Professional Documents
Culture Documents
is
tight
Binding
with
TUEATJLSifi
1
-
AL
APPENDICES
BY
l)u
SStm
PHVsIl'K
AT THK t'NlVKRSTn
OF
ll)VISEI>
EDITION
BY
KOBBET
B.
BAYNES,
M.A.
J
READER
IN PHYSICK
CO.
AUTHOR'S PREFACE
THIS EDITION
THE
first
edition of this
book appeared
in 1877, at the
time
of the theory,
expansive
tendency
of
gases,
explained the
new
laws, hitherto
undreamt
tested
quickly
won
to the theory
many
and adherents.
new
But a deeper
especially to chemists
whom
endeavoured,
AUTHOR'S FKEFACE TO
vi
TITItf
EDITION
by
on experiment. I therefore
collected together, as completely as I could, and summarised,
the observations by which the admissibility of the theory
I dare to think
might be tested and its correctness proved.
that I have not failed in this endeavour for not only did my
it
it
its
way
1
the purpose I intended.
wrote it only for that time, and did not seriously think that
But
it might reappear in a new edition after long years.
In this
my
book
fulfilled
demand for the book did not cease, and, as the supply
was nearly exhausted more than ten years ago, I had to
determine on a revision, although the necessary preliminary
work had not yet been done,
the
As
chemistry!
therefore preserved
it
character,
of the reaction of
gation, also
jet,
vil
pisoous theory.
many
With
the present
limitation to the old range it has cost very much trouble
and very much time to work up the literature of the subject,
that has grown mightily in these more than twenty years,
The
first
large sale in
German
edition of this
This
I confidently hope,
still
worked
he has enabled
me
to
my
Baynes
has
at the translation.
For
him
warmest thanks.
September 189U
TEANSLATOE'S PEEFACB
book of groat use, I
have willingly undertaken the preparation of an English
translation of this greatly enlarged and improved second
HAVING found
the
first
edition of this
I have embodied in
edition.
it
all
German
text,
as
to
also
mo by
has been no part of my intention to add any commentary upon the book, but I have appended a low footnotes that seemed advisable, and I have added an Index
It
which I hope
will
prove useful
EGBERT
OXFOBD
E.
BAYNE8
October 1899
Wmt
P. 46,
1.
13
,,
,,
CONTENTS
PAET
CHAPTEE
1.
2.
........
.
t*l
BEHAVIOUR OF GASKS
4*
0.
CHAPTEE
II
PKBSSUBE OP GASES
7*
BOYLE'S LAW
THE ADMISSIBILITY
8,
9.
6*
OF THE HYPOTHESIS
12
13
15
...
17
10.
MEAN
11.
12.
18*
14.
TEMPREATUER
15.
ABBOLTO ZERO
16.
17*
I)
18,
HEATING BY COMPRESBION
19,
COOLING BY EXPANSION
36
20.
VAPORISATION
37
21.
39
18
19
21
.
23
26
28
OF TEMPERATURE
80
.
31
88
CONTENTS
CHAPTBE
III
MAXWELL'S LAW
SECTION
22.
23.
KINETIC THEORY
24.
2,5.
26.
.........
.........
.......
......
........
MAXWELL'S LAW
PROOF OF MAXWELL'S LAW
FULLER EXPLANATION OF MAXWELL'S LAW
30.
31.
AVOGADRO'S LAW
82.
COEFFICIENT OF EXPANSION
MAXWELL'S LAW FOR A GAS IN MOTION
PRESSURE OF A GAS IN MOTION. KESISTANOK
29.
88.
84.
35.
86.
87.
88.
89.
^
**!
*">
-10
27.
28.
,41
TO THK
***
****
.........
......
.....
.
0*2
tl **
r>;>
<<>
... ......
...
....
EEACTION. CROSS-PRESSURE
PROPAGATION OF SOUND
EFFUSION OF GASES
*7
7;l
77
THERMAL EFFUSION
HEAT EFFECTS ACCOMPANYING EFFUSION
CHAPTER
H7
IV
41.
42.
48.
44.
45.
46.
47.
48.
49.
51*
W>
LAW
HlBN'S AND YAN DBR "WAALS'S CORRROTION
LAW
OF
f!f>
...........
,
1(1*1
1114
107
I
OB
THE
..ill
100
o*
50.
Mi
lift
114
CONTENTS
xi
CHAPTEE V
MOLECULAR AND ATOMIC ENERGY
f>t>,
57.
5H.
flii.
f*;$.
r
.
-l.
55.
POLYATOMIC: MOLECULES
59.
127
.
60.
PAET
180
182
187
139
II
OHAPTEE
"VI
6J1
04*
65,
.......
...,,..,
157
159
160
161
08.
00,
MOLECULAR PATH-VOLXJMM
70.
FltBQUBNCY OF (lOLLISIONS
EBLATIONS OF THK FBKE PATH TO TEE PRKKBURK
TKMPBIATIJRE
.
ON THK ABSOLUTE VALUE OF THE FREE PATH
07.
HFBEDB
7L
72.
162
16B
153
156
AHHUMPTIONH
PROBABILITY OF PAETIOULAR FREE PATHS
FRBK PATH IF ALL THK MOLECULES HAVE EQUAL SPEED
MOLKOULAR FllKE PATH WITH AN UNEQUAL DISTRIBUTION OF
OB.
149
151
AND
.
164
169
CONTENTS
xii
CHAPTER
VII
VISCOSITY OF GASES
J'M.K
SECTION
78.
ON THE CHARACTER
74.
75.
76.
77.
OF INTERNAL FRICTION*
FORMULA FOR THE COEFFICIENT OF VIROORITV OF A GAS
THEORETICAL LAWS OF GASEOUS FRICTION
OBSERVATIONS ON THE FRICTION OF GASES AT DIFFKKKST
.
171
172
174
178
87.
1H1
PRESSURES
NUMERICAL VALUES OF THE FREE PATH AND COLLISIONMH
FREQUENCY OF PARTICLES OF AlR
FREE PATH AND COLLISION-FREQUENCY OF THE PARTICLES m
.101
DIFFERENT GASES
200
VISCOSITY OF GASEOUS MIXTURES
205
FRICTION OF GASES ON FIXED BODIES
207
THEORY OF EXTERNAL FRICTION
210
COMPARISON OF THE THEORY WITH EXPERIMENT
212
PHENOMENA IN VERY RAREFIED GASES
215
INFLUENCE OF THE TEMPERATURE ON THE VISCOSITY
221
VISCOSITY OF VAPOURS
22J$
DEPENDENCE ON THE TEMPERATURE
88.
DISSOCIATION
89.
78.
79.
......
.
80.
81.
82.
83.
84.
85.
86.
90.
91.
....
92.
CHAPTEE
228
21 M)
2:W
2ttU
244
VIII
DIFFUSION OF GAHM8
98.
OBSERVATIONS
247
94.
2r0
95.
96.
97.
98.
COEFFICIENT OF DIFFUSION
99.
100.
101.
102.
......
251
250
*2H
21*0
20*2
204
HIIH
27*1
CONTENTS
CHAPTER
IX
CONDUCTION OF HEAT
*r
" ION
UK*.
104.
PAGE
LOW
('ONDCCTIYITY OF CAKES
KINETIC TllKOUY OF CONDUCTION
107.
JOB.
105.
100.
Hi'RJtw
288
..,...
CIKNT T OF VISCOSITY
PAKT
ON
TllM
377
279
282
287
291
111
CI1APTEE X
ON THK IUBECT PKOPKBTIKS OF MOLECULES
109.
HO.
111.
1
12.
11
IJ.
114.
.......
115.
CULES
116.
117.
118*
f*
119.
120.
121.
12*2.
12*1
124.
299
300
302
304
309
315
317
319
323
327
332
332
335
335
341
348
CONTENTS
XVI
APPENDIX V
DIFFUSION
......
.......
SEf TJON
51*.
52*.
53*.
54*.
"
<>
4fH>
4.*s
4M*
4.VJ
APPENDIX VI
CONDUCTION OF HEAT
55*.
56
4
.
57*.
TRANSFERENCE OF ENERGY
CONDITIONS FOR THE STATIONARY STATE
CONDUCTIVITY
.
INDEX
......461
,
46:3
465
4B7
PAHT I
ITS
ENEEGY
CHAPTEE
THE
is
of energy, is no longer considered a hypothesis, but the expression of a certainly proved fact. Experiment shows that
heat
is
energy.
From
it
itself in
nature.
amount
But
The term potential energy was applied byBankino (Phil. Mag. [4], v.
Leibniz (Acta Erud.
1858, p. 100) to the magnitude called vis mortua by
Gerhardt's ed. vi. 1860, p. 238), and
Lips. 1695, p. 140; collected works,
later called
by Helm holt z.
1
Sjwwikraft
The term kinetic energy was first employed by Thomson (Lord Kelvin)
and Tait (Treatise on Natural Philosophy, Oxford 1807) in place of actual
hall of the magnitude called vis viva by L e i b n i z .
energy, B a n k i n e s name for
This was termed simply energy by Thomas Young (Lectures on Nat. PhiL
loot, -nil London 1808, .p. 78 ; new ed. 1845, p. 59).
3
'
B 2
ITS
ENERGY
at least
energy.
Bum ford's
We
ex-
we may assume
and the
further rotate about an axis of its own
motions
these
of
of the kinetic energies
represents
mechanical energy of the contained heat.
may
sum
the,
In the mechanical theory of heat we extend this speculation, as a rule, no further, so as not unnecessarily to make
our reasonings and conclusions depend on doubtful hypotheses. In this connection physical investigation haB npeeial
reason to avoid hypothesis, as the high value and great
significance of the mechanical theory of heat rest on the
general and unconditional validity of its proposition^
whereby we are enabled to measure forces of unknown
nature equally with known forces, and to subject them to
calculation with equal certainty.
It would, however, be a censurable restriction of investigation to follow out only those laws of nature which have
a general application and are free from hypothesis; for
mathematical physics has won most of its succefwes in
the opposite way, namely, by starting with an unproved and
unprovable, but probable, hypothesis, analytically following
out its consequences in every direction, and determining its
value by comparison of these conclusions with the results of
experiment.
For the mechanical theory of heat, too, this method has
already borne good fruit.
By ascribing special forms to
1
Phil Trans,
properties
known
experiment.
There has thus arisen from the joint labours of the
workers in this field a special theory of the gaseous state
which was formerly known as the dynamical, 1 but is now
better called the kinetic, 2 theory of gases.
In this work I
have endeavoured
tributions
periodicals of
2.
all
to collect
individual
of
kinds.
"with regard to
Hypotheses
of bodies
Heat-motion
We may therefore
distinguish
two kinds
of heat-motion,
[4] xix.
So far as I know, this name was first used by Lord Kelvin (Six W.
Thomson) In an address before the British Association at Edinburgh (B.A.
Maxwell afterwards adopted it (Nature, viii. 1873, p. 298 ;
flap. 1871, p. 93)
*
ITS
ENERGY
$ -*
breaking up
includes^lherotake place within the
fore, not only the oscillations that
about tin*
molecule, but also the rotation of the atoms
centroid of
tlie
molecule.
more
of the
mechanics
of molecules.
bond.
is
cohesion, since
when
the
medium
is
in
equilibrium the
:>
forces
due to
this
motion are
equilibrated
by those
of
cohesion.
The
difficulty of this
ease which
is
Behaviour of
G-ases
found.
the tendency of gases to expand, and it has a further support, derived from experiment, in the thermal behaviour of
gases when changing volume. For if a gas expands without
overcoming pressure, and therefore doing work if, for inits temperature falls
stance, it streams into vacuous space
so little that for long it was admitted, on the evidence of
Gay-Lussac's experiments, that under such circumstances
no fall of temperature at all occurs.
This behaviour would not be possible if on expansion a
gas had to overcome any considerable cohesion, since for
this an expenditure of energy, and therefore of heat, would
be requisite. Just as little can the assumption of repulsive
forces between the molecules be reconciled with this experiment, since such forces would come into play during
expansion and generate energy in the shape of an increase
of heat in the gas.
invalidated, since
in
overcoming
its
own
ITS KNEJNiY
?>
Herein
each other.
hypothesis,
since
in
is
face of
this
there remains as the only admissible hypothesis the opposite view that in a motion of its molecules consists the
4. Character of the
It is
now
Heat-motion In Gases
is removed from
such as gravity, for instirnivand we then introduce the further assumption, which is
sufficiently, though not strictly, accurate, that there is no
consists in a
from each
Two
parabola.
gases as rectilinear,
free
if
a gas
is
is far
from the point of liquefaction, we are justified in representing an overpowering majority of its molecules as far enough
apart to be nearly always outside the range of fcheir mutual
and we may therefore assume that the small
attraction
amount of cohesion which does come into play is to be put
to the account of the rarely occurring cases when two molecules now and then come very near each other.
;
If
we
gas as
practically the
first
on
collision.
6.
When
by A,
2
GrundBilgc,
dn&r Theori
W. Hayn, and
10
ITS
ENERGY
$S
hypothesis,
well as in later papers, especially attracted attention, and
to
many physicists were induced by these investigations
it
to
and
experimental
putting
help in developing the theory
proof.
