~98 Wood.. [ 1- r. 1.

THERMODYNAMICS.

BY PROF. DE VOLSON WOOD.

(Continued .from ~age 209.)

W e next consider the article on intrinsic energy, being article
~47 of The Steam En~ne, by Rankine. T h e intrinsic energy of
a substance is the external work which it is capable of doing in
being expanded from a given state as to temperature and v o l u m e
(or pressure and volume) to that of total privation of heat and in-
definite expansion. Since indefinite expansion is impossible in an
actual machine, the quantity is ideal, but its value m a y be corn-

Fro. 3.
puted for the ideally perfect gas, and the algebraic form of an
expression for its value be found for imperfect fluids.
It will 1~. necessary, in the first case, to find an expression for
the heat absorbed during a change of temperature, as well as of
pressure and volume. In our preceding articles, explaining article
24I , we found that the expression r d P d v was a measure of the

heat absorbed for an expansion in which the temperature, ~-, was

constant ; but this is only one condition out of an infinite n u m b e r
which m a y occur in practice. Suppose that heat be absorbed, so
that during expansion the relation between the pressure and
volume m a y be represented b y the line A B,/~'~. 3," it is required
to find a differential expression for the amount of heat absorbed in
passing from a to e, when a and o ~j'e consecutive points. To find
April, I887.] T/termodynamics. 599

it, we proceed on the same principle t h a t is used in m a n y cases in

g e o m e t r y - - i n s c r i b e a p o l y g o n , whose properties are k n o w n I a n d
pass to the limit. Beginning at a, d r a w an isothermal a t , and
t h r o u g h e the vertical e v', intersecting a r in b ; so that, a t first,
the passage from a to e is along the isothermal a b and the vertical
be. T h e abscissa of b in reference to a will be v v t ~ - d v . Through
a, b and e pass the adiabatics aral, b m 2, em 3, conceived to be
extended indefinitely to the r i g h t ; then will

area ml a b ra s ~ r dd-Pr d v "

T h e area m s b e m s represents the heat absorbed while increasing

the t e m p e r a t u r e at the constant volume v p, the pressure being
increased an a m o u n t b e. In this case, the t e m p e r a t u r e will be
increased an indefinitely small a m o u n t , represented by d r . T h e
determination of the heat necessary to increase the t e m p e r a t u r e b y
one unit (say I ° F ), depends upon an e x p e r i m e n t and a calculation,
the process of which we will assume is known by the reader ; the
result of which is called the specific kcat at constant volume. If
the specific heat is c o n s t a n t l y varying, it is only necessary to con-
sider its rate at b ; that is, it will be such a q u a n t i t y as would
increase the t e m p e r a t u r e t ° , if the specific heat of the substance
remained uniform t h r o u g h o u t the degree.
L e t Kv, = the specific heat of the substance at the state b,
measured in foot-pounds ; then will K,., d v be the heat necessary
to raise the temperature from that at b to that at e, and
area m s b e m 3 ~ Kv, d z-.
U#imatcly, when a and e are consecutive, the area a b c vanishes
and we have, if d H be the heat a b s o r b e d in passing from a to o

dPdv
d H-~- area rai a e m3 ~--~-Kv d r q- ":-dr (1)

which is, substantially, R a n k i n e ' s e q u a t i o n (2), p a g e 3 I o ; and is

precisely the form given b y Clausius for h o m o g e n e o u s bodies.
If, however, the substance is capable of an indefinite expansion,
and if the a g g r e g a t i o n of the substance does not c h a n g e d u r i n g
such expansion, and if b y an isothermal expansion the substance
a p p r o x i m a t e s m o r e and m o r e nearly to a perfect gas, as assumed
by Rankine, t h e n m a y the value of /(v be transformed as follows:
3OC Wood: [J. F. I.,

FIG. 4.

