that determined by the magnitudes p^ and T^, and performs in so doing the work Pi%~Pi"v ^ therefore recur to equation (2), in which the symbol w, denoting the external work, must be replaced by l^snce we have P-2^2 ~Pi''i
>
Q=U,-U^+p,v^-p,v^
(10).
ON NON-REVERSIBLE PROCESSES.
Here we need only
purpose
223
to determine U^ U^, for which again use one of the differential equations for set forth in the last chapter. In this case it is convenient to choose the differential equation in which and p are the independent variables, i.e. equation (30) of Chapter IX.
we can
In
this equation
we may put
J3
^^2'+
p^
cii?
=pdv = d{pv)-vdp.
dU=C,dT-(T^-v)dp-d{pv)
T^, p^, to
(11).
This equation must be integrated from the initial values the final values T^, p^. The integration of the last term can be performed at once, and we may write
:
U,-U,==j^C^dT-
{t^-v)
dp']^
-p,v,+p,v,
....(12).
U^-
U^ in equation (10),
we
Q^j^O,dT-{T^^-v)dp'j
ential of a function of
(13).
Here the expression under the integral sign is the differT and p, since C^ satisfies equation (6)
c
(?<,
of Chapter YIII.
d'v
d^~
And
by the
initial
dr-
thus the quantity of heat Q is completely determined and final values of T and p. If we now introduce the condition corresponding to Thomson and Joule's experiments, viz. that Q = 0, then the difference between the initial and final temperatures is no longer independent of the difference between the initial and final pressures, but on the contrary the one can be found
224
from the
nitely
small,
differential equation
If we here put dQ = 0, we obtain the equation which expresses the relation between dT and dp, and which may be thus written
%-ki'TT-')
If the gas were a perfect gas, and therefore pv should have dv _V
(-)
= It T, we
dT
dp
Thus in this case an indefinitely small difference of pressure produces no difference of temperature; and the same must of course hold if the difference of pressure is Hence one and the same temperature must exist finite. before and behind the porous plug. If on the contrary some difference of temperature is observed, it follows that the gas does not satisfy the law of Mariotte and Gay-Lussac, and by observing the values of these differences of temperature imder various circumstances, definite conclusions may be formed as to the mode in which the gas departs from that law.
7. Results of the Experiments, and Elasticity for the gases, as deduced therefrom.
Equations of
The experiments made by Thomson and Joule in 1854* shewed that the temperatures before and behind the plug, were never exactly equal, but exhibited a small difference, which was proportional to the difference of pressure in each
* Phil. Trans., 1854, p. 321.
.
ON KON-EEVERSIBLE PROCESSES.
case.
225
With air at an initial temperature of about 15V losses of temperature were observed, which, if the pressure were measured in atmospheres, could be expressed by. the equation
With
greater
;
with an
initial
satis-
...
two equa-
The
^=0-26
dp
and
^=
dp
1-15
(15.)
^ '
In a later series of experiments, published in 1862*, Thomson and Joule took special pains to ascertain how the
cooling effect varies when different initial temperatures are chosen. For this purpose they caused the gas, before reaching the porous plug, to pass through a long pipe surrounded by water, the temperature of which could be kept at will to anything up to boiling point. The result showed that the cooling was less at high than at low temperatures, and in the inverse ratio of the squares of the absolute temperatures. For atmospheric air and carbonic acid they arrived at the following complete formulae, in which a is the absolute temperature of freezing point, and the unit of pressure is the weight of a column of quicksilver 100 English inches high:
^=0-92^
If one atmosphere formulae become
is
and
^=
4-64
(yj.
dp
andf.l.a(-;)dp \TJ
C).
in their
later researches that a slight heating effect took place instead of cooling. They have however deduced no exact formula
* Phil: Tram., 1862,
C.
.p.
579,
13
226
accurate.
dT
-.
given in
(16),
we
substi-
tute for the numerical factor a general symbol A, combine into one general formula, viz,
they
f=^'
Substituting in equation (14),
(")
obtain
(18).
we
r^,-t; = ^a(j)
dv
According to Thomson and Joule, this equation should be employed for gases as actually existing, in place of the equation referring to perfect gases,
=Q T-v dT
^
if
"^
"'
we wish to express the relation which exists between the change of volume and. temperature when the pressure is kept constant. If G^ is taken as constant, equation (18) can be integrated immediately. Now it is only for perfect gases that it has really been proved that the specific heat C^ is independent of the pressure; and similarly it is only for perfect gases that the conclusion derived from Regnault's experiments G^ is al'feo independent of the is strictly true, viz. that temperature. If however a gas differs very slightly from the condition of a perfect gas, 0^, will have values differing very slightly from a constant, and these differences may be taken as quantities of the same order. Since in addition the whole term containing G^ is only another- small quantity of the same order, the differences produced in the equation by the differences of G^ will be small quantities of a higher order, and such may in what follows be neglected thus we may take C?, as constant. Then multiplying the
equation by
dT
-^ and
,
integrating,
we have
ON NON-REVERSIBLE PROCESSES,
227
=pr-i^a,(jy
where
case
(19),
P is the constant
the
may be
According to
law of Mariotte
and
Gay-Lussac
we should have
=f2^and
it
(20);
is
the function
the form.
P=:^ + 7r, P
where
ir represents another function of p which however can only be very small. Equation (19) then becomes
v=^R^+^T-iAC^{^^J
as fbllows.
(21).
This equation Thomson and Joule further simplified The mode in which the pressure and volume of a gas depend on each other varies less from the law of Mariotte according as the temperature is higher. Those terms of the foregoing equation which express this variation must thus become smaller as the temperature rises. The last term is the only one which actually fulfils this condition the last but one, irT, does not fulfil it. Accordingly this term should not appear in the equation, and putting
;
TT
= 0, we
obtain
.
= i2|-iJC7,(jy
.(22).
This is the equation which according to Thomson and Joule must be used for gases actually existing, in place of equation (20) which holds for perfect gases. An exactly similar equation was previously deduced by Rankine*, in order to represent the variations from the law of Mariotte and Gay-Lussac, found by Eegnault
* Phil. Trans.,
18.54, p, 336.
152
228
simplest
,an\
(23),
pv = BT--^
is constant. in which a like If we divide this equation by p, and in the last term, which is yery small, replace the product pv by the yery nearly equal
product
BT, and
we
,
obtain
which
is
(22).
On
As a
further
example of the
different
results-
which
may be
consider the behaviour of saturated vapour. will assume two conditions (1) that the vapour expanding has to overcome a resistance equal to its" whole force of expansion ; (2) that it escapes into the
produced by expansion,
we will
We
atmosphere, and thus has only to overcome the atmospheric pressure. Under the last condition we may make a distinction according as the vapour is separate from liquid in the
vessel from which it escapes, or is in contact with liquid, which continually replaces by fresh evaporation the vapour which is lost. In all three cases we will determine the
quantity of heat, which must be given to or taken from the vapour during expansion, in order that it may -continue throughout at maximum density. First then let us suppose a vessel to contain a unitweight of saturated vapour, and let this vapour expand, e. g. by pushing a piston before it. In so doing let it exert upon the piston the whole expansive force which it possesses at each stage of its expansion. For this it is requisite only that the piston should move so slowly that the vapour which follows should always be able to equalize its expansive force to that of the vapour which remains behind in the vessel. The quantity of heat Q, which must be imparted to this vapour, if it expands so far as that its temperature falls from a given initial value T^ to a value T^, is simply found by the equation ..
g^r^j^dT,.,,
(24).
ON NON-REVERSIBLE PROCESSES,
Here
h,
229
named the
the iiiagnitude introduced ia Chapter VI., and Heat of Saturated Vapour. If, as is the case with jpaost vapours, h has a negative value, the foregoing integral, in which the upper limit is less than the lower,
is
Specific
by formula
(31)
of
Chapter VI.,
viz.,
n = I'OU
8003
rp
Applying this formula it is easy to calculate the value of Q for any two temperatures 7', and T^. For example let us assume that the steam has an initial pressure of 5 or of 10 atmospheres, and that it expands until its pressure has fallen to one atmosphere; then by Regnault's tables we must put T^= a+152"2, or = a -1- 180".3 respectively, and T^ = a + 100; we thus obtain
the values Q = 52'1 or = 74"9 units of heat respectively. In the second case we suppose that a vessel contains a unit-weight of saturated vapour apart from liquid, and at a temperature T^ which is above the boiling point of the liquid; and that an opening is made in the vessel, so that the vapour escapes into the atmosphere. Let us proceed to a distance beyond the opening such that the pressure of the vapour is there only equal to the atmospheric pressure. To insure that the current, of vapour shall expand in the proper manner, let the vessel be fitted at the opening with a trumpet-shaped mouth KJPQM (Fig. 22.) This mouth is not actually needed in order that the equations which follow may hold, but merely serves to facilitate the be a surface conception. Let within this mouth, such that the pressure of the vapour is there only equal to atmospheric pressure, and velocity so small that its vis its We will viva may be neglected. further assume that the heat generated by the friction of the vapour against the edge of the .opening and the surface of the mouth is not dissipated, but again imparted to the vapour. Now to determine the quantity "
KLM
230
of heat wliich must be imparted to the vapour during expansion, if it is to remain throughout in the saturated condition, we will again apply the general equation (2) ; which gives, if in this case we denote the heat by ^',
Q'^U,-U,+ W.
(25);
here U^ is the energy of the vapour in its initial condition within the vessel, If^ the energy of the vapour in its final condition at the surface KLM, and the external work done in overcoming the pressure of the atmosphere. The energy of a unit-weight of saturated vapour at temperature T is given by the value of in equation (37) = M=l, It is of Chapter IX., if we there put
First give to
T the
initial
value
T^,
and
let
yjj
be the values of ^,
ture.
-^
and p corresponding to
final value T^,
tempera-
Again
let
have the
and let^^,
(]
and Pj be the corresponding values. Then subtracting these two equations from each other we have
^^-^-/I;(^-^S)^^-^^^
1--4^^
-Pi
1-
A
^.(S.
.(26).
The
external
s,
work which
s^,
the atmospheric
pressure p^,
is
volume
to
volume
ON NON-EEVERSIBLE
PEOCESSE'S.
231
We will give another form to this expression. If, as in -Chapter VI., we put s = u + a; where cr is the specific. volume of the liquid, the equation becomes
Substituting for
tt
F=i>,
+pA'^.-'^^)
(27).
^m. <^\
Now
iand of
substituting in (25) the value of U^~ U^ from (26), from (27), we arrive at the equation
+P^iF.-<^d
(28).
expressed in mechanical units. To is in ordinary heat units the right-hand side must
be divided hy E.
As
before
o-
we wiU put
^ = c;^ = r.
(ff*
At
is
we
will
o-^).
Thus we obtain
g=\^cdT + r,-r,^--^^[p,-p,)
AdT),
(29).
This equation is well adapted for the numerical calculation of Q, since the quantities which it contains have all been determined experimentally for a considerable number
of liquids.
0+1^=0-305;
whence
'"''"
(2;
T^j.
232
quantities in the last term of equation (29) are also known, so that the whole calculation is easy. For example if we take the initial temperature at five or ten atmospheres, we have Q' = 19'5 or = l7"0 units of heat
sufficiently
The
respectively.
this case also heat taken from it, if no part of it is to be allowed to condense which condensation might take pilace not only at the opening, but equally well inside the vessel. The quantity o? vapour so condensed would however be less than in the first case, because Q' is less than Q. It may easily happen that the above equations give a larger quantity of heat for .an initial pressure of five than of ten atmospheres. The reason is that at five atmospheres the volume of the vapour is already very small and tlie diminution of volume, when the pressure is raised to ten
Since
must be imparted
;-
atmospheres,
is
work during the escape of the vapour is more than balanced by the excess of the free heat in the vapour at 180"3 over
that in the vapour at 152'2. L Lastly let us take the third case, in which the vessel contains liquid as well as vapour. Let the vessel (Fig. 23) be filled to the level
vapour.
PQ
of escape, fitted, as
with
the
mouth
the spreading out of the current of vapour. Let there be some source of heat which keeps the liquid at a constant temperature T^, so that it continually gives off new vapour to replace that which escapes, and thus the conditions of the escape remain always j,. the same. This last circumstance makes an important distinction between this case and the foregoing. The pressure, which the vapour newly given oif exerts' on that alrea,dy existing.
to regulate
.
KPQM,
ON KON-BEVBRSIBLE PROCESSES.
'
SSS
performs work during tlie escape of the vapour, wMch must be brought into the calculation as negative external work. Let OHJ be a sui-face at which the vapour which passes through has still the same expansive force p^ teijipefature T^, and specific volume s, which exist within the vessel, and at which the new vapour is given off. Again let be a surface at which the vapour passing through has simply the expansive force equal to the atmospheric pressure p^. At both surfaces we shall assume the velocity to be so small that its, vis viva may be neglected. In its. passage from one surface to the other, the vapour must continually have just that measure of heat given to it or taken from it, which is necessary in order to keep it wholly in the gaseous condition, and completely saturated, and also- in order that at the surface it may have the temperature T^ corresponding to the pressure p^ {i.e. the boiling temperature of the liquid), and the specific volume s^ belonging to that temperature. We have now to enquire how large this quantity of heat Q" must be for each unit-weight of the escaping vapour. To determine this we may proceed as in the last case, remembering that we have now a different value for the external work. This value is the difference between the work done at the surface GHJ, through which passes a volume of vapour s^ at pressure p^, and that done at the surface KLM, through which passes a volume s^ at pressure It is thus given by the equation p^.
KLM
KLM
'W=p.,s^-p,s^.
dT
we have
If
we now form
(25), (26),
for Q" an equation of the same form as and in it substitute for t/^ E/j the expression given in and for TTthe expression given above, the main terms in
234
these
and there
re-
mains
Q"=Jl{(^-P%)
If
^^+ P^ - P^ +P.'^,-P^<r,....i^l).
we transform this equation so. that it refers not, to mechanical but to ordinary units of heat, and neglect the terms containing cr, we arrive at the simple equation
Q"
= r'cdT + r,-r,
forni
(32).
Q" =
and
if
- 0-305(7,-2;);
Q"
for
we
an
initial
we
obtain
Q" =
15"9 or
= 24f5
Since the values of Q are negative, it follows that in this case heat must be taken out of the vapour, not imparted to it. If this withdrawal of heat does not take place to a
any place under consideration, then the there hotter than 100 and therefore superheated. Here it is of course assumed that nothing but steam passes through the first surface GHJ, and thus that there are no particles of liquid mechanically carried off by the steam, as
sufficient extent at
is
steam
'
may happen
CHAPTEE
XI.
1.
Necessity of
a new Investigation
into the
Theory of
the Stewm-Engine.
Since the altered views as to the nature and action of Heat, which are comprised under the name of the Mechanical Theory of Heat, had their first origin in the known fact that heat can be applied to produce mechanical work, it might have been at once expected that the theory so formed would conversely serve to place this application of heat in a clearer light. In particular the more general point of view thus obtained would make it possible to pass a more certain judgment upon the particular machines used for this application, as to whether they abeady completely fulfilled their purpose, or whether and how far they failed to do so. To these reasons, which apply to all thermo-dynamic machines, are joined in the case of the most important of them, the steam-engine, certain special grounds, which make it desirable to undertake a new investigation into its working, derived from the mechanical theory of heat. This theory in fact, in the case of steam of maximum density, has brought to light certain important departures from the laws previously assumed as correct, or at least generally used for purposes of calculation.
236
On this head we need only refer to two results given in Chapter VI. In most of the recent writings on the steamengine, amongst others the excellent work of de Pambour, the foundation of the theory has been taken to be the law of Watt, viz. that saturated steam when contained in a non-conducting vessel remains during all changes of volume steam of maximum density. In some, later writings, after the publication of Eegnault's researches on the heat required to evaporate water at different temperatures, the assumption is made that steam partly condenses during compression, and during expansion cools in a less degree than corresponds
to the reduction of density, and therefore passes into the superheated condition. On the other hand it is proved
steam must behave in a way which assumption and the exact opposite of the second assumption, viz. that it is superheated during compression, and is partly condensed during expanin Chapter VI. that
is
different
from the
first
sion.
Further it is assumed in the above writings, in default of more accurate means of determining the volume of a unitweight of steam at different temperatures, that steam even at its maximum density still follows the law of Mariotte and Gay-Lussac. On the other hand it is shewn in Chapter VI. that it departs widely from that law. These two points have naturally an important influence on the quantity of steam which passes from the boiler into the cylinder at each stroke, and on the behaviour of this steam during expansion. It is thus obvious that they are themselves sufficient to make it necessary that we should calculate in a different way from that hitherto adopted the amount of
steam-engine.
2.
On
the
In order to illustrate more clearly the series of processes which make up the action of a condensing steam-engine, and to bring out clearly the fact that they form a cyclical process, continually repeating itself in the same manner, the imaginary diagram (Fig. 24) may be employed, A is the boiler, the contents of which are kept uniformly at
237
part of the steam passes into the cylinder B, and drives the piston a certain distance upwards. Then the cylinder is shut off from the boiler, and the enclosed steam drives the piston still higher by expansion. The cylinder is now put in connection with the vessel C, which represents the condenser. * It will be supposed that Pig. 24. this condenser iskeptcold, not by injected water, but by cooling from without this makes no great difference in the results, but simplifies the treaitment. The constant temperature of the condenser we may call T. During the connection of the cylinder with the condenser, the piston returns through the whole distance it has previously traversed and thereby all the steam which has not of itself passed into the condenser is driven into it, and there condenses into water. It remains, in order to complete the cycle of operations, that this condensed water should be brought back again into the boiler. This is effected by the small pump D, the working of which is so regulated, that during the upward stroke of its piston it draws out of the condenser exactly as much water as has been brought into it by the condensation during the last stroke ; and this water is then, by the downward stroke of its piston, forced back into the boiler. When it has here been heated once more to the temperature y^' all is again restored to its initial condition, and the same series of processes may begin anew. Thus we have here to deal with a complete cyclical process. In the common steam-engine the steam passes into the cylinder not at one end only, but at both ends alternately, The only difference thereby produced is, however, that during one up and down stroke of the piston two cyclical processes take place instead of one, and it is sufficient in this case to
: ;
238
determine the work done during one process, in order to be able to deduce the whole work done during any given time. In the case of an engine without a condenser, we have only to assume that it is fed with water at 100", and we may then suppose it replaced by an engine with a condenser, the temperature of the condenser being 100,
3.
For the purpose of this investigation we will assume, as has usually been done, that the cylinder is a non-conducting vessel, and so neglect the exchange of heat which takes place during each stroke between the walls of the cylinder and the steam. The vapour within the cylinder can never be anything but steam of maximum density with a certain admixture of water. For it is evident from the conclusions of Chapter VI., that during the expansion which takes place in the cylinder after it is shut off from the boiler the steam cannot pass into the superheated condition, because no heat is imparted to it from without but must rather partially condense. It is true that there are certain other processes, to be mentioned later, which tend to produce a slight super-heating ; but this is prevented from taking place by the fact, that the steam always carries with it into the cylinder a certain amount of water in the form of spray, with which it remains in contact. The exact amount of this water is of no importance ; and since for the most part it is diffused through the steam in fine drops, and therefore readily participates in the changes of temperature which the steam undergoes during expansion, no important error will be introduced if at each moment under consideration we assume that the tenaperature of the whole mass of vapour in the cylinder is the same. Further, to avoid too great complexity in the formulae, we will first determine the whole work done by the steam pressure, without examining how much of this is actually useful work, and how much is expended on the engine itself in overcoming friction, and in actuating the pumps required, besides the one shewn in the figure, for the proper working of the machine. This latter part of the work may be subsequently determined and deducted from the whole, in the manner shown later oh; It may further be remarked with regard to the friction between
;
239
the piston and cylinder, that the work expended upon this is not to be regarded as wholly lost. For since heat is generated by this friction, the inside of the cylinder is thereby kept hotter than it otherwise would be, and the power of the steam increased accordingly. Lastly, since it is desirable to understand the working of the most perfect machine possible, before enquiring into the influence of the various imperfections which occur in practice, we will in this preliminary investigation make two further assumptions, which may afterwards be withdrawn. The first is that the inlet pipe from the boiler to the cylinder, and the outlet pipe to the condenser or to the atmosphere, are so large, or else the speed of the engine so slow, that the pressure within the end of the cylinder connected with the boiler is always equal to the pressure in the boiler itself; and similarly that the pressure within the other end is always equal to that in the condenser, or to the atmospheric pressure as the case may be. The second is that there are no clearance or waste spaces sufficient to affect the result.
4.
Determination of the
Work
stroke.
conditions just enumerated the amount of cyclical process corresponding to a single stroke may be written down by help of the results obtained in Chapter YI. without further calculation, and their sum comprised in a simple expression. Let be the whole quantity of vapour which passes from the boiler to the cylinder during one stroke ; of this let the
Under the
m^ part OTj be in the form of steam and the remainder in the form of water. The space which this mass occupies will be (by Ch. VI., 1) m, , + Mfr, where w, represents the value of u corresponding to T^, while o- is taken as constant and therefore has no suffix. The piston has therefore been raised so far that this amount of space is left under it and since this takes place at the pressure p^ corresponding to T the work done during the first pr6cess, which we may call W^, is given by the following equation
;
W, = m^u^p, + McTp^
(!)
240
Let the expansion whichi succeeds to this continue so far that the temperature of the vapour enclosed in the cylinder The work done during falls from the value T^ to a value T^. this expansion, which we may call W^, is given directly by equation (62) of Chapter VI., if we there take T^ for the final temperature, and corresponding" values for the other quantities involved. Thus
F, = m,
(p,
(2).
On the return stroke of the piston, which now begins, the vapour, which at the end of the expansion occupied the space m^ u^ + Mix, is driven out of the cylinder into the condenser, overcoming the constant resistance jp^ The pegative work thus performed is therefore given by the equation F, = -,Mji)-if<rp (3).
the piston of the small pump rise until it leavesthe space Ma; the pressure then continues to be the pressure p^ of the condenser, and the work done is
let
it
Now
under
W.^Map,
.-.....(4).
Finally, during the descent of this piston, the pressure pj of the boiler has to be overcome, and we have therefore the
negative
work
W, = -M<Tp,
(5).
Adding these five equations we have for the whole work done during the cyclical process by the steam pressure, or in other words by the heat, which work we may call W, the
following expression
W = m,p, - m,p, + MG
From
this equation
{T,
. .
(6).
we must
by equation
then
wijj
241
we
substitute therein p
and
C for
aad
c,
the expression
Substituting this expression we obtain an equation in which all the quantities on the right-hand side are known, since the masses and m^, and the temperatures 2\, T^, and 1\ are supposed to be known directly, and the quantities p, p, and
are
assumed
to
5.
Special
Forms of the
section.
If in equation (6) we put T^=T^, we obtain the work done in the case where the machine works without expansion, viz.
W' = m^u,{p,-p)
(7).
If on the other hand we assume that the expansion continues until the steam has cooled by expanding from the temperature of the boiler down to that of the condenser (an assumption which cannot be realized in practice, but forms
the limiting case to which we may approach as near as is practicable) we have only to put 1\ = T^ ; whence we
have
(8).
Eliminating m^p^ from this equation by means of tha (55) Chapter VI., in which we must also put
= T, we
have
W'=m,p,'^j,^^''+
6.
Mc{t,-T, + Tm'^^
(9).
Imperfections in the
Engiiie.
With all steam-engines as actually constructed the expansion falls much below the maximum value given at the end of the last section. If for example we take the temperature of the boiler at 150, and that of the condenser at 50", then if the temperature of the steam in the cylinder is to be lowered by expansion to the temperature of the condensei-,
c.
,
IG
242
the steam must be expanded (by the table given in 13 In of Chapter VI.) to twenty-six times its original volume. practice, on account of the many evils attending too high an expansion, steam is not expanded beyond three or four times Such its volume in general, or ten times at the very utmost. an expansion, with an initial temperature of 151, is shewn by the same tables to lower the temperature to 100, or 75 at the utmost, instead of to 50. Besides this imperfection, which has already been taken account of in the above investigation and included in equation (6), the steam-engine is subject to several others, two of which have been already expressly excluded from consideraThese are, first the fact that the pressure in one end tion. of the cylinder is less than that in the boiler, and in the other greater than that in the condenser, and secondly the presence of waste spaces. must now extend our previous investigations so as to include these further imperfections. The influence upon the work done of the difference between the boiler and cylinder pressures has been investigated most fully by Pambour in his work TMorie des Machines d Vapeur. The author may therefore be allowed, before himself entering on the subject, to reproduce the most important of these investigations, only making some changes in the notation and omitting the quantities which refer to It will thus be easier to shew how far Pambour's friction. results are no longer in accordance with our present knowledge on the subject of heat, and at the same time to connect with it the new method of investigation, which in the author's opinion must take its place.
We
the
relation
between
Pambour's theory has its foundation in the two laws already mentioned, which at that time were generally applied to the case of steam. The first is the law of Watt, viz. that the sum of free and latent heat is always constant. From this law, as already mentioned, was drawn the conclusion that if a certain quantity of steam at maximum density were inclosed in a non-conducting vessel, and the contents of this vessel then increased or diminished, the steam would neither .be superheated nor partially condensed, but would remain
243
precisely at maximum density; and that this would take place without any reference to the way in which the change
volume was produced, or to whether the steam had to overcome a resistance corresponding to its expansive force, or not. This was the condition of things which Pambour supposed to exist in the steam cylinder; and he also assumed that no clfange of importance would be occasioned by the presence of the particles of water which are always mixed with the steam. Next, to determine more closely the connection between volume and temperature, or between volume and pressure, in the case of steam at maximum density, Pambour applied further the law of Mariotte and Gay-Lussac. If we take the volume of a kilogram of steam at 100 and at maximum density to be 1*696 c. ipa., as given by Gay-Lussac, and if we remember that the corresponding pressure of one atmosphere represents a weight of 10333 kilograms to the square metre and if for any other temperature T we call v the volume and p the pressure then the law of Mariotte and Gay-Lussac leads to the following equation:
of
; ;
= ,.1-696
,,. (10).
Here we have only to substitute for p the value given by the table of pressures, and we can then determine the volume corresponding to any temperature, provided the foregoing assumptions are correct. Since in the formulae for the work done by the steamengine
I
part, it
was necessary
for
the
convenient treatment of this integral, that the simplest The possible relation between p and v should be used. equations which would have been obtained, if it had been from the foregoing sought to eliminate the temperature equation by means of one of the ordinary empirical fortnulae
for p, would have been too complicated Pambour accordinglypreferred to form a special empirical formula for the purpose, and following the example of Navier he gave it the following general form
:
"-u^--.
("'
-162
244
wtere and h are constants. These constants he endeavoured to fix so that the volumes calculated by this formula might agree as nearly as possible with those calculated by the formula given above. As however sufficient accuracy cannot thus be arrived at in the case of all the pressures which are met with in the steam-engine, he made use of two different formulae in the cases of engines with and without condensers.
The
first,
for
condensing engines,
is
as follows
20000
"=1200+^
(l^'^)'
This agrees best with the formula (10) for pressures between and 3^ atmospheres, but may be applied within somewhat wider limits, say ^ and 5 atmospheres. The second, for non-
condensing engines,
is
as follows
21232
^
= 3020+^
,,,,, (11^>-
This is moSt accurate between 2 and 5 atmospheres, and be applied anywhere between 1 J and 10 atmospheres.
8.
may
Pambour's Determination of
the
single stroke.
The quantities needed for determining the work done, and depending on the dimensions of the engine, will be here denoted in a manner somewhat differing from that of Pambour. The whole space within the cylinder, including the waste space, which is left open for steam during a single stroke, we shall call v'. The waste space we shall call ev' and the space swept through by the piston (1 e) v. That part of the whole space, which is left open for the steam up to the moment when the cylinder is shut off from the boiler, again
we shall call ev'. Then the space swept through by the piston during the entrance of steam will be denoted by (e e) v, and that swept through during expansion by (1 e) v. First to determine the work done during the entrance of the steam. For this purpose we must know the actual pressure in the cylinder at this time, which must be less than that in the boiler, otherwise there would be no flow from one into the other. The ampunt of this difference cannot howinclusive of the waste space,
243
ever be fixed in general terms, because it depends not only on the construction of the engine, but also on how widely the valve in the steam pipe has been opened by the engineer, and on the speed with which the engine is moving. By a change in these circumstances the difference may be made to vary within wide limits. Again the pressure in the cylinder may not remain constant during the whole time the steam is entering, because the speed of the piston, and also the opening left by the valve, may be made to vary during this
time.
With reference to this latter point Pambour assumes that the mean pressure, to be used for determining the work done, may with sufficient accuracy be taken to be the same as the final pressure which exists in the cylinder at the moment when it is shut off from the boiler. The author does not think it desirable to introduce into the general formulae an assumption of this kind, although in the absence of more exact data it may fairly be resorted to for the purpose of actual
calculations
;
but he
is
bound
to follow
Pambour's method, in
order to complete the exposition of his theory. The actual pressure at the moment when the steam is shut off, Pambour determines by means of his equation, as given above, between volume and pressure ; assuming that special observations have been made to determine the quantity of steam which passes from the boiler to the cylinder during an unit of time, and therefore during each stroke. the whole quantity which passes wUl, as before, denote by the portion into the cylinder during one stroke, and by which is in the condition of steam. As this quantity M, of which Pambour only recognizes the part which is in the condition of steam, fills at the moment when the cylinder is closed the space ev', we have by equation (11),
We
e^==^
where p^
is
(12),
Hence
P.-'^-i
If
(12^).
we
(e
e)
?;',
which
is
the
246
space swept through by the piston up to this moment, we have the following equation for the first part of the wort
done:
W.^mBx^-^-i/ie-e)!)
The law
is also
..,,(13).
given by equation
of Tariation of the pressure during the exp|,nsion If v is the volume and p the (11).
pressure at any
moment, then
mB
This expression we must substitute in fpdv, and then Thence we obtain for integrate this from v^ev to v = v'. the second part of the work done
W^ = mB\og^-^v' 0.-e)l
(14).
Next, to determine the negative work done by the resistance during the return stroke, we must know the value of that resistance. Without entering at present into the question how this resistance is related to the pressure in the condenser, we will denote the mean pressure by p,; then the work done will be given by.
Tf,
= -'(l-6)^
(15).
Finally there remains the work which must be expended in forcing back into the boiler the quantity of liquid M. Pambour has taken no special account of this work, but included it with the friction of the engine. Since, however, for the sake of completing the cycle of operations, it has been included in the author's formulae, it will be investigated here in order to facilitate the comparison. If p^ be the pressure in the boiler, and p in the condenser, then equations (4) and (5) show, as- in the example already considered, that this
work
is
W, = -M<r{p,-p,)
For the present condenser itself, but
.to
(16).
where p^ is not the pressure in the in the end of the cylinder which is open
case,
the condenser, this equation is not quite exact ; but since on account of the smallness of a the value of the whole
247
almost too small to be taken into account, neglect an error which is small even in comparison with that value; and we shall therefore retain the expression in the form given above.
we may
Adding these four several quantities of work together, we obtain the following expression for the whole work done during the cyclical process
W = mB (iZL%iogl) _^'(l_e)
9.
(J,
the
Unit-weight
of Steam.
If instead of the work done in one stroke, during which the quantity of steam us^d is m, we prefer to find the work done per unit-weight of steam, all that is needed is to divide the foregoing value by m. Let us denote by I the fraction
.
ratio of the
the cylinder to that part of it which is in the form of steam, and is therefore somewhat greater than 1 ; by F the fraction
v'
.
i.e.
is
occupied by the
and by
W the fraction
Then wo
i,e.
have
W=B(^-^ + log'^yril-^e)ib+p,)-la(p,-p,)
(18).