It was, indeed, quickly
in 1851,
which had
1821.
also
presented to the
Royal
lished
it
on account
A whole
series of writers
who
Bois-Beymond 5
has directed the attention of his contemporaries by a German translation of a fragment of his
Abhandlungen ilber die mechanised W&nwtheoritt Brunttwiek
2nd part, p. 229 transl. PML Mag. [4] xlv. 1857, p. 108.
Memoirs of the Manchester Lit. and PML Sac. [2] ix, 1851, p. 107;
1857, p. S53
1867,
1
reprinted later
2
8
PMl. Trans.
clxxxiiL 1892, p. 1.
Pogg. Ann.
evil.
P.
& J.
1859, p. 490.
Y.
Bernoulli,
1752.
11
The
works.
1
interest,
The
come
We
notion
is
&.
cxv. 1862, p. 2,
AbMndl
Warterbuch,
ii.
p. 230
pt.
iv.
;
Lothar Meyer,
'
it
it
was in constant
remained forgotten by
all,
who
12
CHAPTER
ITS EXKRCtY
$6
Tl
PEMSSUES OF GASES
Boyle's Law
which we have described,
6.
THE
hypothesis,
of the to-and-fro
hypothesis,
we have
to see
whether
generally
enunciated
Marietta's
head of
law,
because
*
De
Mariottu
Nature, do
TAir/ which first appeared in 1679. As, however, there can
be no doubt that the discoverer 1 of this law is li abort
2
Boyle, who determined it seventeen years earlier, 1 shall
follow the English custom of calling it Boyle's Jaw*
it
at the
his essay
la
iv.
Hydrodynamics Argentorati
Pogg, Ann.
cvii.
1859, p. 400.
Oorman,
PRESSURE OF GASES
Bernoulli supposed
movable but
vessel with a
13
of an air-pump
this gas can be compressed by
increasing
the pressure on the lid or piston. If, now, the gas consists
of a large number of moving particles, and the pressure
exerted by it on the walls of the vessel arises -from the
;
results
lirst,
there
is
immediately
ticles are
a larger
number
more crowded
hurled back by
walls.
If, by the compression of the gas, the volume is
diminished in the ratio 1 s 3 the distance between any pair
:
s ; the number of
of particles is diminished in the ratio 1
particles, therefore, in the bounding layer, which is in
:
ratio s 2
between
ratio s
"the
and in
this
same
is
increased in the
number
of
in the ratio
the volume
3
:
which
1,
is
The pressure,
therefore,
Boyle's law
is
molecular impacts.
7.
Since
also
consequence
from quite
dif-
of the theory is
14
no proof
hut
it
ENERGY
ITS
results of the
circumstances,*that
is,
at meiui
as
if,
we
not
did
alter.
The
P=2
Winkelmann,
p.
SOU
Oat w a Id,
PRESSURE OF GASES
15
This table shows that Boyle's law does not hold exactly
any gas, but that for the gases named it holds with sufficient approximation to be considered for most purposes an
for
Her wig
by
be
of
of the differences
Defects of
tlie
Hypothesis
We
tion,
shall, however, not rest content with this approximabut try to get nearer the truth by inquiring into the
16
EXEROV
ITS
$8
If Bernoulli's theory
of these variations.
possible causes
that is only approxilaw
a
gives as its necessary consequence,
the
theory cannot In*
mately exact, the hypothesis underlying
dof
octavo, own if only
in every respect, but must be
quite true
S Ig
are
made up
is
Boyle's
In the
law.
first
place,
if
PV
pv/PV
PV
PRESSURE OF GASES
all
hydrogen.
Probably both
the
17
of
7,
except
influences
in
occur
when higher
PV
The product
increases with the
noted with hydrogen.
pressure P, because on account of the dimensions of the
molecules the volume V cannot diminish so much as the
law requires.
These considerations, which we shall repeatedly have
4
again to take up and extend, show that the departures from
Since
the strict law can also be explained by the theory.
the probability of the theory is, therefore, in no respect
prejudiced
entitled to
all,
9.
for
is communicated.
Acpressure of a gas to which heat
is
heat
laws
of
the
to
energy an
thermodynamics
cording
in
an
consists
of
increase
augmentation of
heat, therefore,
;
Pogg. Ann.
latii.
[0]
Comptea E&ndm,
Mix. 1893, p. 68.
a
G.
JR. hex.
1870,
Ixxxviii. 1879, p.
p.
H81,
p, 267.
*
40-51,
336
Ann. Chim.
[5] xrx.
1880, p. 345,
Journ* de Phys.
viii.
1879,
18
ENEKGY
ITS
10.
of the
Energy
between the
investigation of the relation
its molecules
of
kinetic
the
and
a
energy
gas
temperature of
molecules
the
that
the
circumstance
difficult
is rendered
by
The
have not
energy.
closer
all
on collision.
other, that the energy of each particle changes
of the
action
resultant
the
that
however, we can say
If,
We
CO
= 4/(^ -f
V*
in
of
W*)
components
w, t\ to
PRESSURE OF GASES
10
19
all
all of
and Kronig
started in
11.
may
Mag,
<J
in
Mem.
Berlin 1856
many
3
and Phil.
Soc. [2]
ix.
1851, p. 107
Phil
oilier periodicals.
Kronig
Clausing, Pogg.
Xn.c.
1857, p.858;
[4] six.
p.
ia>;
v.Lang,
Saalsehttss,
.
ibid.
ScJvr.
p. 45.
d.
20
511
ITS ENE11C1Y
will
solid wall
For
choosing
also to
assume
for
each one of
We
struck
impact.
To
we have
shape
According to Joule's
as was proved in
which,
representation
10, may be used in the calculation instead of the real
are Nafiy
molecules
of
circumstances,
molecules,
altogether.
the
case,
move
between them.
fyNafty
*y
per-
PRESSURE OF GASES
11
21
The number
time-unit
is
The product
of this
number
&&
face
in the
is,
on
it,
on the surface
is
= $NmG*.
viz. that
volume (6).
12.
We
on the walls
and
this
22
ENERGY
ITS
l-?
P~7
Appendix No. 1,
the
Consider the space occupied by
gas to be divided by
a left half, and
and
half
a plane into two halves, a right
in the Mathematical
"'.
right
exerted on
it
by the
left half, if it
measure a continuous
equal and opposite pressure. Now, we
force by its impulse in a unit of time in the meaning of
our theory, therefore, the pressure is nothing else than the
momentum which is transferred in unit of time through
the unit area of the plane from one half of the. gas to the
other, or, rather, as need hardly be specially specified, it i
the component of this momentum in the direction of the
;
is
whole
left to right,
while
left to
whose base
is
by the speed
equal to
(3%
right in a time-unit
is
cylindor,
measured
unit area of the plane in unit time from the loft half of the
gas to the right is $NG if N, as before, represents the
number of molecules in the unit of volume.
9
PEESSUEE OE GASES
13
23
by the amount
x
while simultanoously the same number of molecules cross
the area from right to left, diminishing thereby the oppomomentum of the right
sitely directed -from right to left
the pressure
2
p = JJNwG
which
is
other half.
This formula
is
collision.
13.
Absolute Value of
Nm
tlie
Molecular Speed
water
of
Nm = p.
Consequently the formula
may be
written
This
defini-
in
ENERGY
13
viz. that
24
which form
its
ITS
to its density, is
proportional
is
directly seen.
In this new form, however,
it
teaclics
us
more
much more
Mm
to explore an unswz*
along the path thus trodclen
world.
Though measured by the height 7* of a column of mercury,
the pressure p is not identical with this height, but with the
action of gravity on the column when taken of unit area,
and <j the acceleraIf, then, q denotes the density of mercury
we have
tion of gravity,
and therefore
is
given by
n
Let us make this calculation for the temperature O 0.
and the pressure of one atmosphere, i.e. of a column of
1
mercury 0*76 metre high. We will take BegnaultV
value, q = 13-596, and his values for tho density of tlio
various gases
for Paris,
we must
= 9-80896
Mem.
and PHI
Boc. [2]
ix.
1851, p. 107
Phil Mag,
p.
Pogg. Ann.
254
c.
transl. Phil.
&.
ii.
tB07
PRESSURE OF GASES
13
given circumstances
4 C., we must put
is
25
where
= s/773-3,
"We obtain in
manner, according to Clausius's calculation, a general
formula for the value of the mean molecular speed of a
gas at 0C., which we will denote by <g, viz.
s is
this
employed in
Prom
its calculation.
this formula
0. in
.........
........
........
........
Taluesof
Nitrogen
485
461
492
Hydrogen
1844
Air
Oxygen
as
of these
in
of
4, which is there
they further justify the assertion in
not proved, that the speeds produced by gravity in short
periods are too small in comparison with these speeds to
cause any sensible parabolic curvature in the paths of the
molecules.
lutely neglected.
3
Pogg. Ann. c. 1857, p. 377 ; AbJi. &. Wfimnethevrie, pt. ii. 1867, p. 256 ;
Me&haniscke W&rmctheorie, 1889-91, 2nd ed. iii p. 35, edited by Planck and
Fulfrich.
26
ITS
EXKI^Y
air
must have
at least 11,000
tion
the
second.
From
this
we may
possesses an atmosphere at
a
it can in any case have only a very thin onc.
its
14.
Temperature
at
C.,
and
it
viL p. 528.
xi.
I8|4 f p*
d& /%*
7f
[S]
PRESSURE OF GASES
14
27
is
= kp(l +
oS),
= &G\
we
0,
where
denotes the
mean molecular
ture
and
it
<2
(1
aS).
a gas,
the
"kinetic
given by
This law
is
in complete
9 from
Bernoulli's
obtained in
Atmaks de Chime
137
Gfilb.
p. 257,
2
Mm.
of the
1802, p. BIO.
8
See
46.
Qilb.
Ami. xii
28
15.
The law
ITS
ENERGY
K>
determined.
If heat is nothing else than the kinetic energy of
molecular motion, the temperature at which a gas possesses
no more heat must be identical with that at which its
molecular motion has disappeared, and all atoms and mole-
The
speed
(?
+ a3 =
0.
l
,
Begnan.lt/ Jolly,
with
all in agreement
each other, we take 0*00367 for the value of in the cane of
air
when
last equation,
we
find
=-
is
272-5
the*.
G.
=
in the general case, for
or,
272-5
all
+3
scales in use
of the constant
to be taken
is
p. 73;
expansion
is
wo have
Pogg. Ann.
3L
1877,
on the Fahrenheit
as tho
Iv.
and
Ivii.
81,
scale, for
PRESSURE OF GASES
15
29
we have
for the
mean molecular
where
A real
meaning, however,
is
so that there
there
is
273 C.
this law loses its validity at some higher point, and another
takes its place. The absolute temperature that has been
introduced has therefore more the signification of an auxiliary
mathematical function than of a physical reality.
the freezing-point of water
-
x 0*003670
32 F.
viz.
and
to
&
=
1
=s
is required on all scales ; so that for air on the F. scale we must take
0-002089 * (1 - 0-002039 x 32) = 0-002181. TB.]
~ v'(272 5 x s) = 29"38/ Vs metres per second for the
[In fact, JET = 485
<
Centigrade scale,
2
See 32.
3
TB.]
On other determinations
Pty/sikaUsches WSrt$rbuch,
x. p. 115.
Gehler's
30
16.
ITS
ENERGY
10
viz.
ff
= 3p,
8,
and
to the absolute
K = Jtt,
<*
where
jp
J?
temperature
is
272-5.
PRESSURE OF GASES
16
On an
31
is
the co-
efficient of
p
The
+ o), and
jp () (l
latter formula,
which
Is
K = Jf (l
a$).
measure
increases
of
perature.
Balton's
17.
Law
Mixed
Gases
From
of
a gaseous mixture.
For such a mixture the calculation of the pressure
exerted would be carried out in exactly the same way as
was done in 11 in the special case of a simple gas. The
pressure on a surface
Is,
in the
of a
value
up
is
present.
= KiT,
4-
...),
p p a>
2,
'
19
32
of the pressures
ENEEGY
ITS
17
separately occupying
= f"
pTp p 2
2,
whence
or the pressure exerted by the mixture is equal to the sum of
the pressures separately exerted by its several components.
1
Thus the law found by Dal ton for the pressure of
mixed
gases,
and
and confirmed
defended
Henry
by
That
this
8].
Dal ton's
And
this reading
of
Dal ton's
Mem.
of the
Manch.
Lit,
hypothesis
and
is
also
in
Gilt.
Ann.
xil.
1802, p. 385.
2
Nicholson's Journal, viii. 1804, p. 297 ; G-ilb. Ann. xxi. 1805, p. 393.
8
Begnault, M&m. de VAcad. xxvi. 1862, p. 256 ; Andrews, Phil. Trans.
clxxviii. 1887, p. 45; Cailletet, Journ. de Phys. [1] ix. 1880, p. 192;
G-alitzine Weber das Dalton'sche Gesetz, Inaug. Diss. Strassburg, 1890 Wied.
Gott. Nadir. 1890, No. 1 U. L al a Comptes Rendus,
xli. 1890, pp. 588, 770
188.
cxi. 1890, p. 819, cxii. 1891, p. 426 Nalurw. Rundschau, vii. 1892, p.
;
Ann.
PRESSURE OF GASES
17
33
by mani-
or
18.
