EULER'S

INTRODUCTIO IN ANALYSIN INFINITORUM VOL. 2

Appendix 5 On Surfaces.
Translated and annotated by Ian Bruce. page 744

CHAPTER V

SURFACES OF THE SECOND ORDER

101. Therefore with the establishment of surfaces with the order following the number of
the dimensions, which the sum of the three coordinates x, y, z taken jointly shall fulfill, if
an algebraic equation may be proposed for the surface, the order to which that surface
must refer will be able to be assigned at once. Since therefore all the surfaces of the first
order shall be shown to be planes, in this chapter I will make the subject to be examined
surfaces of the second order. But in these [of higher order] a greater diversity will be
apparent at once, than in lines of the second degree, which it will be apparent for each to
be attended to easily. Therefore I shall provide the aid, so that I may explain these
different kinds distinctly. Truly with higher orders the diversity will increase by so great
a multitude of kinds, that we must abstain completely from these developments.

102. Because the nature of surfaces of the second order are expressed by an equation of
the second order, in which the variables x, y and z rise to two dimensions, the cylinder
and the cone both right as well as scalene and the sphere, the properties of which we have
now described, will be contained in this second order. Truly all the surfaces belonging to
this order are contained in this general equation

 zz   yz   xz   yy   xy   xx   z   y   x    0.

For whatever three coordinates may be taken, the equation will always be contained in
this form. Therefore the diverse kinds of related surfaces will depend here on the
different mutual relations between the constants, which, even if the same surface may be
expressed by an infinitude of equations, yet the equations will support an infinite number
of variations in the surface.

103. Just as with plane curved lines we have chosen the first division thence, so that they
may be extended to infinity or enclosed within a finite space, thus in a similar manner all
surfaces belonging to some order may be divided into two classes, for the one of which
we may refer to these surfaces which depart to infinity, and to the other truly, which will
be enclosed in a finite space. Thus the cylinder and the cone will be counted in the first
class, the sphere in the second. Indeed in the second class no surfaces will be given with
odd orders ; for since any surface of an odd order shall have plane sections of the same
order, but all curves of odd orders may be extended to infinity, and it is necessary that the
surfaces of these orders are extended to infinity.

104. But as often as a certain surface is extended to infinity, it is necessary, that at least
one of the three variables x, y and z shall go off to infinity. Whereby since it shall be
likewise, whichever in this case shall be assumed to become infinite, we may make z to
 EULER'S
INTRODUCTIO IN ANALYSIN INFINITORUM VOL. 2
Appendix 5 On Surfaces.
Translated and annotated by Ian Bruce. page 745
become infinite, if indeed the surface may be extended to infinity. Therefore we may put
z   to investigate the nature of this part going off to infinity, and now the first term
 zz must be examined mainly, whether it shall be present or truly missing. Therefore this
term shall be present in the equation and before that the terms  z and  vanish, this
equation will be had for the part going off to infinity :

 zz   yz   xz   yy   xy   xx   y   x  0,

from which in turn all the terms vanish, which are not infinite or at least less than  zz .

105. We may put in place all the terms to be present, in which the variables maintain two
dimensions ; indeed whatever surface it should be, all the terms of the sum of the
dimensions will be present always in its most general equation, nor on account of this
hypothesis, by which we may put all the terms of two dimensions to be present, is any
strength of the general solution inferred. But when the terms yz and xz are present, before
these the terms  y and  x vanish, and this equation will be left :

 zz   yz   xz   yy   xy   xx  0,

from which there may be elicited :

  y   x  ((   4 ) yy  (2  y  4 ) xy  (  4 ) xx)

z .
2
Therefore the nature of the part extending to infinity is expressed by this equation.

106. Therefore if as the surface may have a part extended to infinity, that will agree with
the infinite part of the surface, which is expressed by this equation :

 zz   yz   xz   yy   xy   xx  0,

thus so that this surface shall be as if the asymptote of the surface expressed by the
general equation. Because truly in this equation the three variables everywhere have two
dimensions, this will be the equation for the a conical surface, but having the vertex at the
start of the coordinates, where all vanish at the same time ; therefore it will always be
possible to show the conical surface, which will be the asymptote of the proposed
surface, if indeed it may be extended to infinity, or of which the infinite part will agree
with the proposed surface either completely, or only an interval from which the finite part
has been removed. Therefore as we may distinguish the branches of curves going off to
infinity by right lines, thus it will be allowed to distinguish the parts of surfaces extending
to infinity by the asymptotic surfaces of cones.
 EULER'S
INTRODUCTIO IN ANALYSIN INFINITORUM VOL. 2
Appendix 5 On Surfaces.
Translated and annotated by Ian Bruce. page 746

107. Therefore as often as the asymptote conic surface shall be real, so also is the surface
itself extended to infinity and thus indeed, so that the infinite parts of each shall be in
agreement; and thus from the nature of the asymptote surface the nature of the proposed
surface itself will be able to be deduced. But if the asymptote surface were made
imaginary, that surface will have no part extended to infinity, but the whole will be
enclosed in a finite space. Therefore for surfaces of the second order, which may be
contained in a finite space, there is only a need for investigating, as we may consider, in
which cases the asymptotic equation for the surface becomes imaginary ; which comes
about, if this whole surface shall vanish in single point. And so that if any extension may
be had or a point be placed beyond the vertex, by necessity it must be expanded to
infinity, therefore as above we have shown a whole right cone, which may be drawn
through the vertex and one point of the surface, is to be put on that surface itself.

108. Therefore when a conical surface expressed by this equation

 zz   yz   xz   yy   xy   xx  0,

will be changed into a single point, all its sections made through the vertex must vanish
equally in the same point. Therefore in the first place by making z  0 the equation
 yy   xy   xx  0, must be impossible, unless both x  0 and y  0 , which comes
about, if 4 were greater than  . Then likewise it must arise on putting either
x  0 or y  0 ; therefore 4 will be greater than  and 4 greater than  .
Therefore, unless in the equation for the surface of the second order

 zz   yz   xz   yy   xy   xx   z   y   x    0,

4 should be greater than  , 4 greater than  , 4 and greater than  , the
surface certainly will have parts extending to infinity.

109. Indeed nor are these three conditions sufficient for the surface to be enclosed in a
finite space; it is required in addition, that the value of z elicited from the above equation
for the asymptote shall be made imaginary ; which will come about, if that expression

(   4 ) yy  2   y  2  xy     4  xx

may maintain a negative value always, if indeed some values other than 0 may be
substituted for each of the variables x and y. Which comes about, since
  4 and   4 shall be negative quantities, if (   2 )2 shall be made less
than (   4 )    4  ; that is, if  2   2   2 were less than   4 , if
indeed  had a positive value, because we have divided that equation by  . But if
indeed  may have a positive value, on account of the above equations ( 4 greater
 EULER'S
INTRODUCTIO IN ANALYSIN INFINITORUM VOL. 2
Appendix 5 On Surfaces.
Translated and annotated by Ian Bruce. page 747
than  , 4 greater than  and 4 greater than  ) the coefficients  and  will be
positive.