In
a factor. Since this that the work, which can be obtained from one unit-Tsreight of steam, is the greatest, other things being equal, when the volume which that steam occuThe least value of pies in the cylinder is the least possible. this volume, to which we may continually approximate, but can never exactly obtain, is that which is given by the assumption that either the engine goes so slowly, or the steam-pipe is so
this contains
the volume
and
F as
term
is
negative,
it
follows
2i8
large, that the pressure in the cylinder is as great as that in the boiler. This then gives the maximum quantity of work. If with an equal admission of steam the speed of the engine is greater, or if with equal speed the admission of steam is less, then in each case a less quantity of work is obtained from' the same quantity of steam.
its
passage
from
the
Before we pass on to treat the same connected series of processes on the principles of the Mechanical Theory of Heat, it will be advantageous to consider one of them, which requires a special investigation of its own, in order to fix beforehand the results which refer to it. This process is the flow of the steam into the clearance or waste space and into the cylinder, in the case when it has a smaller pressure to overcome than that which forces it out of the boiler. The steam as it comes from the boiler passes first into the waste space there it compresses the steam of less density which remains over from the last stroke, fills up the space thus obtained, and then acts upon the piston ; this, according to the assumption, on account of its comparatively lighter load recedes so fast, that the steam cannot follow it quickly enough to keep the density in the cylinder the same as in the boiler. Under such circumstances, if nothing but saturated
;
steam escaped from the boiler, this would become superheated in the cylinder, inasmuch as the vis viva of the flow would be transformed into heat but since the steam always carries with it small particles of water, the superabundant heat goes to vaporize a part of these, and the steam thus remains
;
in the saturated condition. must now consider the following problem : Given the initial condition of the whole mass under consideration, as well that already found in the waste space as that which is newly received from the boiler; given also the amount of
We
work which is done during the entrance of the steam by the pressure which acts on the piston lastly given the pressure which exists at the moment when the boiler is shut off from the cylinder ; then to determine what proportion of the mass within the cylinder is at that moment in the condition of steam. Let fi. be the mass which exists in the waste space
;
249
before the entrance of the steam ; for the sake of generality we will assume this to be partly in the liquid, partly in the gaseous condition, and will call the latter part /i^. The pressure of this steam may for the moment be denoted by p^, and the corresponding absolute temperature by T, without implying thereby that these are exactly the same values as hold for the same quantities within the condenser. Let p^ and Tj be, as before, the pressure and temperature in the boiler, the mass which flows from the boiler into the cylinder, and jWj the part of which is in the condition of steam. The pressure exerted on the piston during the entrance of the steam need not, as already explained, be constant. will call p the mean pressure, by which the space swept through by the piston during the entrance of the steam must be multiplied, in order to obtain the same amount of work as is actually done by the varying pressure. Let ^2 be the actual pressure in the cylinder at the moment when the steam is cut off, T^ the corresponding temperature. Lastly let m^ be the magnitude we have to determine, viz. the part of the whole mass fi within the cylinder, which is in the condition of steam. To determine this quantity, let us suppose the mass M+fi to be brought back to its initial condition in any way whatever, e. g. the following. Let the gaseous part m^ be condensed in the cylinder by the fall of the piston, it being assumed that the piston can force itself even into the waste space; at the same time let the mass have. such a quantity of heat continuously imparted to it, that its temperature T^ remains constant. Then let the part ilf of the whole liquid mass be forced back into the boiler, where it again takes up its have now within the boiler original temperature T,. the same condition of things as before the flow of the steam,
We
M+
We
and therefore the proportions of liquid and steam must be the same as at the commencement. Whether the individual molecules, which are in the gaseous and in the liquid condition, are exactly the same as at the commencement does not concern us ; we make no 'distinction between these, and never enquire what molecules, but simply how many molecules, are in each of the two conditions.*
* If it
250
ON THE
Let the mass
]ilECHANICAI<
THEORY OF HEAT,
fi, which is not forced back into the boiler, cooled in the liquid condition from T^ to T, and then at this temperature let the paxt /* be transformed into steam the piston receding so far, that this steam can again occupy
be
first
M+
/j,
will
principle of all the quantities of heat taken in during a cyclical process must be equal to the total external work done. In this case the following quantities of heat have been taken in
to which
that the
sum
In the boiler, where the mass has been heated (1) from T^ to T^, and at the latter temperature the part m^ transformed into steam, we have the quantity of heat
m,p,
(2)
+ MCiT,-T;),
at temperature T^,
we
have
(3)
/t
from T^ to
T^,
we
have
(4)
fi^
at temperature
T^,
we have
The whole quantity Q, is thus given by
of heat taken in, which
we may
call
work are obtained as follows To determine the space swept through by the (1) piston during the entrance of the steam, we must remember
quantities of
gaseous condition at tlie end as at the beginning, we need only assume that the water forced back into the boiler is not only in quantity, but also in its actual molecules, the same as that which left the boiler previously; and that when this water takes up the temperature Tj, the quantity m, which was formerly vapour, is again vaporized, whilst an equal quantity of the existing steam is condensed. Por this purpose there is no need that the whole mass in the boiler should take in or give out any heat, because that required for the vaporization, and that generated by the condensation, exactly balance each other.
The
251
up hj
M+
fi
at the
end of
this
From
is filled
Since this subtract the waste space. fi at temperature T^, of which the part fi^ is in the condition of steam, it may
this
we must
at the
commencement by a mass
be represented by
fi^%
+ fia-.
If we subtract this expression from the former, and multiply the remainder by the mean pressure p\, we obtain for the first part of the work done
(2)'
Wjis
(3)
in forcing the
mass
boiler is
-Map,.
(4)
Adding these four quantities together following expression for the whole work done
we
obtain the
(20).
equate the expressions in (19) and and bringing to one side of the equation the terms (20) containing m^, we obtaiA
Q = W,'we may
By means of this ecjuation the quantity m^ in terms of quantities assumed to be already known.
given
In cases where the mean pressure p/ is much larger than the final pressure ^j, e.g. if we suppose that during
the greater part of the time of admission of the steam.
252
the pressure in the cylinder is nearly- the same as in the boiler, and that it is only by the expansion that .the pressure is finally brought down to the value p^, it may happen that the value found for m^ is less than m^ + ji,^, and accordingly that a part of the original steam must have condensed. On the other hand if p^ is only a little greater or even a little smaller than p.^, then the value This, latter is in general of wij, will be greater than m^ + fi^. true of the steam-engine, and in particular holds for the
case assumed by Pambour, in which p^ =p^. Here as in Chapter VI. we have arrived at a result widely diverging from the views of Pambour. Whereas he assumed for the two different kinds of expansion, which
succeed each other in the steam-engine, one and the same law, according to which the original quantity of steam can neither grow less nor greater, but must always remain
exactly at maximum density, we have been led to two separate equation", which indicate an opposite condition of things. In the first expansion, during the admission of sbeam, there must by equation (21), be a continual generation of fresh steam in the further expansion, after the steam is cut off, and while it is doing the full work corresponding to its force- of expansion, there must by equation (36) of Chapter VI. be a condensation of some part of the existing steam. Since these two opposite processes of increase and diminution of steam, which must also exert an opposite influence on the amount of work done by the engine, partly go to cancel each other, the final result may in certain circumstances be nearly the same as under the simpler assumption of Pambour. We must not however on this account cease to take into consideration the difference we have discovered, especially if our object is to determine what effect a change in the arrangement or speed of the engine will have on the amount of work done.
;
Determination of the'Work done during one stroke, 12. taking into consideration the Imperfections already noticed.
We may now return to the complete cyclical process performed by the steam-engine, and treat its several parts one after another in the same way as before.
253
flows from the boiler, in which the pressure the cylinder. Of this mass the part m^ is in the condition of steam, the remainder of water. The mean pressure in the cylinder during this time we will as before call p^, and the final pressure p^. The steam now expands, until its pressure has fallen from p^ to p^, and its temperature from jTjj to Ty Then the cylinder is opened to the condenser, the pressure in which is p^, and the piston moves back again through the whole length of the stroke. The back pressure which it has to overcome is on account of its comparatively rapid motion somewhat larger than^^, and the mean value of this back pressure we will for the sake of distinction call p^. The steam which remains in the waste space at the end of the return stroke, and must be taken into account for the next stroke, has a pressure which may not be equal either to p^ or to p', and therefore must be denoted by />". It may be greater or less than p^, according as the steam is shut off from the condenser a little before or after "the end of the return stroke ; for in the former case the steam will be compressed still further, while in the latter it will have time to escape partially into the condenser and so to must be returned from expand again. Finally the mass the condenser into the boiler, during which as before the pressure p assists the operation, and the presstire p^ has to be overcome. The quantities of work done during these processes will be represented by expressions very similar to those in the simpler case already considered a few obvious changes must be made in the suiBxes to the letters, and the quantities thus obtain the added which refer to the waste space. following equations During the admission of steam we have, as in 10, only writing ," instead of w,
is
p,,
The mass
into
We
W, =
(22).
During the expansion from pressure p^ to p^, we have as in equation {^^) of Chapter VI., writing M-^yi, instead of M,
TF',
254
HiiAT.
During the return stroke of the piston, where the whole, space swept through by the piston equals the whole space occupied by ilf + /* at pressure p,, less the waste space which is represented by /AoMj" + /mt, we have
F3 = - Km, +M<r^ fiX') Po'During the forcing back of the mass have
.(24).
W, = -Mcr(p,^p^
Hence
for'
we have
{T, - T,) + WA (p/ p,) + m^u^ (p^ -p,') -Mo- (p. -p^ +p: - p,) - i^X' (p:
-Po')
-(26).
In this equation the masses m^ and m, are given by equation (21) and by equation (55) of Chapter vl. respectively; substituting in the former p^" for p^, and changing T^, r^, m in the same way, while in the latter + fi is to be sub^ stituted for M, p for r, and G for c. The elimination of m^ and wij, is thus rendered possible: here however we shall only make the substitution in the case of one of them, m^, as it is more convenient for purposes of calculation to combine the equation so obtained with the two named above. Thus the system of equations serving to determine the work done by the steam-engine wiU, in its most general form, be as
follows
W=
m,p, -m^p^
+ MG {T, - T,) + MoPo" -VC (T, - T:) + m,u, (P3 -p^) + {Po-Po") - M<T (p; -j5)
mX
'^i\P2-^^^ipl-P.)]=in,p,
..(27).
+ l^.<{p:-pn+Ma(^,-p;)
Y==^+(M+;.)aiog^
255
13. Pressure of Steam in th6 Cylinder during the different Stages of the Process, and corresponding Simplifications of the Equations. To obtain numerical values from equations (27), we must first determine more accurately the quantities p^,
No general law can be laid down as to the mode in which the pressure within the cylinder varies during the admission of steam, because the opening and closing of the steam pipe is performed with different engines in different ways. For the same reason no fixed general value can be laid down for the relation between the mean pressure p^' and the final pressure p^ taking the latter in its strictest meaning. This is however possible if we make a slight change in the signification of ^2-
The shutting off of the cylinder from the boiler cannot of course be instantaneous: the necessary motion of the valve or cock must consume a certain time, less or greater according to the arrangement of the gear. During this time the steam in the cylinder expands somewhat, because as the opening narrows the quantity of fresh steam which enters is may less than corresponds to the speed of the piston. therefore assume in general that at the end of this time the pressure is already somewhat less than p^. Again, if we do not limit ourselves to taking the end of the time required to close the valve as the moment of cut-off, but allow ourselves some freedom in fixing this moment,
We
other values
may be
had so chosen that, if at this moment the whole mass already been admitted, a pressure would then exist exactly equal to the mean pressure calculated for all the time up to
If we substitute for the actual cut-off the instantaneous cut-off as thus determined, the resulting error in the calculation of the work will be quite insignificant. With this modification therefore we can fall in with Pamr hour's assumption, that _Pi' = p, but we must then for each particular case make a special investigation of the circumstances, in order to determine accurately the moment to be fixed for the cut-off. As regards the back-pressure p^, which exists during the return stroke, the difference Po'Po is obviously smaller,
this instant.
256
other things being equal, as p^ is smaller. It is therefore smaller for a condensing engine than for a high-pressure In the case of the engine, where jo equals one atmosphere. most important high-pressure engine, the locomotive, there is generally a special circumstance which tends to increase this difference, viz. that the steam is not usually led into the air by the shortest and widest channel possible, but is led into the chimney, and there allowed to escape through a somewhat contracted blast-pipe, in order to increase the draught. In this case an accurate determination of this difference is of importance to obtain a trustworthy result. It must also be remembered that this difference is not constant even with the same engine, but depends upon the speed, and the law of this dependence must therefore be investigated. This subject, with the researches made upon it, will not however be entered on here, since it has nothing to do with the present application of the Mechanical Theory of Heat. In the case of engines where the exhaust steam is not thus employed, and especially with condensing engines, p' differs so little from p^, and therefore varies so slightly with the speed, that it is usually sufficient to assume a mean value for^d'. Moreover, since p^ enters into equations (27) only in terms which are multiplied by a; and therefore have but smg,ll influence on the value of the work done, we may also substitute for p,, the most probable value for p'. The pressure p^', which exists in the waste space, depends, as already mentioned, on whether the shutting off from the condenser takes place before or after the end of the stroke, and may therefore have very different values. But this pressure, and the quantities depending on it, occur in equations (27) only in terms containing the small factors /j, and fj,, so that an approximate value is sufficient for our purpose. In cases where there are no special circumstances which cause p^' to vary widely from p^', we may neglect this difference, as we did that between p and p/, and may thus assume as a general value for all three quantities the most probable mean value for the back-pressure within the cylinder. This value we may call p^. These simplifications change equations (27) into the following
APPLICATIDir TO
THE STEAM-ENGINE.
]
257
W =mj>, - m,p, + MG
m^. = .?. + MO (T, -
{T,
(28).
14. Substitution of the Volwme for the porresponding Temp&rature in certain cases.
In the above equations it is assumed that, besides the masses M,m^, fju, and /*,,, of which the two first must be found by direct observation, and the two latter can be approximately determined from the size of the waste space, we have also given the four pressures p^, p^, p and p, or, which is the same thing, the four temperatures T^, T^,- 1\, and T^, In practice however this condition is only partially fulfilled, and we must therefore bring in other data to assist us. Of the four pressures here mentioned, two only, p^ and p^, may be taken as known of these the first is given directly by the gauge on the boiler, and the latter can be at least approximately fixed by the gauge on the condenser. The two others, p^ and p^, are not given directly ; but we know the dimensions of the cylinder and the point of cut-ofi", and can thence deduce the volume of the steam at the moment of cut-off and at the end of the stroke. We may then take these volumes as our data in place of the pressures p^ and p^. For this purpose we must throw the equations into a form, which will enable us to use the above data for calculation. Let us now, as before in the explanation of Pambour's theory, denote by v' the whole space within the cylinder which is open to the steam during one stroke, including the waste space; by ev' the space opened to the steam up to the moment of cut-off; and by e/ the waste space. Then in the same way as before we have the following equations
'
:
mijj ttj
-f-
(i-t- /t)
o-
= ev',
c.
17
258
The
duct
may
be at once neglected
= ev /*o
Further,
we have by equation
P = Tug,
writing the single letter
for
-^, which
Vyill
occur very
frequently in what follows. can therefore replace p^ and Then the masses m, and Pj in equations (28) by u^ and u^. TWj occur only in the products vn^u^ and m^u^, for which we may substitute the values given in the two first of equations (29). Again, by the last of these equations we may eliminate the mass fi^; and as regards the qther mass n, though this may be somewhat larger than fi^, yet, since the terms which contain (i are generally insignificant, we can without serious error use the value found for /t: in other words we may drop for the purposes of calculation the assumption made for the sake of generality, that the original mass in the waste space is partly liquid and partly gaseous, and consider the whole of it to be in the gaseous form. The above substitutions may be effected in the more general equations (27), as well as in the simplified equations This substitution presents no difficulty, and we will (28). here confine ourselves to the latter set, in order to have the equations in a form adapted for calculation. With these changes they read as follows
:
We
+ SV
{ev'
,p^-C(T,-T,)
259
15.
To adapt the above equations, which give the work per stroke or per quantity of steam w,, to determine the work per unit-weight of steam, we must apply the same method as before in transforming equation (17) into (18): viz. to divide the three equations by m,, and then put
^ m=
?,
m,.
^=F,andI^=F; TO.
+er (p,-C{T,-T,) f
(V-I<T)g, = {er-la)g,+ (l+')
(31).
Clog^
16.
The application of the above equations to calculate the work done may be effected as follows. Assuming the pressure of steam to be known, and also the speed at which the engine works, we can thence determine the volume V of one
unit-weight of steam. By help of this value we first calcu^ late the temperature T^ from the second equation, then T^ from the third, and finally apply these to determine the work done from the first equation. Here however we are met by another difficulty. In order to calculate T^ and T, from the two latter equations, these must be solved for those temperatures. But they contaio those temperatures not only explicitly but also implicitly, If to eliminate since p and g are functions of temperature. these quantities we use for p one of the ordinary empirical expressions for the pressure as a function of temperature, and for g the differential coefficient of the same expression, then the equations become too complicated for
172
260
It would probably be possible to adopt the same method as Pambour, and to form new empirical formulae, more convenient for the present purpose, and sufficiently accurate, if not for all temperatures at least "withiA certain limits. This however will not here be attempted, but another method will be adopted instead, which makes the calculation somewhat extended; but easily performed in its
further treatment.
individual parts.
17.
Determination of -^ or
g,
If the range pf pressure of the vapour at different temperais known with sufficient accuracy ,for any liquid, then the values of g and Tg can also be calculated for the same temperatures, and collected in tables, as usually done with the values of p. For steam, which has hitherto been the only vapour used for the steam-engine, the author has performed such a calculation, by help of Keguault's tables, for temperatures from 0 to 200. For this purpose he differentiated according to t the formula used by Regnault to. calculate the values of p under and above 100; and then calculated 5' by means of the new formulse thus obtained. But since these formulae did not seem to answer the purpose so well as to repay the great labour involved, and since the forming and calculating of other more suitable formulae proved still more tedious, the author was contented to use the numbers already calculated for the pressure to determine approximately the differential coefficient of that pressure: e.g., if the pressure for the temperatures 146 and 148 were
tures
p,^,, it
^'" would
represent with sufficient accuracy the value of the differential coefficient for the mean temperature ]47.
Above 100 the same numbers were employed as were used by Regnault*. With regard to values under 100 it has been recently pointed out by Moritzf that the formula used by Regnault is somewhat inaccurate, especially in the neighr
*
AKm.
t Bulletin de
Vol. XIII. p. 41.
de VAcad. des Sciences, Vol. xxi. p. 625. la Classe physico-mathSmatique de VAcad.' de St Pitershcmra
'
261
bourhood of
100, owing to his having employed for the calculation of the constants logarithms to seven places of
decimals only. Moritz has therefore recalculated these constants from the same observed values, but using ten places of logarithms, and has published the values oip derived from this improved formula, so far as they differ from Regnault's values, which first takes place at 40. These are the values used by the author*. When g has been calculated for the various temperatures,, the product Tff can be calculated without further difficulty, since is given by the simple equation
The values of g and Tg thus found are given in a table at the end of this Chapter. To complete this the corresponding values oip are added, those from 0 to 40 and above 100 being Regnault's numbers, and those from 40 to 100 being Moritz's. By the side of each of these three columns are given the differences between each two successive numbers, so that this table enables the values of these thrqe quantities to be found for any given temperature, and conversely the temperature corresponding to any given value of one of these
three quantities.
18. Heat.
One
other remark
is
to be
made
as to the
mode
of using
this table. In equations (31) it is assumed that the pressure p and its differential coefficient g are expressed in kilograms
per square metre, whereas in the table the same unit of pressure is employed as in Regnault's tables, viz. a millimetre of mercury. Hence, if in the formulae which follow
we are to consider p. and g as expressed in these latter iinits, we must alter equations (31) by multiplying p and g by the number 13"596 which expre3ses the specific weight of mercury;
inasmuch as
the units.
stitute
this,
by Chapter
yi,, 10,
is
we denote this number by A;, we must subhp and hg for p and g, whenever these occur in the
If
Similarly for the quantities G and p, which express the specific heat and the heat of vaporization in,
above equations.
262
mechanical units, we will use c and r which, refer to the ordinary unit of heat. For this purpose we must writers andEr for G and p. If we now divide the equations by h, so as to bring the conand k together as much as possible, these equations stants
-j-
and
~ = ~[r, + h{T,-TJ\-{V-l<T){T,g,-p,+p,)
k
M
E,
(32).
{V-W)g,= {eV-la)9,+
The quantity
(l
+ '-Pl^\og^
E has the
y
following value
^_
19.
&-l3^ = ^^^^2^
Tj,
423-55
.(33).
and
T
(34),
The second
as follows
_E
6=
eV-la-
(34a).
ev-y
263
Of the three terms on the right-haud side of (34) the first by far the most important and it is thus possible to determine T^g^, and thereby also the temperature t^ by successive approximations. To obtain the first approximation to Tg, -which, we may. call T'g', we may write t^ for t^, and Then we have Pi for^j.
is
;
T'g'^G (35). The temperature t' corresponding to this value of Tg must be found from the table. Then to obtain the second approximation to Tg we may write the value H thus found
for t^ in equation (34), and the corresponding value p of the pressure for p^. Thus, remembering equation (35), we have
for the
second approximation
(35a). T"g"= Tg' + a{t,~ t') - h {p~p') The temperature t" corresponding to this value of Tg must be found as before from the table. If a second ap-
proximation be not sufficient, the process must be repeated. Writing in equation (34) t" andp" in the place of i, andp,,
and remembering equations (35) and (35a), we have T"'g"'=T'g" + a{t'-t")-b{p' -p") (356). The new temperature t'" can again be found from the table. In this way we may proceed to any degree of approximation but the third approximation only differs by about -^ of a degree, and the fourth by less than -j^Vir ^ ^ degree, from the
;
true value of
t^.
The treatment of the third of equations (32) is very similar. If we divide by F la, and for convenience of calculation replace the natural by common logarithms (for which purpose we have to divide by the modulus M) the equation
takes the form
g=G+a'Log^....
where
(36),
C and
a have the
eV-la.(3.6a).
a = T- X k
V-W) J
264
In equation (36) the first term is again tlie mosi imand we can therefore proceed hy successive approximations. First write T^ for 1\; we then have as the first approximation for cr
portant,
g'=C
(37).
The corresponding temperature t' can be found from the table, and from this we can easily obtain the absolute temperature T'. Write T' for T^ in equation (36); then we have
Sr"=5r'
aLogp
obtain the equation
(37a).
Similarly
we can
g"'=g" +
-virhich
a\og^
(376).
again gives T"; and so on. Here again a very few approximations will give a value which very closely agrees with the actual value of 2^. 20.
,It
Determination, of c
and
r.
to determine the quantities c and r, before proceeding to the numerical application of equations
The quantity e, or the specific heat of the liquid, has hitherto been treated as constant. This is not perfectly correct,If since it increases slightly as the temperature increases. however we take as its general value the correct value for the mean of the temperatures which occur in the investigation, the divergencies may be neglected. This mean temperature in the case of engines driven by steam may be taken at 100, which for an ordinary high-pressure condensing engine is about a mean between the tempeirature of the boiler and that of the condenser. Taking therefore the value of the specific heat which Regnault gives for water' at 100, we may put
c
= l-013
(38).
quantity of heat required to heat a unit-weight of water from 0 to the temperature t, and to transform it at that temperature into steam. This equation is
A,
= 606-5 + 0-305^
2&5
we
substitute for
cdt
+ r, we have
f cdfc Jo
Jo
= 606-5 + 0-305f-
In the integral we must use for c the function of temperature exactly detetmined by Eegnault*, if we are to obtain for r the precise values which Regnault gives. For the present purpose it is perhaps sufficient here also to use for c the constant above-mentioned. This gives
rt
odt=l-01St;
Jo la
and the two terms involving t in the last equation can now be combined into one, viz. O'lOSt. We must at the same time make some change in the constant term of the equation. This we will so determine, that the same value of r which is probably the most accurate of all given by observaNow tion, shall also be correct as given by the formula.
at 100 Regnault found for \, as the mean of 38 observations, the value 636 67. If from this we subtract the quantity of heat required to heat the unit-weight of water from 0 to 100, which according to Regnault is lOb'5 heat-units, there remains to one place of decimals only
r^ = 536-2t.
Applying
this value
we
r = 607-0-708
This formula is already laid down in Chapter VII., 3 and a short table is there given, which exhibits the close accordance between the values of r as calculated by this formula, and as given by Regnault in his tables.
21. Special Form of Equation (32) for an Engine working without expansion. In order to distinguish the effects of the two different kinds of expansive action, to which the two latter of equa* Relation dee expiriencei. Vol. I. p. 748. 536-5 instead of the f Begnault himself used in his tables the nnmher ahdve; this simply arises from the fact that he has used the round number 637 for X at 100", instead of 636-67 as given above.
266
tions (32) refer, it seems desirable first to consider an engine will therefore in which only one of the two occurs.
We
begin with an engine that works without expansion. In this case we may substitute for e, or the ratio of the volumes before and after expansion, the value 1 ; and may also put The equations (32) are thus greatly simplified. Tj = T^.
The
last
becomes an
identity,
and
is
therefore useless.
In
the second the right-hand side remains unaltered, whilst the Finally the first left-hand side becomes {Vla)T^^. equation takes the form
Mj
If we here replace {V-la) T^g^ by the expression on the right-hand side of the second equation, all the terms con. .
tainmg
taining
are
other, as
as factor,
collected together as
and the terms which remain can be two products. Then the two equations
W= F
(1
(V-l<r)T,ff,
(40).
+ ^^{f^
^'^
"^Z
+P^-Po}+^(P-P^)_
The first of these equations is exactly the same as is given by Pambour's theory, if in equation (18) we put e = 1, and then by means of equation (12). (having first put e = l
and
v'
=F
by the volume
V.
The only
difierence therefore is in the second equation, which replaces the simple relation between volume and pressure
assumed by Pambour.
267
Numerical Values qf the Constants. e which occurs in these equations, and represents the ratio of the waste space to the whole space left open to the steam, may be taken as 0"05. The quantity of water which the steam carries over with it into the cylinder
22.
The quantity
in different engines. Pambour considers that it averages 0"25 of the whole mass entering the cylinder in the case of locomotives, but much less, 005, in the case of stationary engines. will here take the latter value, which gives 1 : 0"95 as the ratio of the whole mass entering the cylinder to that part of it which is in the form of steam. Further, let the pressure in the boiler, or p^, be five atmospheres, to which corresponds the teiiiperature 152'22; and let it be supposed that the engine has no condenser, or, which is the same thing, a condenser at the pressure of one atmosphere. The mean Isack-pressure in the cylinder is then greater than one atmosphere. With locomotives this excess of backpressure may be considerable, for the special reasons already mentioned ; with stationary engines it is insignificant. Pambour, in his tables for stationary non-condensing , engines, has altogether neglected this excess; and since our present object is to compare the new formula with his, we will follow his example and put^^ = one atmosphere. have then in this case the following values to insert in equations (40)
is different
We
We
= 005
.(41).
'=m=^'''
p,
3800
^0
= 760
k
<r
There now remain only the quantities V and jp, undetermined in equations (40), in addition to the quantity W, of which we are seeking the value. Of these V is the volume
of steam in the cylinder per Unit-weight, and p^ pressure within the cylinder.
is
the final
268
23. The least possible Value of V, and the corresponding amount of Work.
must first enquire what is the least possible value of V. This value corresponds to the case in which the pressure in the cylinder is the same as in the boiler, and we have therefore only to substitute p^ for p^ in the last of equations (40). Thus we obtain
We
269
density a volume greater than is allowed by the mechanical theory of heat; this view finds its expression in equation (116). If by means of this latter volume we determine the work from Pambour's equation upon the two assumptions that 6 = and that e =^'0'05, the values obtained are 16000 and 15200. These quantities, as follows immediately from the greater volume, are both greater than those found above, but not in the same ratio, because the loss of steam due to the waste space is smaller accojrding to our equations than according to Pambour's theory.
24.
Calculation of the
Work for
other Values
ofY.
"With an engine of the kind here considered, the efiSciency of which was investigated by Pambour, the speed which the engine actually assumed bore a proportion in one case of 1*275 1, and in another, with a lighter load, of 1"70 : 1, to the minimum speed as calculated by his theory for the same intensity of evaporation, and for the same To these speeds correspond in our case the boiler-pressure. will now, as an volumes 0"495 and 0"660 respectively. example of the mode of determining the work done, choose a speed which lies between these two, say in round numbers
:
We
F=0-6.
Our first business is now to find the value of T,^ corresponding to this value of V. For this purpose we use equation (34), which takes the following special form
T,g^ = 26-577 + 56'42
If
is
(,
- ^ - 0-0483
(p,
-p,)
...
(43).
by means
t^,
mations to
of this equation we obtain successive approxias described in section (19), the series of values
tf
as follows
=13301,
=134-43,
=134-32,
if'
r
decimals, and if
we
270
OB*
HEAT.
t^t
takB the last number as the true value of sponding pressure is ^, = 2308-30.
The
corre-
If we substitute in the first of equations (40) these values and p^, together with the approximate values of conof stants as fixed in 22, we obtain
pr= 11960.
O'C
W= 12520.
In order to exhibit more clearly the dependence of the work upon the volume, and at the same time the divergence between Pambour's theory and the author's on this point, calculations similar to that for volume 0"6 have been made for a Series of other volumes increasing by equal intervals.
results are given in the following table. The first horizontal row of numbers, which is separated from the rest, The contains the values for an engine without waste space. further arrangement of the table is easily understood.
The
271
going on during the admission, the amount of steam real-v^ays the same as it was at the commencement; while, according to the author's theory, a part of the priming water is afterwards evaporated, and this quantity is greater the greater the expansion.
mains
hy an Engine
We will now treat in the same way an engine working with expansion, and will choose as our example a condensing engine. To fix the degree of expansion, we will assume that the steam is cut off at one third of the stroke. We then have, to determine e, the equation
6-6=^(1-6);
whence, giving to
e
we have
= ?:J = 0-3666.
o
The
condenser-pressure can with good arrangements be kept below -j^jj atmosphere. But since it is not always so small, and in addition the back-pressure in the cylinder somewhat exceeds the pressure in the condenser, we will assume the mean back-pressure p^ to be in round numbers \ atmosphere, To this corresponds the temperature or 152 millimetres. 60-46. If lastly we give to I the same value as before, ta the constants to be used in our example will he as follows
(44).
^^0=152
able to calculate the work done, we only now need the value of V. To assist us in choosing this, we must first know the least possible value of V. This is found exactly as in the case of the engine without expansion, by putting p^^ for p^ in the second of equations (32), and
id
To be know
272
altering the otter magnitudes, whicli' are in combination thus obtain in the present with Pj, in the same way. 1-010. from this, we will first Starting value the case suppose that the actual speed of the engine exceeds the least
We
2.
We
numbers
speed.
.
work done
t^
in round at this
t^,
We
first
and
by
in the two last of equations (32). substituting this value of Of these t^ has already been approximately found for the non-condensing engine; and since the present case only differs from the former inasmuch as the quantity e, which was there taken equal to 1, has here a different value, we need not go through the process again, but will merely give
is
t,
137-43,
(36),
To determine
this case the
tg
we have equation
51-515
which takes in
form
^3
= 26-604 +
Log
(45).