Heating
"by
Compression
Remembering that the pressure, energy, and temperature of a gas increase together in constant ratios, we have
at once an explanation of the fact that the
temperature of
a gas
is
raised
the
it is
not
difficult to see
We
obtain a deeper
plained by this general consideration.
into
the
nature
the
of
when
we invesphenomenon
insight
the
the
more
nature
of
molecular
in a
motion
closely
tigate
that
is
or
gas
being compressed
expanded.
To keep
exerted upon
it
which
it.
fore,
1
18GB, p.
72,
ix.
34
ENEHOY
ITS
18
when
fixed.
But
A mathematical
explained by Kro'nig and Clausius.
theory of the phenomenon has been given by Woldcmar
2
Voigt.
It has been shown by 01 a u si us that the heat produced by pressure can easily be calculated on the grounds
of our theory, and that it is equal to the work clone.
Tn
a rather later memoir 4 he gives a proof which we here re-
produce.
We
cedure, described in
10,
molecules possess the same mean speed C?, but also that
only one-third of the molecules are to be taken into account
all
in
on a wall
the containing
the compression
takes place so slowly that the disturbance of the equilibrium
has always time at once to subside. With this suppoHition
the number of molecules which in unit time meet unit area
of the wall of the vessel is j*NG, by
32, and the number
therefore in unit time which strike the surface
of the
This assumption
vessel.
is
admissible
of
if
piston
2
3
Grundzilgaeiner Theme d&r <7<w?<?, 18SG ; also Pofjg* Ann. xoix. 1850, p.
Pogg. Ann, c. 1857, p. 865 ; Abfandl. pi 2, 1807, p. 211
Qtitt. Nachr.lSti, No. 6, p. 228.
See Natanson, iried Ann.
xxx viL
4
is
p. 841.
Mcch, Wlirmetheorie,
ill
1889-01,
14, p. 29,
$1.1,
PRESSURE OF GASES
18
35
which the
This
since
is
during
the piston must do to overcome the pressure of the gas.
D 2
36
ITS
ENERGY
19
Cooling by Expansion
19.
work
done.
If
speed
by
its
diminishes
speed of a molecule which impinges on the piston
from the
therefore
there
and
from Q to G
2<x,
passes
which
2maG
is
the volume
The
sion of a gas and the fall that results from its expansion can
from this be easily calculated if the value of tho specific
each other.
and
this holds
The negative
PRESSURE OF GASES
19
S&
o
The change
= - 4l ^1
V'
cp
Here
37
That
this is so
we
shall
Without
easily
by again analysing the procedure.
the
the
law
of
conservation
of
energy we can
transgressing
see
move
ceasing to
Many
brought
much
liquid.
There
is
now
38
ITS ENFJR1 Y
iH)
in general
move
and
is
i)i tJic
mtpour*
this
energy.
first
energy
PRESSUEE OF GASES
20
39
21,
its
it
On
As the
latter is
equal to the
sum
gas from the liquid which has absorbed it. Hence there is
no doubt as to the existence of an attraction of the gas by
the liquid.
If the force with which the liquid, when in the state of a
saturated solution, retains the gas were as great as if the
liquid were pure, the number of the molecules of gas
absorbed by the liquid would be exactly proportional to the
number of molecules remaining above it in the gaseous
state, and Henry's law, that the mass of gas absorbed
increases proportionally to the acting pressure, would hold
But as this supposition cannot be accurate when
Strictly.
are absorbed, Henry's law can only approximasses
large
teaately represent the truth.
rise of temperature also increases
40
ITS KNKUfiY
il
body, and
we must
necessary
to
phenomenon
introduce
as Adsorption
by spongy platinum.
41
CHAPTBE
III
MAXWELL'S LAW
Unequal Distribution of Molecular Speed
THE mean value of the molecular speed given in
22.
13
not the arithmetic mean of all the different values of the
speeds with which individual molecules move. The magnitude G by which the mean value in question is denoted
10 to be such that the mean energy
has been defined in
of the molecules which strike against the wall would be
is
unaltered
if all
of
them
possessed this
same speed
From
this
it
instead
results that
we
of
in
place
lished.
of
moving
particle
42
ITS
ENERGY
l>i>
more
slowly, since
it
must
lose speed.
that in such
an aggregation of
This example shows
molecules as we assume in gaseous bodies in our theory
the speeds of the individual molecules cannot be equal in
the state of equilibrium. Equilibrium can, therefore, only
consist in a condition of continuous interchange of speed
The attempt
to deduce a law for something that in Kiibchance may seem singular and strange, but this
should not deter us from undertaking a research which
touches the very core of the kinetic theory.
This theory, indeed, seeks for the cause of regular phenomena and regular properties of gaseous media in irregular
tumultuous motions of the molecules. We have here to
look at the observed facts, not as direct neeeKBtyry consequences of unchangeable laws as is usual, but an the result
ject only to
of a large
to
number
MAXWELL'S LAW
23
43
first
new
long periods.
In
all
these cases, and in our theory as well, the regufrom the great number of the elementary
absolutely and fixedly one and the same, but there would
be different results conceivable of more or less probability.
But the
by the play
of
motion
of
In a gaseous mass consisting of a finite number of molecules this condition of simple regularity will not be attained,
as in the case of an infinite number at least not at every
In a finite system this law shows itself only if all
instant.
the states which occur with continual change in the course
The
of longish periods are taken into account together.
of regularity which is exhibited at every
condition
particular
moment by an infinite swarm determines also for a finite
number of molecules its mean state during a considerable
period.
44
ITS
ENERGY
i>3
we may
24.
Maxwell's Law
speeds which hitherto had been only estimated, and a knowledge of which was necessary for the development of the
theory of gases.
Maxwell's law
the molecules in question according to the same law as the possible errors of observation are by the method of least squares
distributed
among
the observations.
motion in
PHI. Mag.
P<Z*-** =s 0-74682,
ffec"**"
= 0-18625,
MAXWELL'S LAW
4 24
45
If 10,000 molecules
nent.
The small
values* therefore predominate in remarkably large proportion, and the probability of larger values of
a component of the molecular velocity, just as that of large
errors of observation, is vanishingly small.
In this form the law expresses the frequency of occur-
We
First of
all it tells
us
how many
of
a certain
at
to
with other
its
encounters
particles.
25. Proof of
Maxwell's Law-
law
is
collisions.
PHI. Mag.
PMl
[4] xix*
Scientific Papers,
ii.
p. 43.
46
ITS
ENEKGY
give a
new
proof,
law in his
Further, Kirchhof has given a proof of the
4
Lectures ; but against this, too, according to a remark of
Boltzmann, 5 objection may be made.
In a
in the
different
first
way
wan attempted
The weak points of this
and a varied
attempt were removed by N. N. Pirogoff/'
form of Pirogoff's proof is given in the second of tho
Mathematical Appendices.
These mathematical proofs cannot be repeated hero, nor
should we attempt here to give them I will only indicate ft
from them.
point that arises
;
striking
knowledge
of
the encounters
might bo
and apparently
behave
molecules
whether the
we must
therefore
know
Wiener Sitmngsb&r.
Mil
1868, p. 517
kvl
1872, p. 275.
herau**
Vorkwngen, 'liber mathematiscJie Physik, iv. (Theorio der Wttrnw,
142
von
(14th
1894,
Lecture).
p.
Planck),
geg.
4
5
6
7
is
given in Chapter X,
MAXWELL'S LAW
25
47
We
proofs.
Maxwell's
the law
is
Boltzmann
of
Maxwell's law it is
For
therefore sufficient,
Wiener Sitmmg&b&r.
Ixiii.
48
ITS
ENERGY
$25
energy makes it
which, indeed,
necessary to admit certain assumptions,
arc
which
only hypotheses,
would be contested by no one, but
afforded by a mathesince they fail to possess the certainty
when we apply thia
matical proof. In the particular case
with
to a system of material particles endowed
proposition
The
proposition
of the conservation
of
of these
mutual attraction or repulsion, the most essential
another
is
QH
is that tJw action of otic particle
hypotheses
exerted on the first ly tlte second.
equal to the reaction
no further hypothesis is
If this assumption is admitted,
the proposition
needed for the proof of Maxwell's law save
theorems o
of the conservation of energy, the other general
mathematical
mechanics which are used for the
analytical
and reaction.
assumption
This applies especially to the
addition
to
its
in
possesses a general forward velocity
when it is in a
state of flow.
called theorem
The same
of
Maxwell's
especially
is
admissible
was
first
2.
MAXWELL'S LAW
25
49
which
all
the
For the
part.
side,
60, between the free motion of a molecule and the constrained motion of the atoms that is due to affinity.
We
physical researches, but only in meteorological investigations ; it is therefore sufficient to mention that, in
in
Maxwell's law
Maxwell's
law.
Maxwell's Law
the occurrence of the
might be
most
fre-
quently occurring case would be that in which each component, and consequently the resultant velocity, is vanishingly
about.
It is easy to explain the apparent contradiction between
well' slaw and this undoubtedly true fact.
Max
When
3
8
4
50
ITS
ENERGY
We
ponents
is zero.
We
itself,
ponents of the velocity, but to the actual speed
can then explain the law as follows
There is a most probable value of the speed which occurs
more frequently than any other. Other values of the speed,
whether greater or less, occur the oftener or are the more
probable the nearer they approach to equality with the most
:
values
probable. Infinitely great as well as infinitely small
at
Molecules
small
of the speed have infinitely
probability.
to
met
with.
be
rest, therefore, are infinitely seldom
different
forms
of
the
<71
'
On
Chem.
xiii.
1875, p. 4S8
Nature,
3d, p.
S57
Scientific
Papm>
II p. 418.
MAXWELL'S LAW
26
51
from the
centre.
The
lie
*.
*****
-** * ****
*
*
v.
*
*
*
*
and
of the
components
of the curve
is
of the ordinate
whose
52
unit
ENERGY
26
magnitude
of
ITS
is
probable.
y
10
f
O 6
0*4-
0-Z
FIG. 2
assume that
the values of the molecular speed which really occur lie
between fixed limits, both the very large and the very small
1
Pirogoff has
we
easily see
MAXWELL'S LAW
26
53
limits that the variable values lying between them give the
same mean kinetic energy as if all the values are possible.
have merely to take one limit on the ascending and the
We
The
27.
Mean Value
of tlie Speed
54
ITS
ENERGY
27
mean
the
speeds
is
mean
the
a, 6, <?...;
then
n = (a
mean
-f
E=
//*(?*,
mean
we
so that the
mean
value
obtain
5 2 -f c 2
E=:j>m(a +
+ ,..)/X
by Joule
signification
(0,*
&2
C2
~f-
0/W-
(a* -f 6* 4- 0*
known
a2
H- 2Z>c
-f-
2ca
-f
relation
4-
leads to
or, since
we
<(a2
~h
&
-f c
...)/i
find
12J;
that
the
the
is,
^Ae arithmetic
mean
which holds
in
MAXWELL'S LAW
$27
simple form that
of the
is
Mathematical Appendices)
all
55
gases, viz. (see
19*
= 0-9213 G,
which, with extreme approximation,
may be
written
a = &Q.
Joule and Clausius' values are therefore greater than
the arithmetic means of the molecular speeds by about a
twelfth part.
The
latter
may
of
given by
p=
gas,
viz.
be replaced by
may
iTrpfl
1866.
IF
Maxwell's
theory,
we have
the formula
the value
is,
therefore, smaller
which
most probable value is a third
which is called the value of
mean probability, or, more shortly, the mean probable value.
The signification of this value, which I denote by in 19*
stands to
them
in a ratio
is
of the
greater
have
= 1-09
1
Inaugural dissertation,
TP
= 0-96 fl.
DQ Gasorum
56
ITS
ENERGY
L>8
28.
when
at
high
at
The
= 0-0012576
nitrogen,
G = 461*2
= 491*7
m. per
sec. for
for the
oxygen
nitrogen,
= 424-9 m.
= 453-0
per
values at
= 409-5 in.
= 436*6
sec, for
per
oxygen
nitrogen.
,,
are
sec, for
oxygen
nitrogen,
for
are
oxygen
nitrogen.
G =
=
=
W=
fl
The law
1
Abh.
of the
d. AJcad.
1838*2 m, per
1693-6
1632-2
1600-9
eu Mttnchen,
sec.
xxiii.
xiii. 2.
1888, p. 356.
Abth.
IB
MAXWELL'S LAW
shown by the
oxygen
following numbers.
57
Of 1,000 molecules of
at
13 to
81
166
214
202
sec.
215
300
151
203
152
91
92
400
500
600
76
77
400
500
600
700
above 700
Since this one example will suffice to give a clear representation of the nature of the law, I shall limit myself to
giving only the mean values of the molecular speeds for
other gases and vapours, and these I shall tabulate together
in their calculation.
r
density p of a gas to that of w ater as unity, but instead of
have introduced, as in
13, the specific gravity s
referred to atmospheric air.
This is a procedure which
this I
Values at
1.
Bayleigh's
C.
observations,
'
1888, p. 356.
2.
Tk Thomson,
8.
4.
Gay : Lussac
1
Abh.
d.
{
>
Ann.
of Phil.'
i.
1813, p. 178.
1806, p. 320.
1'Inst.' vxi.
Akad. zu Mil/nehen,
[2] i.
xiii.