110. Therefore surfaces of the second order will be contained in a finite space, if in its
equation the four following conditions may be put in place, clearly if there shall be :

4 greater than  , 4 greater than  , 4 greater than 

and

 2   2   2 shall be less than   4 .

And hence we have defined the first class of surfaces of the second order, to which all
these kinds belong, which do not run off to infinity, but which may be enclosed in a finite
space. Therefore the sphere belongs to this class, of which the equation is

zz  yy  xx  aa ,

for since here there shall be   1,   1,   1,   0,   0,   0 , it will satisfy the four

conditions found. Truly here this more general equation will pertain

 zz   yy   xx  aa ,

which, if  ,  ,  were positive quantities, always is for a closed surface, unless one or
two coefficients may vanish.

111.With these four conditions examined, by which the surface is restrained in a finite
space, if some determined equation of the second order may be proposed, it will be able
to be judged at once, whether the surface expressed by that equation may have parts
extending to infinity or not. But if in fact one of these four conditions may be absent, the
surface certainly is extended to infinity. But in this case some subdivisions are required to
be made, in which the individual variety in the parts extended to infinity is introduced.
The first subdivision therefore may be put in place, if

     were greater than   4 ,

in which case the surface may be extended to infinity and it will have a conical surface
for an asymptote, as we have now shown previously. And this case is the opposite
direction from the preceding diameter, by which the whole surface is contained in a finite
space.
 EULER'S
INTRODUCTIO IN ANALYSIN INFINITORUM VOL. 2
Appendix 5 On Surfaces.
Translated and annotated by Ian Bruce. page 748

112. But in addition certain intermediate cases are given, in which the surface will depart
to infinity, yet in a similar manner maintains a place between the preceding two cases, in
which a parabola may be contained between the ellipse and the hyperbola. This case
arises, if there were

        4 ,

and therefore there will be

 z    y   x  y (   4 )  x (  4 ).

Therefore the asymptotic equation

 zz   yz   xz   yy   xy   xx  0

will have two simple factors, which will be either real or imaginary, or equal to each
other. Therefore that threefold difference provides three kinds of surfaces extending to
infinity, and thus altogether we have arrived at five kinds of surfaces of the second order,
which now we will pursue more diligently.

113. Because with a change in the position of the three axes, to which the coordinates are
parallel, it will be possible to reduce the general equation to a simpler form, we may use
that reduction, so that we may reduce the general equation for surfaces of the second
order to the most simple form, which still includes all the kinds equally and general
among themselves. Therefore since the general equation for surfaces of the second order
shall be

 zz   yz   xz   yy   xy   xx   z   y   x    0,

we may seek an equation between three other coordinates p, q and r, which indeed will
cross each other at the same point as the three prior coordinates. Towards this end from §
92 the equations may be put in place :

x  p (cos.k  cos.m  sin .k  sin .m  cos.n)  q(cos.k  sin .m  sin .k  cos.m  cos.n)  r  sin.k·sin.n,

y   p (sin.k  cos.m  cos.k  sin.m  cos.n)  q  sin.k  sin.m  cos.k  cos.m  cos.n   r·cos.k  sin.n,

and
z   p  sin.m  sin.n  q  cos.m  sin.n  r  cos.n,

from which this equation will result :

 EULER'S
INTRODUCTIO IN ANALYSIN INFINITORUM VOL. 2
Appendix 5 On Surfaces.
Translated and annotated by Ian Bruce. page 749
App  Bqq  Crr  Dpq  Epr  Fqr  Gp  Hq  Ir  K  0.

114. Now these arbitrary angles k, m and n can be defined thus, so that the three
coefficients D, E and F may vanish. Nevertheless the calculation indeed may become
exceedingly involved, in order that the determination of these angles may actually be
shown, still, if perhaps this may be doubted, or that the elimination may lead to real
values of these angles always, this certainly must be conceded that perhaps two of the
coefficients D and E can be reduced to equal zero. But if this were effected, the position
of the third axis, to which the ordinates r are parallel in the plane normal to the ordinates
p , thus can be changed easily, so that the coefficient F shall vanish also. For there may
be put in place

q  t  sin. i  u  cos. i and r  t  cos. i  u  sin. i ,

thus so that in place of the term qr the new term tu may be introduced, the coefficient of
which with the aid of the angle i will become equal to zero. Therefore in this manner the
general equation for surfaces of the second order may be produced according to this form
:

App  Bqq  Crr  Gp  Hq  Ir  K  0 .

115. Now in addition the coordinates p, q, r will be able to be increased or diminished by

some given amounts, so that the coefficients G, H and I shall vanish ; which may be done
only by changing that point, from which all the coordinates have their beginning. And in
this way all surfaces of the second order will be contained in this equation

App  Bqq  Crr  K  0 ,

from which it is understood each one of the three principal planes drawn through the start
of the coordinates bisects the surface into two similar and equal parts.
Therefore all the surfaces of the second order not only have one plane through the
diameter, but thus three, which will mutually intersect each other through the same point ;
which point therefore will constitute the centre of the surface, even if in some cases this
centre may be at an infinite distance, evidently in a like manner, by which all the conic
sections are said to be provided with a centre, even if in the parabola the centre may be
infinitely removed from the vertex.
 EULER'S
INTRODUCTIO IN ANALYSIN INFINITORUM VOL. 2
Appendix 5 On Surfaces.
Translated and annotated by Ian Bruce. page 750

116. Therefore on being led to an equation of the simplest form, in which all the surfaces
of second order are contained, this equation will show the first kind of these surfaces :

App  Bqq  Crr  aa,

if indeed all three coefficients A, B and C may possess

positive values. Therefore surfaces belonging to this
first kind not only will be enclosed in a finite space, but
will have a centre also, in which the three diametrical
planes cross each other mutually at right angles. Let C
be the centre of this figure (Fig. 143) and CA, CB, CD
these principal axes normal between themselves, to
which the coordinates p, q, r are parallel, the three
diametrical planes will be ABab, ADa and BDb, by
which this body will be cut into two equal similar parts.