From
tions
:
this
we obtain the
t'
=99-24
t"
t'"
=101-93
=101-74
t""
= 101-76 J
last of these, from which further approximations would only diverge in higher places of decimals, we will consider to be the correct value of <j, and will apply it, together with the known values of t^ and i, to the first of equations (32), Thus we obtain 31080.
The
W=
If,
we
calculate the
work
done by Pambour's equation (18), remembering to determine B and b not from equation (115) as with the non-condensing engine, but from equation (lla) -which refers to condensing
engines,
we
find
TF= 32640.
273
26. Swmmary of Various Cases relfiting to the Work-', ing of the Engine.
The work done for the volumes 1*2, 1'8 and 2'1 has been calculated in the same way as described above for the volume 1'5. Further, in order to make clear by an example the influence, which the different imperfections of an engine have upon the amount of the work done, the following cases
are subjoined.
The case of an engine which has no waste space, (1) and in which the cylinder pressure is equal to the boiler pressure, and the expansion carried so far that the pressure has fallen from its original value p^ to p. If we only assume
further that p^ is exactly equal to the condenser pressure, this case is that to which equation (9) relates, and which gives the greatest possible amount of work for a given quantity of heat, when the temperatures at which the heat is taken in and given off are also given.
(2)
space,
The case of an engine, where there is again no waste and the cyhnder and boiler-pressures are equal, but
:
is not as before complete, but is only in the proportion of e 1. This is the case to which equation (6) refers ; except that there, in order to determine the magnitude of the expansion, the change of temperature in the steam caused thereby was assumed to be known, whereas here the expansion is fixed by the volume, and the change of temperature must be calculated therefrom.
The case of an engine having waste space and in(3) complete expansion, in which the only one of the above favourable conditions which remains is the fact that the steam in the cylinder has during admission the same pressure as in the boiler, so that the volume has its least possible value. To this are appended the cases already mentioned,
in which this last favourable condition is also absent, and the volume has some other given value instead of the least
possible.
For purposes of comparison Pambour's theory has also been applied to all these cases except the first, for which equations (11a) and (llS) do not hold, inasmuch as even the equation which he has determined for low pressures can
C.
18
274
only be applied to ^ or at the lowest ^- of an atmosphere, whereas here the pressure has to descend to ^ of an atmoFor this first case the numerical values given by our sphere.
equations are as follows
Volume
before expansion
after.
= 0-3637,
=6-345,
Work
The
;
done
= 50460.
foUpwing table in which a line is drawn between the values for an engine without waste space and the remaining values. For the volume are given only the values after expansion, because those before expansion, can be at once derived from,
these by diminishing
them
in the ratio 1
e or 1
0-36667.
275
Friction.
said in conclusion on tlie subject of This will be confined to a justification of the course adopted in leaving friction altogether out of account in the equations hitherto developed. For this purpose we will shew that instead of introducing, friction at once into -the first general expressions for the work, as Pambour has done, the same principles will allow us to take it into account subsequently, according to the method already adopted by other
Something must be
authors.
The
stroke
forces
its
distinguished as follows (1) The resistance which it meets with from without, and in overcoming which consists its useful work. Pambour has named this the load (charge) of the engine. (2) The resistances which originate within the machine itself, so that the work expended in overcoming them produces no useful effect. These latter resistances we combine under the name of Friction, although in strictness they comprise other forces than those of friction, especially the resistances of the various pumps driven by the engine, with the exception of the feed-pump, the effect of which has already been considered.
may be
Both these kinds of resistances Pambour takes into account as forces which oppose the motion of the piston ; and in order to be able to combine them readily with the pressures of the steam on both sides of the piston, he made his method of denotation agree with that used in the case of the steam pressures, so that the symbol denotes not the whole force, but the force referred to one unit of surface of the piston. On this understanding let the load be represented byJi. yet further distinction must be made with regard to the friction. This has not the same constant value for any one engine, but increases with the load. Pambour divides it therefore into two parts, that which exists when the engine runs unloaded and that which is brought into action by the load. The latter he takes to be proportional to the load He accordingly expresses the friction per unit of itself. BR, where / and S are quantities which surface by depend on the construction and dimensions of the engine,
f+
18--2
276
may
according to
Pambour be con-
wort done by the engine to these moving force The negative work done by the former must of the steam. equal thi positive work done by the latter, because otherwise
refer the
forces,
We may
now
an acceleration or retardation of the speed woiild take place, and this is against our original assumption that the speed is uniform. Whilst one unit-weight of steam enters the cylinder, the piston passes through the space (1 e) V, and we
thus obtain the following expression for the work
W:
'
F=(l-e)
The
call
V[{l
+ B)B+f].
we may
:
{W),
(W) = {l-e)rxB.
If
we
eliminate
R from
this equation
by means
of the one
above,
we obtain
^^^_w^a^
By help
of this equation, since
V is
assumed to be known,
we can deduce the useful work (W) from the whole work W, as soon as / and S are given us. The manner in which these are determined by Pambour will not be entered upon
here, since that determination rests upon insecure grounds, and friction is not the special subject of this chapter.
29.
Cyclical Process.
have now so far completed our treatment of the steam-engine, that we have followed out all the processes which take place, determined the several positive or negative quantities of work performed therein, and combined these into one algebraic expression. will proceed to
We
We
consider thermo-dynamic engines from a more general point The expression that an engine is driven by heat of view. must not of course be understood of the direct action of the heat, but only that some substance within the engine sets the working parts in motion through the changes which it
277
undergoes by means of heat. This substance we will call the transmitter of the heat's action. If a continuously working engine is at uniform speed, all the changes which occur must be periodic, so that the same condition, in which the engine and aU its parts may be found at any given time, will recur at regular intervals. Therefore the substance transmitting the heat's action must also at such recurring moments exist within the engine in the same quantity and under the same conditions. This may be accomplished in two different ways. (1) The same original quantity of this substance may always remain within the engine; in which case'the changes which it undergoes during the motion must be such, that at the end of each period it
returns to
its
original
condition,
changes then begins anew. (2) The engine may get rid at the end of each period of the substance which during that period has served to produce its action, and may then take in the same quantity of the same substance from without. This latter method is the more usual in practice. It exists, e.g. in the hot-air engine, as hitherto constructed,, since after every stroke the air which has driven the piston in the working cylinder passes out into the atmosphere, and is replaced by an equal quantity of air drawn from the atmosphere by the feeding cylinder. The same is the case with the non-condensing steam-engine, in which the steam passes from the cylinder into the atmosphere, and is replaced by It fresh water, pumped from a reservoir into the boiler. also applies, partially at least, to the ordinary condensing steam-engine. In this the condensed water is indeed in part pumped back into the boiler, but not as a whole, because it mixes with the cold water, a part of which also enters the boiler. The part of the condensed water which is not used again, together with the remainder of the cold water, must go to waste. The first method has hitherto been applied in a few engines only, amongst others in engines driven by two difIn these the steam ferent vapours, e.g. steam and ether*. is condensed by contact with metal pipes filled inside with liquid ether, and is then wholly pumped back into the
*
S78
boiler.
Similarly the ether vapour is condensed in metal pipes surrounded by cold water, and is then pumped back To keep up into the vessel used for evaporating the ether. a uniform working, it is therefore necessary merely to add so much water and ether as is lost by leakage. In an engine of this kind, in which the same mass of vapour is continually used anew, the different changes which this mass undergoes during each period must, as shewn above, form a complete circuit, or in the terms of this treatise a cyclical process. On the contrary the engines in which a periodical admission and emission of vapour takes place are not necessarily subject to this condition. Nevertheless they may fulfil it, provided that they emit the vapour in the same This is the case with a condition in which they received it. condensing steam-engine, in which the steam passes away from the condenser as water and at the same temperature at which it passed from the condenser into the boiler. The waste water, which enters the condenser cold and leaves it hot, need not be taken into account, because it does not appertain to the substance transmitting the heat's action, but merely serves as a negative source of heat. With other engines the conditions at the emission and admission are different. For example, hot-air engines, even if fitted with a regenerator, send the air back to the atmosphere at a higher temperature than that at which it enters and non-condensing engines take in water and emit steam. In these cases no complete cyclical process takes place but it is always possible to conceive that the actual engine has another attached to it, which receives the vapour from the first, restores it in some way to its original condition, and then lets it escape. The two can then be considered as one engine satisfying the above condition. In many cases the process can be completed in this way, without making the investigation too complicated ; e.g. we may suppose that a non-condensing steam-engine, assuming that the feed water is at 100, is replaced by an engine having a condenser, the
;
temperature of which is 100. We may thus apply to all thermo-dynamic engines the principles which hold for a cyclical process, if we only suppose those engines which do not of themselves fulfil the condition, to be completed in the manner described and we can
:
279
thereby arrive at certain conclusions whicK are independent of the special nature of the process in each particular engine.
30. Process.
For any cyclical process the two equations hold which have been previously developed (Chs. II. and IV.) as analytical expressions of the two fundamental principles, provided the second is further extended so as to embrace non-reversible changes. These equations are
n\
(")
denotes the uncompensated transformation occurring during the process, which is always positive, and has for reversible processes the limiting value zero. If we apply these equations to the cyclical process which takes place during a period of the thermo-dynamic engine, we see at once that if the whole quantity of heat taken in by the substance transmitting the heat's action is given, then the work also is immediately given by the first equation, without our needing to know the nature of the process itself. By combining the two equations we may also determine the work from other data in an equally general manner. We will suppose that the "quantities of heat, which, the variable body receives successively, and its temperature at the moment of receiving each, are given, with the exception of one temperature T, at which the body receives a certain quantity of heat (or if it is negative generates it) the magnitude of whicb is not at present known. Let Q^ be the sum of all the known quantities, and let Q^ be the unknown Now let us divide the integral in the quantity of heat. second of equations (47) mto two parts, one of which relates to the known quantity of heat Q^, and the other to the unknown quantity Q. In the latter T has the constant value T it is therefore immediately integrable, and gives the
;
where
expression
W'
280
/,
whence
Again, since
Q= Q^ + Q,
the
first
equation gives
or,
W=^Q,-T,j^'^-T,N
If
is
..(48)..
we make the
then
reversible,
N=
W=Q,-T,j^'^^9
This equation diflfers from the former only in the term T^N. Since iVis always positive, this term must be negative ; whence we see, as is easily proved directly, that under the conditions fixed above with respect to the imparting of the heat, the greatest possible amount of work is obtained when the whole cyclical process is reversible ; and that every circumstance, which causes the changes in a cyclical process not to be reversible, diminishes the amount of work done. Equation (48) therefore leads to the required value of the work done in a way which is the exact opposite to the ordinary for we do not, as in other cases, determine the several quantities of work done during the several changes, and then add them together, but we start from the maximum of work, and then subtract from this the losses of work due to the various imperfections of the process. This method may therefore be called the method of subtraction. If we limit the conditions under which the heat is imparted, by supposing that the whole quantity Qj is imparted at a given temperature T,, then the other part of the integral also
;
281
Equation
:
for
the
maximum work
T -T
31. Application of the above Equations to the Limiting Case in which the Cyclical Process in a SteamrMngine is
Among the cases considered above with regard to the action of the steam-engine, there is one which cannot indeed be attained in practice, but which it is desirable to approach as nearly as possible, viz. the case in which there is no waste space, in which the cylinder-pressure is the same as in the boiler or condenser respectively, and in which the expansion extends so far, that the steam is thereby cooled from the temperature of the boiler to that of the condenser. In this case the cyclical process is reversible in all its parts. Let us suppose that evaporation takes place in the condenser at the temperature T^ ; that a mass M, of which the part m^ is in the in the liquid form, passes into gaseous, and the part the cylinder and drives the piston upwards ; that by the fall of the piston the steam is first compressed until its temperature rises to T^ and then at that temperature forced into the is again brought out of boiler ; and that lastly the mass the boiler in the liquid form into the condenser, by means of the small pump, and there cools to the initial temperature T. Here the steam passes through the same series of conditions The various quantities as before, but in the reverse order. of heat are imparted or withdrawn in the opposite sense, but to the same amount and at the same temperature of the mass and all the quantities of work have the opposite signs and the same numerical values. Hence it follows that in this case there is no uncompensated transformation in the may therefore put iV=0 in equation (48), process. thereby obtaining an equation similar to (49), but in which, instead of W, for the sake of distinction, we will write
Mm^
We
viz.
^'=.-<'f-
282
Here Q, is in our case the heat imparted within the boiler to is heated as liquid from T^ to Z\, the mass M, whereby and then the part m^ transformed into steam. Hence
Q,
= m,p, + MC{T,-T,)
|
(51).
To determine
To
-^ we must
dQ
(T^ T^ and ,Pj. Q^, viz. integrate for the first of these, let us write the element in the form MCdT: then this part of the integral
MG
becomes
MO
^ = MClog^.
Whilst the latter quantity of heat is being imparted, the temperature remains constant at Tj, and the integral in this case is simply
W becomes
the exact expression given in equation (9), and found and 5 by the successive determination of the several quantities of work done during the cyclical process.
This
is
in sections 4
32,
Anothei"
Expression.
It was observed in the last section with reference to equation (51) that the two quantities of heat which make up (T, r), must be treated differently Q^, viz. m^p, and in the calculation of the work done ; inasmuch as the one is imparted to the substance which transmits the heat's action at a fixed temperature T^, and the other at temperatures rising continuously from T^ to Tj. Similarly these two
MG
283
quantities of heat occur also in different forms in the expression for the work, as is more clearly shewn if we write our last equation as follows
W = m,p,^^'' + MG{T,-T,)
Here
m^pj^ is multiplied
(l
^^^^bg Jo)...(52).
by the
factor
plied
(50),
while
MO (T^ T^
is
multi-
T T + TrfvTlog-^.
we we
will
In order to compare these factors better with each other, throw the latter into another form. If for brevity
write
=-V-"
T -T
(^3)'
= l-^(l + 2+-3 +
Z
,
zfz
1^
z^
z^
\ ^*-)
,
7+ etc. + + 1x22x33x4
75
iT
j^
z^
Equation (52) or
(9)
thus becomes
F' = m^,x.-Filfa(T,-r)x.(i+2^H-3?^+etc.)
(54).
The value of the infinite series in the bracket, which makes the difference between the factor oiMC (T^ Tq) and the factor of m^ p,, varies, as is easily shewn, between ^ and 1,
whilst z increases from
to
1.
284
S3.
Heat.
Since, under the assumptions we have made, the cyclical process passed through periodically by the engine during its motion is reversible in all its parts, as shewn in 31, and since a reversible process gives the maximum work attainable, we are able to lay down the following principle If the temperatures at which the substance transmitting the heat's action takes in the heat from the source, or gives it out externally, are taken as given beforehand, then the steam-engine, under the assumptions made in the development of equations (9) or (52), is a perfect engine, inasmuch as with a given quantity of heat imparted to it, it performs the greatest quantity of work, which according to the. Mechanical Theory of Heat it is possible to perform at those temperatures. It is otherwise, if we consider these temperatures not as given beforehand, but as a variable element which must be taken into account in our estimate of the engine. The fact that the water, whilst being heated and evaporated, has a much lower temperature than the fire, and there:
imparted to it passes from a higher temperature an uncompensated transformation which is not included in JV of equation (47), but which greatly diminishes the useful effect of the heat We see by equation (54) that the work obtained by the steam-engine from the quantity of heat m,p, -1(T, T^) = Q,, is somewhat
to a lower, produces
less
than Q^
T
i
MG
-"
But
if
we assume
transmitting the heat's action could have, throughout the time when it is taking in heat, the same temperature as those parts of the fire which furnish that heat; and if we denote the mean value of this temperature by T, and the temperature at which the heat is given off, as before, by T^; then under these circumstances the work which it is possible to obtain from the quantity of heat Q, will by equation (50) be expressed
^y Qi
50,
T T
let
Tpr-^
To compare the
and
f, the condenser temperature, be fixed at us take for the boiler temperatures 110, 150, and
285
180; the two first of wliich about correspond to those of the low-pressure and ordinary high-pressure engines, while the last may be considered the limit of temperature, which has as yet been practically attempted for the steam-engine. In
these cases
case the
T T ^
When
'
number 273
T,
....150
0-236,
180"
0-287.
On
if
for
example we assume
t'
at 1000,
T'
is
0-746.
Hence it is easy to see, as already observed by Carnot and subsequent authors, that in order to get the greatest advantage from engines driven by heat, the most important
point
is
to increase
the
temperature interval
JJ,
For example the hot-air engine can only be expected to shew a decided advantage over the steam-engine, when the possibility is established of its being worked at decidedly higher temperatures, such as are forbidden to the steam-engine for fear of explosion. The same advantage may however be obtained by the use of superheated steam, since, as soon as the steam is separated from the water, it
may be
heated with no more risk than a permanent gas. Engines which use the steam in the superheated condition ought to unite many advantages of the steam-engine with those of the air-engine, and may therefore be expected to achieve a practical success much sooner than the latter. In the class of engines mentioned above, in which besides the water another more volatile substance is used, the interIt has val T^ Tq is widened, inasmuch as T^ is lowered. the interval might be widened that suggested already been at the higher limit in the same manner, by adding yet a third liquid less volatile than water. The fire would then
directly evaporate the first of the three only ; this in condensing would eva,porate the second, and this in like manner the third. As far as principle goes this combination would
286
undoubtedly be advantageous the practical difficulties, whicli might hinder its being carried out, cannot be conjectured
beforehand.
34;
traction.
Example of the
application of the
Method of Sub-
The ordinary steam-engine, besides the imperfection described above, which is inherent in its nature, possesses many other imperfections, which are rather due to its construction in practice some of these have already been taken into account in our investigations to determine the work done. will now in conclusion exhibit the method in which, with engines possessing such imperfections, the work done may be determined by the method of subtraction given in 30. In order that we may not extend this investigation, which is merely an example of the mode of applying this method, to too great a length, we will only take into account two of these imperfections, viz. the existence of the waste space and the difference of the pressures in the cylinder and boiler, during the admission of steam. On the other hand we will suppose the expansion to be complete, so that the final temperature Tj equals the condenser temperature T^; and we will also take T' and y" as equal to ?". The process is founded on equation (48), which, calling the work W, is as follows
:
We
Here Q^ T^\
-^
represents the
maximum
work, cor-
responding to the case when the cyclical process is reversible; represents the loss of work due to the imperfections and which arise from the process not being reversible. For this maximum work we have in the case of the steam-engine the expression already given at the end of 31, viz.
T^
It only remains to determine JV, the uncompensated transformation which occurs in the cyclical process. This transformation arises when the steam is passing into the waste space and into the cylinder, and the data for its deter"
287
mination are already given in 10. There, by assuming that the mass of steam admitted was at once forced back again into the boiler, and also that everything was brought back by a reversible path to its initial condition, we arrived at a special cyclical process, for which we determined all the quantities of heat imparted to the variable mass, and to which we can now apply the equation
^=-jf
These quantities of heat, some positive some negative, are as
follows
7n,p
-mj>
first
f,,p,
MG{T,are
T,)
and -/*C7(2;-
T,).
The
three
of
these
temperatures
T^, T^
and T^ respectively
^\ -"^ and ^.
T - fiC log -^^
The
two
last are imparted at temperatures which vary contmuously between T^ and T^, and between T^ and T^ respectively; and the corresponding parts of the integral are
T MC log-'
If
and
we substitute the sum of these quantities for the integral, the last equation becomesjVr
Multiplying this expression for JV by Tj, and subtracting the product from the expression given above for the maximum work, we obtain finally for W' the equation
W = mj>,-^ m^, + MC
To compare
first
(T,
- T,) -
(M+fj) GT log
|+
^L,p
(56).
with that given in the this expression for of equations (28), we have only to substitute in the latter the value of m^pg given in the last of those equations,
The expression thus obtained agrees r. exactly with that in equation (56).
288
In the same way we may also deduct the loss of work due to incomplete expansion. For this purpose we must calculate the uncompensated transformation which arises during the passagejof steam from the cylinder to the condenser, and then include this in the expression for W. By this calculation, which we shaiU not here perform at length, we arrive at the same expression for the work as is given in equation
(28).
important to know how far the convenient method of determination employed by the author is allowable ; and for this purpose a few values are here placed side by side for comparison. The, formula employed by Eegnault to calculate the pressiures of steam in his tables, for temperatures above 100, is the following
upon steam,
it is
where Log denotes the denotes the temperature from -20" to the value under consideration, so that a;=t + 20, and the five constants have the following values
From
this formula
we obtain the
where
.a
and
/3
and
B have the
following values
^
dt
for the
\dtjw
= 90-115.
.289
we obtain .from
^i6=3214'74,
whence
This approximate
with the exact value deduced above, that it may be safely used for it in all investigations upon the steam-engine. With regard to temperatures between 0" and ,100, the formula used by Begnault to calculate the pressures of steam between these' limits is the following
log^=a + 6a'-c|3'.
Here the constants, by Morjtz's improved tables, have the following
values
a=4-7393707,
From
this formula
we can
where the values of a and ^ are the same as given above, and those of
and
B are as follows
Log.4 = 6-6930586-10,
we
calculate
corresponding to a
temperature of
we
obtain
=10-1112.
By the
This result again agrees sufficiently closely with that obtained from the more exact equation.
c.
19
Table
its
differential coefficient
-^ =
g,
and
different temperatures:
of
mercury.
t in degrees
Oiinti-
grade.
291
in
292
293
294'
lU
centi-
grade.
159
CHAPTER
XII.
ox THE CONCENTRATION OF KAYS OF LIGHT AND HEAT AND ON THE LIMITS OF ITS ACTION.
1.
The principle assumed by the author as the ground of the second main principle, viz. that heat cannot of itself, or without compensation, pass from a colder to a hotter body, corresponds to everyday experience in certain very simple To this class belongs the cases of the exchange of heat. conduction of heat, which always takes place in such a way that heat passes from hotter bodies or parts of bodies to" colder bodies or parts of bodies. Again as regards the ordinary radiation of heat, it is of course well known that not only do hot bodies radiate to cold, but also cold bodies conversely to hot nevertheless the general result of this simultaneous double exchange of heat always consists, as is established by experience, in an increase of the heat in the colder body at the expense of the hotter. Special circumstances may however occur during radiation, which cause the rays, instead of continuing in the same straight line, to change their direction and this change of direction may be such, that all the rays from a complete pencil of finite section meet together in one point, and there combine This can he accomplished, as is well known, by their action. and several the use of a burning-mirror or burning-glass mirrors or glasses may even be so an-anged, that several pencils of rays from different sources of heat meet together iii
; ; ;
one point.
296
of this kind there is no experimental proof that impossible for a higher temperature to exist at the point of concentration than is possessed by the bodies from which the rays emanate. Rankine* accordingly, on a special occasion, of which we shall speak in another place, has drawn a particular conclusion, which rests entirely on the assumption that rays of heat can be concentrated by reflection in such a way, that at the focus thus produced a body may be raised to a higher temperature than is possessed by the bodies which emit the rays. If this assumption be correct, the principle enunciated above must be false, and the proof, deduced by means of that principle, of the second fundamental principle of the Mechanical Theory of Heat would thus be overthrown. As the author was anxious to secure the principle against any doubt of this kind, and as the concentration of rays of heat, with which is immediately connected that of rays of light, is a subject which, apart from this special question, is of much interest from many points of view, he has attempted a closer mathematical investigation of the laws which govern the concentration of rays, and of the influence which this concentration can have on the exchange of heat between bodies. The results are contained in the following sections.
it is
I.
Reasons why the ordinary method of determining THE mutual radiation OF TWO SURFACES does NOT EXTEND TO THE PRESENT CASE.
2.
and
to
Limitation of the treatment to perfectly black bodies, homaffeneous and unpolarized rays of heat.
When two bodies are placed in a medium permeable to heat rays, they communicate heat to each other by radiation. Of the rays which fall on one of these bodies, part is in general absorbed, part reflected, part transmitted;
and
it is
known
that the power of absorption stands in a simple relation to the power of emission. As it is not here our object to investigate the differences between these relations and the laws to which they conform, we wiU take one simple case, viz. that in which the bodies are such that they completely absorb all the rays which fall upon them, either actually on the
P/iiJ.
On the Ee-concentration of the Meehanioal Energy of the Universe, Mag., Series iv., Vol. it. p. 358.
297
neg-
its
thickness
may be
Such bodies have been named by Kirchhoff, in his well-known and excellent paper on the relation between emission and absorption, "perfectly black bodies*." Bodies of this kind have also the maximum power of emission. It was formerly assumed to be beyond question
that their intensity of emission depended only on their temperature; so that all perfectly black bodies, at the same temperature and with the same extent of surface, would radiate exactly the same quantity of heat. But as the rays emitted by the body are not homogeneous, but differ according to the scale of colours, the question of emission must be studied with special reference to this scale and Kirchhoff has extended the foregoing principle, by laying down that perfectly black bodies at equal temperatures send out not only the same total quantity of heat, but also the same quantity of each class of ray. As the distinctions between the rays according to colour have no place in our investigation, we will assume throughout that we have to do with only one
;
class of ray, or, to speak more accurately, with rays whose wave-length only vaaries within indefinitely small limits. Whatever is true of this class of rays must similarly be true of any other class; and thus the results obtained from homogeneous heat may easily be extended to heat which contains a mixture of different classes of rays. With the same object of avoiding unnecessary complications, we will abstain from discussing polarization, and assume that we have only to do with unpolarized rays. The mode
known
Let Sj and s, be the surfaces of two perfectly black bodies of the same given temperature ; and let us consider two elements of these surfaces ds^ and ds^, in order to determine and to compare the quantities of heat which they mutually send to each other by radiation. If the medium, which surrounds the bodies and fills the intervening space,
Pogg. Ann., Vol.
oix. p. 275.
298
is
homogeneous, so that the rays simply passin straight lines from one surface to the other, it is easy to see that the quantity of heat which ds^ sends to ds^ must be the same as that which ds^ sends to ds^ if on the contrary this medium is not homogeneous, but there are variations in it which cause the rays to be broken up and reflected, the process is less simple, and a closer investigation is needed to prove whether the same perfect reciprocity holds in this case also. This investigation has been performed in a very elegant manner by Kirchhoff; and his results will be briefly stated here, so far as they relate to the ease in which the rays on their way from one element to the other suffer no diminution of in other words in which the breakings up and restrength flections of the ray take place without loss of power, and there is no absorption during its passage. A few variations will alone be made in the denotation and in the system of coordinates, to produce a better accordance with what follows. If two points are given, only one of the infinitely large number of rays sent out by one point can in general attain the other*; or if the rays are so broken up and reflected, that several of them m^et in the other point, yet they form in general only a limited number of separate rays, each of which can be treated by itself. The path of such a ray from one point to the other is determined by the condition that the time expended in traversing this path is a minimum, compared with the times which would be expended in traversing This all other neighbouring paths between the same points. minimum time is determined, if where, several separate rays meet we investigate only one at a time, by the position of the two points between which it passes and we will denote it, as Kirchhoff has done, by T.
;
* The form of expression tliat a point sends out an infinitely large number of rays is perhaps, in the strict mathematical sense, inaccurate, since heat and light can only be sent forth from a surface, and not from a mathematical It would be more accurate to refer the sending out of the heat or light point. not to the point itself, hut to the element of area at the point. As, however, the conception of a ray is itself only a mathematical abstraction, we may, without fear of misconception, retain the statement tiat an infinitely large number of rays proceed from each point of the surface. it were our objeet to determine quantitatively the heat or light radiated by a surface, it is evident that the size of the surface must be taken into account, and that its elements must he considered, not as points, but as indefinitely small surfaces the area of which must appear as a factor in the formula expressing the quantity of heat or light radiated from ah element of surface.
299-
Returning now to the elements of surface ds^ and dSj, we will suppose a plane tangential to the surface to be drawn'
through one point of each element; and we will treat ds^ and ds^ as elements of these planes. In each of these planes let us take any system of rectangular co-ordinates, which we will call aTj, 2/1 in the one case, and os^, y^ in the other*. If we now take a point on each plane, the time T^, expended by the ray in passing from one point to the other, is determined, as stated above, by the position of the two points; and this time may therefore be treated as a function of the four coordinates of the two points. On these assumptions Kirchhoffs expression for the quantity of heat, which the element ds-^ sends to the element
rfsjj
is
as follows
i";
d'T
dy^dy^
d'T
dx^dy^
d'T dy^dxj
'
"'
where ir is the well-known ra.tio between the circumference, and diameter of a circle, and e, is the intensity of emission
of the surface Sj, in the locality of the element ds^, so that e^ dsj^ represents the whole quantity of heat radiated by ds^ per unit, of time.
Conversely to express the quantity of heat which ds^ sends to ds^, we need only substitute for e^ in the above, expression the quantity e,, which is the intensity of emission of the surface s,. Everything else remains unaltered, as being symmetrical with regard to the two elements for the time T, which a ray expends in traversing the path between two points of the two elements, is the same in whichever If we now assume that the surfaces, direction it is moving. which are supposed to be at the same temperature, radiate eqi^al quantities of heat in the same time, then 61 = e^ and therefore the quantity of heat sent by ds^ to ds^ is exactly the same
; ;
as that sent
by
ds^ to ds^.
* Kirohhoff has chosen two planes at right angles to the directions of the 00-' rays in the neighbourhood of the two elements; he has taken axes of ordinates in these planes, and has projected the elements of surface upon
them.
oix. p. 286.
300
of
the
We have already observed that in general only one ray, or a limited number of separate rays, can pass between two given points. In special cases, however, it may happen, that an indefinitely large number of rays, forming a pencil in either two or three dimensions (i.e. in the latter case forming a cone), and starting from the first point, may again unite in the second. This of course holds with rays of light as well as of heat, and it is usual in Optics to call such a point, in which all the rays of a certain conical pencil sent out from a given point unite agairi) the image of the given point or, since conversely the first point may become the image of the second, the two points are called conjugate foci. When what
;
here described in the case of two particular points holds of all the points of two surfaces, so that every point of the one surface is the conjugate focus of some poinff on the other surface, then the one surface is called the optical image of the other. may now ask how the exchange of rays takes place
is
We
between the elements of two such surfaces; whether the above-mentioned reciprocity holds, so that at equal temperatures any element of the one surface sends to any element of the other exactly the same amount of heat as it received from it, and therefore one body cannot heat the other to a higher temperature than its own or whether in such cases the concentration of the rays makes it possible for one body to heat another to a higher temperature than it possesses
;
itself.
To
For
be an optical image of the surface s-^. the rays, which a point p^, lying upon s^, sends out within a certain cone, unite on some point j)^ of the surface s^, and all the surrounding points of s^ receive no rays whatever from^i. Hence if the co-ordinates x^, y^ of the point ^^ are given, then the co-ordinates x^, 3/, of the point p^ are no longer to be taken at pleasure, but are also fixed and
let the surface s^
all
Then
then
x-^,
where
x^ is con-
301
sidered as variable when difFerentiating according to !, whilst the second co-ordinate y^ of the same point, and the co-ordinates x^, y^ of the other, are assumed to be constant, and wh^re similarly in differentiating according to x^, this is taken as the only variable can thus represent no real quantity of finite value. Therefore in this case we must find an expression of somewhat dififerent form from Kirchhoff 's ; and for this purpose we must first consider some questions similar to those considered by Kirchhoff in arriving at his expres-
sion.
II.
Determination of corresponding points and coRr RESPONDING ELEMENTS OF SURFACE IN THREE PLANES CUT BY THE RAYS.