1816, p. 218
Wied. Ann.
ii.
vi.
p. 135.
1879.
'
6.
7.
ENERGY
5.
8.
ITS
150.
p. 72.
i,
1811, p. 510.
'
Be"rard,
12.
13.
H. Eose,
'
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
Phys.*
24.
[1] Ixi.
26.
[1] xcv.
1815, p 177.
'Ann. Chim.
1807, p. 294.
Berzelius, AfbandL*
*
25. B.
27.
Schwoigger
1868, p. 2S2.
[2] i, 1810, p. 218,
Ann. Chim. Phys/ [2] xxxiii. 1820, p. 858.
i.
'
Dumas,
28.
29.
H. Jahn,
MAXWELL'S LAW
28
I could with
little
since the
considerably,
59
tables
of
But
much than
too
temperature, so as to
other.
29. Equality of
Temperature of Different
Gf-ases
rest
the molecules
of molecules in a
Pogg. Ann.
See
60
same law
as that
a simple gas.
Maxwell's law
to single
ITS
ENEBGY
i>9
of distribution
There
and mixed
is,
is
applic-
therefore,
able equally
gases.
in gaseous mixtures, too, such a distribution of molecular
values of the kinetic energy a
energy that of all possible
which will be found in any
be
definite value can
assigned
given
particle
whether larger
particle
a particle of another kind.
as
in
probability
From this follows directly a proposition
first
given by
Clausius, that in a mixture of two or w&re gases the molecules of both kinds possess on the average equal Jdwtic
if m l9 m^ are
or, to express this by a formula,
energy
the molecular masses of the two gases, (? G a the moan
values of the corresponding speeds as calculated from the
;
or,
if
of the
the arithmetic
magnitude
mean speed
be introduced instead
(?,
,,'.
We
If wo further
the mechanical measure of temperature.
remember that two gases must attain equal temperatures
when mixed together, the conclusion follows that tke equality
of the temperature of two gases consists in equality of the
mean energy
MAXWELL'S LAW
29
61
if we consider the
gases, not as already mixed, but
as initially separated.
And here we employ a law that is
the result of experiment, viz. that if two gases of the same
means
in
laid
mean energy
of rectilinear motion of
gases.
It is not, perhaps, superfluous to observe that the law
laid down for the state of a gaseous mixture applies only to
the condition of equilibrium. Hence those cases are excluded in which the mixture of gases is accompanied by
Mag.
*
Pogg. Ann.
[4] xiv.
c. 1857, p. 370
1857, p. 108.
Phil. Mag.
[6] xxxvii.
Abhandl.
1894, p. 245.
2.
of
the
burnt
and
transl. Phil.
62
Law
3O. Gray-Lussac's
ITS
BNEBGY
30
of Gaseous Densities
As simple consequences
empirically
tioned.
same temperature
gases are at the
of each,
molecule
kinetic energies of a
Two
E =
fynfi^ and E%
when
the
mean
iw.C?/,
S =
l
ip
and
&&*,
Ji 3
Hence
ft
it
>m'i
"V
This proposition deduced from our theory agrees subwith Gay-Lussac's law, that the quantities of
two gases which can chemically combine with each other
occupy volumes which, when the measurements are made
at the same temperature and under the same pressuro, aro
More
either equal or in the ratio of simple integers.
runs
the
theorem
demitictt
two
this
of
simply expressed,
stantially
gives
Pi
/> a
= n^
n2 Q2
a condition which
,
de
la,
is
WB^W^
obviously
:%Q,J
fulfilled if
Gilb*
Dalton*s atomic
xvi 1810, p. 4
Ann,
MAXWELL'S LAW
30
63
The
be made useful in
either, with
of the given propor-
may
We may
two ways.
by means
theoretical chemistry in
G-ay-Lussac, calculate
tion
Avogadro's
graph,
which
law,
is
discussed in
is of service.
31.
Avogadro's Law
=m
P*
theory, viz.
m*,
l9
is
For two
we must
different gases
sss
Nm
l
= Nm.
and
write
= JV w
/> 2
2,
we
NI
X,,
This
is
called
Avogadro's
law,
after
its
discoverer.
64
ITS
ENERGY
31
arrived
Although the considerations by which Avogadro
at it are closely bound up with views which were then
viz. with the asuniversally accepted, but are now rejected,
the
a
material
of
caloric, yet
experimental results
sumption
from which he
can be
easily
started,
deduced with
On
this
From
the laws of
Gay-Lussac
its
For
if
in the formula
= JVw
number which
we can
p.
[4J xiv.
1857, p. 108.
MAXWELL'S LAW
31
65
32. Coefficient of
Expansion
Prom
i.e.
equal,
when
IM&* = w
a (? a
2
,
iwj! 2
But,
if
From
x ,
a.2
2
a
we
a
or the
= im
two
gases,
=a
2,
all gases
are
the same.
1884
6.
Aufl.
I.
Mem.
xii. p.
310.
5.
AufL, Breslan
1895.
Gilb.
Ann.
66
33.
Maxwell's Law
Maxwell's law
in
ITS
ENERGY
for a G-as in
33
Motion
it,
mass
any
atoms -were endowed with independent motions, which they
executed without disturbing the equilibrium of the gas as
a whole.
If
we
and
its
there
is
each other.
If
we compound them by
the
known
rale of
of a molecule
deduced from
we
of dis-
by the velocity
See
MAXWELL'S LAW
33
67
molar velocity
is
at
mass
If the gaseous
places
law.
Resistance
different directions
distinction.
The pressure, too, of a streaming gas will,
In
therefore, no longer be equally great in all directions.
the direction of the flow the velocity and pressure will
be greater than in any other direction ; the pressure is
no
remember
consists in a transference of
molecules.
at right
is transferred.
Through a surface
in
unit
time a
there
of
flow
the
to
direction
passes
angles
volume Fa and a mass pFa, if a denotes the velocity.
also
mass
momentum
pFa*.
In
pFa\
Since,
now,
according
to
our
theory,
the
pressure
p 2
is
68
34
p +
pa*
a
/0
(^7rn
-f
a 2 ).
which expresses the law that the resistance increases proportionally to the square of the velocity.
That
Newton
been proved by
experiments on
Hawksbeo
and
by means
of
it
"
Him
MAXWELL'S LAW
34
69
velocity a.
(see
account.
In another
Hutton
to their plane,
AF
where A,
or
AF(1 +
Bq),
are constants,
q the circumference of
easy to indicate. Part of the air which is pushed in front
of the disc turns off sideways, and the resistance is thereby
diminished the theoretical expression has therefore to be
This fraction, however,
multiplied by a proper fraction A.
slide
depends also on the size of the disc, since the air cannot
of a
front
in
aside so easily in front of a large surface as
is
35. Reaction.
Cross -pressure
that
Pogg. Ann.
M&n.
Edm.
il 1790.
exliii 1871, p. 1.
de VAcad. de %elgigue>
xiii.
1882.
Mtm.
70
ENERGY
ITS
33
For
this
we
is
better explained.
in the consideration
start, as
12) of the
motion
of the gaseous
molecules so operates as
if
a third
to this surface.
moment have
mow
the
move
in the opposite
a.
served speeds of flow are for the most part considerably ICKH.
The greatest velocity which the wind attains that, for
instance, of the
most
fearful
storm
may
be taken at about
INF(G + a),
them momentum
equal to
INF(G
1
2.
Bulletin de I'Acad. de
AufL
2
iii.
The
Belgium
- a)
[3] xi.
1886, p, 180
Zfeeh.
p. 248.
calculation
Appendices.
is
Maxwell's
MAXWELL'S LAW
35
71
in unit time,
and these
momentum
a)
The two
is
directed,
an amount of
is
we can
momentum
equal to
+
while
it
loses
momentum
The
$NmF(G -
a)
amount
of right-directed
momentum, and
momentum which
a)
%NmF{(G
The excess
-h a)*
(G
a)*}.
momentum
which
%NmF(G - a)
right-directed momentum equal
for this
and
loses
amount
momentum
momentum
%NmF{(G - aY +
to
in the
left
half will
(G -ha)
by the
}.
In
72
left
ITS
KXKKfn
33
The
momentum which
right half
is
as
directed
time
is
given by
iF{ (G
Just
directed
large
a)*
is
the
(<?
left-directed over
excess of
momentum which
for the
a}*}
occurs in the
right-
left half in
xmifc
former increases by
Now, according to
is
else
nothing
than the
is
oxpivssod by
the formula
-
i7\
-j-
and this formula holds good equally well for the direction
which the gas flows and for the opposite direction. As
for the former direction, the added term
in
expresses the stress exerted by the stream, so for the opposite direction it represents the
equal re act ion -strips
which
is
For a
vessel.
by the formula
p = prwG
which holds good
2
,
magnitude
An
MAXWELL'S LAW
35
78
mass
of the
of gas
remains unaltered.
If,
then, the
gas
when
at rest,
and
its
volume were
therefore
the
molecular
speed
G?
pressure
of the gas
We have,
as before.
flowing
is
p = ^(?
when
therefore,
is,
less
The
than
cross-
when
Sound
36. Propagation of
of
See
W, Kdnig,
Wieti.
Ann.
xlii.
1891, p. 353
Z&itschr. f.
#%s.
u. chem.'
74
36
we
in alternate
propagated. If sound consists
the
the
air,
rarefactions and condensations of
speed of its
the
speed with which
propagation cannot be different from
at any place
arises
that
any inequality of the pressure
a sound
wave
is
1
Now, according to
motions
to-and-fro
the
our theory the pressure arises from
one
from
on
carried
of the particles, and is exerted and
with
layer to another by the same cause; the velocity
which a pressure- or sound-wave is propagated must therefore be just as great as that with which the particles of gas
air-filled space.
propagation of the
wave. The value of the component of the molecular motion
in the given direction, and not the resultant velocity of the
calculation of
particles, conies therefore into account in the
the velocity of sound and hence it follows at once that the
move
to
and
fro
in the
direction
of
speed of sound is about 332 metres per second, and is consequently considerably less than the mean molecular speeds
447 (28).
G 485 and
How much smaller the speed of sound is we may easily,
Q=
and with
in
1
Beibl.
2
iii.
p. 155.
More mathematically
strict calculations
MAXWELL'S
36
of the
third
motion
of
part
LA.
75
we might
look on the
magnitude
as the
mean
We
is less
than
G we
;
= IV
also
have
we
Newton's
of
obtain
But, as
tion.
Laplace
The
first
oscillations
constituting
With more
cxviii. 1863, p. 494), Boiti (Mem. dell' Accad. dei Lincei [3] i. 1876, pp.
762; Nuovo Cimento [2] xvi. 1876; [3] i. 1877, p. 42), and Brusotti
(Ann, Scient. del 1st. Tecnico di Pavia* 1874-5, p. 171); further by
Ann.
39,
Hoorweg
Mees and H.
76
done equally
ITS
ENERGY
36
easily.
of
p + dp represent the values
ratio
the
compression,
p +
__
dp
dp
__
dp
dp
is
_ &p
dp
greater value than Boyle's law gives it when the temperaThus the formula must be
ture is not taken into account.
is
which
a
factor
greater than 1, and, accordcompleted by
heat
"~
~~
c
of
/>
sound
V
That with
is less
than
example.
1,
this
~~
itself is
'
given by
Hxy/(7T<7/8c).
I will
= 1-405
c,
MAXWELL'S LAW
36
77
as is
detail in
55,
we
obtain
v
= 0-74 O,
its
= 332 metres
28,
per second.
The ratio O74 calculated for air holds good equally for
the other so-called permanent gases, and is also approximately admissible for those that are condensable, so that in
general we may assume the ratio of the speed of sound to
the
experiment.
= 332 + 0-603 3
= 332 (1 + 0-00182 3),
is
= 447^(1 + 0-00367 S)
= 447(1 + 0-00183 S),
when
3- is
small enough.
37. Effusion of G-ases
One
We
have,
openings in a thin wall.
the
phenomenon designated effusion
therefore, to consider
will issue
Oes.
through
fine
Benzenberg,
78
ITS
also the
ENERGY
name
37
diffusion has
We
employing
it
The
difference
only in
consists
this,
34, p. 224.
Sydrodynamica, Argentorati 1738, sect. 10
Trans. Boy. Soc. Edin, xii 1834, p. 222 Phil. Mag. ii, 1833, p. 175
Fogg. Ann. xxviii. 1833, p. 331 PML Trans. 1846, p. 573 ; 1863, p. 385.
3
Gasometrische Methoden, Braunschweig 1857, pp. 138, &G. ; 2nd eel 1877,
1
MAXWELL'S LAW
37
79
hand the
it
The
first
disturbance, that
as possible
For
wall,
as
small,
otherwise
on the mean
The manner
of the
80
easily
answered
if
we
ITS
ENERGY
37
of pressure and,
this assumpWith
are
small.
of
efflux
the
therefore,
speed
tion we can make the calculation just as for a gas at rest.
"We need only determine the number of molecules which
are forced per unit time out of the interior of the vessel into
To simplify the calculation we shall further
orifice.
the
provisionally
assume that
all
the particles
arrive at the
reach
it
orifice, let
PIG. 3
let
AB
number
of particles
make
AB
we wish
DB
of
AB in unit
GABD by taking the
through
O.