W e have already found, /rzg. 4,

Vrt

m, be~,---- r ( ~ d v ,
.~d r
m~bem3= Kvdr.
Similarly
m4edm~-~- K v , d r
v q

v - - d ~ ...... v

(,l p
d d r2 dd r dr d~Tr ~ d v d r

dp ~'d~ p
= r f~av + r.) dr, ~ d v d r q- f d p el v,

the last term of the preceding expression being omitted, it being

an infinitesimal of a lower order. We also have

,]dr
Substituting,

K v d r + r ('d~P d v + r ~ ' d 2 P - d v . d r q- d r ~'~-!~dv--Kv ,,

dd r d d v2 dd v

_ r ('Je av = a r f dP v
dd r Jd r "
Cancelling and reducing,
v tt v
~'d 2p •
-hv = K v , , - - r ~"d2p d v = K,., + r ] ~ : ~ - a v .
April, t887.] Thermodynamics. 3ol

The difference between the specific heats at the volumes v and

v" is due to the internal work during the expansion along ed,
exceeding that along b e.
If v " be removed so far from v that the gas becomes perfect,
then Kv,. will, according to Rankine's hypothesis, be constant, and
we will represent it by Cv !n place of his old English k. Further,
the expansion necessary to give this result is assumed to be inde-
finite, so that vpl will be infinite, and also the relation between p
and r is assumed to be a continuous function, for otherwise the
integration could not be performed, so that we finally have

d H - ~ C,. d r + d-~ ~ -dr ,

Oo

which is Rankine's equation (2), page 312.

It is well to observe that d v in the second term is involved in a
different operation from that in the third term ; the latter being a
part of the actual operation in passing from a to e, Fig'. 3, while the
former is a purely ideal one for determining a value for the variable
specific heat at v. In order to reduce the second term, the equation
of the gas must be known. Thus, if it be
p=Rr a
V 7" ~2

as it is for carbonic acid gas, then considering v as constant during

the differentiation,
d2 jo 2 a
-d- ~ -r -- r s d'

then integrating, considering r as constant, since the integration is

along the isothermal b e,
v

fd~Pdv= 2af dr_ 2a

ff-z ~ - - -r~ v~ r s v"
and the second term of equation (2) becomes, for this particular gas,

(~)dr

The value of the expression __2a is so small for known gases,

T~ V
 302 Wood : [ J. F. I.,

compared with Ov that it may generally be omitted. Thus, for

carbonic acid gas
a ---- 3"42 X 17264 X 8"1572,
and if the temperature be that of melting ice, 493°.2 F., at the
pressure of one atmosphere, then
2 a _ 2 X 3"42× 17264 X 8"1572 : 0"42.+,
r2 v (493"2) 2 X 81572
while C',, is about 13.). Since it is impossible to find the value of
' Cv with perfect accuracy, it is apparent that 0"42 may be omitted
without sensible error, since it falls within the limits of the errors
of observation.
The integral of equation (2) between the limits belonging to the
conditions of the problem, would give the heat absorbed. But it
is impossible to find the general integral; it may, however, be
found in the following cases:
(I.) When the gas is perfect and the equation of the path
described is known,
(2.) When the gas is imperfect, its equation being known, and
the expansion isothermal.
(3.) W h e n the pressure and temperature are constant during
expansion.
CASE Io

For a perfect gas we have

pv=Rr;

(?Iv
@p

which, substituted in equation (2), gives

dtt_--_Cvdr+ Rr-dv
V

: Cv d r + p d v
which, integrated, gives
Vs
Ha, b ---- Cv (% - - rL) + f p d v. (3)
Vl

The several terms of this equation may be illustrated by means

April, 1 8 8 7 . J Thermodynamtcs. 303:

FIG. 5"

of Fig. 5" L e t A and B be respectively the initial and terminal'

states of the gas~ the line A B being the path of the gas in expand.
ing from v, to %. The last term, or
V~
fpdv
VI
is the external work done, and equals v, A B vs, and since the gas
is perfect, the heat absorbed will be represented by the same area.
Through A draw the isothermal A C, and through A, (7, B, respec-
tively, the adiabatics; then, according to the second law, the heat
absorbed, while gel~erating the path A C, will be
area m 1A C m 2 = vl A C v.,.
The area A C 32 represents at the same time a portion of the
external work and a corresponding equal quantity of heat; hence
ml A B C m s ~--- vl A B %.
The term
Cv (~-~- - ~-,)
will be represented by the area m.2 C B m3, vs being the temperature
of the substance at the state B, and z'l at the states A and C;. hence
the second member will be represented by the area ml A B m3,.
which represents the heat absorbed in passing from the state A t o
the state B, and indicated b y / / a , b.
Let the equation of the path of A B be known, or
p = ~ (v),
in which case the area vtA B y z may be found. To be more
 304 Wood: [ J. F. I,,

specific, let the path be a straight line, whose equation is

p=a+bv,
then
V~

fpdv=f(a+bv)a,,, v1
which is easily reduced.
CASE II.