5. Equations between the co-ordinates of the points in which a ray cuts three given planes.
Let there be three given planes each of them let there be a system
of rectangular co-ordinates, which we may call respectively oo^ya, x^y^, and x^^. Let us take a point p^ in plane a, and a point p^ in plane b,
a, b, c (Fig. 25),
and in
302
three points of section, which are so related to each other that any one of them is in general determined by the other The equations which serve for this determination may two. be easily deduced from the above-mentioned condition. Let us first suppose that, points p^ andp are given beforehand, and that the point is still unknown in which the ray cuts the intermediate plane h. This point, to distinguish it from other points in the plane, we will call p\. Take any point whatever p^ in this plane, and consider two rays, which we may call auxiliary rays, passing the one from p^ to p^, and the other from p^ to p. In Eig. 25 these rays are shewn by dotted lines, and the primary ray, which goes direct from p^ If, as before, we call the times exto p^, by a full line*. pended by these two rays T^^ and T^^, the value of the sum of these times Tj + T^^ will depend on the position of jjj, and therefore, since the points p^ and p^ are assumed to be given, it may be considered as a function of the co-ordinates x^y^ Among all the values, which this sum may of point Pj. assume if we give to point p^ various positions in the neighbourhood of p 5, the minimum value must be that which is obtained by making p^ coincide with p\, in which case the auxiliary rays merely become parts of the direct ray. therefore have the following two equations to determine the co-ordinates of this point p\ :
We
i(2kt^ = 0-
^i^kt^U^O
(1)
As T^ and T^^, in addition to the co-ordinates of the unknown point Pj, contain also the co-ordinates of the known points p^ and p^, we may consider the two equations thus
established as being simply two equations between the six co-ordinates of the three points iu question. These equations, therefore, can be applied, not merely to determine the coordinates of the point in the intermediate plane from those
* These lines are shewn canred in the figure, to indicate that the path taken by a ray between two given points need not be simply the straight line drawn between the two points, but a different Une determined by the refractions or reflections which the ray may undergo: it may thus be either a broken bne, made up of several straight lines, or (if the medium through which it passes changes its character continuously instead of suddenly) a continuous curve.
303
of the two other points, but generally to determine any two of the six co-ordinates from the other four. Next let us assume that the points p^ a,ndp^ (Fig, 26),. in which the ray cuts planes a and h, are given, and that the point is yet unknown in which it cuts plane c. This point let us call p""^, to distinguish it from other points in the same plane. Take any point p^ in plane c, and consider two auxiliairy rays, one of which goes from p^ to p^, and the other from Pi to p^. In Fig. 26 these are again shewn dotted, while the primary ray is shewn full. Let T^^ and T^ be the times of passage of these auxiliary rays. Then the value of the differ-
ence T^^ Tj^ depends on the position of p^ in plane c. Among the various values obtained by giving to p various positions in the neighbourhood of p\, the maximum must be that obtained by making p^ coincide with p\. For in that case the ray passing from j?^ to p^ cuts the plane b in the given point jp^, and is therefore made up of the ray which passes from p^ to p^ and of that which passes from p^ to p^. Accordingly we may put
by
T T =T
'as
-'Sc
-^aS*
If on the other hand p^ does not coincide with p'^, then the ray which passes from p^ to p^ does not coincide with the two which pass from p^ to p^ and from p^ to p^ and since the direct ray between p and p^ travels in the shortest
;
time,
we must have
304
This
thus generally smaller than in is where p^ lies in the continuation of the which passes from 'p^ to 'p^, and this special value of Hence we have difference thus forms a maximum*. two following conditions
special case
d^.
'
dy,
^>-
If we lastly assume that the points p^ and p^ in the planes h and c are given beforehand, while the point in which the ray cuts the plane a is still unknown, we obtain by an exactly similar procedure the two following conditions
d \Tm ~ TJ) _ ^
dx^
a[I^~dy.
l^i)
f,^
'
In this way we arrive at three pairs of equations, each of which serves to express the corresponding relation bet-^een the three points in which a ray cuts the three planes a,h,c; so that if two of the points are given the third can be found, or, more generally, if of the six co-ordinates of the three points four are given the other two may be determined.
face.
6.
We
will
following case.
Given on one
of the planes, say a, a point p^, and on another, say b, an element of surface ds'^; then if rays pass from pa to the different points of ds^, and if we suppose these rays produced
till
they cut the third plane c, they will all cut that plane in general within another indefinitely small element of surface, which we will call dsg (Fig. 27). Let us now deter-
mine the
sidered,
relation
between
ds^
and
dso.
* In KirchhofE's paper (p. 285) it ia stated of the quantity there conwhich is essentially the same as the difference here treated of, except that it refers to four planes instead of three, that it must be a minimum. This may possibly be a printer's error, and in any case an interchange of maximum and minimum in this place would have no further influence, because the principle used in the calculations which follow, viz. , that the dierential coefficient =:0i holds equally for a maximum or a minimum.
305
case, of the six co-ordinates which relate to each those of the three points in which the ray cuts the three planes) two, viz. x^ and y^, are given beforehand. If we now take any values we, please for 00^ and y^, the co-ordinates x^ and y^ are in general thereby determined. Thus in this case we may consider x and y^ as two functions of x^ and y^. As the form of the element ds,, may be any whatever, let it be a rectFig. 27, angle da?5, dy, and let us find the point in plane c correspondisng shall to every point in the outline of this rectangle. then have on plane c an indefinitely small parallelogram which forms the corresponding element of surface.
ray
We
of this parallelogram is determined as Let \ be the length of the side which corresponds to the side dx of the rectangle in plane h, and let (Xa;) and (\y) be the angles which this side makes with the axes
follows.
The magnitude
of co-ordinates.
Then
and we have
(>*%)
li
(jiy^
be the other side of the parallelogram, and let be the angles it makes with the axes. Then
cos
{^uc,)
= -j
dy, ;
fi
cos
{fiyc)
'^3'^
X and
/*.
Then we
(V)
= cos {\x,)
'dx,
cos
i/ix,)
dx
dy,
\dt dyt
C.
306
Now
may
of the parallelogram,
we
write
ds^
and
fi"
:
,_{fdfoX
Then
several terms under the root caricel each other, and the remainder form a square as follows
:
\dx^ dy^
dy^
dxj
"
This quadratic equation can therefore be solved at once. But it must be observed that the' difference within the brackets may be either positive or negative, and, as we have only to do with the positive root, we will denote this by putting the letters v.n. (valor numericus) before this can then write difference.
We
To
ascertain
how
a;
and
depend upon
ir^
and
y^,
we
of the three pairs of equations in 5. If we differentiate those will first choose equations (1). equations according to ss^ and ^j, remembering that'eacli of
We
T contains two of the three pairs of co-ordinates x^, y^, x^, y^', x^, y, as denoted. by the indices; and if in differentiating we treat x, and .y.-as functions of
the quantities denoted by
307
and y -whilst we take x^ and y^ as constant; then obtain the following four equations
we
{dx,f
dx^dx^
_L
tfoj
dx^dy,
(fr..
dXi
dy.
+
dx^dy^
dx
dy^
dx^dy,
dyi
.(5).
dy^dx
dy^dx^
dx.
dy^dy,
dy^dy^
dxt
d'(T^.+T,:)
(.dy>T
dy^
x^ =
dy^
If
by help
of these equations
,
we determine the
,
four
differential
coefficients -j^
dx^
(4),
-p
^
dy,,
and
subsititute
relation
between
more
briefly,
we
we obtain the required able to write the result will use the following symbols
ds^
and
ds.
To be
A = v.n.(^
d^Z,
dy^dy,
d'Z,
dx^dy.
dx^dx^
dy.dxj
W.
E= v.n.
.
idx,r
""
{dy,Y
dx^dy^
in
relation
may be
written as follows
E
dsg
'A'
(8).
Again, if we suppose in like manner that a point p^ is given on plane c (Fig. 28), and find on plane a the element da^, which corresponds to the given element ds^ on plane b, then the result can be derived frofla that last given by simply interchanging the indices a and c. If for brevity we write
a = vn
fi^x^^
Kdx^dXi
dyjiy
dxj,y^
dyM\
-)...(9).
202
308
then we have
ds,
.(10).
ib
Pig. 28.
Fig. 29
Let ds^ on plane a. us suppose that rays from different points of this element pass through p^, and that they are produced to the plane c. Then the magnitude of the element of surface dse, in which all these rays meet plane c, is found, using the same symbols as before, to be as follows
A'
From
this
.(11).
we
see that the. two corresponding elements in same relation to each other, as the
309
two elements which are obtained when we have an element rfSj given in plane b, and then, having assumed as the origin of the rays a point first in plane a, and secondly
in plane c, determine in each case the element of surface in the third plane corresponding to the element da^,
V. Fractions formed out of six qucmtities to express the Relations between Corresponding Elements.
In the last section we have only employed the first of the can however employ the three pairs of equations in 5. two other pairs (2) and (3) in the same manner. Each pair leads us to three quantities of the same kind as those already denoted by A, G, and E. These quantities serve to express the relations between the elements of surface. Of the nine quantities thus obtained, hoTvever, there are four which are equal to each other, whereby the actual number is reduced to six. The expressions for these six are here placed together for the sake of convenience, although three of them have been already given.
We
'{dx^dx^
dyjiy^
dx^dij,
dydxj
'\dx^dx^
dyjiy^
dxjiy^
dyJ^J
\dxjx^
dyj,y^
^^
n ^ ^~
{^{T^-TJ)
l~~W^
[dx:f
{dx:f
(dyj^
L
L
^ady.
dx,dy,
J I
^-^^i
""
.
{dy,r
J J
^ = ^-i
(dyj
dx^y,
j]
(I-).
By help of these six quantities every relation between two elements of surface can be expressed by three different fractions, .as may be shewn in "tabular form as foUows
310
311
we
a and
c (Fig.
ds^. For this purpose let us suppose the intermediate plane h to lie parallel to plane a at a
distance p, which
is
so small, that
the part which lies between these two planes of any ray passing from ds^ to ds^ may be considered as a straight line, and the medium through which it passes as homogeneous. Let us now take any point in element ds^, and consider the pencil of rays which passes from this point to the Fig. 30. element ds^. This pencil will cut plane h in an element ds^ whose magnitude is given by one of the three fractions in the uppermost horizontal row of equations (II.). Choosing the last of these we have the equation
,
^ds.
(14).
The quantity may in this case be brought into a specially simple form, on account of the special position of plane h. For this purpose let us follow Kirchhoff in choosing the system of co-ordinates in h so as to correspond exactly with that in the parallel plane a; i.e. let the origins of both lie in a common perpendicular to the two planes
let each axis of one system be parallel to the corresponding axis of the other. Let r be the distance between two points lying on the two planes, and having co-ordinates x^,
and
y^ and
x^, y^ respectively.
Then
.(15).
r^Jp'+ix.-xS + iy-yJ'
Let us now suppose a single ray to pass from one of these points to the other ; then, since its motion between the two planes is supposed to be rectilinear, the length of its path will be simply represented by r ; and if we denote by v^ its velocity in the neighbourhood of plane a, which by the assumption, will remain nearly constant between a and b,
312
then the time which the ray expends in the passage will be given by the equation
"'
v.-
The
expression for
1 /
C may
d'r
therefore be written
cPr
^^
^V
dx^dy^
^^
CPr
~^'^' v^Kdx^dx^
Substituting for r
its
dy^dy^
dy^dxj'
we
obtain
G=
Hence equation
(14)
l-^^'?
W
.(17).
becomes
ds,
= v:x-,Bds^
r
If
we denote by
the normal at that point, then cos ^ = tion also takes the form
""'
'
and the
last
equa-
"'""''^.Bds,
(18).
cos=^^
9.
When the magnitude of the element of surface ds^ is determined, the quantity of heat which ds^ sends to ds, can be easily expressed. ^ From every point of ds^ an indefinitely small pencil of rays goes to .d&/, and the solid angle of the cone made by the pencil from each of these points may be taken as the same. The magnitude of this angle is determined by the magnitude and position of that element ds^ in which the cone cuts plane b. To express this angle geometrically, let us suppose that a sphere of radius p is drawn round the point from which the rays start, and that within this sphere we may consider the path of the rays as being rectilinear. If da is
313
cut
by the
But
since the element ds^ is at the distance r from the vertex of the cone, and since the normal to the surface at ds^, which is parallel to the already mentioned normal to the surface at
ds^, forms
angle
6,
If
_ cos
X
i
ds^
\1"J'
we
substitute for
?,
its
we
obtain
(20).
d^^^Bds^
We
have now to determine the magnitude of that part of the heat sent from ds^ which corresponds to this indefinitely small cone; or in other words, how much heat ds^ sends through the given element da- upon the spherical area. This quantity of heat is proportional (1) to the magnitude of the radiating element ds^, (2) to the angle of the cone, or to
-^, and (3), according to the well-known law of radiation, to r the cosine of the angle 6, which the indefinitely small cone makes with the normal. It may therefore be expressed by
^ -^ ds^, P where e is a factor depending on the temperature of the To determine this factor we have the condition surface. that the whole quantity of heat which tfe radiates, or which it sends to the whole surface of any hemisphere above plane a, must equal e^ds^ where e is the intensity of emission from plane a at the position of ds^. Hence we have
e cos
cos 0da?!'
^
!
= e^
The
and gives
314
Substituting this value of e in the above expression, obtain for the quantity of heat which ds^ sends through the formula
we da
cos a 2
TT
ft
ds^.
We
for
J given
and we
sion for the quantity of heat which ds^ sends to ds^, viz.:
If conversely we require the quantity of heat which ds^ sends to ds^, and if we denote by e the intensity of emission from plane c at the position of ds^, and by t'^the velocity of the rays in the neighbourhood of ds^, we obtain the expression
in
10.
The expressions obtained in the last section are in general the same as Kirchhotf's expression given in 3, and differ only inasmuch as they contain as factor the square of the velocity, which does not occur in Kirchhoff's expression, because' he considers nothing but the velocity in vacuo, and takes this as unity. Since however the bodies, whose mutual radiation we are considering, may often be in different media, where the velocity of the rays is different, this factor is not without importance; and its introduction leads also to a
'
special conclusion of
some
As mentioned
in 2, it has
assumed that
with perfectly black bodies the intensity of emission depends only on the temperature, so that two such bodies of equal temperature would radiate equal quantities of heat from equal areas of surface. As. far as the author knows, the question whether the surrounding medium has also an influence on the intensity of emission has never been considered. Since however the two expressions given above for the mutual radiation of two elements contain a factor which
315
depends on the nature of the medium, it becomes necessary to consider this medium, and the method of determining its influence. If from the above two expressions we omit the factor
which
is
common
to both, viz.
ds^ds^, we
that the quantity of heat, which the element ds^ sends to the element ds^, bears to that which ds^ sends to ds the ratio If w^e now assume that at equal temperatures ^a^a ^a"the radiation is always equal, even when the media in contact with the two elements are- different, then for equal temperatures we must put e^=e^', and the quantities of heat, which the two elements radiate to each other, would then not be equal, but would be in the ratio u/ : v^. It would follow that two bodies which are in different media, e.g. one in water and the other in air, would not tend to equalize their temperatures by mutual radiation, but that one would be able by radiation to raise the other to a higher temperature than that which itself possesses. If on the other hand we maintain the universal correctness of the fundamental principle- laid down by the author, viz. that heat cannot of itself pass from a colder to a hotter body, then we must consider the mutual radiations of two perfectly black elements of equal temperature as being themselves equal, and must therefore put
'
eav:
= e,v:^
(21).
Hence
e,,
:c, ::^;/^^^/
(22).
Since the ratio of the velocities is the reciprocal of that of the coefficients of refraction, which we m&y call n, and m, this proportion may be written
e.:c
::/:<
(23).
Hence the radiation of perfectly black bodies at equal temperatures is different in different media, and varies inversely as the squares of the velocities in those media, or Thus directly as the squares of the coefficients of refraction. the radiation in water must bear to that in air the ratio
{^y
:
1, nearly.
316
>
have also to temember that in the heat radiated frofia perfectly black bodies there are rays of very different wavelengths; and if we assume that the equality of mutual
radiation holds, not merely for the heat ais a whole, but also each wave-length. in particular, we must have for each of these a proportion similar to (22) and (23) but in which the quantities in the right-hand ratio have somewhat different values. Lastly, instead of perfectly black bodies, let us consider bodies in which the absorption of the rays received is partial only. must then introduce in the formula, in place of the emission, a fraction having the emission as numerator and the coefficient of absorption as denominator. For this fraction we can obtain relations similar to those obtained for the emission alone. This generalization, in which the influence of the direction of the rays upon the emission and absorption must also be taken into account, need not here be entered
for
We
We
upon.
IV.
Determination of the mutual radiation of TWO ELEMENTS OF SURFACE, IN THE CASE WHEN ONE IS THE OPTICAL IMAGE OP THE OTHER,
11.
Relations between B,
T>,
and E,
far as
have hitherto assumed that the planes a and c, so are concerned with them, give out their rays in such a way, that one ray and one only, or at most a limited quantity of individual rays, pa^ss from any point in the one to any point in the other. We will now pass on to the case in which this does not hold. The rays which diverge from points in the one plane may be made to converge by reflections or refractions, and to meet again on the other plane so that for any point p^ on plane a there may be one or more points or lines on plane c, in which an indefinitely large number of the rays coming from p^ cut that plane, whilst other parts of the same plane receive no rays whatever from p^. The same of course holds of the rays which start from plane e and arrive at plane a, since the rays passing to and fro between the same points describe the same paths. Among the innumerable cases of this description we will,
We
we
317
the extreme case illustrated in Fig. 31. In this case all the rays sent out by J3 within a certain definite cone meet again in a single point p^ of plane c. This case occurs for example, when the deflection of the rays is effected
fixst
We
318
HEAT.
limited number of rays, pass from any point on plane b Then for all the points with to any point on plane a or c. which we are concerned, the quantities T^ and Tj and their differential coefficients have real values not indefinitely large. are then as applicable in this case The quantities A, C and as in the former. One of these quantities, E, takes in this case a special For -the three points value, which may at once be found. in which the ray cuts the three planes a, h, , the two equa^ tions given in (1) must hold, viz.
In the present case the position of the poirCt in which the ray cuts plane h is not determined hj the position of the two points p^ and p^, but plane h may be cut in all points of a certain finite area. Hence the two equations above must hold for all these points, and therefore the equations obtained by differentiating these according to ajj and y^ must also hold,
viz.
cP (?:.+
da,:
rj _..
d' (T,,+
-"'
dx,dy,
TJ _ -"'
d'
(T^+ TJ _ ^
d^^
-U...C^4;.
If
tions
(I.),
E in equa--,
(25).
E=Q
The two other
values,
quantities
and
C have
in general finite
which
and which we
'
investigation,
O to
Let us suppose that the element ds^ on plane a has an image ds^ on plane c, so that every point of ds^ has a -point on ds^ as its conjugate focus, and vice vers^. We have now to enquire whether the quantities of heat, which these elements send to each other, when taken as elements of the surfaces of two perfectly blaclc bodies of equal temperatures, are equal or not,.
319^
First, to determine tlie position and magnitude of the image ds corresponding to the given element ds^. Take any point p^ on the intermediate plane h, and consider all the rays starting from points on ds^, which pass through
Each of these rays strikes plane c in the conjugate focus p^. to that from which it starts ; and therefore the element of
surface in which this pencil of rays cuts plane c is pre-i cisely the optical image ds of the element ds^. Therefore, to express the ratio between the ^reas of ds and ds^, we may use one of the three fractions in the lowest row of equations (II.)> which express the relation between the two elements of surface, in which an indefinitely small pencil, passing through a single point ^j of plane b, cuts the two planes a and c. Of these three fractions the first is alone applicable in this case, since the other two are undetermined.
equation
1=^
.
.
-f^^)-
This equation is interesting also from an optical point of view, as being the most general equation which can be given to determine the ratio between the area of an object and that of its optical image. It should be remarked that and G the intermediate plane h, to which the quantities are related, may be any whatever, and can therefore in any particular case be chosen as is most convenient for calcula-
tion.
13.
ds^
Quantities of
Heat which
and
Having thus determined the element of surface ds^ which forms the image of ds^, let us take on plane b, instead of a single point, an element of surface ds^, and let us con-^ sider the rays which the two elements ds and ds^ send through this element dsy All the rays, which start from one point of ds^ and pass through ds^, unite again in on^ point of ds^ and thus all the rays which ds sends through? The dsj exactly strike the element ds,i and vice yers^. two quantities of heat which ds^ and ds send to ds^ are thus the same as the quantities of heat. which ds^ and ds^ sea^
;
320
These to each other through the intermediate element ds^. quantities of heat are therefore given at once by what has Thus for the quantity of heat which ds^ gone before. sends to ds^, the same expression will hold, as held in 9 for the quantity of heat which ds^ sends to ds^, provided
we
substitute
for
B and ds^
eji^^
IT
for ds^.
The
expression thus
becomes
ds^ ds^. A
Similarly for the quantity of heat which ds^ sends to ds^, the expression will be the same as that for the quantity of heat which ds^ sends to ds^, provided we substitute for 5
and
be
e/
IT
dsjds^.
(26)
Gds^ = Ads^,
we
ratio ew/
e,v'.
obtain precisely the same result, if we take any other ds,, on plane b, and consider the quantities of heat which the two elements ds^ and ds^ send to each other through this new element. These will always be found to Since the quanstand to each other in the ratio e^t;/ ei;/. tities of heat, which ds^ and ds send to each other on the whole, are made up of those which they send to each other through all the different elements of the intermediate plane, the same ratio "must hold for the whole ; and we thus obtain the final result, that the total quantities of heat which ds^ and ds^ radiate to each other, stand in the ratio ej!^ e^vf. This is the same relation as was found in sections 8 and 9 for the case where there is no concentration of rays it thus follows that the concentration of rays, however much it alters the absolute magnitudes of the quantities of heat which two elements radiate to each other, leaves the ratio between them exactly the same. It was shewn in 10, that if in the case of ordinary unconcentrated radiation the principle holda that heat cannot pass from a colder to a hotter body, then the radiation must
We
element
321
media, and must be^uch that for perfectly black bodies of equal temperatures
then in the present ease also, where the elements ds^ and ds^ are images one of the other, the quantities of heat which they mutually radiate must be equal; and therefore, in spite of the concentration of the rays, one element cannot raise the other to a higher temperature than its own.
V.
Relation between the increment of area and the RATIO or THE TWO SOLID ANGLES OP AN ELEMENTARY PENCIL OF RAYS.
14.
Statement of the Proportions for this Case,
result of the foregoing we may here develope a proportion, which appears to have some general interest, inasmuch as it illustrates a peculiar difference in the behaviour of a pencil of rays in the case of an object and of its image. This difference' must always exist and have a determinate value, when the object and the image have different areas. Consider an indefinitely small pencil of rays, which starts from a point on ds^, passes through the element ds^ on the intermediate, plane, and then unites again in a point on cfo. may compare the divergence of the rays at their starting point with the convergence of the same at their point of reunion. This divergence and convergence (or, to use the ordinary phrase, the solid angles of the indefinitely small cones, which the pencil forms at its points of starting and re-union) are given directly by the same method which we have used ' in 9, d,g follows: Suppose that round each of the points there is described a sphere of so small a radius, that we may consider the i ays' as going in straight lines as far as the surface of this sphere: and then consider the element of surface in which the pencil Let do- be this element and let p be bif rays cuts the sphere. the radius of the sphere; then the angle of the indefinitely small cone, which contains the rays so far as they are recti-
As an immediate
We
c.
21
322
linear, will'
This fraction
we have
de-
termined for a similax case in 9 by equation (20), and in the expression there given we have only to alter the letters in order to transform it into the expression for the present In order to express the angle of the cone for that case. starting; point of the rays which lies 4n pl^ie a, we have only In to substitute in that expression ds^ for ds^ and G for B. addition,, ;the symbol ^,. which expresses the angle between the pencil of rays and the normal to the surface of ds^, may be changed to 6^ so as to express more clearly that it relates to plane a; and for the same reason the suffix a may be
,
added to
j,
.,
Thus we obtain
^4]=^Gds,
To
(27),
obtain the other equation, which gives the angle of the cone at the point of re-union on plane c, we have only to change the S'uffia^ a into c throughout, and also to substitute
A for
G.
Thus we have
!da ^'^
""
Ads^
./)V
cos^,
(28).
From
we
cos 5 (d<T\
since
by equation
(26)
Gds^= Ads^.,
If
we
this proportion
becomes
:
n^ cos
e [ -?)
K cos
0,
(-^ j
::
ds,
ds
(30).
the.
Here the ratio on the right-hand side is the ratio between area of an element of surface of the image, and the area of
element of the object, or in short the proportionate increment of area ; and we thus obtain a simple relation between this increment and the ratio of the angles of the
the. corresponding
is easily
seen
323
that it is not necessary for the truth of this proportion that the rays should finally converge, and meet at one point, but they may also be divergent, in which case their directions meet in one poiiit when produced backwards, and form what is called in optics a virtual image. If we take the special case in which the medium at the point of starting and of re-union is the same (e.g. where the rays issue from an object which is in air, and, after certain refractions or reflections, form an image which really or virtually is also in air), then v^ = v and w^ = whence
;
we have
cos0f-^j
If
:
cos0j-?j
::
ds^
ds^.
we add
we have
{?)At);-
'''''
In this case the angles of the cones formed by the pencil of rays at the object and at the image stand simply in the inverse ratio of the areas of the corresponding elements of object and image. In the valuable demonstration which Helmholtz has given in his "Physiological Optics"* of the laws of refraction in the case of spherical surfaces, he seeks to connect with these the case of the refractions which take place in the eye ; and he' finds in page 50, and extends further in page 54, an equation which expresses- the relation between the size of the image and the convergence of the rays, for the case in which the change of direction of the rays takes place by refraction or by reflection at the surface of co-axial spheres, and in which the rays are approximately perpendicular to the planes which
contain the object and the image. The author however believes that the relation' has not before been given in its general form, as in proportions (29) and (30).
Karsten's Universal Eneyelopedia of Physics.
212
32-l!
VI.
General determination of the mutual radiation BETWEEN TWO SURFACES, IN THE CASE WHERE' ANY CONCENTRATION WHATEVER MAY TAKE PLACE.
15.
We must now extend our investigation so as to embrace not only the extreme case, in which all the rays which issue from a point on plane a within a certain finite cone unite again in one point forming a conjugate focus on plane c, but also every conceivable case of the concentration of rays. To obtain a closer view of the phenomena of concentration, we may use the following definition. If rays issue from any point p^ and fall on plane c, and if these rays when close to that plane have such directions that on one part of the plane the density of the impinging' rays is indefinitely great compared to the mean density, then at this part there is concentration of the rays which issue from p^. With this definition we may easily treat mathematically the case of concentration. Between point p^ and plane c take any intermediate plane h, which is so placed that there is no concentration in it of the rays issuing from p^ and also that its relation to plane c is such, so far as we are concerned, that the pencils of rays issuing from points on one of those planes suffer no concentration on the other. Now consider an indefinitely, small pencil, which starts from p^ and cuts the planes h and c. Let us compare the areas of the elements ds^ and ds^ in which these planes are cut. If ds^ vanishes in comparison to ds^, so that we may put
;
|-
this is
(5').
a sign that there is a concentration of rays, in the sense defined above, at plane c. Let us now return to equations (II.) of 7. The equations in the first horizontal row are those that refer to the present case and of the three fractions, which there represent the ratio of the elements of surface, the first applies to our case, because under the assumption made as to the position of the intermediate plane we may- determine and in the ordi:
325
A'
This fraction can only equal zero if the numerator is zero, since under the assumption made as to the position of the intermediate planes the denominator A cannot be indefinitely large. We have then as a mathematical criterion, whether the rays issuing from p^ suffer concentration at plane c or not, the condition
E=Q
(32),.
which must be fulfilled where there is concentration. Now assume conversely that on plane c a point p^ is given, and that we have to decide whether the rays issuing from this point suffer concentration at any part of plane a. Then
we have
(II.)
in the same
way the
ds condition t-"
and
since
by
-T^
=^
we
arrive at the
same
final condition
i;=o.
It is in fact easy to see that when the rays issuing from a point on plane a suffer concentration at plane c, then conversely the rays issuing from the latter point must suffer concentration on plane a. Since equations (13) express the relations which hold between the six quantities J, B, G, 3, E, F, we may apply and F, those equations to ascertain what becomes of B, = 0, whilst and have finite values. in the case where By those equations
^=-'
Hence
^=?'
^=i"
must
(^^)-
in the present
16. Mutual Radiation of an Element of Surface and of a Finite Surface, through an Element of an Intermediate
Plane.
We will now attempt so to determine the ratio between the quantities of heat which two surfaces radiate to each
326
other, that the result must hold in all cases, independently of the question whether there is any concentration of heai
or no.
will substitute for the planes a as hitherto considered, two surfaces of any kind, which we may call s and s. Between them let us take any third surface Sj, which need only fulfil the condition that the rays
and
s to s, s
Now
dsj,
choose in
ment
it
when produced strike the surface s. Then we will first determine how much heat ds^ sends through ds^ io the surface s, and how much heat it receives back from s through the same element of the intermediate, plane. To ascertain the first mentioned quantity of heat, we have only to determine how much heat ds^ sends to ds^, since, by our
from ds^
will
assumption as to the position of ds^, all this heat after passing dsj must strike the surface s. This quantity of heat may be expressed at once by means of our previous formulae. Suppose a tangent plane to be drawn to s^ at a point of the element ds^, and similarly a tangent plane Sj at a point oids^; and consider the given elements of surface as elements of these planes. If in these tangent planes we take two systems of co-ordinates a;, y^, and x^, y, and if we form the quantity G by means of the third of equations (I.), then the required quantity of heat, which ds^ sends through ds^ to the surface s^, is given by the expressions
ej]^
ds^dSi.
IT
Next with regard to the quantity of heat which ds^ receives through dsj from the surface s, the relations of the points in s^, from which these rays issue, are not in general
so simple as that which holds in the special case, where ds, has an optical image ds lying on s, and therefore is also itself the optical image of ds. If we choose a known point p^ on the intermediate element ds^, and consider the rays which pass through this point from all points of ds^, we have an indefinitely small pencil of rays, cutting s^ in a certain element of surface. It is this element which sends rays to ds^ through the selected point p^. But if we now choose
327
.another point of ds^ as the vertex of the pencil, we arrive at a somewhat different element on the surface s^. Thus the rays, which ds^ receives from s^ through different points on ds^, do not all issue from one and the same eleinent of s,. Since however the area of ds^ may be any whatever, nothing hinders us from supposing it so small, that it is an indefinitely small quantity of a higher order than ds^ In this case, if the vertex of the pencil changes its position within dsf, then the element of which corresponds to ds^ will change its position through a distance so small that in comparison with the dimensions of the element it is indefinitely small and may be neglected. Hence in this case the 'element ds^, which we obtain when we choose any point whatever p^ on element dsi and make it the vertex of the pencil of rays issuing from,ds, may be considered as the part of ds^ which exchanges rays with ds^ through ids^. The area of this element ds^ is easily found from what precedes. Let us suppose as before that a tangent plane to the surface s, is drawn at p^, and that tangent planes to the surfaces s and s are drawn at points on the elements ds^ and ds^ respectively; and let us consider the two latter elements as elements of the tangent planes. Take systems of co-ordinates on these and C, as three tangent planes, and form the quantities given by the first and third of equations (I.). Then by
equation
(II.)
we have
'
ds,=
-^d8^.