For
it is
AB
of time.
come,
is
If
parallel to each other.
of particles that pass
of time, we have to limit the space
by
number
DB
AB
from
which are
to determine the
by such
GA. A B
FQ, cos
$,
MAXWELL'S LAW
37
where
put
for
s is
GA
cos
81
number
s.
determine.
direction given
by
GA
It represents
in the
of particles
This
to begin their
brought to one and the same point A, and
then at a small area
rectilinear paths from this point
of a sphere described about A with unit radius there will
arrive
particles
for the
element
radii
AC,
to be the zone
AD
particles.
If
we
GDD'G' obtained by
K=
AG
4^, will be
is
take the
rotating the
as axis, then
27rrA,
4TT
EN =
82
ENERGY
ITS
37
We may
when we
great
reach
we
alter
If
we
we
obtain the
number
area
particles which in unit time meet the
tion
s,
which
at
of all the
an inclina-
therefore
is
cos
5.
A cos 5,
CD = A of
This expression
which
may be
EA F,
by the
line
HJ =
H'J'
this gives
for the
area
number
of particles
From
which
this
expression for
all
the zones,
i.e.
tude
To
all
calculate the
to have the
sum
same value
is 1,
we
is
a very
HJ = 8 =
i.
AH
= (AH + AJ),
mean
of the
MAXWELL'S LAW
37
we can
express
it
as a multiple of S
AH =
where
is
and
1)8
(v
for
83
if
AJ =
v$,
q,
then
from 1
2.(2*>~l)
to q
g(g
so that
S.rS
but
we have
+ l)-~g=g
= i,
for the
the
number
orifice
from
of
particles
the vessel.
we have found
represents
to the
The formula
shown
portion will
o 2
84
ITS
ENEEGY
37
uji-i
and consequently
flj
12
OnJkAn
VX V5
:
i>
MAXWELL'S LAW
37
The two
series of
85
"brass
Graham
The
greatest deviation from the law is exhibited by hydrogen, and, doubtless, for the reason assigned by Graham,
that for this light gas, which undergoes very great friction
in
narrow tubes, the plate was not thin enough and the tube
by St.
later
which
is
less
densely
filled
coming
Journ. de
VEcoU
ENERGY
ITS
37
them but
much
not
The equation
p = frrDs^*
both cases.
"7rD$ a
Q2
be the same in
p.
2
3
GfasometriscJie Methoden,
[This
is
decomposing water.
TE.]
MAXWELL'S LAW
37
The
87
38.
Graham's
experiments.
Thermal Effusion
tube
is
rises.
temperature
From
this it
follows that
by warming the
it
must be
warming
we
shall
it.
be
plate.
in this
flow, but
it
is
sufficient of itself,
according to
Maxwell's 1
suggestion,
is
called thermal
effusion.
air
He
*
3
PHI, Trans,
Pogg. Ann.
88
ITS
ENERGY
38
is
as in a formula developed in
37,
If in unit volume on one side of the partition there are
then
first
direction than in
Phil Trans,
727
MAXWELL'S LAW
38
We
89
of
N and O
have then
and consequently
if
or
The
side
is
as in
of
ordinary effusion
rarefaction
is
it is effected
by warming.
from a colder region to a warmer is
a result of the theory no less than of experiment.
The pressure therefore rises on the warmer side, and a
force opposing the motion is brought into play by which a
state of equilibrium is finally set up
and we have now to
under
what
circumstances
this
will happen. The
investigate
of a gas
In
when
this case,
of the pressure
on the two
= ^TrJY^flt!
and
_? 2
= ^7rJV" mri
2
the relation
Pi
a;
must
hold.
into
Introducing
temperature
defined in
15,
fi
in general,
and in
this
O2 _
f)
iz
O 2 2 _ iz
O
^ p iz
2 <H>
we
JPa
2 <H)
vy 2 ,
Pi :jp s
90
ITS
ENERGY
38
The thermal
of
39.
As
it
from which the flow takes place, while the other space into
which the gas streams is warmed. The same action therefore occurs which has already been described in
18, 19.
Where the gas expands it cools, and where it is condensed
it gets warmed.
The explanation
of this behaviour
is
also essentially
that which has been given for the case in which it was
assumed that the condensation was caused by the pushing
On
partition the
particles that
L.
Nat an son,
MAXWELL'S LAW
39
stream
91
of
the particles
ITS ENEKG-Y
40
CHAPTEE IV
IDEAL AND ACTUAL G-ASES
40. Inexactness of the Theoretical La.ws
ALTHOUGH
all
the
yet
same
deficiency as
There
pressures exerted by the components separately.
must therefore be present several causes of different kinds
which act together and cause the deviations from the theolaws in either the one direction or the other.
Just as incompletely do the experiments on the effusion
of gases agree with the conclusions of theory.
Neither
does the speed of flow, as determined in Graham's and
retical
Bun sen's
retical
law, nor do
1890
Wied. Ami.
xli.
1890, p. 088
40
93
39.
Joule and Lord Kelvin
have measured with great care the alterations of temperature that occur in the vessels from and to which the
flow takes place, and by these observations have proved that
the whole amount of work done by the gas in the receiver,
when it flows out, is not to be found in the increased energy
of the gas in the outer vessel which has been warmed by
compression, and in the heat that has been produced by the
exactly as represented in
of the heat
of
obeyed.
Pogg. Ann.
Iv.
1842, p.
5.
94
distinguish in
increase and the
to
ITS
ENERGY
the coefficient
a gas, viz.
of
40
volume-
the
pmsw/0-increment
former determines the increase of volume that occurs with
rise of temperature when the pressure remains constant, and
of
coefficient
is
1
alone showing the reverse property. Eegnault further
observed that the two coefficients are not entirely independent of the pressure and temperature of the gas, but that
they increase with the pressure and diminish when the
temperature rises.
This behaviour of gases
may be
whether
as
air,
this gas
the pressure.
ties of
values
air,
2*481 at
20 C.
2-478
50
2*468
100
150
200
2*461
2-450
The falling-off
more than air.
Mm.
4-0
its
41.
With vapours
95
expansion-coefficient
of air,
while at lower
Vapours
developed
from
liquids
by heat
the
much
larger than
with gases proper. On the whole, vapours comport themselves as gases.
They obey Boyle's law approximately,
so that in their case, too, the pressure and volume vary
very
nearly in inverse proportion and further, both pressure and
volume increase with rising temperature in almost equal
Dal ton's law also holds good
ratio, just as with gases.
for mixtures of vapours and gases with approximate exact;
We might,
correspondingly, the vapour-density approaches constancy in its value as the pressure falls off.
and,
96
ITS
ENERGY
41
As regards
atmospheric
air.
Absolute Boiling-point
this direction
vapour which
by this, that the
equilibrium is maintained by vaporisation and condensation
The molecules of the vapour
at the surface of the liquid.
which in their to-and-fro motion strike the liquid surface
will not all bounce back, but a part will be retained by the
For the
lies
above
liquid
is
characterised
measurement
of the
maximum
42
97
which
their
cohesion
temperature and
is
diminished.
If
we
increase the
motion
lost
98
ITS
ENERGY
42
Conformably to
Mendele Jeff's
tion, therefore,
we must
otherwise approved
1
itself in so
many
[Voigt
Gott.
42
99
An
Modification
necessitated
is
nagel allowed
by
it,
by assuming
We
other particles.
may, therefore, assume this diminution
to be directly proportional to the density of the gas, or inversely proportional to its volume, and thereby obtain for
pv
where
and
Note in a Memoir by
Thomson
exliv.
p. 336.
2
PHI. Tram.
clii.
v. 1871, p. 563.
1862, p. 588.
H2
1854 >
100
ENERGY
ITS
43
this
we have
bound together in
is,
same
8),
viz.
p =
1
XL
Thtorie
Mteamque de
la Chaleur,
ii.
1864, p. 215
Ann.
CMm.
Phys.
[4]
1867, p. 47.
2
ProefscTwiftj
44
which may
also
101
be written
if
tion of the
of energy to the
On
of
part which
where
is left
free
The assumption
of
is
also
102
ITS
ENERGY
44
between the
permissible, namely, that the forces acting
molecules
molecules drive two encountering
away from each
actual
contact.
an
of
the
moment
other even before
We
envelopes as spherical, we might justify the name molecular sphere, which we will retain until in our investigation of the free path ( 63) we introduce the term sphere
of action, used by Clausius, for a sphere with a similar
meaning.
way
as
Laplace
magnitude
of similar
actions of attracting and attracted particles, it is proportional to the number of attracting particles on the one hand,
and of attracted particles on the other ; and it is conse-
and not
44
103
law becomes
av-*)(v
Boyle's
5),
(p
where
av~*)(v
- 6) = E(l +
aty,
a constant, and a, 9- denote, as before, the coefficient of expansion of gases and the temperature.
Only
is
45.
form
in
with
Eegnault
this
the
formula
of
interpolation
by which
~ = 1 - A(m -
in
1)
+ B(m -
I)
m=
Vjv
while
and
are
constant
coefficients.
is
If,
as
in
taken as unity
gases in a
Morozowiczin
2
3
M&m.
104
ITS
ENERGY
45
or
- Bv~* =
1
pv + (A + 2B)v-
4-
A +
B.
In
this
l
pv + av~
bp
aZw~ 2
= JS,
it
is
PV =0 _!
p =
for then
(a
S)^"
runs
<zfo?~
= jB,
ab
J2
=A +
2JB,
=5,
=i+A +
B.
B found by Eegnaultwe
shall afterwards be able to calcuthe numerical values of a and 6, which represent the
magnitude of the cohesion and the extension in space of the
molecules. 2
late
46. Pressure-
and Volume-coefficients
Van
Van
See Chap. X.
xi.
1880, p.
46
105
termed the
pressure-coefficient, is the
more
easily obtained.
The
remains unchanged
(p
is
av~*)(v
when we
we get
give to v the
6)
= E(l +
aty
same value
in both.
to
By
0,
subtraction
or
whence we obtain
formula
also
observation
theoretical
Since
is
now
to be taken as constant
the f ormulse
(p
av~*)(v
Z>)
and v as variable,
106
ITS
ENERGY
46
the value
let
the cohesion
vanishingly small.
is
we have
ay
(1
-ft^o-V,
so that
<a
= a.
then
we have
a w > &p>
Ct>.
*,
= 0-003661,
af
= 0-003667
but for all other gases the relation between the values of
the coefficients required for the second case is fulfilled.
Van der Waals's theory thus agrees in all points
with experiment in so far as it rightly expresses the general
This agreement speaks for the fact that to a certain
laws.
the
degree
assumptions on which the theory rests correspond
If we should consider van der Waals's
to reality.
theory
also as probably not yet perfect, we are yet justified in the
view that in it the first step is taken along the
which
we
shall arrive at a
path by
completely satisfactory kinetic theory of
46
107
it
Completions of
47.
{p +
differs
that
a(v
from that
in
this:
the
which
in
o is a
constant and
perature
@=
(B)
a"
1
.
= C4-*
J3)JB(1
aS),
xiv.
Mechanische
108
in which. A, B,
are constants.
= (A~
~~
ENERGY
ITS
47,
1
Similarly Battelli put
B@ m)E(l +
a9),
Gr.
@-y
/e
,
where y is constant.
These formnlse exhibit much better agreement with experiment than the simpler formula of van der Waals, as
is to be expected from the greater number of disposable
3
constants.
Yet, as Kprteweg has remarked, the formula
of Clausius deviates from the observed behaviour of gases
in many other regards more largely than that of van der
Waals. For our theory the more complicated formulae
are less valuable than the original simpler one on account
of the difficulty of their interpretation.
Amagat
by giving
it
has completed
equation
the form
{p + Av~\v
which contains
- s)}{v - b +
B(v
n
b)
= B(l +
o9),
This formula
constants, A, B, s,
has proved itself good in a comparison with the observed
behaviour of hydrogen.
Bolt-zmann and Mache 5 assume the formula
(p
b, n.
five
av~ l )(v
48.
6)
= E(v +
25) (1
a9).
it
Memorie di Tormo
[2] xliv.
1893, p. 27.
p. 87.
48
109
moment
of a collision
the deviations as an
hypothesis might just as well cause
two hypotheses.
can frame for ourselves no idea of the cohesion of
We
gases that
is
strength
customary supposition
We
attractive forces of
110
cohesion,
we
ITS
ENEKGY
48
off.
shall
have
to
an
effect
moments
force, that
of
it
by
nothing
cohesion might be able, at least now and then under favourable circumstances, to bind together two colliding particles
so fast that they traverse the next stretch of their path
together as a double molecule. By this a state of equilibrium would be produced in the gas in which, among the
In his memoir on Temperature and the Measure of Temperature (Pogg.
Ann. Erg.-Bd. vi. 1874, p. 275) Eecknagel also considers effects of two kinds
attraction of molecule on molecule and actions within molecules. The latter
1
'
'
manner explained
later.
48
111
49. Playfair
The hypothesis
last
of the
of the
has been given by Playfair and Wanklyn, 1 an explanation which embraces all the other deviations of actual gases
from the theoretical laws.
According to what has already been said regarding the
hypothesis, it is only necessary to remember the mechanical
definition of temperature, given before in
14, to see at once
the possibility of the explanation of all anomalies. Accord29, the mean value of the kinetic energy of a
ing to
molecule, even when of different kinds, forms the measure
of temperature.