When the temperature is constant, we have d r ~ o, and equa-

, tion (2) becomes
dr~
dH=r~- r d ~,
Let the equation of the gas be
a
pv=Rr-- --, (4)
TV
then
(5)
=p+2a
7"~2 ;
2a
" . d t I = (P + -~v~ ) d v ;
hence for an isothermal expansion
V2

.
V2

7" V1 V2
v1
the lasteterm of which is the internal work done during the
expansion, and may be represented by area A a b B , Fig'. 6.

FIG. 6.
April, I887.] 7 ]l(rtJlOdJ~lltlTllet 3. 3o5

The external work will be

v~

vx A 13 v 2 = f pd v V T V2
v1

R r log v~
Vl
_a (1 _ :2 )
2" V1

which, substituted in (7), gives

Ha~=Rvlog v~ + a (1 1)

We note here an interesting fact in regard to the actual and the

virtual pressures. Let A B be the isothermal of the gas whose
equation is (4) above. If now it were possible to remove the im-
perfection, all other qualities remaining the same, the ordinate p
would gradually increase, and when a becomes zero, we would
have the equation of a perfect gas, or,
p~R 2-
V

Let c d be the isothermal in this case. Or, to put it another

way, assume a ~ o, and find the corresponding isothermal c d ;
then, from equation (4),
Ao~ a
TV 2.

But the virtual ordinate doing the internal work, is, from equa-
tion (6),
2a
2" V2

from which it appears that

ae~--cA.
As a physical explanation of the manner of producing this
result, conceive that the gas is perfect, then at the volume Vx the
pressure will be eva. Now conceive that the gas undergoes such
a change, so as to become imperfect ; not only would the pressure
fall, as shown by equation (4), but because internal work is being
done, the temperature would fall, and hence to maintain the initial
temperature, the substance must absorb heat, to represent which
the v i r t u a l pressure must be increased, so that while A falls below
e, a will rise above. The distance A e will equal c a only when the
equation of the gas is of the form of equation (4).
306 Ifgod: [ J. F. I.,

CASE Ill.

W h e n p and r are both constant, equation (2) becomes

V2

I1= r d P(d v
d r dV1 -
dp
= ~ a ~- (v~-- ,,~, (8)

which is applicable to cases of fusion and of evaporation.

, Returning now to the general equation (2), we observe that in
dp
t h e third term, the factor v - may, in all cases, be consi~lered as
dr
a virtualpressure, such that being multiplied by d v, gives an ele-
ment of the work, both external and internal, for the expansion
.d v. Subtracting the external work, p d v, gives the internal work
due to expansion; hence adding and s u b t r a c t i n g p d r , equation
~2) becomes
V2

cvd + dv.d + ( J P - - p ) +pdv

•J~d v ~ \ d r
,or, integrating,
V2 ~ V 2

v1

the last term of which represents the internal work done due to
changes of temperature and volume. The resultant internal work
is dependint only upon the in#ial and f i n a l states of the gas and not
upon the intermediate path.
T o illustrate this point, suppose that in passing from A to B,
Fig. 5, the temperature at some part of the path were made to
e x c e e d that at B by 5o °, more internal work would be done in
raising the temperature this amount, than if it had not been made
t o exceed that of B ; but in the remainder of the path, in bringing
the temperature down from the 5o ° a b o v e / 3 to that of /~, the
excess of internal work that had been performed will be undone,
and the final result upon reaching B will be t h e same as if the
temperature had in no part of the path exceeded that of B.
April, I887.] Thermody~lamws. 307