The quantity of heat which <i5, sends to ds and which, as mentioned above, may be considered as the quantity which ds^ receives from the surface s, through ds is expressed, by
IT
S28
If we Compare this expression with that found above, which expresses the quantity of heat sent by ds^ through ds, to s, we see that the two stand to each other in the ratio If we now suppose that and s are the sure^v^ efi^. faces of two perfectly black bodies of equal temperature, and make for such surfaces the assumption (which we have already seen to be necessary in the case of radiation without concentration), that the two products ev/ and e^^ are equal, then the quantities of heat given by the two expressions above
:
Mutucd Radiation of Entire Surfaces. on the intermediate surface s^ we take, instead of the element hitherto considered, another element which is also supposed to be an indefinitely small quantity of a higher order, then the element of s^, which exchanges rays with ds^ through this element of s^, has a different position from that but the two quantities of heat thus exin the last case changed are again equal to each other and the same holds of all other elements of the intermediate surface. To obtain the quantity of heat which ds^ sends to s, not through a single element of the intermediate surface, but as a whole, and similarly the quantity of heat which as a whole it receives from s, we must integrate the twp expressions found above over the surface s so far as this surface is cut by the rays which pass from ds^ to s, and vice verset. It is evident that if for each element of surface ds^ the two differentials are equal, then the whole integrals must also be equal. Lastly, to find the quantities of heat, which the whole surface s^ exchanges with s, we must integrate both these expressions over the surface *,. This process again will not disturb the equality, which exists for each of the separate
17.
If
;
elements
ds^.
principle previously discovered in a special case, perfectly bla,ck bodies of equal temperatures exchange equal quantities of heat with, each other, so far as the
equation ejD^ ^^^ holds for them, appears also as the result of an investigation, which in no way depends upon whether the rays issuing from s, suffer a concentration at s and vice vers^, or not; since the only condition was, that the rays
329
issuing from s^ and s suffer no concentration at the interinediate surface Sj, a condition which may be always fulfilled, since this surface may be chosen at pleasure. It follows further from this result that, if a given black body exchanges heat not only with one but with any number of black bodies, it receives from all of them exactly the same quantity of heat as it sends to them.
18.
The
made throughout
and refractions
reflections
take place without loss, or that there is no absorption of heat. can, however, easily go on to shew that the results are not altered, if this condition is dropped. For consider any one of the different processes, by which a ray may be weakened on its way from one body to another say that at a place where the ray cuts the boundary of two media, one part passes into the further medium by refraction, and the other is reflected. Then whether we consider the one part or the other as the continuation of the original ray, we have in both cases a weakened ray to deal with. The same holds if a ray be partially absorbed by entering a medium. But in each of these cases we have the law that two rays which traverse the same path in opposite directions are weakened in equal proportion. The quantities of heat, which two bodies mutually send to each other, are therefore weakened by such processes to the same extent so that, if they would be equal without such weakening, they will also be equal when thus weakened. Another circumstance may be considered in connection with the processes above mentioned, viz. that a body may receive from the same direction rays which proceed from different bodies. For example, a body may receive from a point, which lies on the bounding surface of two media, two rays coinciding in direction, but issuing
We
and 0. One of these may come from two different bodies from the bounding medium and be refracted at the point, whilst the other is already in the bounded medium, and is reflected at the point. In this case, however, the two rays are weakened by refraction and reflection in such a way, that, if both were before of equal intensity, their sum afterwards is
330
of the
same intensity as either one of them had beforehand. suppose a ray of the same intensity to. issue from the in the opposite direction,, this will be divided, at the body same point, into two parts, of which one enters the bounding medium, and passes forward to the body B, while the other is reflected and passes to the body C. The two parts which thus reach are exactly as great as those which and from receives from and C. The body A thus stands to each of the bodies and G in .such a relation, that, assuming equal temperatures, it exchanges with them equal quantities of heat. The equality of the modifications which two rays undergo, when passing in opposite directions in any path whatever, must produce the same result in all other cases
Now
bla>ek bodies,
we
consider
such as only partially absorb the rays falling on them; or if instead of homogeneous heat we consider heat. which contains systems of waves f different lengths or lastly, if instead of taking all the rays as unpolarized we include the phenomena of polarization; still in all these cases we have to do only with facts, which hold equally for the heat sent out by any one body,. and for that which it receives from other bodies. It is not necessary to consider these facts more closely, since they also take place with ordinary radiation without concentration, and the object of the present investigation was only to consider the special actions which might possibly be produced by concentration of the rays.
;
19.
Sumrmiry of
Restdts,
The main
may be
briefly
stated as follows
In order to bring the action of ordinary radiation, (1) without concentration, into accordance with the fundamental principle, that heat cannot of itself pass from a colder to a hotter body, it is necessary to assume that the intensity of emission from a body depends not only on its own composition and temperature, but also on the nature of the surrounding medium; the relation being such, that the intensities of emission in different media stand in "the inverse ratio of the squares of the velocities of radiation in the
331
CHAPTER
XIII.
DISCUSSIONS ON THE MECHANICAL THEOET OF HEAT AS HERE DEVELOPED, AND ON ITS FOUNDATIONS.
Different Views as to the Relation between
1.
Heat and
Work.
The papers of the author on the Mechanical Theory of Heat, as reproduced in all essential particulars in this volume, have frequently met with opposition; and it may be desirahle to give here some of the discussions which have taken place on the question, since many points are raised in them, on which the reader may even yet be in doubt. The removal of these doubts may be facilitated by his learning what has been already written upon these points. As already mentioned in Chapter III., the first important attempt to reduce to general principles the action of heat in doing work was made by Carnot. He started from the assumption that the total quantity of heat existing was invariable, and then supposed that the falling of heat from a higher to a lower temperature brought about mechanical work, in the same way as the falling of water from a higher to a lower level. But simultaneously with this conception the view gained ground that heat is a mode of motion, and that in producing work heat is expended. This view was set forth at intervals from the end of the last century by individual writers, such as Rumford, Davy, and Seguin*; but it was not. tin 1840 that the law corresponding to this view, viz. that of
* In a paper by Mohr, publiBhed 1837, heat molion, in otfieis a force.
is
333
tVie
equivalence of neat and work, was definitely laid down by Mayer and Joule, and proved by the latter to be correct by means of numerous and brilliant experiments. Soon
afterwards the general principle of the conservation of energy was laid down by Mayer* and Helmholtzf (by the latter in a specially clear and convincing manner), and was applied to various natural forces. fresh starting point was thus given to researches on the science of heat but in carrying these out great difficulties presented themselves, as was natural with so widely extended a theory, which was intertwined with all branches of natural science, and influenced the whole range of physical thought. In addition, the wide acceptation which Carnet's treatment of the mechanical action of heat had won for
especially after being brought by Clapeyron into an elegant analytical form, was unfavourable for tlie reception of the new theory. It was believed that there was no alternative but either to hold to the theory of Carnot, and reject the new view according to which heat must be expended to produce work, or conversely to accept the new view and reject Carnot's theory.
itself,
2.
Papers on
the Subject hy
Thomson and
the Author.
very definite utterance on the then position of the question was given by the celebrated English physicist, now Sir William Thomson, in an interesting paper which he published in 1849 (when most of the above-mentioned researches of Joule had already appeared and were known to him), under the title, "An Account of Carnot's Theory of the Motive-Power of Heat, with numerical results deduced from Kegnault's ex;periments on steam J." He still maintains the position of Carnot, that heat may do work without any change He however points in the quantity of heat taking place. out a difficulty in this view, and goes on to say, p. 545 " It might appear that this difficulty might be wholly removed, if we gave up Carnot's fundamental axiom, a view which has
:
Mr
Joule,"
He
adds,,
<'
but
if
we do
Stoff^
weeheel.
Heilbronn, 1845. + Ueber die Erhaltung der Kraft, 1837. T Trans. Royal Soc, of Edin., Vol. xvi.
p. 541,
834
this we stumble over innumerable other difficulties, which are insuperable without the aid of further experimental researches, and without a complete reconstruction of the Theory of Heat. It is in fact experiment to which we must look, eitber for a confirmation of Carnot's axiom, and a clearing up of the difiSculty whicb we have noticed, or for a cornpletely new foundation for the Theory of Heat."
At the time when this paper appeared the author was writing his first paper " On the Mechanical Theory of Heat," which was brought before the Berlin Academy in 1850, and printed in the March and April numbers of Poggendorff's Annalen. In this paper he attempted to begin the reconstruction of the theory, without waiting for further experiments; and he succeeded, he believes, in overcoming the
mentioned by Thomson, so far at least as to leave plain for any further researches of this character. He there pointed out the way in which the fundamental conception, and the whole mathematical treatment of heat,, must be altered, if We accepted the principle of the equiva^' lence of heat and work ; and he further shewed that it was not needful wholly to reject the theory of Carnot, but thatwe might adopt a principle, based on a different foundation from Carnot's, but differing only slightly in form, which might be combined with the principle of the equivalence of heat and work, to form with it the ground-work of the new theory. This theory he then developed for the special cases of perfect gases and saturated vapour, and thereby obtained; a series of equations, which have been universally employed in the form there given, and which will be found in Chap' ters II. and VI. of this volume.
difficulties
the
way
3. On Rcunhine's Paper and Thomson's Second Paper. In the same month (February, 1850) in which the author's papej* was read before the Berlin Academy, a valuable paper by Eankine was read before the Royal Society of Edinburgh, and was afterwards published in their Transactions (;VoL 20, p. 147)*.
Rankine there proposes the hypothesis, that heat consists in a rotary motion of the molecules ; and thence deduces with
* In 1854 it was reprinted -with Series 4, Yol. vn. pp. 1, HI, 172.
some
Mag,,
DISCtrSSIONS
ON THE THKORT.
335
much
skill a series of principles on the action of heat, which agree with those deduced by the author from the first main principle of the mechanical theory*. The second main principle was not touched by Rankine in this paper, but was treated in a second paper, which was brought before the Eoyal Society of Edinburgh a year afterwards (April, 1851). In this he remarka-f- that be had at first felt doubtful of the correctness of the mode erf reasoning by which the author had arrived at this second principle ; but that having communicated his doubts to Sir W, Thomson, he was induced by him to investigate the subject more closely. Ke then found that it ought not to be treated as an independent principle in the theory of heat, but might be deduced from the equations, which he had given in the first section of his former paper. He proceeds to give this new proof of the principle, which however, as will be shewn further on, is in opposition, for certain very important, cases, with his own views, as elsewhere expressed. This paper of 1851 Rankine added as a fifth section to his. former paper on account of the connection between them. Thence has arisen with some authors the mistaken idea that this new paper was actually part of the former one, and that Rankine had therefore given a proof of the second main From the foreprinciple at the same time as the author. going it will be seen that his proof (waiving the question how far it is satisfactory) appeared a year later than the
,
author's.
same year, 1851, a second paper by Sir on the Theory of Heat was laid before the Royal Society of Edinburgh f. In this paper he abandons his former position with regard to Carnot's theory and assents to
In
of the
March
W. Thomson
the author's exposition of the second main principle. He then extends the treatenent of the subject. For whilst the author had confined himself in the mathematical treatment of the question to the case of gases, of vapours, and of the process of evaporation, and had only added that it was easy to see how to make similar applications of the theory. to other
Edinb. Trans., Yol.,xx. p. 205; Phil. Mag., Series i, Vol. vii. p. 2i9. Phil. Mag., Vol. vii. p. 250. j Edin. Trans., Vol. xx. p. 261 ; Phil. Mag., Series 4, Vol. iv. pp. 8, 105,
t
168.
330.
cdses,
Thomson developes.a series of general equations, whieh are independent of the body's condition of aggregation, and only then passes on to more special applications. On one point this second paper still falls short of the For here also Thomson holds fast by the law of author's. Mariotte and Gay-Lussac in the case of saturated vapour, and hesitates to accept an hypothesis with respect to permanent gases, which the author had made use of in his investigation (see Chapter II., 2, of this work). On this he remarks* " I cannot see that any hypothesis, such as that adopted by Clausius fundamentally in his investigations on this subject, and leading, as he shews, to determinations of the densities of saturated steam at different temperatures, which indicate enormous deviations from the gaseous laws of variation with temperature and pressure, is more probable, or is probably nearer the truth, than that the density of saturated steam does follow these laws, as it is usually assumed to do. In the present state of science it would perhaps be wrong to say that either hypothesis is more probable than the other." Some years later, after he had proved by his joint experiments with Joule that this hypothesis is correct within the limits assigned by the author, hfe used the same method as the author to determine the density of saturated
:
vapour "I".
Rankine and Thomson, so far as the author knows, have always recognized most frankly the position here assigned to the first labours of themselves and the author on the meThomson remarks in his paper J: chanical theory of heat. " The whole theory of the moving force of heat rests on the two following principles, which are respectively due to Joule and to Carnot and Clausius." Similarly he introduces the second main principle as follows: "Prop, II. (Carnot and Clausius)." He then proceeds to give a proof discovered by himself, and goes on: "It is with no wish to claim priority that I make these statements, as the merit of first establishing the proposition on correct principles is entirely due to Clausius, who published his demonstration of it in the month
p.
iv. p. 111.
Phil. Mag., Vol. iv; p. 11; Phil, Mag., Vol. iv. pp. 14, 242.
'
'
337'
May
4,
last year, ia
other quarters repeated and in some cases violent objections were raised to the author's first paper, to which, in the same and following years, a series of other papers, serving to complete the theory, were added. The earliest of these objections came from Holtzmann, who had published in 1845 a short pamphlet* on the subject. In this it would at first appear as if he wished to treat the question from the point of view, that for the generation of work there was necessaiy not merely a change in the distribution of heat, but also an actual destruction of it, and that conversely by destroying work heat could be again generated. He remarks (p. 7): " The action of the heat which has passed to the gas is thus either a raising of temperature, combined with an increase of the elastic force, or a certain quantity of mechanical work, or a mixture of the two ; and a certain quantity of mechanical work is equivalent to the rise in temperature,. Heat can only be measured by its effects; of the two above-named effects mechanical work is especially adapted for measurement, and it I call a unit of will be chosen accordingly for the purpose. heat that amount of heat, which by its entrance into a gas can perform the mechanical work a, or, using definite measures, which can raise a kilograms to a height of 1 metre." Further on (p. 12) he determines the numerical value of the constant a by the method previously used by Mayer, and explained in Chapter II., 5; the number thus obtained corresponds perfectly with the mechanical equivalent of heat, as determined by Joule by various other methods. In extending his theory however, i.e. in developing the equations by which his conclusions are arrived at, he follows the same method as Clapeyron; so that he tacitly retains the assumption that the total quantity of heat which exists is invariable; and therefore that the quantity of heat which a body takes in, while it passes from a given initial condition to its present condition, must be expressible as a function of the variables, which determine that condition.
* Ueber die
From
Holtzmann.
Q.
Warme und Etasticitdt der Gase und Dampfe; von Mannheim, 1845 also Pogg. Am., Vol. lxxii a.
;
C.
22
338
method was pointed out, and the question treated in another way.; on which Holtzmann wrote an article*, in which he endeavoured to shew that this method of treatment, and specially the assumption that heat was expended in producing work, was inadmissible. The first objection which he raised was of a mathematical character. He carried out
an investigation similar to that in the author's paper, in order first to determine the excess of the heat which a body takes in over that which it gives out, during a simple cyclical process consisting of indefinitely small variations, and secondly But iii such a to compare this excess with the work done. process both the work done and the excess of heat must be indefinitely small quantities of the second order and therefore in the whole investigation, care must be taken that all quantities of the second order, which do not cancel each other, shall be taken into account. This Holtzmann neglected to do and he was thus led to a final equation, which contained a self-contradiction, and in which he therefore imagined that he had found a proof of the inadmissibility of This objection was easily disposed of the whole method.
; ;
by the author
in his reply. further brought forward as an obstacle to the theory, that, according to the formulae given, the specific heat of a perfect gas must be independent of its pressure, whereas the experiments of Suermann, and also those of De la Roche and B^rard, shewed that the specific heat of gases increased as the pressure diminished. On this conflict between his
He
own theory and the experiments which were then known and supposed to be correct, the author remarked in his reply as follows "On this point I must first point out that, even
:
these observations are perfectly correct, they yet say nothing against the fundamental principle of the equivalence of heat and work, but only against the approximate assumption which I have made, viz. that a permanent gas, if it expand at constant temperature, absorbs only so much heat, as is required for the external work which it thus performs. But besides it is sufficiently known how unreliable are in general the determinations of the specific heats of gases ; and all the
if
339
more
made
in those few observations -whicli have been hitherto at varying pressures. I did not therefore conceive
myself bound to abandon the above assumption on account of these observations, although they were well known to me at the time when I wrote my former work; because the other grounds, which may be alleged for the correctness of the assumption within the limits which I had there laid down, are not wholly destroyed by the grounds which may be alleged against it." This remark found its full confirmation in Eegnault's Mesearches, published some years a,ftervvards, on the specific
heats of gases, which actually led to the result that these earlier, observations were inaccurate, and that the specific heat of permanent gases is not visibly dependent on the
pressure.
5.
Decker's Objections.
222
340
the differential equations there formed, which, though not generally int^grable, become so as soon as one further relaIn spite of all tion is assumed to exist among the variables. which the author has said, he has throughout treated the quantities to which these equations refer, viz. the quantities of heat taken in by a body in passing from a given initial condition to its present condition, as mere functions of the variables which determine the condition of the body. After quoting the author's equation for gases, viz.
dt\dv)
dvKdt)
'^
''
is the heat-equivalent of the unit of work, i.e. the where reciprocal of E, he remarks, page 243: "In equation (1)
(
-^1 and
a known function of v and t, viz- Q, taken according to V and t respectively as sole variables ; and in whatever way this function may be formed, aud whatever relation may be supposed to exist between v and t, the right side of the equation must always equal zero." This incorrect conception, thus formed by a professed mathematician, convinced the author that the meaning and treatment of this kind of differential equation, although long before established by Monge, was not so generally known as he had supposed; accordingly in his reply*, after a brief notice of some other points raised by Decher, he treated the subject more fully, giving a mathematical explanation, which seemed to him sufficient to obviate any such misunderstandings in future. This was afterwards prefixed to the collection of the author's papers as a mathematical introduction ; and the essential part of it has been imported into the mathematical introduction to the present work.
efficients of
the Author^s
Proof
The more recent objections to the author's theory, and the departure from his views in more recent treatises, chiefly
Dinglei's
ci. p. 29,
341
refer to "his method of proving tte second main principle of the theory. This proof rests, as shewn in Chapter III., on the following fundamental principle: Heat cannot of itself (or
without compensation) pass from a colder to a hotter body. This fundamental principle has been very variously received by the scientific public. Some appear to consider it so self-evident that it is needless to state it as a specific principle, whilst others on the contrary doubt its correctness.
7.
of the two modes of viewing the question mentioned in the last section appears in Zeuner^s valuable paper of 1860 " On the Foundations of the Mechanical Theory of Heat." Here Zeuner gives the author's proof of the second main principle essentially in the same form as it has also been given by Reech*. The two differ only in one point. Reech gives the principle, that heat cannot of itself pass from a colder to a hotter body, expressly as a fundamental principle laid down by the author, and bases his proof upon it. Zeuner on the contrary does not mention this principle at all he shews that if for any two bodies the second main principle of the theory did not hold, then by means of two cychcal processes performed with these two bodies in opposite directions, heat could be made to pass from a colder to a hotter body without any other special change, and he then goes on " as we may repeat both processes as often as we please, using the two bodies alternately in the way described, it would follow that we might, with the aid of nothing and without using either work or heat, continually transfer heat from a body of lower to one of higher temperature, which is an absurdity." Few readers would probably assent to the opinion that the impossibility of transferring heat from a colder to a hotter body is so self-evident, as is here indicated by the short remark "which is an absurdity." Taking the facts of conduction, and of radiation under ordinary circumstances, we may undoubtedly say that this impossibility is established by daily experience. But even with radiation the question
:
The
* Secapitulatian tris-succincte des recherches algebriqties faites sur la theorie des ejfects m^caniques de la chaleur par diffSrents auteurs : Jmirn. de Limville, Ser. n. Vol. i. p. 58.
Sii
arises, whether it is not possible to concentrate the rays of heat artificially by means of mirrors or burning-glasses, so as to produce a higher temperature than that of the radiating and thus to effect the passage of the heat into a , bodies, hotter body. The author has, therefore, thought it necessary to treat this question in a special paper, the contents of which are given in Chapter XII. Matters are still more complicated in cases when heat is transformed into work, and vice versi, whether this be by effects such as .those of friction, resistance of the air, and electrical resistances, or whether by the fact that one or more bodies suffer such changes of condition, as are connected partly with positive and partly with negative work, both internal and external. For by such changes heat, to use the common expression, becomes latent or free, as the case maybe; and this heat the variable bodies may draw from or impart to other bodies of different temperatures. If for all such cases, however complicated the processes may be, it is maintained that without some other permanent change, which may be looked upon as a compensation, heat can never pass from a colder to a hotter body, it would seem that this principle ought not to be treated as one altogether Self-evident, but rather as a newly-propounded fundamental principle, on whose acceptance or non-acceptance the validity of the proof depends.
8.
The mode of expression employed by Zeuner was criticized by the author on the grounds stated in the last section, in a paper published in 1863. In the second edition of his book, published in 1866, Zeuner has therefore struck out another way of proving the second main principle. Assuming the condition of the body to be determined by the pressure p and volume v, he forms, for the quantity of heat d Q, taken in by the body during an indefinitely small variation,
dQ = A{Xdp+7dv)....
where
(2),
and
Y are functions
equivalent of work.
343
be integrated so long as p and v are independent variables. He then proceeds (p. 41) "But let /S be a new functioti of jo and v, the form of which may be taken for the present to be known as little as. that of and 3^, but to which we will give a signification, which will appear immediately from what follows. Multiply-
we have
dQ = ^/S&2> + 1^^1)1
(3).
S, so that
is
may be
the
integrating factor, or B the integrating divisor, of the expression within brackets of equation (2)." From this it follows that in the following equation derived
from
(3),
(4),
is
a perfect
differential,
and there-
cyclical process
we must have
/f-o
In
(7)
(*
this
fT
The resemblance, however,
= 0.
is merely external. The essence of this latter equation consists in this, that t is a function of temperature only, and further a function which is independent of the nature of the body, and is therefore the same Zeuner's quantity 8, on the contrary, is a for all bodies.
function of both the variables, f and v, on which the bodies' and Y, condition depends ; and further, since the functions in equation (2), are different for different bodies, it must be true of S also that it may be different for different bodies. So long as this holds with regard to B, equation (5) has done nothing for the proof of the second main principle
344
since
must in general be an
denoted by
-q,
may be
and by
which the expression within brackets in equation (2) may be converted into a complete differential. Accordingly, in Zeuner's proof, as he himself concludes, everything depends on the fact that iS* is a function of temperature only, and a function which is the same for all bodies, so that it may be taken as the true measure of the temperature. For this purpose he supposes a body to undergo different variations, which are such that the body takes in heat whilst 8 has one constant value, and gives out heat whilst 8 has another constant value and which together make up a cycliThis procedure cal process, shewing a gain or loss of heat. he compares with the lifting or dropping of a weight from one level to another, and with the corresponding mechanical work ; and he proceeds (p. 68) "A further comparison leads
;
:
we may
-^ as
a weight
in what follows therefore I shall call the above value the Weight of the Heat." Since a name has here been introduced for a magnitude containing 8, in which name there is nothing which relates to the body under consideration, it appears that an assumption has here been tacitly made, viz. that 8 is independent of the nature of the body, which is by no means borne out by the earlier definition. Zeuner then carries still further the comparison between the processes relating to gravity and those relating to heat, and .transfers to the case of heat some of the principles which hold
for gravity
;
in so doing he treats
8 as
a,
height,
and -j^
as a
weight, just as before. Then, having finally observed that the principles thus obtained are true if we take 8 to mean the temperature itself, he proceeds (p. 74) " are therefore justified in taking as the basis of our further researches the hypothesis that 8 is the true measure of temperature."
:
We
It appears from this that the only real foundation of the reasonings, which in bis second edition Zeuner puts forward
is
the analogy
345
between the performance of work by gravity and by heat and moreover that the point "which has to be proved is in part tacitly assumed, in part expressly laid down as a mere
hypothesis.
9.
Subject.
turn to those authors who have considered that the fundamental principle is not sufficiently trustworthy, or even that it is incorrect. Here we must first examine somewhat more closely the mode of treatment which, as already mentioned, Bankine considered must be substituted for that of the author. Rankine, like the author, divides the heat which must be imparted to the body, in order to raise its temperature, into two distinct parts. One of these serves to increase the heat actually existing in the body, and the other is absorbed in work. For the latter, which comprises the heat absorbed in the internal and in the external work, Rankine uses an expression, which in his first section he derives from the hypothesis that matter consists of vortices. Into this method of reasoning we need not enter further, since the circumstance that it rests on a particular hypothesis as to the nature of molecules and their mode of motion, makes it sufficiently clear that it must lead to the consideration of complicated questions, and thus leaves much room for doubt as to its trustworthiness. In the author's treatises he has based the development of his equations, not on any special views as to the molecular constitution of bodies, but only on fixed and universal principles; and thus, even if the above fact were the only one which could be alleged against Rankine's proof, the author would stiU expect his own mode of treating the subject to be finally established as the most correct. But yet more uncertain is Rankine's mode of determining the second part of the heat to be imparted, viz. that which serves to increase the heat actually existing in the body. Rankine expresses the increase of the heat within the body, when its temperature t changes by dt, simply by the product Kdt, whether the volume of the body changes at the same time or not. This quantity K, which he calls the real specific heat, he treats in his proof as a quantity independent of the specific voltime^ Any sufficient ground for this
We may now
346
procedure will be sought in vain in his paper ; on the contrary, data are found which stand in direct opposition to it. In the introduction to his paper he gives, under equation (13), an expression for the real specific heat K, which contains a factor k, and of which he speaks as follows*: "The coefficient Ic (which enters into the value of the specific heat) being the ratio of the vis viva of the entire motion impressed on the atomic atmospheres by the action of their nuclei, to the vis viva of a peculiar kind of motion, may be conjectured to have a specific value for each kind of substance, depending in a manner yet unknown on some circumstance in the constitution of its atoms. Although it varies in some cases for the same substance in the solid, liquid, and gaseous states, there is no experimental evidence that it varies for the same substance in the same condition." Hence it appears to be Rankine's view that the real specific heat of the same substance may be different in difi:erent states of aggregation and even for the assumption that it may be taken as invariable for the same state of aggregation he produces no other ground than that there is no experimental proof to the
contrary.
In a later work, A Manual of the Steam Engine and other Prime Movers, 1859, Rankine speaks yet more distinctly on
change of real specific accompanies the change between any two of those conditions" (i.e. the three conditions of aggregation). How great a difference Rankine conceives to be possible between the real specific heats of one and the same substance in different states of aggregation, is shewn by his remark at the same place, that in the case of water the specific heat as determined by observation, which he calls the apparent specific heat, is nearly, equal to the real specific heat. Now Rankine well knew that the observed specific heat for water is twice as great as that for ice, and more than twice as great as that for steam. Since then the real specific heat for ice and steam can never be greater than the observed, but only smaller, Rankine must assume that the real specific heat of water exceeds that of ice and steam by 100 per cent, or more. If we now ask the question how on this supposition the
this point as follows (p. 307): heat, sometinles considerable, often
*
"A
347
increase of heat actually present in a body, whicli occurs when its temperature t increases by dt, and its volume v by dv, is to be expressed, the answer will be as follows When
:
change of volume, suffers no change in its state of aggregation, we shall be able to express the increase of heat, as Rankine has done, by a simple product of the form Kdt but the factor must have different values for different states of aggregation. In cases, however, where the body during its change of volume also changes its state of aggregation (e.g. the case we have treated so often, where we have a certain quantity of matter partly in the liquid and partly in the gaseous condition, and where the magnitude of these two parts changes with the change in volume, either by the evaporation of part of the liquid, or by the condensation of part of the vapour), we can then no longer express the increase of heat connected with a simultaneous change in temperature and volume by a simple product Kdt but must use an expression of the form
its
;
Kdt + K^dv.
the real specific heat of a substance were different in different states of aggregation, it would be necessary to conclude that the quantity of heat existing in it must also depend on its state of aggregation ; so that equal quantities of the substance in the solid, liquid, and gaseous condition would contain different amounts of heat. Accordingly, if part of the substance change its state of aggregation without any change of temperature, there must also be a change in the quantity of heat contained in the substance as a whole. Hence it follows that Bankine by his own admission can only treat the mode in which he expresses the increase of heat, and the mode in which he uses that expression in his proof, aS being allowable for the cases in which there are no changes and, therefore, his proof holds in the state of aggregation For all cases where such changes occur for those cases only. the principle remains unproved and yet these cases are of special importance, inasmuch as it is chiefly to these that the principle has hitherto been applied. In fact we must go further, and say thtii the proof thus loses all strength even for cases where there is no change in If Rankine .assumes that the real !the state of aggregation.
if
; ;
For
348
specific
heat may be different in different states of aggregathere seems no ground whatever left for supposing that it is invariable in the same state of aggregation. It is known that with solid and liquid bodies changes may occur in the conditions of cohesion, apart from any change in the With gaseous bodies also, in addition state of aggregation. to their great variations in volume, we have the distinction, that the more or less widely they are removed from their condensation-point the more or less closely do they follow the law of Mariotte and Gay-Lussac. How then, if changes in the state of aggregation may have an influence on the real
tion,
heat, can we refuse to ascribe a similar, even if a smaller, influence to changes like the above? Thus the
specific
proposition, that the real specific heat is invariable in the same state of aggregation, is not only left unproved by Rankine, but, if we accept his special assumption, becomes in
a high degree improbable. To this criticism on his proof, which appeared in a paper of the author's, published in 1863*, Rankine made no reply; but in a later article on the subject-f* he expressly maintained the truth of his view, firequently before stated, that the real specific heat of a body may be different in different states of aggregation; whereby the force of his proof is limited to the cases in which no change in the state
of aggregation takes place.
10.
Hvriis Objection.
yet more definite attack upon the author's fundamental principle, that heat cannot of itself pass from a colder to a hotter body, was made by Him in his work, published in 1862, Exposition ATudytique et Eocperimentale de la theorie mecanique de la chaleur, and in two subsequent articles in Cosmosl. He has there described a particular operation which gives at first sight an altogether startling result. After a reply irom the author , he explained his attack as having for its object only to mark an apparent objection to the principle, whilst in reality he agreed with the author; and he has expressed himself to the same effect in the second and third editions of his valuable work.
||
* Pogg. Ann., Vol. cxx. p. 426. Phil. Mag., Ser. 4, Vol. xxx. p. 410. J Tol. XXII. pp. 283, 413, Vol. xxii. p. 560.
||
VoL
xxii. p. 734.
349
In spite of this it seems worth while to state here the , attack and the reply, because the conception of the subject there expressed is one very near the truth, and which might easily hold under other circiimstances. An objection thus raised has a real scientific value of its own ; and when it is put in so clear and precise a light, as Hirn has done in this case by means of his skilfully-conceived operation, it can only be advantageous for science since the fact that the apparent objection is defined and placed clearly in view will greatly facilitate the clearing up of the point. In this way we shall attain the advantage that a difficulty, which otherwise might probably lead to many misunderstandings, and necessitate repeated and long discussions, will be disposed of once and for ever. In^ thus referring once more to this question, the author is far from wishing to make the objec-. tion a ground of complaint against Hirn, but rather believes that this objection has increased the debt which the Mechanical Theory of Heat owes to him on other accounts. The operation alluded to, on which Hirn has based his and observations, is as follows : Let there be two cylinders (Fig. 32) of equal area, which are connected at the bottom by a comparatively narrow pipe. In each of these let there be an air-tight piston; and let the piston-rods be fitted with teeth engaging on each side with the teeth of a spur wheel, so that if one piston descends the other must rise through the same distance. The whole space below the cylinders, including the connecting pipe, must thus remain invariable during the motion, because as the space diminishes in one cylinder it increases in the other by aa equal
:
amount.