Consequently, in a gas whose molecules
are either wholly or in part bound together to form larger
aggregates, the kinetic energy contained in unit volume is
less
441
112
The
ITS
-vapour-density
is
ENERGY
49
smaller at higher
is explained
larger at lower temperatures
the bond
loosens
heat-motion
the
for since
temperatures and
in the same way
between the molecules the molecules will be lighter, and
therefore the gas or vapour will be specifically lighter at
;
heating
therefore at
vapours and condensable gases must
same thermal coefficients of
high temperatures attain the
while at lower
expansion as the so-called permanent gases,
more
largely.
temperatures they expand
If this explanation of the deviations is really true, a con-
gases
all
further separation
is
possible
ground
asEegnault
M6m. de
50
5O.
113
Horstmann's Explanation
Horstmann
by
hypotheses described in
Horstmann
48, although
gives
He
it
theory,
the hypothesis
necessary consequence of
rectilinear
that
motion of
fore, of a
straight paths.
the
is
the
increases,
increased,
therefore,
pressure
density
If,
not in the same ratio, but in a less degree, because the
increase of the
number
a diminution
the other hand, by an in-
of collisions causes
On
These views,
schrift,
vi.
1868, p. 53
also in
Ms
Habilitatiom-
Heidelberg 1867.
I
114
ITS
ENBEGY
50
their condensing point, but also the fact that their volumeand pressure-coefficients are greater than those of ideal
gases.
51.
Claims of the
Two Explanations
of the
of the
two explanations
of the
anoma-
distinction
pursuit
probabilities offers
as
was
would be only
speeds.
law of disthis
not only
form of combination
It
of the molecules,
of explanation
of
effusion
at
different
At present we
115
CHAPTEB Y
MOLECULAB AND ATOMIC ENERGY
Dulong and
52.
THE
theories of
Petit's
Law
for G-ases
and
II.
One
in
kind of
bnt the
entitled
of those
of the
energy
which
14, 16)
its
is
all gases.
On
Dulong
and Petit
amount
for this
of all
we
generally say
bodies is the same.
more
we remember
116
ITS
ENEKGY
52
on the contrary, in
graph we
shall
examine
its
meaning more
closely.
Prom
it
question
its
of molecular motion.
We
53
unity,
of
117
where
equivalent of this
energy
amount
is
With
volume
of heat in
terms of mechanical
H = Jcp.
this value of the
of the gas
we have
-5T
= f^
which we obtained in
tion
the gas.
The
formulae
are
given
sufficient
to
determine
the
ratio.
Mayer
256
Pogg. Awn. c. 1857, p. 377 Abhandl. liber Warmetheorie, 2. Abth. p.
Mech. W&rmet"heorie, iii. 1889-91, p. 35 ; transl. Phil. Mag. [4] xiv. 1857,
1
p. 108.
2
xlii.
1842, p. 239
p. 28.
118
ITS
ENERGY
53
which
p,
volume needs
at constant
cp<&
for the
same
rise of
Cp
is
needed, where
pressure.
The
(0
- c)p
is
volume
v,
is
is
pv =jp.
The production
was zero
of this
work, the
initial
value of which
at the
= J(G
J here
we
K_3C-e
JET~"2~T"~'
is
deter-
58
119
wherein the
,
The
kinetic
of
expansion.
energy
thereby simultaneously increases by
J5: o
must be added
to
without
where p
p Q is the resulting increase
change of pressure the relation
of pressure.
Since
for this
holds good,
we have
the
K - JT
Jff-'Ho
-- c
'
54.
is
quite similar.
Monatomic Molecules
To prove whether or not this theoretical formula corresponds to the truth, we may apply it to the special case of a
gas whose molecules consist each of a single atom. To this
class of gases, which we may call monatomic, if we may use
this word in a different sense from the term single-valued,
belong the vapours of mercury, cadmium, and perhaps zinc.
Eor these monatomic vapours the possibility of assuming
proper motions of the atoms, in addition to the motion of
the molecules, falls to the ground we shall therefore have
to suppose that in these vapours the kinetic energy J3T is
identical with the total energy H, provided that, like gases,
they may be considered free from cohesion.
;
120
For
ITS
ENERGY
54
K =H
jH"0>
and thus
C-_c__
"
~~3'
c
or, if this
assumption
is
two
specific
heats must be
(75
c
Led by
3'
Kundt
this consideration
and
Warburg
have
= 1-67,
c
if
after
Bontgen's
1-405.
C/c
= If
for a
monatomic
gas,
we must assume
established
that the
by observation,
theoretical hypothesis which led to its discovery corresponds
also to the truth.
In mercury vapour, therefore, and in the
= that is, there is no other
other monatomic gases,
of
kind
but
that
of the progressive motion of
any
nergy
is
H K
the molecules.
1
viii.
1875, p. 945
Pogg. Ann.
clvii.
1876,* p. 353.
2
Pogg. Ann.
Bams ay,
cxlviii.
1873, p. 580.
Proc.
Roy. Soc.
Gases of the Atmosphere, London 1896.
ML
1895, pp.
282,
286;
54
121
H H
We
and 124.
That monatomic gaseous molecules also may be capable
of oscillatory motions in their interior we may look upon
as probable, since in their spectra whole series of different
But these motions, as we may assume in
lines are found.
accordance with E. Wiedemann's observations, require
1
come
55.
Polyatomic Molecules
In the case
of a
Comp.
123.
122
ITS
ENERGY
55
greater than
or,
more
gives
generally,
limits
for
or the value
1
and f
of
of
the two
lie
between
the ratio
must
in
all
cases
the
K_
C-c
specific
At
least, for a
H
will not
be contested
HQ
c"
formula
mined
number
of
gases,
though un-
55
123
this
value
1-405
this
we
obtain from
2
Dulong's and Masson's observa3
of sound, if we apply a correction 4 that
1
tions
the energy
therefore of the molecular motion in atmoair
stands
to the whole energy
contained in
spheric
the gas in about the ratio 3 5. Thence it follows that
the two parts into which we may break up the whole
H K
of the mole-'
energy JEZ", viz. the internal energy
cule (which we may distinguish as the atomic energy] and
the energy
of its progressive motion, must bear to each
other nearly the ratio 2:3; or, more exactly, we have
-
?=
The
values of
unit volume.
If,
however,
is
K have
Ibid. [3]
liii.
refer
their
Compare
36.
Wied. Ann.
1878, p. 321.
oxlviii. 1873, p. 580.
d.
W miner's Lehrbuch
it
we may
volume, and
1
and
0-646.
iv.
Pogg. Ann.
Journ. de Phys. Ixxxix. 1819, pp. 321, 428.
Inaug. Diss. Breslau 1882 ; Wied. Ann. xviii. 1883,
Pogg. Ann. Ixxxv. 1852,
p. 1.
p. 94.
iii.
p. 522.
124
ITS
ENERGY
55
Hence if,
as, for instance, the molecular volume.
21* of the Mathematical Appendices, we denote the
mean energy of forward motion of a single molecule by J3,
and the mean value of its internal atomic energy by @, the
volume,
as in
two magnitudes
ratio of these
above, viz.
is
H-K
c==
@
n'
Since
air is
it
composition
this formula to
But
it.
and
J-
The energy
the energy
This ratio
of
H
is
= 0-646,
an atom
is
0-323.
j-
vapours.
The observations
of
Dulong
(D)
and
Mas son
Atti del
1st.
Yen, [5]
vii.
1880-1, p. 491;
55
(MF),
3
(0&),
Lnmmer
and
less exact.
and
also those of
From
125
H, @
E, and
E.
Ivii,
1895, p. 322.
[3] hrvi.
1862, p. 206.
p. 565.
126
ITS
ENERGY
cent.
55
56
Approximate Calculation
56.
of the
127
Atomic Energy
in Polyatomic Molecules
A
ratios
r=
where p
is
GP,
= cp,
Then
unity.
cules to their
H= 2
This
when we
r--y
*
new form
p = J(G - c)p
developed in
53.
If
we
write
it
in the
form
it
all gases.
It follows for our purpose, from this law, that the value
in different gases is inversely proporto
of the ratio of
for
Ann.
volume
of the gas at
128
constant volume.
ITS
ENERGY
56
Hence the
known.
heat at constant volume is with much
to observation and measurement
greater difficulty accessible
On this ground
constant
at
heat
than the specific
pressure.
been made.
have
former
the
of
no direct determinations
values
measured
the
from
calculated
be
But their values
But the
specific
may
are strictly exact only for perfect gases, the calculation can
indeed be admitted without hesitation only for such gases as
theoretical calculation
Lussac's laws
or of
Avogadro's
law.
served by
4
3
2
Eegnault, E. Wiedemann, Winkelmann,
specific heat of
many
gases and
made by
3?.
Weber, 6 with
carbon.
With
1
[Dr.
Joly's
not be ignored.
direct determinations
See
PML
961.
TB.]
M&m.
4
5
6
by means
p. 73,;
must
Pogg. Ann,
p. 580.
of his steam-calorimeter
p. 82
civil.
1876, p.
Pogg. Ann.
1.
oliv.
1875,
56
129
Clan si us
of the ratio
of
the
kinetic
energy K of
1.
130
ENEBGY
ITS
56
l
The number
perhaps ethyl ether
and possibly seven 1 chlorine compounds, is so small that we
shall be inclined to consider the rule, that in gases with
of exceptions, including
atom is smaller
polyatomic molecules the mean energy of an
be a veritable
to
a
than the translatory energy of molecule,
other laws of
the
and
law of nature, which, like Boyle's
under certain circumstances.
gases, admits exceptions
The possible grounds for such exceptions will appear
from the following considerations.
57. Dissociation
and Disgregation
By
We
state exert
other,
cules,
magnitude @
is
the
sum
of the
amounts
of
both kinds of
way
the
mean energy
energy and
its
of
an atom
is
made up
of its kinetic
forces of affinity.
From this it
first
we
55.
57
131
calculated values of
translatory motion
as
of
energy E.
strength of chemical affinity from the atomic energy c.
But from what has been said regarding the character of
the atomic energy we further draw this as a necessary con-
an atom, cannot,
clusion, that
speed
would
we
choose.
that the
mean energy
of
of translatory
cules.
That an atom can be actually loosened from the moleby an increase in its speed cannot be
cular combination
doubted
by a
amount
is,
phenomenon
up
the
up
an increase in the
With a moderate
favourable circumstances.
Sitzungsber. d. Berl
A'ka.d,.
1894, p. 295.
132
57
is
of
of
ways
comes
it
the letter
which we
will denote
by
e.
ment
of
of the
maximum
an atom
58.
is
capable of giving
rise.
is not in
general true, that the
heat of gases at constant volume is
independent of
the temperature.
cannot well test by direct observation of the
specific heat at constant volume whether this
is
specific
We
assumption
53.
58
133
admissible.
known
if
laws of
the
gas in
question
p =p
which
is
x const,
that
we have
in
which
J"is
volume has
also a constant
Now Begnault
magnitude.
has experimentally shown
"Wiedemann 2
for
for
air
carbon
at constant pressure
does not depend on the temperature. From this we may
probably assume that all gases whose molecules contain
acid,
and B.
nitrous
Wiedemann
oxide,
Pogg. Ami.
clvii.
1876, p. 1.
134
ENERGY
ITS
58
ana
dependence of the specific heat <7*on temperature. L u s s
found that for a series of gases it varies also with the
pressureincreasing, in fact, with the pressure. For these
1
gases,
therefore,
at
constant volume
and 100
temperature
rises,
thus
atom
to the molecular
and Gray-Lnssac.
1
Nuovo
AM del 1st.
dm.
[3]
Veneto
1894, xxxvi.
[7] viii.
[4]
1895,
i.
1896,
1897, vi.
1898, viL
58
135
of the molecause
was there
possible
indicated in the circumstance that the atomic energy is not,
like the molecular energy, simply kinetic, but partly consists
cular
of the potential
energy of chemical
affinity,
latter is
And
this
For
because
overcome.
cule,
now
of
act exactly
136
ITS
ENEBGY
Wii liner.
Mtlller
is
attained.
augments the
number
made by
Wtillner.
At the present stage of our knowledge we shall have
therefore to assume that the law discovered byClausius,
according to which the molecular and atomic energies
should bear to each other a ratio that is always constant,
holds good in its full strictness only for diatomic gaseous
for other molecules, however,
molecules
that ratio is
variable with the temperature, and may, indeed, according
;
to
perature
But
rises.
58
137
for
these, then,
true.
exact-
exactly
chlorine
chlorine
the striking circumstance that, of the possibly eight substances for which the ratio of the atomic energy e to the
molecular energy
chlorine compounds.
The
in the
same way.
If the
at
energy
to the
Dependence
of the Specific
of Atoms
Heat on the
Number
From
energy
energy
it
On closer examination
Among them the fact is
CO,
N H
them
NO
@ is
138
ENERGY
ITS
59
law
of
Dulong and
it is
the ratios
-=
and
1-4
E?=0-33
(see
54) in giving
1-67
and
= 0.
The
idea
ratio of
to
indicate this;
at
least the
mean
= 1-27
numbers
values
and
1 =
0-5.
59
law
of
139
specific heats
on the
all gases.