Similarly, the resultant internal work done by expansion will be

the same, whether the expansion increases directly from v~ to v2,
or whether there be expansions and compressions. Thus suppose
that some part of the path in passing from A to B should be five
feet to the right of the v~ B, more internal work would be done in
producing this five feet of expansion than if the expansion were
limited to vv but in reaching the state ~ , there must be a com-
pression of these five feet, so that the previous extra internal work
will be undone, the final result being the same as if the path did
not cross the ordinate v2 B. The same reasoning would apply if
the path passed to the left of v 1A. The external work, v~ A B v~,
is dependent upon the path A B. This being the case, we may
choose the path along which we may conceive the substance to be
worked in order to determine the internal work, which path may
have no relation to the one actually described in doing external
work ; and such a path may be chosen as will favor the algebraic
analysis. In [rig. 5, conceive that the fluid is expanded along the
isothermal A C from vl to vv thence at the constant volume v2 the
pressure is raised from C to B ; then will the internal work be
V2

from A to C, f (v dp - - P ) dv, (10)

v1

in which r is constant, and is represented by A a e C,/rig. 5, and

from (7 to B, j r ~--~r,dr dv , (11)

o o

• d~p
in which v is constant in determining )-T2 from the equation of the
gas, but variable in reference to the integral, and r variable
throughout the operation. The last expression is not written in
the order given by the author, but is made to conform to modern
forms involving successive integrations; according to which all
the terms involving ~ are integrated between proper limits, after
which the remaining integral is performed. Since v and r are
entirely independent of each other, the integrations in this case
may be performed in reverse order, with the same final result.
The fact that r is treated as constant in the first integration, makes
it proper to place it before the integral sign, as in equation (9), but
308 Wood.. [ J. F. I.,
it is unnecessary to state that it is equally correct to place it after
the integral sign as in (11), both forms producing the same final
value. The conditions of the solution make ra = re.
As a special example illustrating the use of equations (10) and
(11), let the equation of the gas be as given in equation (4), then
equation (6), substituted in (10), gives--

) 2o ~av 2 d v =
2o(1 r,)
,~, ~ x - - ~ ' (12)
Vl
and from (4) find
r idO2 2a
(/ T 2 -- T2 V2

which in (11) gives

f ( 2. d3dv__ 2. _:2. 1_1 (lz)

J J r 2v 2 v2 J 32- v~-~
Te Oa .re

The entire internal work done in passing from state A to state

/3 will be the sum of the values in equations (l 2) and (13), and is
represented by AS, in the Steam Engine, page 304, and in some
other places simply by S. Hence, in this case,
1 1 +1 1 1
O4)

which admits of a numerical value when a and the initial and final
limits are known.
The author, in the article under consideration, uses Sa and St,
for the vallzgs of the potential energy necessary to overcome mole-
cular forces in expanding unity of weight of a substance from the
states A and B, respectively, to that of a perfect gas; the last
clause of which simply means that the gas is expanded indefinitely.
We may conceive the fluid worked along other paths than
those just indicated in Fiff. 5 ; for instance, it may be worked
along the adiabatic A ml indefinitely, becoming asymptotic with
the adiabatic through 23, and then compressed along ma 23 to •,
in which case, we have

vl
April, I887.] 7herlhodynamws. 3o9

S bd?p(=r ~Tv--p)dc,
Vll
as shown in /~)g'. 7, in which v, p and v are all variable, but the

FIG. 7.
integration cannot be performed under these conditions, even if the
equat~o~ o f the adiabatic were known, for there would be two
variables after ,elimination, one dependent upon the other.
Since all i s o ~ e r m a l s of the same gas are mutually asymptotic,
one may pass from A outward on the isothermal r~ indefinitely,
and back ¢o B on the isothermal r b. In this case, we have

d p --p)d,,
YI

dp --p]d~ (!0)
/
vg

which may be integrated, since r will be constant. In passing

from A to B, thence, indefinitely,along the isothermal rb, and
finallyback to A along the isothermal r., the sum of the internal
works will be zero, or
s+&--&=o;
. ' . 8----- S, - - Sb (17)
In passing from state A to stale B, the total heat absorbed will
equal the total work done, internal and external, added to the
increased energy of the substance ; hence, or by equation (9),
_l-I--fpd,, S-q-
+ Cv (rb-- ri),
WHOLE NO. VOL. CXXIII.--(THIRD SERIES, Vo]. xciii.) ZZ
 31o ll'oad: [J. F. I.,