First, let
to
top and let cylinder with a perfect gas of any be given density and of temperature t,. Now let the piston descend ia A, and that in
^ at the
filled
Fig. 32,
S50
rise in
is
into
is
B.
must pass
kept at a constant temperature t^, which is higher than i, so that the gas in passing is heated to temperature t^, and at that -temperature enters cylinder B. The walls of both cylinders, on the other hand, are non-conducting, so that within them the gas can neither receive nor give off heat, but can only receive heat from without as it passes through the pipe. To fix our ideas let the initial temperature of the gas be that of freezing, or 0, and that of the connecting pipe 100, the pipe being surrounded by the steam of boiling water. It is easy to see what will be the result of this operation. The first small quantity of gas which passes through the pipe will be heated from 0 to 100, and will expand by the cor^ responding amount, i. e. ^^ of its original volume. By this means the gas which remains in A will be somewhat compressed, and the pressure in both the cylinders somewhat raised. The next small quantity of gas which passes through the pipe will expand in the same way, and will thereby com^ press the gas in both cylinders. Similarly each successive portion of gas will act to compress still further not only the gas left in A, but also the gas which has previously expanded in B, so that the latter will continually tend to approach its This compression causes a heating of the gas initial density. and as all the gas which enters B enters in both cylinders at a temperature of 100, the subsequent temperature must rise above 100, and this rise must be the greater, the more the gas within B is subsequently compressed. Let us now consider the state of things at the end of the (pperation, when all the gas has passed from A into B. In the topmost layer, just under the piston, will be the gas which entered first, and which, as it has suffered the greatest subsequent compression, will be the hottest. The layers below will be successively less hot down to the lowest, which will have the same temperature, 100, which it attained in passing the pipe. For our present purpose there is no need to know the temperature of each separate layer, but only the mean temperature of the whole, which is equal to the temr perature that would exist if the temperatures in the different layers were equalized by a mixing up of the gas. This mean temperjiture will be about 120.
;
351
In a later article published in Cosmos, Him lias completed this operation, by supposing that the gas in is finally brought into contact with mercury at 0, and thereby cooled back again to 0; that it is then driven back from to under the same conditions as from to B, and is therefore heated in the same manner ; that it is then again cooled by mercuTy, again driven from to B, and so on. Thus we have a periodical operation, in which the gas is continually brought back to its original condition, and all the heat given off by the source of heat passes over to the mercury employed for cooling. Here we will not enter into this extension of the. process, but confine ourselves to the first simple operation, in which the gas is heated from 0" to a mean temperature of 120; for this operation comprises the essence of Hirn's ob-
jection.
In this operation it is clear that no heat is gained or lost both cylinders is always equal, and therefore both pistons are always pressed upwards with equal force. These forces are communicated to the wheel which and thus, neglecting friction, an gears with the piston-rods indefinitely small force will be sufficient to turn the wheel in one or the other direction, and thereby move one piston up and the other down. The excess of heat in the gas cannot therefore be created by external work. The process, as is easily seen, is as follows. Whilst a quantity of gas, which is a very small fraction of the whole, is heated and expanded in passing through the pipe, it must take in from the source sufficient heat to heat it at constant pressure. Of this, one part goes to increase the heat actually existing in the gas, and another part to do the work of expansion.
for the pressure in
;
'
But since the expansion of the gas within the pipe is followed by a compression of the gas within the cylinders, the same
quantity of heat will be generated in the one place as is absorbed in the other. Thus that part of the heat derived from the source, which is turned into work within the pipe, appears again as heat within the cylinders; and serves to above 0, its initial temperature, and heat the gas left in above 100, the temperature the gas which has passed into words to produce the rise in other in entered at which it temperature already mentioned. Accordingly, without considering the intermediate process, we may say that all the
352
at the beginning, comes from the source of heat Hence we have the sinattached to the connecting-pipe. gular result, that by means of a body whose temperature is 100, i.e. the steam surrounding the pipe, the gas within the cylinders is heated above 100, or, looking only to the mean temperature, to 120. Here then, is a contradiction of the
more than
fundamental principle that heat cannot of itself pass from a colder to a hotter body, inasmuch as the heat imparted by the steam to the gas has passed from a body at 100 to a
body at
120.
If the circumstance however has been forgotten. gas had had an initial temperature of 100 or more, and had then been raised to a still higher temperature by steam, whose temperature was only 100, this would no doubt be a contradiction of the fundamental principle. But this is not the real state of things. In order that the gas may be above 100 at the end of the operation, it must necessarily be below 100 at the beginning. In our example, in which the final temperature is 120, the initial is 0, The heat, which the steam has imparted to the gas, has therefore served in part to heat it from 0 to 100, and in part to raise it from 100 to 120. But the fundamental principle refers only to the temperatures of the bodies between which heat passes, as they are at the exact moment of the passage, and not as they are at any subsequent time. Accordingly we must conceive the passage of heat in this operation to take place as follows. The one part of the heat given off by the steam has passed into the gas, whilst its temperature was still below 100, and has therefore passed into a colder body ; and only the other part, which has served to heat the gas beyond 100, has passed If then we compare this with the fundainto a hotter body. mental principle, which says that, when heat passes from a colder to a hotter body, without any transformation of work into heat or any change in molecular arrangement, then of necessity there must be in the same operation a passage of heat from a hotter to a colder body, we easily see that there is complete agreement between them. The peculiarity in Him's operation is only this, that there are not two different bodies concerned, of which one is colder and the other hotter than the source of heat ; but one and the same body^
One
353
the gas, takes in one part of the operation the place of the colder, and in another part that of the hotter body. This involves no departure from the principle, but is only one special case out of the many which may occur. Dupre has raised similar objections against the fundamental principle but as there is nothing in them essentially new, they will not here be entered upon.
;
11.
Wands
Objections.
years later Th. Wand treated of the same principle in a paper entitled "Kritische Darstellung des zweiten Satzes der Mechanische Warmetheorie*." He gives his conclusions in the three propositions following: (1) "The second principle of the mechanical theory of heat, i.e. the impossibility of a passage of heat to a higher temperature without a conversion into work or a corresponding passage of heat to a lower temperature, is false." (2) "The deductions from this principle are only approximate empirical truths, which hold only so far as they are established by experiment." (3) "For technical calculations the principle may be taken as correct, since experiments on the substances used for the generation of work and of cold shew a very close agreement with it." The placing of such propositions side by side seem.i in itself a doubtful measure. If a principle has been found to agree with fact in so many cases, as to compel us to say that
Some
may be taken as correct, it seems dangerous to conclude nevertheless that it is false, in the face of the probable supposition that the apparent objections which yet remain would be cleared up by closer examination. The following appear to be the chief grounds on which Wand bases his rejection of the principle; excluding those
for technical calculations it
electrical phenomena, because these subjects are not here treated of. " If we suppose," he says on p. 314, " that in the bringing of a certain quantity of heat from a lower to a higher
temperature a certain quantity. of work must of necessity be destroyed, it follows that, if the same quantity of heat falls from a higher to a lower temperature, the same quantity of work must re-appear. Now suppose that this fall takes
Karl's Repertorium der Bxjtenmental-Phynk, Yol.
c.
iv.
23
354
by a non-reversible cyclical not true, because the falling of heat by conduction goes on without any other change whatTherefore for equalizations of temperature by simple ever. conduction there is nothing equivalent to the second prinand this from the logical point of view is one of the ciple weakest points in that principle, and leads to much subsequent inconvenience." The circumstance here mentioned, that compensation is required only in the passage of heat to a higher temperature, and not to a lower, has been frequently stated above and in Chapter X. is expressed in the general form, that negative transformations cannot take place without positive, but that positive transformations can take place without negative. From this circumstance the second main principle becomes doubtless less simple in form than the first, but it would be hard to shew that it is logically implace by simple conduction or
process.
is
; ;
perfect.
The inconveniences mentioned by Wand iri the above paragraph he arrives at by the following considerations. He supposes a simple cyclical process to be carried out, during which the two bodies between which the heat passes, and which he calls the heating and cooling body, have temperatures which are close to 0, and difier from each other by an indefinitely small quantity, which he calls dt. For this symbol, which will appear with another signification in the analysis which follows, we will substitute h, and will call the temperatures of the two bodies, reckoned from freezing-point, and S respectively. Further, Wand supposes the cyclical process to be so arranged, that one unit of heat passes over from the hotter to the colder body, and therefore that the
quantity of heat
:
The process being ended, I will heat the whole apceeds paratus, comprising both the heating and the cooling body, by 100. The difference of temperature between the two bodies will remain unaltered. If we now wish to destroy the
"
heat thus obtained by means of the reverse cyclical process, the colder body the heat f^f. The colder body thus loses the heat ^^, and giv.es it up to the hotter body and if by the reverse process all is cooled back again to the initial temperature 0", the initial condition of things
355
There is no work performed or consumed, and yet a passage of heat has taken place from the second body, which has remained the colder, to the first which has remained the hotter throughout the process. This however is no refutation of the second principle. For to obtain this result there must be a continuous succession of alternate heatings and coolings of the apparatus, i.e. heat must pass from a hotter to a colder body; but this passage takes place by conduction, for which there is no equivalent. Hence it follows from the process here described, that with regard to the distribution of the heat it is by no means the same thing, whether we do nothing at all, or carry out a compound cyclical process as here described." have here the case of two opposite cyclical processes carried out at different temperatures, in which the work done and the work consumed cancel each other, but more heat passes from the hotter to the colder body than vice vers4 ; and "Wand holds that the passage of the surplus heat from the colder to the hotter body has taken place without compensation. He has however neglected certain differences of temperature, which occur in this somewhat complicated operation. For after the first process, in which the hotter body has given off and the colder body taken in heat, he heats the whole apparatus and the two bodies by 100; and he cools them by 100 after the second process, in which the colder body has given off heat and the hotter taken it in. But in giving off and taking in heat the two bodies alter their temperature somewhat, and the reservoirs of heat, which perform the heating and cooling, do not therefore take back the heat during the cooling at the same temperaHence arise ture as they gave it out during the heating. passages of heat of which Wand has taken no account. These differences of temperature are of course very small, since the two bodies must be assumed so large, that the variations of temperature produced in them by the cyclical process may remain small compared to the difference of the original temperatures. But then the quantities of heat, which the bodies take from and give back to the reservoirs during their heating and cooling, are also very large; and since to determine the heat which has passed we must multiply the differences of temperature by the actual
We
233
356
quantities of heat, we arrive at magnitudes wliicli are quite large enough to compensate for the surplus heat which has passed between the bodies. To prove this point we will make the calculation itself.
with regard to the actual surplus heat which has passed and B, and between the bodies, since the temperatures are
First,
2^g ( 273 + 8
first
" 273)
^'
with regard to
^=^ S. We
have now
to determine the equivalence value of the passages of heat, which take place during the heating and coohng of the body by 100. By Chapter IV., 5, we must divide the element of heat taken in by one of the two bodies from a reservoir of heat (reckoning heat given off as negative heat taken in)
by the absolute temperature which the body has at the moment, and we must then form the negative integral for the heating and cooling.
be the mass of each body, and C its specific heat, Let which we suppose constant then the quantity of heat, which it takes in during a rise of temperature dt, equals MCdt, and this we shall take as expressing the element of If for convenience we put heat.
;
^~MG
the element will be expressed by - dt.
e
(6).
We
must consider
MG
as very large, and therefore e as very small, so much so as to be small even in comparison with the small difiference of temperature h. If we now take the first cyclical process, the colder body has, to begin with, the temperature 0, and the hotter the temperature S. During the process the former takes in the
s
quantity of heat
1,
and the
+ ^=^
to obtain
357
the changes of temperature which they produce in the bodies ; thus at the end of the process 'wie colder body has the temperature e, and the hotter the temperature
From
by
now
to
be heated
100.
The
body
is
^"~i.
If
273+T
1
we put T = t e, then
/lOO .100
A
1
h K
H T + e* 273 +
e,
we have
273
+ T + e~273+T
(273+t)'"
^'^-
whence
^ = --J^
273T-r+Jo
(273T^^
dt
f"
.(,,i-).
273
+r
Whence we
way
as before
cZt
.
_ ^
1 /""+
dr
B Ui+
7L
273T^'"l^"^273JJs
373 gs^ and the
,
(273
+ Tf
^''^-
body gives
a quantity of heat =
body receives a
358
quantity ^r^^
+ ^f^
The temperatures
become respectively
100 +
e-||. = 100-|-e,
From
100.
The negative
f
1""
Ann
lOI
1
^
~\
whence
as before
^
100
100
'"o-are"
273+^
'"are'
273+i 273+
we have
^~eJo
273
+ T + 273J0
(273
+ t)^
^''^
^~ejs
By
value of
^^"^
adding together A, B, G, D, we obtain the equivalence all the transferences of heat during the heating and cooling. In this addition the integrals which have the factor - cancel each other, and two of the others
together.
may be combined
Whence we have
373
/I""
A+JB + G + I)--
dr
(273
''27BJo
+ T)''
(273
V273^273/'J8
+ t)'
^^^>
359
1_\
"*"
/373
"^
_S_\
27ii)
V273
273/ [
373
+ S "^ 273 + S.
and
in terms of S to the
the expression for the equivalence value of the transferences of heat during the heating and cooling becomes finally
100
(273)'
s *
This expression fulfils the condition of being equal and opposite to the equivalence value of the surplus heat actually transferred between the two bodies. This transfer therefore is not uncompensated, but fully compensated as the second principle- requires. thus see that the operation suggested by Wand does not give the smallest, ground for objecting to the principle. Another objection is drawn by Wand from the following considerations. He proposes the question whether the principle can be derived by mechanical reasoning from the ideas which we can form of the nature and action of heat. With this object he first applies the hypothesis maintained by the author and others as to the molecular motion of gaseous bodies, and finds that this does actually lead He then says that it ,is to the principle in question. not sufficient to prove that one particular hypothesis leads to the principle, bat that all possible mechanical hypotheses on the nature of heat must be shewn to lead to it. Accordingly as a second example he takes another hypothesis, which he conceives specially adapted to represent the phenomena of expansion, and of the increase of pressure by On this hypothesis a row of elastic balls, any two of heat. which are connected by an elastic spring, vibrate in such a way that all of them are in similar phases. From this hypothesis he arrives at an equation different from that which he has chosen as the criterion of the second principle, and
We
then draws the conclusion, "the second principle cannot therefore be universally derived from the principles of mechanics." But upon such a conclusion the question may
S60
be asked, -whether in reality the hypothetical motion, which he supposes, agrees with the actual motion, which we call heat, in such a way that the same equations must hold for both. So long as this is not proved the conclusion cannot be
considered to be
Finally,
made
out.
considers the process which Occurs in nature, when in the growth of plants, under the influence of the sun's rays of light and heat, carbonic acid and water are absorbed, and oxygen liberated whilst the organic substances thus formed, if afterwards burnt or serving as nourishment to animals, unite themselves again with oxygen to form carbonic acid and water, and thereby generate heat. This transformation of the sun's heat he considers to be in flat contradiction to the second principle. To the analysis which he gives many objections might be taken; but the author considers that a process in which so much is still unknown, as that of the growth of plants under the influence of the sun, is altogether unfit to be used as a proof either for or against the principle in question.
;
Wand
12.
Tait's Objections.
Finally the author has to mention certain objections recently raised against his theory by Tait ; objections which have surprised him equally by their substance and by their form. In an article which appeared in 1872 on the History of the Mechanical Theory of Heat*, the author had observed, that Tait's work, sketch of Thermodynamics, no doubt owed its existence chiefly to the wish of claiming the Mechanical Theory of Heat as far as possible for the English nation ;-^a supposition for which the clearest grounds can be adduced. Further on in the article he had observed that Tait had ascribed to Sir William Thomson a formula due to the author, and had quoted it as given by Thomson in & paper which contained neither the formula itself nor anything equivalent to it. The author expected that Mr Tait, in answering this article, would specially address himself to these two points, of which the latter particularly required clearing up. reply aj)peared indeed -f, and one written in
p. 132.
4, Vol.
xlui.
361
two points were nowljere touched upon, a different turn being given to the whole matter. For whilst in the work previously alluded to the, author's researches on the Mechanical Theory of Heat, even if in his own view they were not put in their right relation to those of English writers, were, yet described at great length and with a general recognition of their merit, their correctness was here at once assailed, by
the declaration that the fundamental principle, that heat cannot of itself pass from a colder to a hotter body, is
untrue.
To prove
are adduced.
this
electric currents
published shortly afterwards * the author was easily able to prove that these phenomena in no way contradict the principle, and that one of them is even so evidently in accordance with it, that it may serve as an example specially adapted to illustrate and establish it. As electrical phenomena are not here treated of, this is not the place to enter further into this subject. Tait further observed in his reply, that by the author's introduction of what he namdd Internal Work and Disgregation, he had done a serious injury to science, offering however nothing to support this, beyond the brief remark " In our present ignorance of the nature of matter such ideas can do only harm." What Tait has to object to the conception of In his first internal work, it is difficult to understand. paper on the Mechanical Theory of Heat the author divided the work during the body's change of condition into External and Internal Work, and shewed that those two quantities of work differed essentially in their mode of action. Since that time this distinction has been similarly made by all writers, so far as he is aware, who have treated of the Mechanical Theory of Heat. As regards the method (which will be described on a future occasion) of calculating the combined internal and external work, and the conception of disgregation introduced, by the author with this object, purely mechanical investigations have recently led to an equation, exactly corresponding with that which the author proposed in the science of heat, and in which the disgregation makes its appearance. If
But in a
36^
these investigations cannot yet be considered as complete, they nevertheless shew, at least in the author's opinion, that the nature of things requires the introduction of this conception.
He therefore leaves the objections of Tait, with as much confidence as those of Holtzmann, Decher, and others, to the good judgment of the reader *.
* For a rejoinder by Prof. Tait to these observations, see Sketch of Thermodynamics, 2nd edition, 1877, p. xv.
Finis.
APPENDIX
I.
SOLIDS.
who examined the thermoInstead of abstracting his investigation {Quarterii/ Mathematical Jmirnal, 1855) it may be well to present the subject as an illustration of the method of treatment by the Adiabatic Function. Consider any homogeneously-strained elastic solid. To define the state of the body as to strain six quantities must be specified, say 11, V, w, X, y, z: these are generally the extensions along three rectangular axes, and the shearing strains about them, each relative to a defined standard temperature and a state when the body is free from stress. The work done by external forces when the strains change by small variations may always be expressed in the form
William Thomson
-was the first
solids.
IJJhi+7bv+
:
...)x
volume of the
solid,
U, V are because the conditions of strain are homogeneous. and of the the stresses in the solid each is a function oi uv temperature, and is determined when these are known.. Let 6 denote the temperature (where 6 is to be regarded merely as the name of a temperature, and the question of how temperatures are to be measured is not prejudged). Amongst other conditions under which the strains of the body may be varied, there are two which we must consider. First, suppose that the temperature is maintained constant; or Then 6 is constant. that the change is efiected isothermally. Secondly, suppose that' the variation is efiected under such conditions that no heat is allowed to pass into or to leave the body ; or that the change is efiected adidbatically. In the latter' are connected by a relation involving a paracase 6,u,v, meter which is always constant when heat does not pass into
364
have now fourteen quantities relating to the body, viz. which defines the temperature, and the parameter ^ the constancy of which
six elements of strain, six of stress, the quantity 6
We
Any
seven of these
may
be
now undergo
1. Let the stresses and strains vary slightly under the sole condition that the temperature does not change. Let the consequent increase of <^ be 8i^. Heat will be absorbed or given out, and, since the variations are small, the quantity will be propor-
tional to
8(^,
say
f{e,u, v,w,
'2.
.'.....) S<f}.
and
let S6
3.
Let the stresses and strains further vary adiabatically, be the consequent increase of temperature,
Let the stresses and strains receive any isothermal variaparameter ^ returns to its first value. Heat will be given out or absorbed, equal to
tion, such that the
(/+8/)8<^.
4.
of
Here we have a complete and reversible cycle. The quantity heat given off SfxS^ is equal to the work done by external
forces.
Now
Carnot's
theorem
(or
the
Second Principle of
Thermodynamics)
or Sfx 8^, divided transferred from the lower to the higher temperaX 8<^, is equal to a function of 6 only (which function is ture, the same for all bodies) multiplied by 89. Thus
work d6ne,
by the heat
^=-^(5)8(9,
log (/)
= 1^0 de + a
function of
<^
the function of <j> being added because the variation was performed under the condition that ^ was constant. By properly choosing the parameter <j) this function may be included in 8^, and we have, as the quantity of heat absorbed in the first
operation,
f
bqj.
365
The mode of measuring temperature being arbitrary, we shall find it convenient to define that temperature is so measured that
F (0) =
-^-j
then
we have
Heat absorbed
in
first
operation
forces
6S(f>
(1);
Si^
= 86 x
(2).
respectively
and the sum of these is equal to S(^Sd. Hence a variety of important relations may be obtained. Let all the strains but one be constant then we have
:
M,=
366
__
,. ,. jj^^^de^u, + ^d6
dll
dU
dU J. vi__
dU^,
,_ae.
Adding
the differentiations being performed -when u and Zf are expressed 6, <j) and the five other strains. The same is true if five stresses are constant, that is if u and and the five other stresses. are expressed as functions of 6, But from (2) the "Work done = d6x d<^. Hence it follows generally (using the well-known theorem as to Jacobians) that
<;!>
dtf)
dO
d<j>
dUd^'d^dU'
<j)
d9 _^
,.
^
''
and being expressed as functions of u, U; and either the five other stresses or the five other strains being constant.
These equations are still true if the independent variables are partly stresses and partly strains, so long as no two are of the same name e.g. if they are vwXYZ. From equation (3) all the thermo-elastio properties of bodies may be deduced, have generally
:
We
^'=fu'^^fu'^
W'
(^)-
''^-'P^-fu'^
867
= 0, we have
dU
^ (when
.,
<j,
IS
coDBtant)
=-
^
dU
J,,
Putting de = 0, we have
-j~ (when 6
OM
is
constant)
-^ 'do
=
dU_
Let
that
is,
denote
is
under the
condition that
<^
is
constant,
expressed as a function of
Au
instead of
Uu
d^^dO^dU du dU du
de^_dO_
du
dU
^
d<^
d4
d6
du
dlf
dU
This is the fourth thermodynamic relation (see Maxwell on Heat, 1877, p. 169). The others are obtained in a similar way thus
:
d^
d^_de__de^ dU
djfa
'
dU~ dU
du
d<f>
d^~
du
d4>
du
de_
de4>_d<l>
d^ du
duV
~du^du~dU dF~~'de^~W dU dU
d$
1 _ de<t> _ d<f d<f> dU _ dU~ dU~ du dO^ d^"
djju
'df'
du
du
368
Take the
stant
When
= 6d<j>.
6 is con-
we have bj
(or dq)
Hence
APPENDIX
II.
ON CAPILLARITY.
The equations obtained in Appendix I. may also be applied to the eqailibiium of the surface film of a liquid in contact with its owu vapour. This subject is generally known under the name of Capillarity, from its having first been studied in connection with Capillary tubes (Maxwell, On Heat, 1871, p. 263). Thus let the body considered be the film of fluid at the surface of a fluid, and let it be so small in volume that its capacity of heat in virtue of its volume may be neglected. Let T be the surface tension, and Then 2* is a function of 6 only hence, from an <S^ the surface. equation analogous to Equation (3), Appendix L, we obtain
:
dT
d6
*'''
""
dS__ d^"
'
(by Equatiion
'
Appendix
negative,
I.),
where q
is
heat absorbed.
dd surface means absorption of heat. The question of the equilibrium of vapour at a curved surThe face of liquid has been treated by Sir Wm. Thomson. following is mainly taken from his paper {Froc. Royal Society of
-Edinburgh, 1870, Vol.
vii., p. 63).
If then
dT
-j^
is
as- is
closed vessel containing only liquid and its vapour, all at one temperature, the liq'iiid' restsj with its free surface raised or depressed in capillary tubes and in the neighbourhood of the solid boundary, in permanent equilibrium according to the same law of relation- between curvature and pressure as in vessels open to the air. The permanence of this equilibrium implies physical equilibrium between' the liquid and the vapour in contact with it at
In a
But the pressure of the vapour at parts of its surface. Hence the difierent levels difiers according to hydrostatic law. pressure of saturated vapour in contact with liquid differs
all
.
c.
24
370
according to the curvature of the bounding surface, being less when the liquid is concave, and greater when it is convex. And detached portions of the liquid iu separate vessels, all enclosed in one containing vessel, cannot remain permanently with their free surfaces in any other relative positions than those they would occupy if there were hydrostatic communication of pressure between the portions of liquid in the several vessels. There must be evaporation from those surfaces which are too high, and condensation into the liquid at those surfaces which are too low process which goes on until hydrostratic equilibrium, as if with free communication of pressure from vegsel to vessel, is attained. Thus, for example, if there are two large open vessels of water, one considerably abpve the other in level, and if the temperature of the surrounding matter is kept rigorously constant, the liquid in the higher vessel will gradually evaporate until it is all gone and condensed into the lower vessel. Or yre may suppose a capillary tube, with a small quantity of liquid occupying it from its bottom up to a certain level, to be placed upright in the middle of a quantity of the same liquid with a wide free surface, and contained in a hermetically sealed exhausted receiver. The vapour wUl gradually become condensed into the liquid in the capillary tube, until the level of the liquid in it is the same as it would be were the lower end of the tube in hydrostatic communication with the large mass of liquid. The effect would be that in a very short time liquid would visibly rise in the capillary tube, and that, provided care were taken to maintain, the equality of temperature all over the surface of the hermetically sealed vessel, the liquid in the capillary tube would soon take very nearly the same level as it would have were its lower end open ; sinking to this level if the capillary tube were in the beginning filled too full, or rising to it if there is not enough of the liquid in it at first to fulfil ''the condition of equilibrium. The following shews precisely the relations between curvatures, differences of level, and differences of pressure with which _ we are concerned. Let jo be the pressure of equilibrium above the curved surface in the tube, rr' the principal radii of curvature of that surface, and T the surface tension then by the principles of Capillarity
is
- + -,
liquid liquid,
ON CAPILLARITY.
Let V be the equilibrium pressure of
371
th.e vapour at the plane surface of the liquid, p the density of the liquid, tr the density of the vapour, h the height at which the liquid stands in the tube above the plane surface then clearly
:
-h^^^y
But
also
IT
+h
(1).
=p + ha(p
(2).
Therefore
TT (p
- 0-) =^
- <r) + r ^- +
-,)
0-,
which gives the relation between the pressure of equilibrium' above the curved surface and that above the plane surface. In strictness the value of a- to be uSed' ought to be the mean density of a vertical column of vapour extending through the given height. But in all cases in which we can practically apply the formulae, according to our present knowledge of the properties of matter, the. difference of densities in this colunm. is very Hence, if .S denote the height smaiU, and njiay be' neglected. above the plaij.e surfaee of the liquid of an imaginary homogeneous
fluid whic|i, if of
would prgduce by
its
the same density as the vapour at that plane, weight the actual pressure jr, we have
j^t^t
by Equations
(1)
and
(2)
ji^ence
by Equation
(3)
/IT
ah = Tr(l -jz.)
is about For vapour at ordinary atmospheric temperatures, 1,300,000 centimetres. Hence in the capillary tube which would keep water up to a height of 3 metres above the plane level, the curved surface of the water is jn equilibrium with the vapour in contact with it, when the pressure of the vapour is less by about j^^"' of its own amount than the pressure of vapour in equilibrium at a plane surfaee pf water ai the same temperature.
J.
24-2
APPENDIX
III.
373
the boundary of
all
DBEGF
Kg.
1.
corresponding to 30'''92, at which their length vanishes. In otherwords, at temperatures above 30''"92 the " gas line " AB, and the
37^!
"liquid line"
portion.
CD
meet each
other,
In the case of steam, for which such isothermal lines were first drawn, similar phenomena occur; and when it was seen that the gas line and the Hquid Une thus continually approached each other as the temperature was raised, the question naturally arose, Do they ever meet? If they do meet, then at that temperature the substance cannot exist partly as a liquid and partly as a vapour, but must be entirely converted, at the corresponding pressure, from the state of liquid to that of vapour 3 or else, ?iiice in this case the vapour and liquid have lihe same density, it may be suspected that the distinction between liquid and vapour has here lost its meaning. The answer to this question, has l^een to a great extent supplied by a series of very interesting researches. In 1822 M. Cagniard de la Tour ' observed the effect of a high temperature upon liquids enclosed in glass tubes of a capacity not much greater than that of the liquid itself. He found that when the temperature was raised to a certain point, the. substance, which till then was partly Ijquid and partly gaseous, suddenly became uniform in appearance throughput, without any visible surface of separation, or any evidence that the substance in the tube was partly in one state a:^d partly ii). another. He concluded that at this temperature the whole became gaseous. The true conclusion, as Dr Ai?drews has shewn, is, that the properties of the liquid a^d those pf the vapour continually approach to similarity, and thafc above * certain temperature, the properties of the liquid are not separated from those of the vapour by any apparent distinction between^ them. In 1823, the year following the researches of Cagniard de la Tour, Faraday succeeded in liquefying several bodies hitherto known only in t^e gaseous form, by p-essure alog-e, and in 1826 he ^eatly extended our knowledge of the effects of temperature He considers that above a certain tempetlflid pressure on gases. r^jtsire, which, ifl. the language of Dr Andrews, "sre may call the (jrijtical temperature for the substance, po amount of pressure will ppodiice the phenomenon which we p^U condensation, and he supposes that the temperature of 166 P. below zero is probably above the critical temperature for oxygen, hydrogen, and nitrogen. Dr Andrews has examined carbonic acid under varied conditions of temperature and pressure, in order to ascertain the relations of the liquid and gaseous states, and has arrived at the conclusion that the gaseous and liquid states are only widely' Annalei de Chimie, 2nd Series, Vols. xxi. and xxii.