On theoretical grounds
the
forward
that
view
the mean energy c
put
of an atom must be equal to the kinetic energy
of the
molecule, but he found this assumption not confirmed by
5
observation.
Pilling obtained a formula that agrees very
well with measurement by starting with the hypothesis that
the atoms exert on each other forces which vary inversely as
the sixth power of the distance between them. Eddy 6 has
made the more general assumption that the action is inbut
versely proportional to some power of the distance
Bdcharz 7 has proved by general theorems of mechanics
that an assumption of this kind respecting the law of action
between atoms is inadmissible, as it does not satisfy the
conditions for the stability of the molecular combination.
A new memoir by Staigmuller 8 on the kinetic theory of
polyatomic gases must also be mentioned.
Boltzmann 4
6O. Degrees of
Freedom
Motion
of the
In another way Maxwell 9 and "Watson 10 have attempted to answer the question by taking into account the
1
Die gesetzmass. Beziehungen zw. d. Dicht. d. spec. Warme u. d. Zusammensetzung der Gase, Grdttingen 1857; Gb'tt. Nachr.1851, p. 165; Ann. d.
p. 301.
Grundriss
xi.
1875, p. 357
Scientific Papers,
ii.
418
p. 242.
10
p. 81.
2nd
ed.
1893,
140
ENERGY
ITS
60
this
It shall be given
then by Eoiti, and later also by Violi.
here on account of its clearness, although very weighty
objections must be raised against its admissibility.
We can assert with regard to a single massive point that
it has a threefold freedom of movability, since its motion is
determined by the values of the three components of its
The above-named English physicists ascribe to it,
velocity.
therefore, three degrees of freedom, which Boltzmann expresses as three kinds of movability (Bewegtichkeitsarten) A.
material system consisting of a multitude of particles possesses
.
as
many
number
of variables
determination of
its state of
motion
if
made up
of atoms,
This mode
and
If
marked by simplicity
therefore leads to very simple results.
added to a gas whose molecules consist
clearness,
heat
and
is
it
^W
same temperature
heat
amount
the
and
If
the
W, be given to
further
the
of heat
2:3;
"
Wjr JfJ
3'
whence
W__Z
~~3*
Wim.
AUi
Lincei
[3]
i.
Pogg. Ann.
1877, p. 774
Nwwo
clx.
1877, p. 175
Cimento
[3]
ii.
1877,
p. 61.
*
AUi delV
p. 183.
[3] vii.
1883, p. 112
Nuovo Cimmto,
xiv. 1884,
60
If
now
a gas
which possesses
141
degrees of freedom
is
heated
much, each
just as
is
needed, where A
is
a constant.
on the contrary,
J(q
is
heat, therefore,
The
needed.
The whole
volume is
amount
Aq)
W+ W =
(;}
&q
2)
T7
must be
in the ratio
shall here
has investigated.
=
q
5,
since
142
three coordinates of
its
ITS
ENERGY
60
which
5=1
o
and gives the value
if
for
= Q.
if
we
considered
But
by
five coordinates.
against this
Pirogoff
about
its axis of
In
we may suppose
that inside a
Abth.
p.
247.
60
143
inside a molecule,
excellently.
It is not for all diatomic gases that the value of the ratio
of the specific heats is independent of the temperature and
equal to 1-4 ; for many of these gases it is smaller and is
than
1*4.
Another, though not essentially different, possible explanation might be found in the assumption that the
number of degrees of freedom q is not 5, but 6. There
is, indeed, a sixth kind of movability if we drop the assumption that the two atoms of the molecule must remain at an
invariable distance from each other.
With the value Ji
0,
the ratio would approach the value
*
1-33,
E.
v.
1890, p. 64.
Helmholtz,
Licht-
Gase,
144
as
from the
table in
55
we
see
is
ITS
ENERGY
60
some
one
of
axis
position of
are given;
its
centroid and
there
are thus
the
more
If, therefore, we
elements required for its determination.
ratio of C to c
the
assume q to be greater than 5 or 6,
becomes smaller, and with increasing q approaches more
,and
more the
value, but
weighted.
of
freedom
Franklin Inst
[S]
60
Boltzmann's
energy
among
objection
PART II
149
CHAPTER VI
MOLECULAR FEME PATHS
61.
THE
theory of gases developed in the foregoing investigahas been shown to be in agreement with experimentally determined laws in a series of important points ;
Boyle's law and Dal ton's law respecting the pressure of
tions
gases
are necessary
consequences of the
theory,
which
As
soon,
indeed,
as
Kronig
and
Clausius 2 had
cili.
p. 240.
p. 315.
2
4
150
Jochmann, and
61
and
28.
On
we
the contrary,
by
fast as sound.
oviii. p.
153.
61
conductors that
life could
which
62.
it
151
exist.
Refutation of
tlie
Objections
of
such a piling
up
us traverses a path of 400 metres without actual disturbance hardly ever occurs but each air particle collides with
some other exceedingly often, indeed many million times a
;
second.
motion
of molecules
each
among
other,
thither in
;
By
'
u.s.w.,*
PMl
Phil Mag.
[4] xvii.
1859, p. 81.
152
distant
is
not at
all
made
62
in the
kinetic theory.
in accordance with this
struck
is
follows that
The
transference of heat
is
there-
development
1
62
153
it.
For, owing to that conflict of conceptions, a new
and interesting side of our theory has been brought forward which is worth a searching investigation. As in
Part I. of this book, the speed of the molecular motion
and the supply of energy therein contained were considered
and calculated, so there remain for this Part II. the
to
collisions,
will result
of
observed in gases.
The
As
in
original
of
compound
interest.
we
Before
which
p.
Pogg. Awn. 1858, cv. p. 239 ; Abhandl. 11. Warmetheorie, 2. Abth. 1867,
MecK Wdrmetheorie, iii. 1889-91, p. 55 ; transl. Phil Mag. [4] xvii.
260
1859, p. 81.
154
direction) is
63
is
idea, viz.
What
and express
density of distribution of the particles at rest
If there are
in terms of their mutual distances apart.
it
cubes,
which Clausius
calls
the
mean
distance
betioeen
is
holds good.
Since the density p may be expressed in terms of the
and the number
molecular weight
by the formula
(13)
p
= Nm,
pX
= m,
Prom
the mean distance X between neighbouring moleClausius deduces the mean probable length of free
path by comparison of that mean distance with the smallest
cules
volume X3
63
155
diameter of a molecule.
molecules cannot come
touch, but that they are
actual contact occurring,
But
mean
possibility
it is
it.
The
radius of
sufficient
might
call
we
its
molecular sphere.
Denoting the radius of the sphere of action by s, and,
2
therefore, the area of its central section by 7r$ we find that
if the moving particle considered advances by the mean
,
7r$ x.
described
is
as
much
smaller than 1 as
7r$>
is less
than X 3
The
by the
ratio
= 7T* /X
2
2
,
156
or
by the
63
it, is
"We may
^/(X
before
that
it
9T*
2
)
-H^/VS:
(X
7r*
)/X
2
-Trs-
)/?
to 1
If we
lies a simple meaning.
from
were
the
struck
that
the
molecule
pushed
suppose
it
into
the
same
cube
the
elemental
of
interior
containing
face through which the moving particle entered, it would
whose area
X2 a
is
The two
64.
From
64
while
M(l
7rs
157
)
particles pass through this layer.
2
2 2
2
2
the third layer M(l
7
/X ) 7r$ /X collide, and
2
2 3
7rs
M(l
/X ) pass through. Proceeding in this manner
we see that in the xth layer there are ptobably
/X
So in
- 7r*
M(l
/X r-
W/X
From
number M,
on
its
65*
is
completion
From the foregoing formulae we can calculate by elementary methods and without great difficulty the probable mean
value of the lengths of the paths traversed by all the
To find this it is only requisite to calculate the
particles.
sum
molecules
paths traversed by the
number, that is, by M.
Of the
particles there remain MirsPfK* in the first
thus the
layer, and these therefore traverse only the path X
of all the different
and
to divide
it
by
their
sum
2
2
2
/X )7r* /X the sum of
2
2
7r* /X )7r$
In this way
whose paths is therefore 21f(l
/X.
we find in general that the sum of the paths of the particles
which collide in the xih layer is xM(l^ Tr^/X 2)*- W/X, and
the total sum of the paths traversed by the whole of the
number M(l
particles is
therefore
S.a?af(l
This formula
is
developed in
158
of
65
for all
by the formula
This
or. in
sum may be
easily calculated.
For
our case,
2.a?(l~-7r
Hence
it
results
from
on the average
L =X
:X
=X
2
:
7T*
2
;
then expresses that the mean free path bears the same
ratio to the mean distance separating two neighbouring
the elemental cube has to
particles as the area of a face of
action.
the
section
the central
sphere of
of
it
From
this proposition
than
X.
shows that
is
considerably larger
65
159
many molecules
like itself
Free Paths
The
is
a shorter path
expression M(I
probable number
than a longer ?
2
T^/X )*, which we have found
path
we wish
traversed,
number
x,
mean
free path.
mean
free
we can put
for the
path
= x\ = qL,
we
shall
have
path
of length
Qq,
particles
qL
is
M(l-l/Q-)
We do
Qq
.
know
not indeed
we may
look
160
upon
it
finitely great,
(1
Q were
If
66
actually in-
we should have
= e~ =
l
Q-I)
1/2-718
0-36788,
where
is,
M(l
- Q~
l
)
Me~.
According to this formula we have calculated the following table. Out of every 100 particles
99
98
90
82
78
72
61
37
14
5
2
1
without
The
collision.
free
path is considerably
only play of chance,
therefore, is to set up in this case too, as well as in that of
the distribution of speeds, a uniformity that is maintained
table
The
67.
Free Path,
if all
Speed
The
still
value
we have found
a correction,
for the
mean
free
path needs
The
value of the
mean
free
path found in
65, viz.
MOLECULAR
FIIEK
OATHS
161
We
L=
for the
medium, and
former value.
this
differs
From
this equation also we can deduce a proformer and of similarly simple meaning, viz.
L
68.
fX'/ir*
=X
But these
The
at a
temperature which
the same.
by Maxwell.
is
everywhere
162
Maxwell
free
68
of this law.
too,
alteration
mean
free
path
L=X
for
/ir*V2.
The
in
[or, still
Path-volume
69. Molecular
The name
of molecular
Loschmidt 3 to
a molecule describes
The magnitude
when
traverses its
it
volume
mean
free path.
is
mula
it is
equal to
3
A,
If
/4V2.
we replace
for-
in this expression
2
3
[4] xix. p.
28
Hi.
Abth.
2,
p. 397.
i.
p. 387.
69
expressions
163
given by N\*
1, we
path-volume is determined by the
is
= X V2 = 1/4A/2JV,
3
frrsPL
its
value
is
7O.
number
all
gases, since,
N has the
same
Frequency of Collisions
From
now
two
is
a straight
line,
particle to
move
in
From
1/27
If
mean
we put
= 0/1,.
ft calculated
path
the value
=X
/7rA/2>
(
68) on the same theory.
on the contrary, in accordance with Clausius' theory,
we assign to all the molecules the same mean energy and
the same value G of the speed which corresponds to it, the
interval between successive collisions in this case would be
as calculated
If,
T'^L'/G,
M
164
where
is
this assumption,
70
67 on
viz.
L'
and where
as in
27.
there
is
the ratio
T=
between successive
and the collisionthan on
larger on Clausius' theory
frequency a
little
Maxwell's.
71.
and Temperature
According to the theoretical formula we have found,
the value of the molecular free path depends only on the
2
volume X3 of the elemental cube and the area vrs of the
central section of the molecular sphere of action; the
molecular speed ii, by which the value of the temperature
of the gas is determined, does not, however, occur in the
formula.
the gas.
If now the size of the sphere of action were not variable
with either the pressure or the temperature of the gas, it
would follow that the molecular free path cannot depend on
the temperature, but only on the density of the gas and,
;
71
165
The assumption
But
this
view
is
Numerous
especially by the phenomena of dissociation.
instances of gases and vapours can be cited wherein the
molecules are composed of more atoms at lower temperatures
than at higher. The vapours and many gases deviate at low
temperatures from the laws of the ideal state of gas, especias has been
ally from those of Boyle and Gray-Lussac,
deviations
these
in
described
being such
already
Chapter IV.,
that the gases have too great a density and an expansibility
These and
is much greater than that of ideal gases.
that the
conclusion
the
to
us
force
other
many
which
irregularities
mass
by
rise of
temperature
the
space, and therewith
smaller
become
both
will
rises.
We
occupies
distance to which the centres, or,
of gravity, of
a collision.
166
71
that is at all
temperatures than at lower entails nothing
or probability. For the speed of
reason
either
to
contrary
the particles increases with the temperature, and, therefore,
also the intensity of the stress during
and
collision,
it
is
We
elastic
are
bodies,
by atmospheres
compressed
the
of ether
which,
more during a
collision the
will be later
allied
phenomena.
W. Sutherland 4
He assumes forces
these facts.
between the gaseous particles, when very near together,
which are not repulsive, as Maxwell takes them, but, on
different explanation
of
2
3
xxxv.
*
p. 133.
clii.
p, 579.
2, p.
339.
clvii. p. 51.
[4]
Scientific Papers,
1862,
certain
of
p.
li.