117. On putting r  0 , the equation App  Bqq  aa expresses the nature of the
principal section ABab; which therefore will be an ellipse having the centre at C, of
which the semiaxes will be
a a
CA  Ca  and CB  Cb  .
A B

If there may be put q  0 , the equation App  Crr  aa will be for the principal section
ADa, which equally will be an ellipse having the centre at C, the semiaxis of which will
be

a a
CA  Ca  and CD 
A C

Moreover on putting p  0 , the equation for the third principal section BDb will be
Bqq  Crr  aa , which also will be an ellipse having the centre at C and the semiaxis

a a
CB  Cb  and CD  .,
B C

But with these three principal sections known or only with the semiaxes of these

a a a
CA  , CB  and CD  ,
A B C
 EULER'S
INTRODUCTIO IN ANALYSIN INFINITORUM VOL. 2
Appendix 5 On Surfaces.
Translated and annotated by Ian Bruce. page 751
the nature of this body is determined and known. Hence it will be agreed to call this first
genus [kind] of surfaces of the second order ellipsoids, because its three principal
sections are ellipses.

118. Within this kind three species are held more noteworthy from the others. The first is,
if all three principal axes CA, CB and CD were equal to each other, in which case the
three principal sections will change into circles and the body itself into a sphere, the
equation of which, as we have observed above, will be

pp  qq  rr  aa .

The second species includes these cases, in which only two of the principal axes are equal
to each other. Without doubt there shall be CD  CB or C  B , and the section BDb will
become a circle, moreover from the equation

App  B(qq  rr )  aa

it is understood that all the sections parallel to this equally will be circles ; so that the
body will be a spheroid or an oblong, if AC shall be greater than BC, or shortened, if AC
shall be less than BC. And finally the species includes these bodies, in which the
coefficients A, B, C are unequal, which thus generally will retain the name of ellipsoids.

119. The following kinds of surfaces of the second order will be contained in this
equation

App  Bqq  Crr  aa ,

and indeed in the first place, if none of the coefficients A, B, C in short may be absent,
but one or two of these may have negative
values. Let one only be negative and we will
consider this equation

App  Bqq  Crr  aa ,

in which now we put A, B, and C to denote

positive numbers. So that it is constrained to
the centre of this body and to the diametrical
plane, everything shall be prepared in the
same way as before. Therefore it is apparent
(Fig. 144) the first principal section ABab of
the body is an ellipse, of which the semiaxis
 EULER'S
INTRODUCTIO IN ANALYSIN INFINITORUM VOL. 2
Appendix 5 On Surfaces.
Translated and annotated by Ian Bruce. page 752
a a
AC  and the other BC  . The two remaining principal sections Aq and BS will
A B
a
be hyperbolas having the centre at C and the conjugate semiaxes  .
C

120. Therefore this surface will represent a kind of funnel, diverging along hyperbolas
upwards and downwards. So that this asymptotic surface will have a vertical cone
expressed by the equation App  Bqq  Crr  0 , with the centre had at C, and the sides of
which are the asymptotes of hyperbolas. Moreover this asymptote cone will stand
between the surface and a right cone, if there were A  B , truly a scale cone, if A will not
be equal to B. But the axis of the right cone will be the normal CD to the plane ABa.
Moreover all the sections normal to the axis CD will be ellipses similar to the ellipse
ABab, truly all the sections normal to the plane section ABab will be hyperbolas ; from
which these surfaces may be agreed to be called elliptic-hyperboloids, circumscribed by
their asymptotic cone. The surfaces of which for us therefore constitute the second kind.

121. Again three species [or sorts] can be noted in this genus, the first of which will be, if
a  0 , in which case the ellipse ABab vanishes to a point and the hyperbolas will change
into right lines, truly the surface will be merged completely with its asymptote, from
which this first species will include all the cones, either right or scalene ; so that a new
subdivision can be made. There will be another species if A  B , in which case the ellipse
ABab is changed into a circle and the surface becomes rounded or turned [i.e. a surface of
revolution]. Clearly this surface will arise, if some hyperbola may be turned about its
conjugate axis. The third species will not be described from that kind.

122. We may define the third genus, if two coefficients of the terms pp, qq and rr
become negative, the equation of which therefore shall be :

App  Bqq  Crr  aa .

Therefore putting r  0 the first principal section (Fig. 145) will be the hyperbola EAFeaf
a
having the centre at C, the transverse semiaxis of which will be  and the conjugate
A
 EULER'S
INTRODUCTIO IN ANALYSIN INFINITORUM VOL. 2
Appendix 5 On Surfaces.
Translated and annotated by Ian Bruce. page 753
a
semiaxis  . The other principal section on putting q  0 equally will be the
B
hyperbola AQ, aq provided with the same transverse semiaxis, but the semiconjugate axis
a
of which will be  ; the third principal axis becomes imaginary. Finally the whole
C
surface will be placed within the surface of the asymptotic cone, from which this genus
can be called a hyperbolic-hyperboloid inscribed in the asymptotic cone. If there be
made B  C , the surface will be round, arising from the conversion of the hyperbola
about it transverse axis, in which case it shall constitute a particular species. But if there
may be put a  0 , a conical surface arises, which now we have considered as a specimen
of the preceding kind.

123. Towards recognizing the following kinds we may put one of the coefficients A, B,
and C to vanish. Therefore let C  0 , and the general equation found § 114 will be

App  Bqq  Gp  Hq  Ir  K  0 ,

in which by increasing or diminishing the ordinates p and q it will be able to remove the
terms Gp and Hq, but not indeed Ir. Therefore the term Ir will remain in the equation ;
truly with the help of which the final term K will be able to be removed, from which we
will have an equation of this kind :

App  Bqq  ar ,

two cases of which are to be considered. The former, if each coefficient A and B were
positive, the latter, if either shall be negative. But in each case the centre of the surface
will be placed on the axis CD, but removed to an infinite distance.

124. At first both the coefficients A and B shall be positive ; in which case the fourth
genus may be put in place contained in this equation :

App  Bqq  ar .

Therefore the first principal section arising (Fig. 146), if there may be put r  0 , will
vanish to a point ; the second on putting q  0 and the
third on putting p  0 each will be a parabola, truly
MAm and NAn. Therefore since all the sections of this
surface shall be ellipses normal to the axis AD, truly
the sections made through this axis itself make
parabolas, and we will call bodies of this kind
elliptic-paraboloids. Two species of which are to be
noted, the one if there shall be A  B , in which case a
rounded body arises that we call a parabolic-conoid ;
 EULER'S
INTRODUCTIO IN ANALYSIN INFINITORUM VOL. 2
Appendix 5 On Surfaces.
Translated and annotated by Ian Bruce. page 754
and truly the other, if a  0 , and there becomes

App  Bqq  bb ,

which give cylinders both right, if A  B , as well as scalene, if A and B were

unequal.