875
separated forms of the same condition of matter, and may be made to pass one into the other "without any interruption or breach of continuity '. The diagram, Kg. 1, for carbonic acid is taken from Dr Andrews' paper, "with the exception of the dotted line shewing the region within which the substance can exist as a liquid in presence of its vapour. The base line of the diagram corresponds, not to zero pressure, but to a pressure of 47 atmospheres. The lowest of the isothermal lines is that of 13'l C. or 55-6 r: This line shews that at a pressure of about 47 atmospheres condensation occurs. The substance is seen to become separated into two distinct portions, the upper portion being in the state of vapour or gas, and the lower in the state of liquid. The upper surface of the liquid can be distinctly seen, and where this surface is close to the sides of the glass containing the substance it is seen to be curved, as the surface of water is in small tubes. As the volume is diminished, more of the substance is liquefied, till at last the whole is compressed into the liquid form. Liquid carbonic acid, as was first observed by ThUorier, dilates as the temperature rises to a greater degree than even a gas, and, as Dr Andrews has shewn, it yields to pressure much more than any ordinary liquid. Prom Dr Andrews' experiments it also appears that its compressibility diminishes as the pressure inThese results are apparent even in the diagram. It is, creases. therefore, far more compressible than any ordinary liquid, and it appears from the experiments of Andrews that its compressibility diminishes as the volume is reduced. It appears, therefore, that the behaviour of liquid carbonic acid under the action of heat and pressure is vet-y different from that of ordinary liquids, and ia some respects approaches to that
of a gas. If we examine the next of the isothermals of the diagram, that for 21'5 C. or 70*7 F., the approximation between the Here conliquid and the gaseous states is still more apparent. densation takes Talsice at about 60 atmospheres of pressure, and the liquid occupiel nearly a third of the volume of the gas. The
dense gas
is
approaching in
its properties to
the
Still there is a distinct separation light liquid. gaseous and liquid states, though we are approaching
This critical temperature has been At this to be 30-92 C. or 87''-7 F. temperature, and at a pressure of from 73 to 75 atmospheres,
temperature.
determined by
Dr Andrews
1
376
No sepacarbonic acid appears to be in the critical condition. ration into liquid and vapour can be detected, but at tbe.same time very small variations gf pressure or of temperature produce such great variations of density that flickering movements are observed in' the tube' "reseflibling in an exaggerated form the appearances exhibited' during the" mixture of liquids of different densities, or when columns of heated' air ascend through colder
strata."
Tke isothermal line for' sr'l C. or 88 F. passes above this During the whole compression the substance is never in two distinct conditions in diiSerent parts' of the tube.
critical point.
than 73 atmospheres the isothermal than that of a perfect gas, resembles it in- general features. From 73 to 75 atmtisphe^res the' vblume diminishes very rapidly, but by no means suddenly, and above this pressure the volume diminishes more gradually than in the case of a perfect gas, but still more rapidly than in most liquids. In the isothermals for 32-5 C. or 90-5 F. and for SB'-S C. or 9 5 '9 F. we can still observe a slight increase of compressibility liear the same part of the diagram, but in the isothermal line for 48"1 C. or 118'6 F. the curve is concave Upw'ards throughout its rt'hole course, and differs from the corresponding isothermal line for a pferfect gas only by being somewhat flatter, shewing thalt for all ordinary' pressdVes the- volilme is somewhat less than that
the pressure
is less
line;-
When
though greatly
flatter
assigned-
by Boyle's
law.
StUlat the temperatui'e of ri8'6 F. carbonic acid has all the firoperties of a gas, and the effects of heat and pressure on it differ from their effects on a perfect gas oiify by quantities reqiiiring
careful experiments to detect them. have no reason to believe that any^ phenomen'on similar to condensation would occur, however great" a pressure were applied to carbonic acid at this temperature.
We
CAllUllIDGK
miKlED BY
C.
J.
CLAY, M.A., AT
Macmillan ^ Co.'s Catalogue of Works in Mathematics and Physical Science; incltiding Pure and Applied Mathematics; Physic's,, Astronomy, Geqlogt^, Chemistry, Zoology, Botany; and of
MATHEMATICS.
Airy i Works hy
sities.
Sir G. B.
Use of Students
With Diagrams.
ON THE ALGEBRAICAL AND NUMERICAL THEORY O^ ERRORS OF OBSERVATIONS AND THE COMBINATION OF OBSERVATIONS.
6s. 6d.
Crown
for the
8vo.
of
designed
Use
ON Sound and
.
6s, 6d.
With
Designed for the Ijse of the Mathematical Elements of IVtusic. Students 'of the University. Second Edition; revised and enlarged. Crown 8vo. gj. Designed, for the!; Use of TREATISE
ON MAGNETISM.
Crown
8vo.
gs. 6d.
'
' '
A.M.)EXPERIMENTAL MECHANICS. A
Course of Lectures delivered at the Royal College of Scie;n,c?,for Ireland. By Robert Stawell Ball, A.M., Professor of Applied Mathematics and Mechanics in the. Royal College of Science for Ireland (Science and Art Department). Royal 8vo. 16s.
" We have
of the sort in English. It elucia teacher of the very highest We most cordially tecomviend, it to, (ill our readersJ' rank. Mechanics' Magazine.
not put with
any
bi>ok
5,000. 3.
79
SCIENTIFIC CATALOGUE.
By Joseph Bayma,
S.J.,
cjotiij;
Sto.nyhurst Cmlege.
Demy
8vo.
PhUosophy,
\\
i'.
'.{
v'Boole.Worts
.
by G. BooEe, D.C.L, F.R.S., Professor Mathematics in the Queen's University, Ireland !-^
>
of
Third.
14J.
Edited by I.
Todhuntek.
Crown 8vo.
cloth.
"yi
treatise
subject with
incomparably superior to anyothef^eletnentArj) book on the which we are acquainted" Philosophical Magazine.
A TREATISE ON DIFFERJENTIAL
mentary Volume.
8j-.
Edited by
I.
TodhunteR.
dd.
CrOwn
8vo.
ON THE
.
; :
By
"^
Clifford.THE
to the study of
Motion and Rest in Solid and Fluid Bodies. By W. K. Clifford, F.R.S., Professor of Applied Mathematics and Mechanics at University College, London. Part I. Kinematic.
Crown
8yo.
"Js.
6a.
,.
"^Cuxnilling. AN
With numerous Examples. By L'inn^us CuJiMlNG, M.A., Assistant Master at Rugby School. Crown 8vo.
ELEpTRICITY.
8j. 6d.
Clithbertson.EUCLIDIAN. GEOMETRY.
School,
'^
By
F.
Cuth,
.
4?. 6j/.;
Ferrers, M, A. F.R. S. ,]^ellow and Tutor of Gonville and Caius College, Cambridge
,
;
Method
SPHERICAL HARMONICS AND SUBJECTS CONNECTED WITH THEM. Crown 8vo. Js. 6d.-
MATHEMATld^.
^ Frost.
^Works
l?y
late
Fellbw of
St.
NEWTOiS^'S PRIN-
Also a Collection ,of Problems, principally intended as Examples of Newton's- Methods. Third Edition. 8vo. cloth. I2,s.
SOLID GEOMETRY.
of the Treatise
Being a
New
Editioh, revised
and enlarged,
I.
8vo.
i6j.
Godfray.
Works
-Mathematical
A TREATISE ON ASTRONOMY,
Schools.
'^
for the
8vo. cloth.
12s. 6cl.
It is
a /vorkmg book," says the puardian, ^'taking Astronomy It is'a book prober place injhe Mathematical Sciences. which is not likely to be got up wninielligeutly."
in
its
. . .
"
AN
ELEMENTARY
TREATISE
ON
THE
LUNAR
THEORY, with a
Brief Sketch of the Problem up to the time.flf Newton. Sepphd Edition, revised. Crovm 8vo. cloth. Jj. D(^. "As an elementary treatise and introduction to the subject, we think London, Edinit may justly claim to supersede all former ones." burgh, and Dublin Phil. Magazine.
the
Fellow, of Gonville and Caius Edited by N. M". Ferrers, M.A., Fellow College, Cambridge. I5r. 8vo. q.nd Tutor of Gonville and Caius College. "It has been for some time recognized, that Green's writings are amongst the most valuable ,pfafhematical productions we possess."
Athenseum.
Hemrning.AN ELEMENTARY TREATISE ON THE DIFFERENTIAL AND INTEGRAL CALCULUS. For the
Use of Colleges and
Fellow of
St.
By G. W. Hemming, M.A,, Schools. John's College, Cambridge.! Second Editieny with 9J. 8vo. cloth. Corrections and Additions.
An
Ele-
mentary Treatise in which the Cpnic Sections are defjiied as the Plane Sections of a Cone, and treated by the Method of Projections. By J.' StuarT Jackson, M.A., late Fellow of Gonville and Cams
College.
Crown
8vo.
\s. 6d.
SCIENTIFIC. CA TALQGUE.
an
introduction TQQUATERj
By p. Kbslland, M.A., ,,With numerous Examples. F.ii.S., and P. G. Tait, M. A., Professors in the department of Mathematics iii the University df Edinburgh. Crown' 8vo. 'js. 'dd.
Grpwn
8vo.
ij.e^/.
,-
'
By H:
"a;
Sadlerian and Mathematical Lecturer of Jesus Croym 8vo. cloth. 6s. 6d. College, Cambridge.
v^Newton's Principia.4to.
'
cVoth.
31s. 6^.
If is a
sufficient guarantee
of the
yyra/iftfw'j
Prilicipia that it has been printed for wtid under the care of Professor Sir William Thomson and Professor Slackbum, of
Glasgoiv University.
'<'
.
,
^Parkinson.
Cambridge.
cloth.
a treatise ON optics.
Fellow and Tutor of
St.
'
By
S.
Parkin-
Phear.ELEMENTARY HYDROSTATICS.
,
With Numerous
,
|Exan?ples. By J. B,jPhear, M.A., Fellow and late Assistant Tutor ofiClare Coll. Cambridge. Fourth Edition. Cr. 8vo. cloth. 5^. to.
Pirrie.LESSONS
Crown
8vo.
bs.
rigid dVnAMICS. By the Rev. G. PiRRiE, M.A., Fellow and Tutor of Queen's College, Cambridge.
ON
Puckle.AN ELEMENTARY TREATISE ON CONIC SECTIONS AND ALGEBRAIC GEOMETRY. With numerous
.Exij.inples
M.A.
and Hints for their Solution. By G. Hale Puckle, Fouth Edition, enlarged. Crown 8vo. "js. 6d.
.
V Rayleigh.
OF SOUND. By Lord Rayljeigh, F.R.S., formerly Fellow of Trinity College, Cambridge. 8vo. [Vol. III. in preparation. Vol. 1. izs. dd. Vol. II. izs. 6d.
;
THE THEORY
Out-
By Professor F. Reuleaux. of a Theory of Machines, Translated' aiid edited by A. B. W. KeNnedy, C.fi., Professor of University College, London. Engineering, Civil and Mechanical
With 450
Illustrations.
Royal 8yo.-
20J.'
MA THEMA TICS.
Routh.;
Works by Edward John Routh, M.A., F,R.S., late Fellow and Assistant Tutor of St. Peter's College, Cambridge; Examiner in the University of London
:
AN ELEMENTARY TREATISE ON THE DYNAMICS OF THE SYSTEM OF RIGID BODIES. With numerous
Examples.
8vo.
21s.
Stability of a given state of motion, particularly STEADY MOTION. The Adams' Prize Essay for
1877.
'
Svo.
8j:
6d.
DYNAMICS OF A
Crown
8vo.
lis.
PAI^TICLE.
With
Thomson.
netism.
Svo.
i&r.
Sir
William Thomson,
M,A.,
F.R.S.
Todhunter
,
Works by
is chiefly
I.
Todhuuter,
:
F.R.S.,
of
Cambridge
"Mr. Todhunter
author of a
known to students of mathematics as the of admirable mathematical text-books, which possess the rare qualities of being clear in style and absolutely free from, mistakes, tyfografhical or other'}' S^-today Review. nV
series
New
New
Crown
Svo. cloth,
e^. 6d.
PLANE CO-ORDINATE'GEOMETRY,
Line and the Conic Sections/
Edition.
Crown
Svo.
^s. 6d.
TREATISE ON THE DIFFERENTIAL CALCULUS. New Edition. Crown Svo. loj. 6d. vh TREATISE ON THE INTEGRAL CALCULUS AND ITS APPLICATIONS. With numerous Examples. New Edition,
vA.
Crown
Svo. cloth.
loj. 6d.
Edition, revised.
Crown
Svo. cloth.
4J.
vA TREATISE ON ANALYTICAL
New
lOJ. 6d.
STATICS.
vliESEARCHES IN THE CALCULUS OF VARIATIONS, Principally on the Theory of Discontinuous Solutions An Essay
;
which the Adams' Prize was awarded in the University of Cambridge in 1871. Svo. 6s.
to
SCIENTIFIC^ CA TALOGUE.
Todhunter
:
cpntmued.
,.,..r
>i/AHrSTORY; OF
'
THE MATHEMATICAL
in
THEORIES'" OP
ATTRACTION,
Newton
and the Figure of the Earth, from the time of '24r. to that of Laplace. Two vols. 8vo.
no
" Probably
man
England ii
'
theme as Mr.'
Tblthkliiter.
'
To
and
so qualified to do justice to the all mathematicians these volumes to all succeeding investigators^ of the
' '
AND
W.
BESSEL'S FUNCTIONS.
Crown
8vo.
lor. '6^.
V. Wilson, M. A., Fellow of St. John's College, Cambridge, and Professor of Mathematics in Queen's Collegfe, Belfast. 8vo.
gj. 6d.
J';'
Wolstenhblme.
SubDivisions Of the Schedule jects included in the First of Subjects for the Cambridge Mathematical Tripos Examination. Devised and arranged by Joseph Wolstenholme, late Fellow of Christ's College, sometime' Fellow of St. John's College, and Professor of Mathematics in the Royal Indian Engineering College.
oii
Svo.
iSj.
Young
a Supplementary Essay on Economy in suspenBy E. W. Young, Associate of King's College, London, and Member pi the Institution of Civil Engineers. Svo.
PHYSICAL SCIENCE.
PHYSICAL SCIENCE.
Airy (G. B.)popular astronomy.
By
Sir G. B. iSipo. cloth.
4P. 6a?.
With
illustrations.
Ifdition.
'
New
Bastian.
&c.
:
Works, by
Origin,
F,R.S.,
Professor of Pathological
Anatomy
:
in University College,
London,
Being some Account of the Nature,. and Traiisfornnations of Lower Organisms, In Two Volufties. With upwards of lOO Illustrations. Crown 8vo. 28.r. "/if is a book that cannot be ignored, and must inevitably lead lo
renewed discussions and repeated observations, and through the establishment of truth." A. R. Wallace in Nature,
these to
LIFE.
Crown
of
8vo.
in information oj interest
to
the student
biological
'
-Daily Nev/s.
,
F.
Heavens."
By John
^.
Blake.
Crown
8vo.
Blanford (H. F.) RUDIMENTS OF PHYSICAL QEOGRAPHY FOR THE USE OF INDIAN SCHOOLS. By
With numerous Illustrations and H. F. Blanford, F.G.S. Globe Glossary of Technical Terms employed. New Edition.
8vo.
2s. 6d.
Blanford ("W.
ABYSSINIA.
an Enharmonic Harmonium exhibited in the Loan Collection of South Kensington, 1876; also of an Enharmonic Organ exhibited to the Musical Association of London, May, 1875. By R. H. Bosanquet, Fellow of St. John's College,
Oxford.
8vo.
6s.
Codl
' '
MiALL, -Thorpe, Ruckek, and Marshall, of the Yorkshire With Illustrations. 8yo. 12s. 6d. College, Leeds. whole subject of Coal Itfurnishes a very com^ehMsive treatise on the
.
,
ITS
HISTORY AND
ITS USES.
By
Professors GREfek,
'from the gealogicttlpchemical,- mechanical, andindMstrig4 points 'of as viewt .co^cltiMn-g it/itfia chapter on the important topic known
the
'
SCIENTIFIC CATALOGUE.
Cooke (Josiah P., Jun.) FIRST PRINCIPLES OF CHEMICAL PHILOSOPPIY. By Josiah P. Cooke, Jun.,
Ervine Professor of Chemistry and Mineralogy in Harvard College.
Crown
8vo.
I2J-.
"-'ly
subject
of which
it treats,
for
many years
:
to come.
"
Stand,ar4..
Dawkins, CAVE-HUNTING
Researches on the Evijience of Caves respecting the Early Inhabitants of Europe. By W. Boyd Dawkins, F.R.S., &c.. Professor of Geology at Owens College, Manchester. With Coloured Plate and WCodcuts. 8vo.< zu. " The'mass of infdrmatioH he has brought together, with the judicious use he has made of his materials, will be found to' invest his book with much of new and singular value." Saturday Review.
Dawson
GEOLOGY. The Geologic (J. Organic Remains, and Mineral Resources of; Nova New Branswick, and Prince Edward Island. By John William Dawson, M:A.', LL^.D., F.R.S., F.G.S., Principal and Vice-Chancellor of M'Gill College and University, Montreal, &c. ^ith a Geological Map and numerous Illustrations, Third Edition, with Supplement. ;8vo. 2 1 j. Supplement, separately,- za'. 6i^.
Structure, Scotia,
i
W.)ACADIAN
" The
a place, in. the library, not 'dnly of the scientific geologist, but fUsp- of all who are desirous of the industrial progress and commercial prosperity of the Acadian i>rovinces." Mining Journal. .. .,, , ,
book
tuill doubtless
fnd
Fleischer.
M.'
by
Illustrated;
M.
By George Forbes,
B.A,, Professor of Natural Philosophy in the Andersonian University of Glasgow. With numerous Illustrations. Crown 8vo. y. 6d.
elements of embryology
Js.
F.R.S., andF. M, Balfour, M.A., Fellow of Trinity College, Cambridge. With numerous Illustra-
Crown
8vo.
6d.
PHYSICAL SCIENCE.
Gallon
continued.
;
HEREDITARY (iENTOS
sequences.
its
Demy
calls it
,,
8vo.
i2j.
The Times
" a most
able
and most
interesting book."
of very great
interest."
Naturei'
Geikie.
Works by ARCHIBALD Geikie, LL.D., F.R.S., MurcHson Professor of Geplogy and Mineralogy at Edinburgh :
ELEMENTARY LESSONS
With numerous
IN PHYSICAL GEOGRAPHY.
4f. 6d. Questions, is. 6d.
8vo.
is.
[Shortly.
Illustrated.
iSmo.
is.
By J. E. Cambridge.
Crown 8vo.
A Popular Intro-
By AM^ofe the Study "of Physical Phenomena. Translated from the French by Mrs. Norman GuiLLEMiN. LofeKYfiR ; and Edited, with Additions and Notes, by J. Norman Lockyer, F.R.S. Illustrated by Coloured' Plates, and 455 WoodRoyal 8vo. 2i,s. cuts. Third ajid, cheaper Edition.
duction
to
The "translator and Editor have done justice .to their trust. text has all iJie force and flow of original writing, combining faithfulness to the authors meaning with purity and independence
'
in regard pervading
idiom ; while the historical precision and .accuracy work throughout, speak of the watchful editorial supervisionwhich has been given to every scientific ddail, 'N(^Hng can well exceed the cifnrness,'av4 delicacy of the illustrative, woodto
the'
'
Altogether, the work may be said to'hflvef no parallel, either in point of fulness or attraction, as, a popular, manual of physical
cuts.
science."
Saturday Review.
.-,>
By A.
With Coloured
Edited with Notes and Additiunsi by, Ji N. Lockyer,' F.R.S. Chea,i)pr Plates and numerous Illustrafions.
;
I.
' Edition.'' Imperial 8vo. clotli, extra gilt 36^. Alsoin'Eigbte^n Monthly Parts, price is. each. Part
1878.
. '
'"'
'
in Novfember,
^
,
'
''^A book uihichr vie can heartily recommend, both on account of the width and soundness of its contents, and also- because of the excellence of its. print. Us illustrations,,, amd external \appecirance."
Westminster'.Review.
<
'
SCIENTIFIC CATALOGUE.
Hanbury.
Botanical.
chiefly .Pharmacological
ajid
LIVltSTG and Application of. the- Principles of Evolution to Reli^on considered as Illustrative of the Wisdom and Benefi-
By
,
the
.6j.
;
Rev.
George Henslow,
C.B.,
M.A,jF.L.S.
CrownSvo,
by
Sir
Hooker.Works
F.R.S., M.D.,
D.C.L. :
J.
D.
HOOKER, K.C.S.I.,
.
of
clearness, brevity,
',
.
Gazette.
'V
and
PRiMER
(i)F
BOTANY.
With
Illustrations.,
i8mq.
is.
New
Edition, revised
corrected.
Hooker and Ball.jpuRNAL OF A TOUR IN MARdCCO AND THE GREAT ATLAS. By Sir J. D. HookER, K.C.S.L,
C.B., F.RiS., &c., and John Ball, F.R.S. With. Appendices, including a Sketch of the Geology of Marocco. By G; MAW,
F.Li.S.,
F.G.S.
2W.
" This
most thoroughly valuable hook to teachers OMd'students which has ever appeared in the English tongue" ' London Quarterly. Review.
is the of biology
.;
AND REVIEWS,,. By T. H, Huxley, LL.D,, F.R.S. New and Cheaper Edition. Crown 8vo. Is. 6d. FoU7 teen Discourses on thefollowing suijeetst^-\ i ) On the Advisableness of Improving Natural ICmrwledge;: (2) EmanHpation Slack and White : (3) A Liberal Education, and where to find
it;
(4)
ScientificEducation:
(5)
On -ithe
Educational' Value of
PHYSICAL SCIENCE.
Huxley
the
(7)
(Prp,fessor)-Tr-coi
Sciences: (6) On the Study of Zoology: the Physical Basis of Li/e:^S) The Scientific Aspects of Positivism: (10) Geological Contem(9) On a Piece of Chalk: poraneity and. Persistent Types, of Life: (11) Geological Reform: '" Origin (i2) The Ori^n of Species.: (13) Criticisms on the of Species:" {14) On Descartes' '' Liiscourse touching tjie Method of using Onis Season rightly and i>f seeking Scientific Truth."
Natural History
On
'
ESSAYS SELECTED
DRESSES, CRITIQUES
Contents:
AND REVIEWS." Second Edition. Crown 8vo. AND ADDRESSES. 8vo. ios,.6d.
'
'
FROM
i.
AdmiHisiraiive Nihilism.
2.
they may do. 3. On Medical Education. On- the Formation of Coal. 6. On Coral 4. Yeast. 5. and Coral Reefs. 7. On the Methods and Results of Ethnology. 8. On some Fixed Points in British Ethnology. 9. Pala;ontology and the Doctrine of Evolution. 10. Biogenesis and Abiogenesis.
what
and what
II.
Mr. Darwin's
Critics.
12.
13.
With numerous
,
New Edition.
Fcap. 8vo.
4s. 6d.
Westminster Review.
8vo.
6/. (>d,
" Unquestionably
AMERICAN ADDRESSES:
Biology.
New
Edition.
With Crown
Jellet
(John H.,
Academy.
THEORY OF
Irish
B.D.) FRICTION.
TREATISE
;
ON THE
Francis Jones, Chemical Master in the Grammar School,- Manchester. With Preface by Professor Roscoe. New Edition. i8ra.o. with Illustr^ttipns.; 2s. dd.
New
Crown
Svo*
6s.
Lahgdon.THE APPLiGATiON of' eLectrjgity *ro of the RAILWAY WORKING,, ByW. E. Langdon, lyiember Ill,us|trations.
'
With numerous
.
.
.12
SCIENTIFIC CATALOGUE.
(J.
Lockyer
N.)
Worksby
J.
merous Illustrations. New Edition. i8m6; 5j. 6d. ''The book is full, clear, sound, and worthy of attention, not only as Athenaeum. a popular exposition, but as a scientific 'Index.' " " The most fascinating of elementary books on the .Sciences."
J.
'.
Nonconformist.
'
''
^THE SPECTROSCOiPE
,1
'.
AND
ITS APPLICATIONS.
'Plate
2^- ^^
.
By
Norman Lockyer,
,
Illustrations.
and numerous
J.
Norman
Popular Account of Inquiries into the Physical Constitutiqn ,of the Sun, with especial reference to Recent II. Communicatioiis to the Royal ^jSpe.ctroscopic Researches. Society of London and the French Academy of Sciences, with
Illustrated by 7 Colc^urd Lithographic Plates ^and 1 75 Woodcuts. Royal 8yo. cloth, extra gilt, price SW. dd. " The book may be taken as an authentic exposition of the present
Notes.
of
spectro-
scope analysis. ... Even, the unscientific public Daily News. information from it."
may
derive much
-
PRIMER OF ASTRONOMY. With Illustrations.. iSmo. u. Lockyer and Seabroke.^STAR-GAZlNG PAST AND
:
PRESENT.
J.
,
Introduction to Instrumental Astronomy. By Expanded- from Shorthand' Notes of a Course of Royal Institution Lectures with, the assistance of iG. M.
An
N- Lockyer, F.R.S.
"
'" A
'
2ij.
'
Athenssum.
book, of
great inieresi
'
and utility
to the
astronomical student."
Lubbock.-^Works by Sir John Lubbock, M.P.jF.R.is.,D.QX.: THE; ORIGIN AND METAMORPHOSES OiF INSECTS.
:With Numerous Illustrations. Second Edition: Crown 8vo. 3J. bd. "As a summary of th'e phenomena Of insect ihetamorphoses Ms little book is of great value, and will be read with interest and profit The, i/uhole chapter on the by all students of natural history. The ijliisira'origin of insects is most interesting and Valuable'. Westminster Review. tions are numerous and good."
.
'
IN RELASecond
i
Numerous
other
IIlustrati6ns.
Crown
8yo.
i^.dd.,
...
For
;
^
Works by
.'
.
the same
.
.
Theological Catalogue.
'
"
;
or, J^amblps.
'
and^Ineid^ts in
:'''
sedrcH 'of Alpine Plants. Globe 8vo, cloth, (is. 7 '"One of the most charmirig' books of its kind' ever written." Litetary Churchman. "'Mrl' Maciriiilan's glo^oing pictures
-"'
of.
Scandinavian scenery."
Saturday Review.
PHYSICAL SCIENCE.
Macmillan (Rev. Hugh)B(,;j^ar, FIRST FORMS F VEGETATION. Second
.
13
;.'
Edition, _ corrected
and enlarged, with Coloured Frontispiece and numerous lUiistrations. Globe 8v(5. 6j. The first edition of this book was published, under the name of "Footnotes from the Page of Nature; or. First Forms afWegetation. Probably the best popular guide to the .studyi of mosses, .^jlichetts, and fungi ever iDritten. fts practical value as a help to
.
>
ihe
student
and
collector
Examiner.
Works
B,.
by the
late C.iB.
Mansfield :
A THEORY OF
Solution.
Treatise on the Constitution of Bipolar (two-membered) Chemical Compounds. Crown 8vo. 14J.
SALTS.
AERIAL NAVIGATION.
,
Edited by R.
M. Ludlow.
the Problem, with Hints 'for its Mansfield. With a Preface by J. With Illustrations. Crown 8vo. loi. (id.
Mayer.
a Series of Simple, Entertaining, and In; expensive Experiment^ in the Phenomena of Sound, for the Use of By A. M. Mayer, Professor of Physics Students of every age. With numerous Illusin the Stevens Institute of Technology, &c. Crown 8vo. 35^. bd. trations.
SOUND
light,
phenomena of Light, for By A. M. Mayer and C. the use of Students of ev.ery age. Barnard. With Illustrations. Crown 8yo. 2J. (>d.
and Useful Experiments
in the
No.
i,
Skull of the Crocodile. A M^^Vial for Students. By L. C. MiALL, Professor of Biology in Yorkshiire College. 8vo. 2J. M. No. 2, The Anatomy of the Indian Elephant. By L. C. MlALt
and F. Greenwood.
With
Plates.
5^.
Assistant Tutor of St. Peters ColSecond Edition, revised and enlarged. Crown
Mivart
( St.
ON THE GENESIS OF
Edition
in reference and, reply to to which; notes have been added Illustrations, pp. Darwin's "Descent of Man." With numerous Q-^' XV' 2Q6 <> -work in the English language lias this great-controversy broad and vigorous p-asp of been treated at once with the same candid ^CTi^<?n" Saturday Review. facts, ntid the same liberal and
1,4
SCIENTIFIC CATALOGUE.
(St. Gfiovge.)cotfimued.
Mivart
With Numerous
.
Illustrations.
Crown
8vo. 3f. 6(/. (Nature Series.) " It is an able monogram of the Frog, and something more. It throws valuable' crosslights over wide portions of anbi^ated nature. Would that such works were more plentiful." Quarterly Journal
of Sdence.
'"
Chal-
made during
the voyage of
^cienti^c, Pjates, and
H.M.S. " Challenger" round the world in the years By H. N. Moseley, M. A.. F.R.S., Member of fhe Staff ,of the "Challenger." With Map, Coloi^red
,
Woodcuts.
8vo.
2ij.!-
/,,
.
,
'
Mui'r. PRACTICAL CHEMISTRY FOR MEDICAL STUDENTS. Specially arranged for the first M. B. Course, By
M. M. Pattison Muir, F.R.S.E.
Essays on
the'
Fcap.
'
'8vo.
\s. dd.
Murphy.
writtfeli.
a Series of Life and Mind. By Joseph John Second Edition, thoroughly revised and mostly reIllustrations.
With
8vo.
16^.
Nature.A
SCIENCE.
Parts, 2j.
and
Vols. Is. 6d. " This able and well-edited Journal, which- posts up the science of the day promptly, and prbmises to be of signal service to students and savants Scarcely any expressions that we can emplj>y wo'iild exaggerate our sense of the moral and theological value- of British Quarterly Rfeview. the work,"
'
'
"^
By Simon New-
With 112 comb, LL.D., Professor U.S. Naval Observatory. Engravings and Five Maps of the Stars. 8vo. iSj. "As affording a thoroughly reliable foundation for more advanced reading. Professor Neiucomb's Popular Astronotiiy is deserving
'
'
of strong recommendation."
Nature.
Oliver.Wprks by Daniel
Oliver, F.R.S., F.L.S., PrcifeSsor of Botany in Uiiiversity College, London, and Keeper of the HerEarium and Library of the Royal Gardfens, Kew LESSQNS IN ELEMENTARY BOTANY. With, nearly Two
:
Hundred
Illustrations. New Edition, Foap. 8vo. 4^. 6d. This book is designed to teaci the. eleihents of Botany on Professor Henslov^ s pla,n of selected Types and by the use of Schedules. The earlier chapters, embracing thielements of Structural and Physiological Botany, introduce us to the methodical stiidy of the Ordinal Types. The mv.clud.ing chapter^ g,re e^f;4kd, " Hirw to D/y
PHYSICAL SCIENCE.
15
OWVQVcontinued.
Plants " and" How to Describe Plants. " valuable Qlossary is apjiende^io the volume. In the preparation of this work free use has been made, of the n^anuscripi materials of the late Professor ftenslaw,
With numerous Extra fcap. 8yp. ds. 6d. It contains a well-digested sumik&ry of all essential knowledge P^aining^ to Indian Botany, wrought out in accordance with the best 'princijiles of scientific arrangement." Allen's lildiali Mail.
Pennington.NOTES ON THE BARROWS AND BONE CAVES OF DERBYSHIRE. With, an account o? a Descent
-into
By RooKE Pennington,
B.A.,
LL.B.,
Penrose (F, C.) on a method of predicting by GRAPHICAL CONSTRUCTION, OCCULTATIONS OF STARS BY THE MO.ON, AND SOLAR ECLIPSES FDR ANY GIVEN PLACE. Together with more rigorous methods
.
for the Accurate Calculation of Longitude. F.R.A.S: With Charts, Tables^ &c. 4to.
By
12s.
F. C.
Penrose,
Perry. AN ELEMENTARY TREATISE ON STEAM. By John Perry, B.E., Professor of Engineering, Imperial College of
With numerous Woodcuts, Numerical ExEngineering, Yedo. i8mo. 4r. dd. amples, and Exercises.
>
^'Mr. Perry has ^in this compact little volume brougkp together an immense amount of information, new told, regarding steam and its application, not the least of its merits being that it is suited to the capacities alike of the tyro in engineeiing science or the better grade of artisan." Iron.