or, in the form given by the author, after substitu.ting from (17),
tt--f pdv-----(Cvr--S)b--(Cvr--S)~ (18)
which is the author's equation (2), page 3 t 3
If A and B are in the same vertical through v,,, no external
work will be done in passing from state A to state B, hence
f pdv =0,
and
= ( -- -- (cv --

If A be at the foot of the perpendicular B vv then

i
va ~- O,
and
H= Cv r b - - Sb -~- S a
=- Cv ~'b + S (19)
in which S is the internal work done in raising the temperature
from zero tO r b. . . .

If the gas were perfect, tt)e entire energy, C~ r, would be e x p e n d e d

in doing external.work by an indefinite adiabatic expansion, a fact,
easily yerified b y showing directly that the area m, A v, X, Fig. I,
indefinitely extended, equals C~r,.; and the entire area under the.
adiabatic represents the intrinsic energy of a perfect gas. But if
the gas be imperfect, a part of the energy possessed by the sub-
stance will be expended in overcoming its own molecular attrac-
tions during "expansion, so that the external work which it can do
is less by the amount of internal work so done ; hence, in the state
/~; the zntrbzsic energy, or the external work which it can do by
virtue of the energy it possesses, by an indefinite expansion, will
be ~"
Cv
and similarly for the state A. This is also the area under the
adiabatic of the imperfect gas, but the area is 4ess for the same
temperature.
T o ascertain if equations (15), (16), (17), will give the same
result as (12) and (13) for the particular gas here considered, we
find in the same manner as in equation (12), which is the same as
(15) when v~ = ~ ,
2a.
Ta Vl
April, I887.] T]zermOdynamics. 31 r

similarly,
Sb __ 2 a
Th Y2
which, in (17), give
(1 1)
S- 2 a rav] vb v2

which is the same as (14) ; hence, the two processes give the same
results in this case.
In the same manner, the two processes may be verified for the
more general equation of imperfect fluids, which, as given b y
Rankine, and verified to some extent by Regnault, is
p v ~ r __ ao <h a: etc.,
jO o v~) TO r V T2 V

in which a 0, a . a 2, are functions of the density, and assuming them

to be inverse functions of the specific volume, and multiplying
through by P0 v0, it may be written,
p : R z Ao Az A2 ete.

From this find

V2
v= (A o + 2Al+3A____~ + e t o , ) ( v 1 1 1J,,
~'a Ta2 " fJ2
gl

f ~('d'P
d r2
ere, 2 At v2
the sum of which is
S=(Ao+2A'+3A2+ete.) 1

--(Ao2[-2AI-juaA~2 -~- et(~o ) 1 (20)

Tb Tb2 VT ~
Equations (15) and (16) give

Sa---- ( Ao + 2 AI-I--- 8-t2 -I- et¢. ) 1- -

Ta Ta2 V1

Tb Tb tl2
which in (17) give precisely the same value for S as just tound i~
equation (20). No further verification of the analysis is needed.
312 Wood. [ J. F. I.,
If, in (20),
Ao=O , A t = a , A2~---O,
it becomes

S'=2a(rlvt r f2" )
which is the same as 04). In this, if
Va Vl ~ rb t'2p
the internal work for this gas in working from state A to state B
will be zero. If external work be done in this case, we have
% > vl;

and if the final temperature exceed the initial, work must have
been done upon the gas since v2 will be less than ~ ; for

V2 ~ ~1 Ta
Tb
the result being produced by compression during the absorption
of heat•
Recognizing the fact that the matter treated in the lingo of the
Strata Engint, here considered, is so greatly condensed as to make
it exceedingly diffi, ult &r the student to st~mount the difficulties
encountered, we have considered it better to err, if at all, on Me
side of unnecessarily minute explanations, rather than assume an
anmunt of" kaowI~lg, of the subject by the reader.