125. The fifth genus will be contained in this equation

App  Bqq  ar ,

the first principal section of which (Fig. 147) on making r  0 will be two right lines Ee,
Ff mutually crossing each other at the point A.
Truly all the sections parallel to this will be
hyperbolas having their centres on the axis
AD and within the right asymptotes Ee and Ff
put in place . Therefore the two planes, which
stand normally to the plane ABC in the lines
Ee and Ff, will be congruent with the
proposed surface at infinity and thus this
surface will have two planes mutually
crossing each other for an asymptote. The
remaining principal sections made in the planes ACD and ABd will be parabolas,
from which we will call the surfaces pertaining to this genus parabolic-hyperboloids
having two planes for asymptotes, the species of which (if a  0 , so that there shall be
App  Bqq  bb ), will be the cylindric- hyperboloid, all the sections of which normal to
the axis AD will be hyperbolas equal to each other ; if in addition there shall be b  0 ,
these two asymptotic planes themselves arise.

126. Finally the sixth genus of surfaces of the second order will include this equation

App  aq ,

which provides the cylindrical paraboloid, all the sections of which normal to the axis
AD will be similar and equal parabolas, thus so that the vertices of the individual
parabolas lie on the right line AD, and the axes shall be parallel to each other. Therefore
all six kinds of surfaces of the second order can be reduced to these, thus so that none
may be shown, which may not be contained in one of these kinds. Moreover, if in the
final kind there is made a  0 , so that there shall be App  bb , this equation will provide
two planes parallel to each other, which will constitute as if a specimen of this kind.
Clearly the similarity used here applies to lines of the second order, where we have seen
two right lines crossing each other constituting a kind of hyperbola, but two parallel lines
constitute a kind of parabola
 EULER'S
INTRODUCTIO IN ANALYSIN INFINITORUM VOL. 2
Appendix 5 On Surfaces.
Translated and annotated by Ian Bruce. page 755
127. Nevertheless we have formed these six kinds from the simplest equation, to which a
surface of the second order is allowed to be reduced, yet now, if some equation of the
second order shall be proposed, it will be easy to assign the genus to which the surface
may belong. So that if indeed this equation were proposed

 zz   yz   xz   yy   xy   xx   z   y   x    0.

a judgment will be desired from the upper terms, in which the dimensions of two
variables occur ; clearly these terms are to be examined :

 zz   yz   xz   yy   xy   xx ,

in which if 4 were greater than  , 4 were greater than  , 4 were greater
than  and      were less than   4 , the surface will be closed and
to belong to the first kind, which we have called ellipsoids.

128. If one or more of these conditions may be absent and nor yet shall there be
        4 , the surface will belong to either the second or the third
genus and there a hyperbolic body will be provided with an asymptotic cone and that will
be either circumscribed in the second genus, or inscribed in the third genus.
But, of there were
        4 ,

in which case the expression

 zz   yz   xz + yy   xy   xx

can be resolved into two simple factors, either imaginary or real, in the former case the
surface will belong to the fourth genus, in the latter truly to the fifth. But if finally this
expression may have two equal factors or it shall be a square, then the sixth genus will
arise. And thus at once it will be easy to decide, to which genus some equation proposed
may belong ; yet it will be more difficult to decide between the second genus and third,
which both thus may be put together into one.

129. Surfaces of the third and of the following orders will be treated and divided into
kinds in a similar manner. Clearly so many terms of the general equation must be
considered and consequently for surfaces of the third order these, in which the
coordinates possess three dimensions, which will be

 z 3   yzz   yyz   xxz   xzz   xyz  etc.

Therefore initially it is required to be seen clearly, whether or not these terms be taken
jointly, or the greatest member [i.e. the terms of the greatest power]of the equation may
 EULER'S
INTRODUCTIO IN ANALYSIN INFINITORUM VOL. 2
Appendix 5 On Surfaces.
Translated and annotated by Ian Bruce. page 756
be able to be resolved into simple factors. If the resolution into factors may be rejected,
the surface will have a cone of the third order for an asymptote. But because the nature of
this cone is expressed by the uppermost member to be equal to zero, several cones of the
third order of this kind will be given, from the diversity of which hence several kinds of
surfaces will be put in place. For just as cones of the second order all refer to the same
genus, because they are either right or scalene, in the third order much greater varieties
still will find a place.

130. Therefore from these kinds put in place, the cases are required to be considered, in
which the greatest member may be resolved into simple factors, whether they shall be
real or imaginary. In the first place it may have one simple factor, which will be real ;
from that the surface will have a plane asymptote. The other factor put equal to zero
either will give a possible equation or not : if the equation were impossible, unless all the
coordinates may vanish, there will be a single plane asymptote, but if that shall be
impossible, the surface will have two asymptotes, the first a plane and the second a cone
of the second order. But if it may have three simple factors, because one is real always, if
the two remaining shall be either imaginary or real, two new kinds will arise. Finally if all
three simple factors shall be real, according as two or all three shall be equal to each
other, now two kinds of surfaces will be able to be put in place. But none is given in this
order of surfaces, which does not extend to infinity.
 EULER'S
INTRODUCTIO IN ANALYSIN INFINITORUM VOL. 2
Appendix 5 On Surfaces.
Translated and annotated by Ian Bruce. page 757
CAPUT V

DE SUPERFICIEBUS SECUNDI ORDINIS

101. Constitutis ergo superficierum ordinibus secundum numerum dimensionum,
quas summae trium coordinatarum x, y, z potestates in aequatione iunctim sumtae
adimplent, si proponatur pro superficie aequatio algebraica, statim assignari potest ordo,
ad quem illa superficies referri debet. Cum igitur omnis superficies primi ordinis ostensa
sit esse plana, in hoc capite superficies secundi ordinis examini subiiciam. In iis autem
maior statim deprehenditur diversitas, quam in lineis secundi gradus, quod quidem cuique
attendenti facile patebit. Operam igitur dabo, ut haec diversa genera distincte exponam.
In ordinibus vero altioribus tantopere multitudo generum increscit, ut ab iis
evolvendis prorsus abstinere debeamus.

102. Quoniam natura superficierum secundi ordinis exprimitur aequatione, in qua

variabiles x, y et z ad duas dimensiones assurgunt, cylindrus et conus tam rectus quam
scalenus et globus, quorum proprietates iam descripsimus, in hoc secundo ordine
continentur. Omnes vero superficies ad hunc ordinem pertinentes comprehenduntur in
hac aequatione generali

 zz   yz   xz   yy   xy   xx   z   y   x    0.

Utcunque enim tres coordinatae acciplantur, aequatio semper in hac forma

continebitur. Varia ergo superficierum huc pertinentium genera a diversa
coefficientium relatione mutua pendebunt, qui, etsi eadem superficies infinitis
aequationibus exprimatur, tamen infinitam variarum superficierum multitudinem
suppeditabunt.