^ Pickering. ELEMENTS OF PHYSICAL MANIPXJLATION. By E. C. Pickering, Thayer Professor of Physics in the Massaichiisetts Institute
.
of Technology,
Part
I.,
medium
8vo.
\Qs. 6d.
Part
II.,,.,
los. 6d.
"When
finished' 'Physical
Nature. wv
,
and most
Prestwich THE PAST AND FUTUllE OF GEOLOGY. An Inaugural Lecture, by J. P?iESTWiGH, M.A., F.R.g., &c.,
Professor of Geology, Oxford.
8yo.
zs.
Radcliffe.^PROTEUS
OR UNITY IN NATURE.
B. RadcLiffe, M.D., Author of "Vital Motion as a Second Edition. 8vo. 7j, dd. Physical Motion.
mode
By. C. of
i6
SCIEI^TIFIC CAT4^0GUE.
Translated by A. Weus, Q.C, Alpine Club. To which are added, the Original Meniciir and Supplementary Articles byPrbfessoirs Tait and RusKiN. Edited with Introductory remarks by George Forbes, B.A., Professor of Natural Philosophy in the Andersonian Universityj ,' '.,. . .'./] \-\ Glasgow. ,'8vo.. "js.Cd.
late President of the
.
,
'
Roscpe.
Professor
of
LESSONS' IN
litho
ELEMENTARY CHEMISTRY, INORGANIC AND ORGANIC. With numerous Illustrations and Chromqof the' Solar- Spectrum, and of the AlkaMs and Alkaline New Edition. Fcap.'SvOi' ^s.Sii.
Earths.'
'
CHEMICAL '^liOBLEMS,'
Thorpe.
"
adapted to
2s.
,
the,
atove by Professor
We
i^ffhesitcfiingly^
treatises
'
PRIMER OF CHEMISTRY. Illustrated. iSma; is.. Roscoe and Schorlemmpr. A TREATISE ON CHE.By PRaFESSQES RpscoE and Schorlemmer'. The Non-metaUic Elements. With numerous Illustrations and Portrait of Dalton. Medium 8vo. 21s. Parti. With illustrations. 8vo. t^s. Vol. II., Metals.
MISTRY.
I.,
Vol.
"Regarded as a treatise oh the Non-rnetallii Elements,' there can bci no doubt that this volume is inc'omparahlv the most satisfactory one of which we are in possession'' Spectator. " It would be difficult to praise the work too highly. All the merits wltich we noticed in the first volume are conspicuous in the second. The arrangement is clear and scientific ; the facts gained by modern
throughout
is
and
the -.,'.
Rumford (Count). THE LIFE AND COMPLETE WORKS OF BENJAMIN THOMPSON, COUNT RUMFORD. With
Notices of his Daughter. Five Vols. 8vo. 4/. "i+r.
By George Ellis.
6(f.
With
Portrait.
By
C.
Schorlemmer, F.R.S.,
College, Manchester.
to beas^ complete
Owens
.
&vo.
,fl,
"Zt appears Jo us
manual, of the metamorphoses, of cfirbon as could be at /present produced, ctiid it must ptoVe eminently Athenieum. useful to the chemical student.''
PHYSICAL SCIENCE.
Shann.
By
17
AN
IN
Shann, M.A.
Crown
8vo,
4r. 6d.
Exposition of the Nature, Number, and Organography of Ferns, and Review of the Principles upon -which Genera are founded, and the Systeias of Classification of the principal Authors, with a new GeneralArrangement, &c. By J. Smith, A.L.S., ex-Curator of the Royal Botanic Garden, Kew. With Thirty Lithographic Plates by W. H. Fitch, F.L.S. Crown 8vo. 12s. 6d. " No one anxious to work up a thorough knowledge of ferns can Gardener's Chronicle. afford to do mthout it."
:
An
"^
Abney, F.RS., Professor S.tokes, BrAmwell, F.R.S., Professor G. Forbes, iH. C. Sorby, F.R.S., J. T. Bottomley, F.R.S.E., Crown 8vo. S. H. ViNESj B. Sc, and Professor Carey Foster.
[Vol. II. nearly ready.
6s.
By W.
the Royal Society. With numerous Second Edition! ,Cr. Svbl.SJ. id. (Nature Series.) Illustrations. " The illustrations are exceedingly well adapted to' assist in making clear, trustworthy the text comprehensible.'" ^-K^enxura. manual." Standard.
"A
Stewart
(B.)
of Natural Philosophy in
Illnstrations
F.R.S., Professor
:
LESSONS IN ELEMENTARY
and Nebulae.
PIIYSICS.
4?. 6d.
With numerous
Stars,
New Edition.
Fcap. 8vo.
scientific text-
PRIMER OF
Questions.
New
Edition, with
iSmo.-
or.
Physical Speculations on a Future State, By Balfour Stewart, Crown 8vb. 6j-. Sixth Edition. F.R.S., and P. G. Tait, M.A. " The book is one which well deserves the attention of thoughtful ana
It is a petfectly sober inquiry, on scientific religious readers. grounds, into the posiiUlilies of a fttture existence." Guardian.
. .
-i
SCIENTIFIC CATALOGUE.
By
F.C.S., Professor of Agricultutal Science, University College, Aberystwith, Examiner in the Principles of Agriculture under the, Government Department of Science. iSmo.
'
'
Taylor.
tise
sound Non-Mathematical Trea MUSIC on the Physical Constitution of Musical Sounds and Harniony including the Chief Acoustical Discoveries of Pirofessor Helm holtz. By Sedley Taylor, M.A., late Fellow of Trinity Col Large crovpn 8vo. %s. dd. ledge, Cambridge.
:
and
"In no previbus
scientific treatise
do
so richly illustrated
description
wave-motion in fluids."
Musical Standard.
:
Thomson. Works by Sir Wyville Thomson, K.C.B., F.R.S. THE DEPTHS OF THE SEA An Account of the General
Results of the Dredging Cruises of H.M.SS. "Porcupine" and " Lightning " during the Summers of 1868-69 and 70, under the scientific direction of Dr. Carpenter, F.R.S., J. Gwyn Jeffreys,; With nearly 100 F.R.S., and Sir Wyville Thomson, F.R.S. Second Edition. Illustrations and 8 coloured Maps and Plans.
Royal 8vo. cloth, gilt. '3,u. 6d. The Athenseum says : " The book is full of interesting matter, and It is is written by a master of the art of popular exposition.
excellently illustrated, both coloured
possessing
Those who have already become interested in dredging operations will of course make a point of reading this work ; those who wish to be pleasantly introduced to the subject, an<l rightly to appreciate the news which artiyes from time to time from the ' Challenger,' should not fail to seek instruction from it."
high
merit..
THE VOYAGE OF THE " CHALLENGER."THE ATLANTIC. A Preliminary account of the Exploring Voyages of H.M.S.
With numerous
.
"Challenger," during the year 1S73 and the early part of 1876. Illustrations, Coloured Maps & Charts, & Portrait of the Author, ^ngraved'.byC. H. JEENS. 2 Vols. Medium 8vo. ,42^. " It is right that the public should have some I%e Times says authoritative account of the general results of the expedition, and that as many of the ascertained data as may be accepted with confidence should speedily find their place in the general body oj No one can ba more competent than the scientific knowledge. accomplished scientific chief of the expedition to satisfy the public in The paper, printing, and especially the numerous this respect illustrations, are of- the highest quality; ... We have rarely, if ever, seen more beautiful specimens ofiiiood engraving than abound Sir Wyville Thomson's style is particularly in this work.
attractive; he is easy
and
and
exceedingly
PHYSICAL SCIENCE.
19
Thomson
continued.
hapjyin the choice of language, andtHraughout the work there%re touches which show that science has not banished sentiment .from hts bosom"
Thudichum and
ORIGIN,
NATURE, AND
Being a Complete Manual of Viticulture and CEnology. By J. L. W. Thudichum, M.D.., and August Dupr, Ph.B., Lecturer on Chemistry at Westminster Hospital. Medium 8vo. cloth gilt. 25^.
OF
WINE.
"A treatise
almost unique for its usejiilness either to the wine-grower the vendor, or the consumer The analyses of wine are of wine. the most complete we have yet Seen, exKibiting at a glance the constituentprinciples oj nearly all the wines inown in this country.
SELECTION.
of Essays.
New
8s. 6d.
Edition,
with
Crown
8vo.
Dr.
ffooker, in his address to the British Assofiatian, spokf thus of the author : "OfMr. Wallace and his many contrOtttions to philosophical biology it is not easy to speak without enthusiasm; for, putting aside their great merits, he, throughout his writings, with a modesty as rare as I believe it to be unconscious, forgets his awn unquestioned claim to the honour oJ hoping originated independently of Mr. Darwin, the theories which he so ably defends." 7%^ Saturday Review says: "He has combined an abundarue of fresh and original facts with a liveliness and sagacity of reasoning which are not often displayed so effectively on so small a 'scale."
'
'^ Altogether it is a wonderful and fascinating whatever objections may be taken to theories founded, upon it. Mr. IVallace has not attempted to add to its interest by any adornments of style ; he has given a simple and clear statepient of intrinsically interesting facts, and. what he considers to be legitimate inductions from them. Naturalists ought to, be grateful to him for hamng undertaken so toilsome a task. The work, indeed, the author, the publishers, the artist is a credit to all concerned unfortunately now no more of the attractive illustrations last but by no means least, Mr. Stanford's map-designer"
SCIENTIFIC CATALOGUE.
Wallace (A. R.)continued. TROPICAL NATURE: with other Essays^
" Nffwhere amid
8vo.
I2s.
the niany descriptions of the irdpUs that have been given is to be found a summary of the past history and actual phenomena of the tropics which gives thai which is distinctive of t^e .phases \of nature in them more clearly, shortly, and impressively." Satui'day Review.
Warington,THE WEEIC OF CREATION; OR, THE COSMOGONY OF GENESIS CONSIDERED IN 'ITS REL4.TION TO MODERN SCIENCE. By Georg:? WaeINGTON, Author of " The Historic Character of the Pentateuch
,
Vindicate^."
Crown
8vo.
,4^.
dd.
Wilson.RELIG16 CHEMICI.
By tiie late George Wilson, M.D., F.R.S.E., Regius Professor of Technology in the University With a Vignette beautifully engraved aftei; a of Edinburgh. Crown 8vo. Zs.dd. design by Sir Noel fATON. "A more fascinating vpmme," the Spectator says, " has seldom
,
a Critique on Shakespeare's Night's Dream." By Daniel Wilson,, LL.D., Professor of History and English Literature in University College, Toronto. 8vo, los. 6d.
:
''
-The whole volutne is most rich in the eloquence of thought and imagination as well as of words. It is a choice contribution at once to science, theology, religion, and literature.'''' British Quarterly Revifew.
&c., Lec-
Extra fcap.
-^
from the
down to the present time. By Ad. Wurtz. Translated by Henry Watts, F.R. Si Crown 8vo. 6j.., " The discourse, as a resume of chemical theory and research, unites singular luminousness and grasp. A fewjudicious notes are added " The treatment of the by the translator." Pall Mall Gazette. sttbject is admirable, and the translator has evidently done his duly
. '
ETC.
21
RHETORIC.
M. Cope,
AN With INTRODUCTION
TO
ARISTOTLE'S
'
By E.
ARISTOTLE ON FALLACIES; OR, THE SOPHISTICI ELENCHI. With a Translation and Notes by Edward Poste,
M.A., Fellow of Oriel College, Oxford.
8vo.
8j. 6(f.
Birks.-
Works by
Svo.
sophy, Cambridge
FIRST PRINCIPLES OF
Crown
ITiis
.
,
MORAL SCIENCE;
.
or,
a Firs
1
'
work treats of three topies all preliminary to the direct exposiThese are the Certainty amd. Dignity 'Hon of Moral Philosophy.
of Mgrai Science, its Spiritual Geography, or relation to vther main subjects of human thought, and its FormaHve Principles, or some elementary truths on whuh its whole development must
MODERN
UTILITARIANISM;
Mill,
Bentham, and
6s.
Paley, 6d.
Boole.
AN INVESTIGATION OF THE LAWS OF THOUGHT, OlNf WHICH ARE FOUi^DED -THE MATHEMATICAL THEORIES OF LOGIC AND 'IPROBy George Boole, LL.D,, Professor ,of Mathematics in the Queen's University, Ireland, &c. Svo. I4r.''
,
B ABILITIES.
'
;.-;
KANT.
iSs.
With
By, E. Caird,
,\
.,
22
SCIENTIFIC CATALOGUE.
Works by the Rev. Henry Calderwood, M.A., LL.D., Professor of Moral Philosophy in the University of Edinburgh PHILOSOPHY OF THE INFINITE: A"^ Treatise on Man's Knowledge of the Infinite Being, in answer to Sir W. Hamilton and Dr. Mansel. Cheaper Edition. 8vo. Is. 6d.
:
Calderwood.
-,-,-
"A
book of great ability > ... written in a clear style, and may be easily understood by,, even those who are not versed in such discussions!' British Quarterly Review-
New, Edition. 8vo. 6^. ., "It is, we/eel convinced, the best handbook on the subject, intellectually and morally, and does inHnite credit to its author. Standard, "A- compact and useful work, going over a great deal of ground in a manner adapted to suggest and facilitate further study. , . . His book Tjoill be an assistance to many students outside his awn University of Edinburgh. Guardian.
.
'
,.''
INearly ready.
Fiske. OUTLINES OF COSMIC PHILOSOPHY, BASED ON THE DOCTRINE OF EVOLUTION, WITH CRITICISMS ON THE POSITIVE PHILOSOPHY. By John
FiSKE, M.A., LL.B., Harvard University. 2
formerly
vols. 8vo.
Lecturer
ajj.
on
Philosophy
at
" The work constitutes a very -ejfecHve encyclopcedia of the evolutionary philosophyi and is well worth the study of all who wish to see at once the entire scope and purport of the scientific dogmatism of
the day."
Saturday Review.
Herbert. THE realistic assumptions of modern SCIENCE EXAMINED. By T. M. Herbert, M.A., late
Professor of Philosophy, &c., in the Lancashire Independent College, Msinchester. 8vo. 14J.
Jardine.THE elements of THE psychology of COGNITION. By Robert Jardine, B.D., D'.Sc, Principal of
the General Assembly's College, Calcutta; and Fellow of the University of Calcutta, Crown Svo. 6j. 6<j?.
;
Jevons.
Professor of Political
Scientific
F.R.S.,
Treatise on Logic
ieviseflj
and
New
Crown
Svo. I2S. 6d. " Ifo one in future can be said to have any true'' knowledge of what has been done in the way of logical and scientific method in England without \ having carefully studied \Iofeisor yevons' Spectator. book."
23
ELEMENTARY LESSONS IN
,
LOGIC, DEDUCTIVE
AND
'
jINDUCTIVE. With
Logical Terms.
New
l8mo.
is.
MaCCOll. THE GREEK SCEPTICS, from Pyrrho to Sextus. An Essay which obtained the Hare Prize in the year 1868. By Norman Maccoll, B.A., Scholar of Downing College, Cambiidge.
Crown
8vo.
3^. 6d.
New
Jersey,
U,S.
himself skilful in that amplication of logic to psychology, in that inductive science of the human mind which is His philosophy as a whole is the fine side of English philosophy. worthy of attention." '&.&s'^e de Deux Mondes.
Physical
its
He certainly shows
being accurate knowledge of its present condition, and by its entering in a deeperand moreunfettered manner than its predecessors upon thediscussion of the appropricUepsychological, ethical, and theological quesThe author keeps aloof at once from the i priori idealism and turns. dreaminess of German speculation since Schelling, and from the onesidedness and narrowness of the empiricism and positivism which have so prevailed in England." Dr. Ulrici, in "Zeitschrift
fiir
Philosophic."
A New Edition.
.
8vo.
J
; '
adfust the claims of the sensational and inand of the i posteriori and h. priori methods, success." is accontpUshedinthis work with a great amount of f I value it for its large acquaintance IBKeSitminster Review, the 'with English Philosophy, which has not led him to neglect sreat Go-man works. I admire the moderation and clearness, as well as. comprehensiveness, of the author's views." Dr. Dbrner, of
to
tuitional philosophies,
Berlin.
AN EXAMINATION OF MR.
additions.
to doit.
J:
S.
MILL'S PHILOSOPHY:
Second
edition, with
needed to be done, and tJuauihor was the man This volume is important, not merely in reference to th
24
SCIENTIFIC CATALOGUE.
'"'-i'!''"'
M' CoshcondnueJ.
mews
of
Mr.
pres'eni, British
Mitt, but of the mhiple. school of writers, ^pmt and and Continental, he so ably refiresents."T-:Pxinf;eton
Review.
Being a Text-
book of Formal Logic. Crown 8vo. ^s, The amount of summarized information which it contains is very great; and it is the only work on the very important subject with which it deals. Never was such a -work so much needed as in '_ // /ra^Kif ii?y."London Quarterly Review.
Serie? of Lectures to
tlie
']s.
Crown
'.
8vo.
;,.;
-.,
..;
Royal 8vo.
i6j.
Review
with Criticisms including some Comments on Mr. Mill's Answer By Da vii) Masson, M.A., Professor to Sir William Hamilton. of Rhetoric and EnglishLiterature in the University of Edinburgh. Third Edition, with an Additional Chapter. Crown 8vo. 6'*. " We can nowhere point to a work which gives so clear an exposition of the^ course of philosophical speculation in Britain during the past century, or which indicates so instructively the mutual inFortnightly Review. fluences ofphilosopihic and scientific thought."
,
Maudsley.
Professdr of Medical
being the First Part of a Third ; Edition, Revised,, Enlarged, and in great part Rewritten,, of, ," The \os.()d.' Physiology and Pathology of Mind; " Crown 8v6.
llnthe'Press.
their Connexion and Mutual to Mental Disorders.
specially with refeirence An Enlarged and Revised edition. To which are added, Psychological
Influence,
Essays.
Crown
8 vo.
ds. dd.
Maurice.
Works by the Rev. Frederick Denison Maurice, M.A., Professor of Moral Philosophy in the University Of Cambridge. (For other; 'Vyorks by the same Author, see Thsopogigal ':''' Catalqgue.)
;
'
SOCIAL MORALITY.
los.dd.
Twenty-one LectuVes deliveired in the University of Cambridge. New and Cheaper Editiori. Crowri 8vo.
25
and prejudice,
extant in theiiitirU, Which aninifltes it /rom one end to tfie other. We gain new thoughts and new ways of viewing things, even more, ferSaps, from being brought for a time under the influence of so
noble
it we are charmed by thefreedomfrom exchisiveness the large charity, the hrfdness of thought, the eagerness to recognize and appreciate whateve)' there is of real worth
Whilst reading
Athenseum.
THE CONSCIENCE
versity of
T'yJ*
Cambridge.
New
5j.
Saturday Review says: "We rise from them with detestation of all that is selfish and mean, and with a living impression that there is such a thing as [goodness after all.
.
Vol.
Ancient Philosophy from the First to the Thirteenth Centuries ; Vol. II. the' FbiTrteenth Century and the French Revolution, with a glimpse into the Nineteenth Century. New Edition and Preface. 2 Vols. 8vo. 25J.
or Researches in the Lines of Progress, from Savagery, through Barbarism to Civilisation. of the National Academy of Sciences. 8vo. i6s.
:
Murphy.
8vo.
'
'
THE
By
The booh is, not without, substantial value ; the writer continues the wofr^ of the best apologists of the last century, it may be with less
forfe tact;
'..
and clearness, iftt still with commendable persuasiveness and and with an intelligent feelingfor the changed conditions oj
theproktem."
Academy.
8vo.
"js.
Paradoxical Philosophy.
verse."
Sequel to
"The Unseen
Uni-
Crown
6d.
Picton. THE MYSTERY OF MATTER AND OTHER ESSAYS. By J. Allanson Picton, Author of " New Theories
;
Cheaper
issue with
New
Preface.
Crown
Contents : ra? Mystery of Matter Tlie Philosophy of Ignorance The Antithesis of Faith and Sight The Essential Nature
of Religion
Christian Pantheism.
26
SCIENTIFIC CATALOGUE.
,
EyjjHENRY
with important
additions and
A SUPPLEMENT to the
" This
First Edition, containing all the important alterations in the Second. 2S. 8vo.
. .
excellent and very mekome volume. .,. J-eaving to metaphysicians any further discussion thai may be needed respecting the already over-discussed problem of the origin of the moral faculty, he tcikes it. for granted' as readily as the geometrician takes spasce for granted, or the physicist the existence of matter^ But htjakes Utile elsefmr granted, and defining ethics as ' th^sciencf of conduct,^ be carefully examines,'not the various ethnical, sysfems that have been propounded by Aristotle and AristotUs followers downwards, but the principles upon which, so far as they confine themselves to the strict province of ethics, they are based." AthenSeum.
Thornton. old-fashioned ethics, and commonsense METAPHYSICS, with some of their Applications.,. By
William Thomas Thornton,
8vo.
los. 6d.
The present volume aeals with problems which are agitating minds of all thoughtjul men. The following are the Contents
/.
>
Ante-Utilitarianism. //. ffistory's Scientific Tretensioiis. JII. Hume as a MetaphymeHn. IV. HuxkyismU V. Recent I'hase of Scientific Atheism. VI. Limits of Demonstrable Theism.
Damd
Thring
8vo.
(E.,
M. A.)THOUGHTS ON LIFE-SCIENCE,
Place),
School.
New Edition,
dd.
:
Foundaand Province of the Theory of Probability, with especial ^Mdtal and reterence to its logical bearings, and its apphcation to Social Science. By John Venn, M.A., Fellow and Lecturer of Gonville and Caius College, Cambridge. Second Edition, rewritten and greatly enlarged. Crown 8vo. lOr. 6d. " One of the most thoughtful and philosophical treatises on any subject connected with logic
this or
any
p. 77.
and evidence which has been prudu^d many years." Mill's Logic, vol.
in
ii.
.,
, ,
,'
,V}:'''.:A
SCIENCE PRIMERS.
27
Editorship of
Professors
Balfour Stewart.
Chemistry
in
By H.
Owens
IJP.
l8mo.
E. Roscob, F.R.S., Professor of Chemistiy College, Manchester. With numerous Illustrations. New Edition. With Questiohs.
Physics. By
Balfour Stewart, F.R.S., Professor of Natural Philosophy in Owens College, Manchester. With numeri8mo. \s. NeW Edition. With Questions. ous Illustrations.
^
Physical Geography.
Archibald Geikik, F.R.S., Murchison Professor of Geology and Mineralogy at Edinburgh. With numerous Illustrations. New Edition with Questions. l8mp. \s.
_By
Geology
trations.
By
With numerous
Illus-
Logic By
Botany
i8mo.
New Edition. l8mo. cloth, is. Physiology By Michael Foster, M.D., F.R.S. With numerous Illustrations. New Edition. i8mo. \s. Astronomy By J, Norman Lockyer, F.R.S. With numerous Illustrations. New Edition. i8mo. \s.
numerous
^By Sir J. D. Hooker; K.C.S.I., C.B., F.R.S. Edition. iSmo. is. Illustrations.
With
New
Professor
New
Edition.
IS.
Political
l8mo.
Economy.By Professor
IS.
In preparation:
&c. &c.
INTRODUCTORY.
JBy Pcofessor
Huxley.
ELEMENTARY SCIENCE
Astronomy TRONOMY.
With
Astronomer Royal.
CLASS-BOOKS.
POPULAR
ASB. Airy,, K.C.B., 41. dd.
New
Edition.
i8mo.
28
SCIENTIFIC CATALOGUE.
Elementary Science Class-books continued. QUESTIONS ON LOGKYER'S ELEMENTARY LESSONS IN ASTRONOMY. For the Use of Schools. By John
Forbes Robertson.
With numerous
i8mo, cloth limp.
is.
Bd.
HUXLEY'S PHYSIOLOGY FOR OIJ SCHOOLS. By T. Alcock, M.D. iSmo. is. (>d. Botany LESSONS IN ELEMENTARY BOTANY. ByD.
QUESTIONS
Oliver, F.R.S., F.L.S., Professor of Botany in University College, London. With nearly Two flundred Illustrations. New,
Edition.
Fcap. 8vo.
'
d^. dd.
Owens
College,
Manchester,
and Chromo-Litho ^of the Solar Spectrum, and of the Alkalies and Alkaline Earths. New Edition.
Illustrations
With numerous
4J-.
Fcap. 8vo.
6d.
'
prepared with Special Reference to the above, by T. E. Thorpe, Ph.D., Professor of Chemistry in the Yorfehire College of Science, Leeds. Adapted for the preparation of Students for the Government, Science, and Society;of: Arts Examinations. With a Preface by Professor RoscoE. Fifth Edition, with Key. i8mo. 3J.
,
Political
i8mo.
GINNERS.
BEI
New.
;
Edition.
r-
Logic
2s.6d.
London.
New
Edition.
Fcap. 8yo.
By
Ne'w
Pi-actical
,
Grammar
School, Manch^ter.
With
RoscoE,
and,. Illustrations.
New
'
SCIENCE CLASS-BOOKS.
Elementary Science Cl^ss-books continued, Anattomy.^LESSONS IN ELEMENTARY ANATOMY.
St.
8y6.
29
By
George Mivart,
6j. 6(1.
at St.
F.R.S., Lecturer in Comparative Anatomy Mary's Hospital. With upwards of 400 Illustrations. Fcap.
Mechanics
iSteam
1
treatise. By A. B. C.E., Professor of Applied Mechanics in University [In preparation. College, London. With Illustrations.
an elementary
W. Kennedy,
an elementary
4T. 6d.
/Professor- sf Engineering, Imperial College of Engineering, Yedo. With numerous Woodcuts and Numerical Examples and Exercises.
l8mo.
Physical Geography.
.Illustrations.,
_ ELEMENTARY
LESSONS IN
PHYSICAL GEOGRAPHY. By A.
son Professor of Geology, &c., 4^.61^. Fcap. 8vo.
By
C. B.
igmo. Properties of Solid and Fluid Bpdies. Sound, iight, and Heat. iSmo. 3j. 6d.
y.
6d.
Part
Sound
AN ELEMENTARY
With
TREATISE.
l8mo.
By W. H. Stone,
[In ihi Press.
M.D., F.R.S.
Illustrations.
Others in Preparation.
By
Dyer, F.R.S., assisted by Sydney Professor Thiselton ViNfeSi B.Sc, Fellow and Lecturer of Christ's College, Cambridge. [In preparation. With numerous Illustrations,
Fawcett. A manual of political ECOIJOMY. By New Edition, revised and enlarged. Professor Fawcett, M.P.
Crown
Svo.
i2J-.
611?.
Translated, with. Notes and Additions, from, the second German With IllustraEdition, by M. M. PaTtison Muik, F.R.S.E. Crown 8vo. 7^. ()d.. tions.
30
SCIENTIFIC CATALOGUE.
Manuals for St\xA^n\.scontinued. Flower (W. H.)_aN INTRODUCTION TO THE OSTEOLOGY OF THE MAMMALIA. Being the Substance of the
Course of Lectures delivered at the Royal College of Surgeons of England in 1870. By Professor W. H. Flower, F.R.S., F.R.C.S. With numerous Illustrations. New Edition, enlarged.
Crown
8vo.
'
los. 6d.
Foster and Balfour._THE elements of EMBRYOLOGY. By Michael :Foster, M.D., F.R.S., and F. M.
Balfour, M.A.
Part
I.
crown 8vo.
^s. 6il.
Foster and Langley._A COURSE OF ELEMENTARY PRACTICAL PHYSIOLOGY. By Michael Foster, M.D., F.R.S., and J. N. Langley, B.A. New Edition; Crown 8vo. 6s. Hooker (Dr.)_THE STUDENT'S FLORA OF THE BRITISH ISLANDS. By Sir J. D. Hooker, K.C.S.I., C.B., F.R.S., M.D., D.C.L. New Edition, revised. Globe Svo. los. 6d, Huxley PHYSIOGRAPHY. An Introduction to the Study of
Nature.
7^.
By
Professor
HuxLEV,
Plates.
F.R.S.
Illustrations,
and Coloured
New
Edition.
'
6d.
New
Crown
Huxley, F.R.S.
assisted
With
Illustrations.
\ln> prepdi^atioH.
Jevons. THE PRINCIPLES OF SCIENCE. A Treatise on Logic and Scientific Method. By Professor W. Stanley Jevons, LL.D., F.R.S., New and Revised Edition. Crown Svo. X2s.6d. Oliver (Professor)._FlRST BOOK OF INDIAN BOTANY.
Professor Daniel Oliver, F.R.S., F.L.S., Keeper of the Herbarium and Library of the Royal Gardens, Kew. With numerous Illustrations. Extra fcap. Svo. 6s. 6d.
By
Parker and Bettany._THE MORPHOLOGY OF, THE SKULL. By Professor Parker and G. T. Bettany. Illus,
trated.
Crown
Svo.
los. 6d.
,
F.R.S.E.
Illustrated.,'.,
Thomson. ZOOLOGY.
F.R.S.
Illustrated.
By
Sir
C.
Wyville Thomson,
\In preparation^
Tyler and Lankester. ANTHROPOLOGY. By E/B. Tylor, M.A., F.R.S., and Professor E. Ray Lankester, M.A.,
F.R.S.
Illustrated.
[In preparation.
NATURE
-^THE SPECTROSCOPE
By
8vo. J.
SERIES.
ITS APPLICATIONS.
Second Edition.
Illustrations.
AND
With
N.
LOCKYER,
F.R>S.
Crown
3J. 6</.
IN-
By
Sir
JOHN LUBBOCK,
Second EdttioH.
M.E., F.R.S.
With
Illustrations.
Crown8vo.l3j.6rf.
THE
TR.ANSIT OF VENUS.
Illustrations.
By
3^-.
G.
Forbes, B.A^,
Univfersity,
GlasgCrw,
With numerous
Crown
8vo.
6^/.
,By St.
a*,
Geqrge
Miyart,
Crown
8vo.
dd.
POLARISATION OF LIGHT.
'
By W. Spottiswoode,
Second Edition.
Illustrated.
Crown
IN
By
Crown
Sis
8vo.
JOHN LUBBOCK,
j[s.
M.P., F.R.S.
Second Edition.
6d.
of the Standards.
Illustrated.
Crown
Svo.
HOW TO
Linkages.
Lecture on
if.
KEMPE,
B.A.
Illustrated.
Crown
Svo.
6d.
'LIGHT
J
A
2J.'
Series
6d.
of Simple,
Experiments in the Phenomena of Light for the Use of Students of every Age.
By ALFRED M.
Crown
""
Svo.
SOUND
Series
pensive Experiments in the Phenomena of Sound, for the Use of Students of every Age. By A. M. MAYER, Professor of Physics in the Stevens Institute of Technology, &c. With numerous Illustrations. Crown Svo^ 3f 6d.
(Others to follow.)
MACMILLAN AND
CO.,
LONDON.
Monthly PaHs
f^s.'
is.
and
td.,
Half- Yearly
Volumes,
nature:
AJV
Nature expounds in a popular and yet authentic manner, the Grand Results of Scientific Research^ discussing
the most recent scientific discoveries,
the bearing of Science
its
upon
civilisation
and
progress,
and
claims to a
more general
recognition, as well as to a
,
all
domain of Science
forming a
Reviews
munication
Serial
medium of Scientific discussion and of intercomamong the most distmguished men of Science
giving the gist
Columns,
of
the most
important
Journals, both
Home
and
Foreign
and Academies of the World, Notes, &c. In Schools where Science is included
what
doing in Science
all
in the
regular
tells
is.
is
is
are. directei^ to
The
various