103. Quemadmodum in lineis curvis planis praecipuam divisionem inde desumsimus,

quod vel in infinitum extendantur vel in spatio finito includantur, ita simili modo omnes
superficies ad quemcunque ordinem pertinentes in duas classes dividentur, ad quarum
alteram referemus eas, quae in infinitum abeunt, ad alteram vero, quae in spatio finito
continentur. Ita cylindrus et conus priori classi, globus vero posteriori annumerabitur.
Posterioris quidem classis nulla dabitur superficies in ordinibus imparibus; cum enim
quaelibet superficies imparis ordinis habeat sectiones planas eiusdem ordinis, curvae
autem imparium ordinum omnes in infinitum extendantur, necesse est, ut etiam ipsae
superficies istorum ordinum in infinitum porrigantur.

104. Quoties autem quaepiam superficies in infinitum extenditur, necesse est, ut ad

minimum una trium variabilium x, y et z in infinitum abeat. Quare cum perinde sit,
quaenam hoc casu infinita fieri assumatur, ponamus z fieri infinitam, siquidem
superficies in infinitum porrigatur. Naturam ergo huius partis in infinitum abeuntis
investigaturi ponamus esse z   , atque nunc potissimum spectari debet terminus primus
 zz , utrum is adsit an vero deficiat. Adsit ergo primum iste terminus in aequatione atque
 EULER'S
INTRODUCTIO IN ANALYSIN INFINITORUM VOL. 2
Appendix 5 On Surfaces.
Translated and annotated by Ian Bruce. page 758
prae eo termini  z et  evanescent habiturque pro parte in infinitum excurrente haec
aequatio

 zz   yz   xz   yy   xy   xx   y   x  0,

ex qua porro omnes termini, qui non sunt infiniti vel infinities minores saltem
quam  zz , evanescunt.

105. Statuamus omnes terminos, in quibus variabiles duas tenent dimensiones,

adesse; quaecunque enim fuerit superficies, in eius aequatione generalissima semper
omnes inerunt termini summarum dimensionum neque idcirco hypothesis, qua omnes
terminos duarum dimensionum adesse ponimus, universaliti solutionis ullam vim infert.
Quando autem termini yz et xz adsunt, prae iis termini  y et  x evanescunt relinqueturque
haec aequatio

 zz   yz   xz   yy   xy   xx  0,

ex qua elicitor

  y   x  ((   4 ) yy  (2 y  4 ) xy  (  4 ) xx)

z .
2

Hac igitur aequatione natura portionis in infinitum extensae exprimitur.

106. Si quam igitur superficies habeat portionem in infinitum extensam, ea congruet cum
portione infinita superficiei, quae exprimitur hac aequatione

 zz   yz   xz   yy   xy   xx  0,

ita ut haec superficies sit quasi asymptota illius superficiei aequatione generali
expressae. Quia vero in hac aequatione tres variabiles ubique duas habent
dimensiones, erit ea pro superficie conica, verticem in initio coordinatarum, ubi omnes
simul evanescunt, habente; semper ergo exhiberi potest superficies conica, quae erit
asymptota superficiei propositae, siquidem in infinitum extenditur, seu cuius portio
infinita cum superficie proposita vel penitus congruit, vel intervallo tantum finito ab eo
est remota. Uti ergo ramos curvarum in infinitum abeuntes per lineas rectas asymptotas
distinximus, ita superficierum partes in infinitum extensas per superficies conicas
asymptotas distinguere licebit.

107. Quoties ergo superficies asymptota conica erit realis, toties superficies ipsa in
infinitum extenditur atque ita quidem, ut utriusque partes infinitae congruant; sicque ex
natura superficiei asymptotae natura ipsius superficiei propositae colligi poterit. Quodsi
autem superficies asymptota fiat imaginaria, ipsa superficies nullam habebit partem in
 EULER'S
INTRODUCTIO IN ANALYSIN INFINITORUM VOL. 2
Appendix 5 On Surfaces.
Translated and annotated by Ian Bruce. page 759
infinitum extensam, sed tota spatio finito includetur. Ad superficies ergo secundi ordinis,
quae in spatio finito contineantur, indagandas tantum opus est, ut videamus, quibus in
casibus aequatio pro superficie asymptota fiat imaginaria; quod fit, si tota superficies
haec in punctum unicum evanescit. Namque si ullam extensionem haberet vel punctum
extra verticem situm, necessario in infinitum expandi deberet, propterea quod supra
ostendimus totam rectam, quae per verticem et unum superficiei punctum ducitur, in ipsa
superficie esse positum.

108. Quando ergo superficies conica asymptota hac aequatione expressa

 zz   yz   xz   yy   xy   xx  0,

in unicum punctum abit, omnes eius sectiones per verticem factae pariter in idem
punctum evanescere debent. Primum ergo facto z  0 aequatio  yy   xy   xx  0,
debet esse impossibilis, nisi sit x  0 et y  0 , quod evenit, si fuerit 4 maior quam
 . Deinde idem evenire debet posito x  0 vel y  0 ; erit ergo 4 maior quam  et
4 maior quam  . Nisi ergo in aequatione pro superficie secundi ordinis

 zz   yz   xz   yy   xy   xx   z   y   x    0.

fuerit 4 maior quam  , 4 maior quam  , 4 maior quam  , superficies
certo habebit partes in infinitum extensas.

109. Neque vero hae tres conditiones sufficiunt ad superficiem in spatium finitum
includendam; requiritur insuper, ut valor ipsius z ex aequatione asymptotica supra erutus
fiat imaginarius; quod fit, si ista expression

(   4 ) yy  2   y  2  xy     4  xx

perpetuo obtineat valorem negativum, siquidem pro utraque variabili

x et y valores quicunque praeter 0 substituantur. Quod, cum   4 et   4
sint quantitates negativae, fiet, si (   2 ) 2 minor quam (   4 )    4 
; hoc est, si fuerit  2   2   2 minor quam   4 , siquidem  habuerit
valorem affirmativum, quoniam illam aequationem per  divisimus. Quodsi vero 
habeat valorem affirmativum, ob superiores aequationes ( 4 maior quam  , 4
maior quam  et 4 maior quam  ) coefficientes  et  erunt affirmativi.

110. Superficies ergo secundi ordinis in spatio finito continebitur, si in eius aequatione
quatuor sequentes conditiones locum habeant, nempe si sit

4 maior quam  , 4 maior quam  , 4 maior quam 

 EULER'S
INTRODUCTIO IN ANALYSIN INFINITORUM VOL. 2
Appendix 5 On Surfaces.
Translated and annotated by Ian Bruce. page 760
et

 2   2   2 minor quam   4 .

Hincque genus primum superficierum secundi ordinis definimus, ad quod eae species
omnes pertinent, quae non in infinitum excurrunt, sed in spatio finito includuntur. Ad hoc
ergo genus pertinet globus, cuius aequatio est

zz  yy  xx  aa ,

cum enim hic sit   1,   1,   1,   0,   0,   0 , quatuor inventis conditionibus

omnibus satisfit. Generalius vero hic pertinebit aequatio ista

 zz   yy   xx  aa ,
quae, si  ,  ,  fuerint quantitates affirmativae, semper est pro superficie clausa, nisi
unus duove coefficientes evanescant.

111. Perspectis his quatuor conditionibus, quibus superficies in spatium finitum redigitur,
si proponatur aequatio secundi ordinis quaecunque determinata, statim diiudicari poterit,
utrum superficies ea aequatione expressa habeat partes in infinitum extensas an nullas.
Quodsi enim unica illarum quatuor conditionum desit, superficies certo in infinitum
extenditur. Hoc autem casu nonnullae subdivisiones sunt faciendae, quibus singularis
varietas partibus in infinitum extensis inducitur. Prima subdivisio ergo constituatur,
si fuerit
     maior quam   4 ,

quo casu superficies in infinitum extendetur atque superficiem conicam pro

asymptota habebit, uti iam ante ostendimus. Hicque casus e diametro est oppositus
praecedenti, quo tota superficies in spatio finito continetur.

112. Praeterea autem dantur casus quidam intermedii, quibus etsi superficies in infinitum
abit, simili tamen modo inter duos praecedentes locum tenet, quo parabola inter ellipsin
et hyperbolam continetur. Casus iste oritur, si fuerit

        4 ,

eritque propterea

 z    y   x  y (   4 )  x (  4 ).

Habebit ergo aequatio asymptotica

 zz   yz   xz   yy   xy   xx  0
 EULER'S
INTRODUCTIO IN ANALYSIN INFINITORUM VOL. 2
Appendix 5 On Surfaces.
Translated and annotated by Ian Bruce. page 761
duos factores simplices, qui erunt vel reales vel imaginarii vel inter se aequales.
Triplex ista diversitas ergo tria genera superficierum in infinitum extensarum praebet,
sicque omnino quinque genera superficierum secundi ordinis sumus adepti, quae nunc
diligentius prosequemur.

113. Quia mutando positionem ternorum axium, quibus coordinatae sunt parallelae,
aequatio generalis ad formam simpliciorem reduci potest, ista reductione ita utamur, ut
aequationem generalem pro superficiebus secundi ordinis ad formam simplicissimam
redigamus, quae tamen omnes species aeque ac generalis in se complectatur. Cum igitur
aequatio generalis pro superficiebus secundi ordinis sit

 zz   yz   xz   yy   xy   xx   z   y   x    0.

quaeramus aequationem inter alias ternas coordinatas p, q et r, quae quidem se mutuo in

eodem puncto, quo ternae priores, se decussent. Ad hoc ex § 92 statuatur

x  p (cos.k  cos.m  sin .k  sin .m  cos.n)  q(cos.k  sin .m  sin .k  cos.m  cos.n)  r  sin.k·sin.n

et

y   p (sin.k  cos.m  cos.k  sin.m  cos.n)  q  sin.k  sin.m  cos.k  cos.m  cos.n   r·cos.k  sin.n

atque

z   p  sin.m  sin.n  q  cos.m  sin.n  r  cos.n,

unde resultet ista aequatio

App  Bqq  Crr  Dpq  Epr  Fqr  Gp  Hq  Ir  K  0.

114. Iam anguli illi arbitrarii k, m et n ita definiri poterunt, ut tres coefficientes D, E et F
evanescant. Quanquam enim calculus nimis fit prolixus, quam ut angulorum illorum
determinatio actu ostendi possit, tamen, si quis forte dubitet, an semper ista eliminatio ad
valores reales angulorum illorum perducat, is certe concedere debebit duos saltem
coefficientes D et E nihilo aequales reddi posse. Hoc autem si fuerit effectum, positio
tertii axis, cui ordinatae r sunt parallelae in plano ad ordinatas p normali, facile ita mutari
potest, ut etiam coefficiens F evanescat. Statuatur enim

q  t  sin. i  u  cos. i et r  t  cos. i  u  sin. i ,

 EULER'S
INTRODUCTIO IN ANALYSIN INFINITORUM VOL. 2
Appendix 5 On Surfaces.
Translated and annotated by Ian Bruce. page 762
ita ut loco termini qr novus terminus tu ingrediatur, cuius coefficiens ope anguli i nihilo
aequalis fieri poterit. Hoc igitur modo aequatio generalis pro superficiebus secundi
ordinis ad hanc formam perducetur

App  Bqq  Crr  Gp  Hq  Ir  K  0 .

115. Nunc praeterea coordinatae p, q, r datis quantitatibus ita augeri diminuive poterunt,
ut coefficientes G, H et I evanescent ; quod fiet mutato tantum puncto illo, unde omnes
coordinatae initium habent. Atque hoc modo omnes superficies secundi ordinis in hac
aequatione continebuntur
App  Bqq  Crr  K  0 ,

ex qua intelligitur unumquodque trium planorum principalium per initium

coordinatarum ductorum superficiem in duas partes similes et aequales bisecare.
Omnis ergo superficies secundi ordinis non solum unum habet planum diametrale, sed
adeo tria, quae se mutuo in eodem puncto normaliter intersecent; quod punctum propterea
centrum superficiei constituet, etiamsi in nonnullis casibus hoc centrum in infinitum
distet, simili scilicet modo, quo omnes sectiones conicae centro dicuntur praeditae,
etiamsi in parabola centrum a vertice infinite removeatur.

116. Perducta ergo aequatione, qua omnes superficies secundi ordinis continentur, ad
formam simplicissimam primum harum superficierum genus exhibebit ista aequatio

App  Bqq  Crr  aa,

siquidem omnes tres coefficientes A, B et C valores

obtineant affirmativos. Superficies igitur ad hoc primum
genus pertinentes non solum totae in finito spatio
includentur, sed omnes quoque centrum habebunt, in
quo tria plana diametralia se mutuo ad angulos rectos
decussant. Sit (Fig. 143) C centrum huius figurae et CA,
CB, CD axes illi principales inter se normales, quibus
coordinatae p, q, r sunt parallelae, erunt tria plana
diametralia ABab, ADa et BDb, quibus hoc corpus in
binas portiones similes aequales secabitur.

117. Ponatur r  0 et aequatio App  Bqq  aa exprimet naturam sectionis principalis

ABab; quae idcirco erit ellipsis centrum habens in C, cuius semiaxes erunt

a a
CA  Ca  et CB  Cb  .
A B
 EULER'S
INTRODUCTIO IN ANALYSIN INFINITORUM VOL. 2
Appendix 5 On Surfaces.
Translated and annotated by Ian Bruce. page 763
Si ponatur q  0 , aequatio App  Crr  aa erit pro sectione principali ADa, quae pariter
erit ellipsis centrum habens in C, cuius semiaxes erunt

a a
CA  Ca  et CD 
A C

Posito autem p  0 prodibit pro tertia sectione principali BDb aequatio Bqq  Crr  aa ,
quae etiam erit ellipsis centrum habens in C et semiaxes

a a
CB  Cb  et CD  .,
B C

Cognitis autem his tribus sectionibus principalibus seu tantum earum semiaxibus

a a a
CA  , CB  et CD  ,
A B C

natura huius corporis determinatur et cognoscitur. Hinc primum istud superficierum

secundi ordinis genus elliptoides appellari conveniet, quia tres eius sectiones principales
sunt ellipses.

118. Sub hoc genere continentur tres species prae primis notatu dignae. Prima est, si
omnes tres axes principales CA, CB et CD inter se fuerint aequales, quo casu tres
sectiones principales abibunt in circulos ipsumque corpus in globum, cuius aequatio, uti
supra vidimus, erit

pp  qq  rr  aa .

Secunda species eos complectitur casus, quibus duo tantum axes principales sunt inter se
aequales. Sit nimirum CD  CB seu C  B , atque sectio BDb fiet circulus, ex aequatione
autem
App  B(qq  rr )  aa

intelligitur omnes sectiones huic parallelas pariter fore circulos; unde hoc corpus erit
sphaeroides sive oblongum, si AC maior sit quam BC, sive compressum, si AC sit minor
quam BC. Tertia denique species ea complectitur corpora, in quibus coefficientes A, B, C
sunt inaequales, quae ideo nomen generale elliptoidis retinebunt.

119. Sequentia genera superficierum secundi ordinis hac continebuntur aequatione

App  Bqq  Crr  aa ,

 EULER'S
INTRODUCTIO IN ANALYSIN INFINITORUM VOL. 2
Appendix 5 On Surfaces.
Translated and annotated by Ian Bruce. page 764
ac primo quidem, si nullus coefficientium A, B, C prorsus desit, eorum autem vel unus
vel duo valores habeant negativos. Sit unus
tantum negativus atque consideremus hanc
aequationem
App  Bqq  Crr  aa ,

in qua iam A, B, C numeros affirmativos

denotare ponimus. Quod ad centrum huius
corporis et plana diametralia attinet, omnia
eodem modo sunt comparata ut ante. Patet
igitur (Fig. 144) huius corporis sectionem
principalem primam ABab esse ellipsin,
a
cuius semiaxis AC  alterque
A
a
BC  . Binae reliquae sectiones principales Aq, BS erunt hyperbolae centrum in C et
B
a
semiaxem coniugatum  habentes.
C

120. Repraesentabit ergo haec superficies speciem infundibuli, sursum et deorsum

secundum hyperbolas divergens. Unde ista superficies asymptoton habebit conum
aequatione App  Bqq  Crr  0 expressum, verticem, in centro C habentem, et cuius
latera sunt asymptota hyperbolarum. Stabit autem iste conus asymptotos intra superficiem
eritque conus rectus, si fuerit A  B , scalenus vero, si A non ipsi B aequabitur. Axis
autem coni erit recta CD normalis ad planum ABa. Ceterum omnes sectiones axi CD
normales erunt ellipses similes ellipsi ABab, sectiones vero plano ABab normales
omnes erunt hyperbolae; unde istas superficies elliptico-hyperbolicas vocari
conveniet, asymptoto suo conico circumscriptas. Huius igitur superficies nobis
constituent genus secundum.

121. Species in hoc genere iterum tres notari poterunt, quarum prima erit, si a  0 , quo
casu ellipsis ABab in punctum evanescit et hyperbolae in lineas rectas abibunt,
superficies vero ipsa cum asymptota sua penitus confundetur, ex quo haec prima species
complectetur omnes conos, sive rectos sive scalenos; unde nova subdivisio fieri posset.
Altera species erit, si fiat A  B , quo casu ellipsis ABab in circulum mutatur et ipsa
superficies fiet rotunda seu tornata. Orietur scilicet haec superficies, si hyperbola
quaecunque circa axem coniugatum convertatur. Tertia species ab ipso genere non
discrepabit.

122. Tertium genus definiamus, si duo coefficientes terminorum pp,qq et rr fiant

negativi, cuius ergo aequatio sit

App  Bqq  Crr  aa .

 EULER'S
INTRODUCTIO IN ANALYSIN INFINITORUM VOL. 2
Appendix 5 On Surfaces.
Translated and annotated by Ian Bruce. page 765

Posito ergo r  0 erit (Fig. 145) prima sectio principalis hyperbola EAFeaf centrum
a a
habens in C, cuius semiaxis transversus erit  et semiaxis coniugatus  .
A B
Altera sectio principalis posito q  0 pariter erit hyperbola AQ, aq eodem semiaxe
a
transversa praedita, sed cuius axis semiconiugatus erit  ; tertia sectio principalis fit
C
imaginaria. Tota denique haec superficies intra superficiem conicam asymptotam erit sita,
unde hoc genus vocari potest hyperbolico-hyperbolicum cono asymptoto inscriptum. Si
fiat B  C , superficies erit rotunda, orta ex conversione hyperbolae circa suum axem
transversum, quo casu species peculiaris constitui posset. Sin autem ponatur a  0 , oritur
superficies conica, quam iam tanquam speciem generis praecedentis sumus contemplati.

123. Ad sequentia genera cognoscenda ponamus unum coefficientium A, B, C

evanescere. Sit igitur C  0 , atque aequatio generalis § 114 inventa erit

App  Bqq  Gp  Hq  Ir  K  0 ,

in qua augendo seu diminuendo ordinatas p et q termini Gp et Hq, non vero Ir tolli
poterit. Relinquetur ergo terminus Ir in aequatione; eius vero ope tolli poterit terminus
ultimus K, unde eiusmodi aequationem habebimus

App  Bqq  ar ,

cuius duo casus sunt perpendendi. Prior, si uterque coefficiens A et B fuerit affirmativus,
posterior, si alter sit negativus. Utroque autem casu centrum superficiei in axe CD erit
situm, sed ad distantiam infinitam remotum.

124. Sint primo ambo coefficientes A et B affirmativi; quo casu constituatur genus
quartum aequatione hac contentum

App  Bqq  ar .
 EULER'S
INTRODUCTIO IN ANALYSIN INFINITORUM VOL. 2
Appendix 5 On Surfaces.
Translated and annotated by Ian Bruce. page 766
Prima ergo sectio principalis (Fig. 146) oriunda, si ponatur r  0 , in punctum evanescet ;
altera posito q  0 et tertia posito p  0 utraque erit
parabola, nempe MAm et NAn. Cum igitur huius
superficiei omnes sectiones ad axem AD normales sint
ellipses, sectiones vero per hunc ipsum axem factae
parabolae, huius generis corpora elliptico-parabolica
appellabimus. Cuius species sunt notandae duae,
altera, si A  B , quo casu oritur corpus rotundum,
conoides parabolicum vocatum; fitque altera vero, si
a  0 , fitque
App  Bqq  bb ,

quae dat cylindros tam rectos, si A  B , quam scalenos, si A et B fuerint

inaequales.

125. Quintum genus continebitur hac aequatione

App  Bqq  ar ,

cuius (Fig. 147) sectio principalis prima facto r  0 erunt duae lineae rectae Ee, Ff se
mutuo in puncto A decussantes. Omnes vero sectiones huic parallelae erunt hyperbolae
sua centra in axe AD habentes et intra asymptotas rectas Ee et Ff constitutae. Duo igitur
plana, quae plano ABC in lineis Ee et Ff normaliter insistunt, in infinitum cum superficie
proposita congruent ideoque haec superficies pro asymptoto habebit duo plana se mutuo
decussantia. Sectiones reliquae principales in planis ACD et ABd factae erunt parabolae,
unde superficies ad hoc genus pertinentes vocabimus parabolico-hyperbolicas duo plana
pro asymptotis habentes, cuius species (si a  0 , ut sit App  Bqq  bb ), erit cylindrus
hyperbolicus, cuius omnes sectiones ad axem
AD normales erunt hyperbolae inter se
aequales; si insuper sit b  0 , oriuntur duo
illa ipsa plana asymptotica.

126. Sextum denique genus superficierum

secundi ordinis complectetur
haec aequatio
App  aq ,
quae praebet cylindrum parabolicum, cuius
omnes sectiones axi AD normales erunt parabolae similes et aequales, ita ut singularum
vertices in rectam AD incidant et axes inter se sint paralleli. Ad haec igitur sex genera
omnes superficies secundi ordinis reduci poterunt, ita ut nulla exhiberi possit, quae non in
uno horum generum contineatur. Ceterum, si in genere ultimo fiat a  0 , ut sit App  bb ,
haec aequatio praebebit duo plana inter se parallela, quae quasi speciem huius generis
 EULER'S
INTRODUCTIO IN ANALYSIN INFINITORUM VOL. 2
Appendix 5 On Surfaces.
Translated and annotated by Ian Bruce. page 767
constituent. Similitudo scilicet hic uti in lineis secundi ordinis obtinet, ubi vidimus duas
rectas se decussantes hyperbolae
speciem, duas autem lineas parallelas parabolae speciem constituere.

127. Quanquam haec sex genera ex aequatione simplicissima, ad quam superficies

secundi ordinis reducere licet, formavimus, tamen nunc facile erit, si aequatio
quaecunque secundi gradus sit proposita, genus assignare, ad quod superficies pertineat.
Quodsi enim proposita fuerit haec aequatio

 zz   yz   xz   yy   xy   xx   z   y   x    0.

iudicium ex supremis terminis, in quibus duae variabilium occurrunt dimensiones,

petendum erit; spectari scilicet debebunt hi termini

 zz   yz   xz   yy   xy   xx ,

in quibus si fuerit 4 maior quam  , 4 maior quam  , 4 maior quam 
et      minor quam   4 superficies erit clausa et ad genus primum,
quod elliptoides vocavimus, pertinebit.

128. Si una pluresve harum conditionum desint neque tamen sit

        4 , superficies vel ad secundum vel ad tertium
genus pertinebit eritque corpus hyperbolicum cono asymptoto praeditum
eique vel circumscriptum in genere secundo vel inscriptum in genere tertio.
At, si fuerit

        4 ,

quo casu expression

 zz   yz   xz + yy   xy   xx

resolvi poterit in duos factores simplices sive imaginarios sive reales, casu priori
superficies pertinebit ad genus quartum, posteriori vero ad quintum. Quodsi denique ista
expressio duos habeat factores aequales seu sit quadratum, tum orietur genus sextum.
Sicque statim facile diiudicari poterit, ad quodnam genus quaevis aequatio proposita
pertineat; difficilius tantum erit iudicium circa genus secundum et tertium, quae ambo
ideo in unum conflari possent.

129. Simili modo superficies tertii et sequentium ordinum pertractari atque in genera
dividi poterunt. Spectari scilicet tantum debebunt aequationis generalis termini supremi et
consequenter pro superficiebus tertii ordinis ii, in quibus coordinatae tres obtinent
dimensiones, qui erunt
 EULER'S
INTRODUCTIO IN ANALYSIN INFINITORUM VOL. 2
Appendix 5 On Surfaces.
Translated and annotated by Ian Bruce. page 768
 z   yzz   yyz   xxz   xzz   xyz  etc.
3

Primum igitur dispiciendum est, utrum hi termini coniunctim sumti seu

aequationis membrum supremum resolvi possit in factores simplices an non.
Si resolutionem in factores respuat, habebit superficies conum tertii ordinis pro
asymptoto. Quia autem natura huius coni exprimitur supremo membro nihilo aequali
posito, plures huiusmodi coni tertii ordinis dabuntur, ex quorum diversitate hinc plura
superficierum genera constituentur. Quamvis enim coni secundi ordinis omnes ad unum
genus referuntur, quia sunt vel recti vel scaleni, tamen in tertio ordine multo maior
varietas locum invenit.

130. Expositis ergo his generibus, considerandi sunt casus, quibus supremum membrum
in factores simplices resolvi potest, sive sint reales sive imaginarii. Habeat primum unum
factorem simplicem, qui erit realis; ex eo superficies habebit asymptotam planam. Alter
factor nihilo aequalis positus vel dabit aequationem possibilem vel non: si aequatio fuerit
impossibilis, nisi omnes coordinatae evanescant, unica erit asymptota plana, sin autem sit
impossibilis, superficies duas habebit asymptotas, alteram planam alteram conum secundi
ordinis. Quodsi habeat tres factores simplices, quia unus semper est realis, si bini reliqui
sint vel imaginarii vel reales, duo nova genera oriuntur. Denique si omnes tres factores
simplices sint reales, prout duo vel omnes sint inter se aequales, adhuc duo genera
constitui poterunt. Nulla autem in hoc ordine datur superficies, quae non in infinitum
extendatur.