357
1860,
cl.
p.
325
71
167
The
amount
is
of diminution of the
much
therefore diminished.
process by which
way *of
Sutherland
calculates
the
35* ;
path
without
this
object
attaining
free
is
given in
calculation.
the encounter.
The number
amount
E of
potential energy
activity
on
energy
particles.
of encounters which a particle undergoes in unit time, by
reason of the attractive forces, is increased by a magnitude
which is proportional to the given potential energy E,
mean
and, on the contrary, is inversely proportional to the
kinetic energy of the gaseous molecules, and thus inversely
proportional to the magnitude
which
is
27).
of encounters
According to a formula of 70, the number
in unit time without reference to the molecular attraction is
168
this
increased in
is
where
Jc
T=
71
may
(1 4-
But now
^mG
4if
E is
then
would indicate
of encounters
that,
is
Jc
1,
since
we can
temperature @, while
we
Jc,
in
which the
it,
collision-frequency is
may
also be
ZJcElmG*:!,
<7/@
1,
Hence the
collision-frequency appears as a
which depends
in a twofold
magnitude
71
factor diminishes as
The
value
169
increases.
path
is
in
which
of melting
It
the factor
from absolute
general.
72.
zero,
(Compare
On
and consequently in a
slighter ratio in
87.)
mentioned in
and therefore
65, that
larger than
L very
larger than X.
a
molecule
passes by
Although, therefore,
?,
much
many
mole-
170
cules like
it,
many
perhaps
72
the phenomena
plaining the apparent contradiction between
at the beginning of Part II. ( 61) and the theory
mentioned
developed in Part
I.
depends.
paths
may
We
conduction in gases.
171
CHAPTEE
VII
VISCOSITY OF G-ASES
*
73.
BY
On
the
name
Newton
of internal friction *
media which is also
termed
property of fluid
denoted a
This
viscosity.
lias
motion and
is
must
speeds.
It is not difficult to see the reason
to
overcome
which
why
force
must be
is
upon
a part of
1
its
172
7S
and
applies
fluid
giving
it
friction
The
up
to
with
its
friction
consists
ternal,
support.
of fluids, therefore, both internal
only in a transfer of motion;
and ex-
but this
a part of the
not
becomes complete
at once
74.
must
of Internal
Friction
The
force
layer to another
when
of
VISCOSITY OF GASES
74
173
which
this
we must
which
its
the motion.
The way in which the magnitude of this pressure depends on the velocity that is present cannot be completely
determined by theoretical considerations. There is nothing
else, therefore, for us but to make up for our defective knowledge by a hypothesis, and afterwards to make trial of its
soundness by comparing the conclusions drawn from it with
that
from a doubling
of the
Such a
greater friction and a thinner fluid a less friction.
dependence on the matter may be expressed by a numerical
factor by which the mathematical expression of the frictional
This numerical factor which gives
pressure is multiplieda numerical measure of the viscosity or the friction of the
fluid is called the coefficient of friction or of viscosity.
174
74
The
and
let its
of
such
all
and
that,
it
to
time, so that the speed in each layer is numerically equal
the height, then the friction on unit area which is exerted
between any two neighbouring layers of this fluid is equal
to its coefficient of viscosity.
The measurement
medium
a constant magnitude
true.
if
Newton's assumption
that
is really
proportional to the relative velocity
Newton's theory will therefore be proved to be
the friction
true
if
if
is
different
made by
different
friction, especially
constant.
of measures has really been found,
for
and
both for liquids
gases, as will be described later
In
in fuller detail.
gases a perfect agreement with experifound
for those properties which, on the
ment has also been
kinetic theory of gases, are connected with the coefficient
Such an agreement
of viscosity.
75.
Formula fox
tlie Coefficient
To
of Viscosity of
Gt-as
VISCOSITY OF GASES
75
175
general terms.
according to that analysis, the internal friction cona transfer of the motion of flow from one layer to
another, and in a transformation of the motion so transferred
If,
sists in
from
internal
that
into
is,
heat.
The phenomenon
by the
of
internal
kinetic theory in
an
unforced manner.
Our
on general
We
To form
are identical.
directly the
measure
of the viscosity
Maxwell, PHI. Mag. 1860 [4] xix, p. 31. 0. E. Meyer, Pogg. Ann.
Maxwell, Phil Mag. 1868 [4] xxxv. p. 209. V. von
Lang, Wiener Sitmngsber. 1871, Ixiv* Abth. 2, p 485; Pogg. Ann. 1871,
1
cxlv. p.
176
75
same
surface, and so that the Telocity is the
parallel to the
the
take
further
we
horizontal
plane;
at all points of a
to be
base
the
above
x
distance
a
at
velocity of flow
numerically equal to x, that is,
a
that each particle passes in unit time over
In
base.
the
above
its
height
distance which is equal to
unit area between
this distribution of flow the friction per
or
we assume
of viscosity.
equal to the coefficient
is the heat-motion
v
there
Besides this forward velocity
this
with
exceedingly rapid
In comparison
of the gas.
a mean speed
at
has
motion, which in air, for instance,
forward
assumed
the
velocity v,
of 447 metres per second,
metre
1
is
metre
1
per second,
only
which at the height x
is
We calculate the
friction
if
lie
We
$NG;
come
that only a third of the
particles in unit volume
as
a
in
and,
into account as regards passage
given direction,
the
value
of
mean
the
speed
before, we take G to represent
as deduced from the
The
particles
mean
forming
kinetic energy
this
27).
their path
average they
and therefore from a layer
equal to the mean free path Z/,
above the base. Their mean
which is at the height x
forward velocity
is
therefore given
t?'
=x
by
L,
VISCOSITY OF GASES
75
177
momentum which
and the
amounts
to
mv
on the whole, then, the
= m(x
L)
momentum
%NmG(x
- L)
is
layers
NmG(x
+).
G(x
- L) =
and
is
this
therefore"
magnitude
is
viscosity
may
To
are
MIr
T~*^
178
free
path
75
measured only
it requires a
complete specification
be laid down. It is therefore well not to
values of the friction in the way usual
express the numerical
a length, but
time unit also to
by
'for its
of the formula,
for values of pressure, but, after deduction
the
of
density p we bring
by
which on replacement
Nm
into the
to refer
form
them
and time.
We
mental
units.
76. Theoretical
Laws
of Gaseotis Friction
The
final
free paths
calculation of viscosity,
we put
here
L=
with Clausius, and find for the viscosity-coefficient the
value
17
In
this
formula
is
=
the
number
of molecules
per unit
3
volume, and X the volume occupied by a single molecule
= 1, as we have already ( 63) several
consequently N\*
;
[4] xix. p.
31
i.
p. 390.
VISCOSITY OF GASES
76
179
the
simpler form
According
the gas increases in rarity. This seems to lead to a
conclusion which, although apparently admissible by the
when
last formula,
contains a contradiction in
itself,
viz.
that a
We
formula
77
it
is
when we
consider
how
a transformation of the
justify
whose meaning
The
friction
is
is
easily
perceived, offers
the quantity of
itself
suitably.
momentum which
is
Compare further
81.
N 2
180
76
place,
which
are
moving
in
all
directions,
law
of
Maxwell
The
will
have
lost
much
of its improbability.
however, independent
only
on the
square root
of the absolute temperature, or that it increases with the
temperature 3 measured on the usual scale in the ratio
*/(!
+a3)
1,
where a denotes the thermal coefficient of expanAs to the free path L, the theory leaves it unit alters with the
temperature or not. The
VISCOSITY OF GASES
76
181
whose
77.
viscosity
is
Observations on
Maxwell
in
Maxwell and
mination
of
the
viscosity
of
invented by
that
liquids,
Coulomb presented
because with it the same pressure is
If a circular disc is
throughout the gas investigated.
to its centre, it
fastened
a
wire
suspended horizontally by
be
of
the
wire,
put into oscillamay, by means of the torsion
the disc is in
If
centre.
that
tion in its own plane about
1
fluid,
of a geometrical progression.
If
we measure
the amount of
decrease,
VInst. National,
an IX,
iiL p. 246.
182
77
air,
altered
it
in
my own
together, like a
single disc, so as to
faces only.
The combination
move the
all
air
close
by two
two ex-
method
each other as Maxwell's theory required, but the differences between the values found were small enough to prove
the law at least within certain limits of the pressure. As
an instance I cite the numbers 2 which I obtained by experi-
temperature
Centigrade
W. Kdnig
(ibid.
VISCOSITY OF GASES
$77
183
Graham
The theory
of these
experiments, which
Ha
are
2
by Poiseuille and
and other liquids in
pendence
place,
everywhere.
5
memoir,
which corresponds
'confirmation
Here
by
Graham
in his
calculation of the
tables
353.
Pogg. Ann. 1866, cxxvii. pp. 253,
Ibid. 1866, exxvii. pp. 279, 365. In this memoir I have given a slightly
has no influence on the
different meaning to these numbers ; this, however,
s
184
77
a given
volume of
in inches.
length of the transpiration tube
Other
series of
results given.
appeared,
air
plates ; they hang therefore together for a long time by reason of the pressure
exerted by the air on their external surfaces.
2
PML
Trans. 1866,
clvi. p.
249.
VISCOSITY OF GASES
77
185
differs
common
axis,
as well as above
fixed discs at
same and
as
little
The
as possible.
first effect of
this
is
to
is,
these reasons
value which is calculated on the assumption of the constancy of the coefficient of viscosity
:
Maxwell
hydrogen
and
further
experimented
carbonic
my memoir
See
on
this
on
the
friction
method
clvi. p.
of
acid.
186
77
G- rah
transpiration experiments of
of
of friction
still
We
optical experiments
first
and Warburg, who repeated Maxwell's experiments and pursued the phenomenon down to extremely
low values of the pressure. Their experiments confirmed
Maxwell's result that the law is valid till the pressure
but from that point onward they
falls to ^V atmosphere
in the value of the logarithmic
diminution
noted a marked
But they explained this diminution not as a
decrement.
2
Kundt
That
We shall return to
of viscosity really
burg
83).
assume,
down
to
extremely small
values
of
the
pressure,
instead of a horizontally oscillating disc. From the oscillations of this leaf he was able, by help of a theory
developed
by Stokes,
1
2
3
to
VISCOSITY OF GASES
77
187
down to
pressures
which are
occurs
a sudden
of
drop in the value of the coefficient
viscosity.
pressures,
Maxwell's law
to
lose
its
of
validity
at moderate
i.e.
Warburg
as
355,
Ann. 1882,
Ber. d. naturf. Oes,
Freiburg *. B. 1882, viii. ; Wied.
the following.
390 Berl. Sitzungsber. 1882, p. 509. Compare 89 and
Graham, Phil. Trans. 1846 and 1849 Pogg. Ann. 1865, cxxvii. pp. 271,
Bull, of
and fol. P. Hoffmann, Wied. Ann. 1884, acri. p. 470. Barus,
xvii. p.
188
77
78.
Numerical "Values of
This confirmation of a
predicted
ourselves
by
that
it
we
fact
theory.
are not
moving
constitutes
We may
in
assure
the doubtful
The formula
established in
viscosity of a gas,
r,
*=
&LG,
now
when
Before
we
follow
Maxwell
we
must point out that the formula for the viscosity is capable
of a slight improvement.
It was deduced in
75 by a confhe
Washington 1889.
p. 195.
De Heen,
VISCOSITY OF GASES
78
which
and the
sideration in
of the speed
cules instead of
189
free path,
molecule.
gases,
slight error.
It is
more exact
9?
to put
= 0-30967
H
L =X
is
the
68).
mean
/7rA/2
bility, to
mean
calculate the
molecules.
coefficient
free
ri
0-000168 to
<yj
0-000174,
von Obermayer, 3
two
Fourth and
cxlviii.
2
0-000179;
by
transpiration experiments
fifth
by
different methods,
T?
also
Abth.
2,
190
Otto Schumann, by
1
well's method,
^,,,0
= 0-000168
T?
78
Max-
Tomlinson, 3 by
with cylinders,
oscillation experiments
77
The mean
0-000171
value of
all
= 0-000172.
these determinations
77
is
= 0*000172.
we
a = 44,700,
C. under
of the density of air at
of 76 cm. as given by Eegnault,
= 0-0012932,
L=
From
this
mean value
of the
parison of
it
how
correct
4
was the expectation which 01 ausius expressed
tude which
1
2
3
is
[3] xiv. p.
4
197.
Compare
62 and 72.
VISCOSITY OF GASES
78
191
l
scopically visible magnitude ; and the vibrations of the air
molecules as they course backwards and forwards take place
for the
2
highest tone audible.
79.
Just as for atmospheric air, the values of the two magnithe free path and the collision-frequency, may be
tudes,
also sufficient
gas,
which G-raham
if
tubes
in the
same
G-raham
atmospheric
air, viz.
= 0*000172,
determined by
and the transpiration-coefficient
the
coefficient of
at
calculate
once
G-raham, viz. 0-899, to
for
air
= 0-000191,
We
then
the other gases experimented on by Graham.
and
collisionfree
the
of
values
the
for
as
path
air,
find, just
these gases.
frequency of the molecules of all
The results of this calculation are given in the following
He
Helmholtz
Depretz, Comptes
for audibility.
3
192
table.
The numbers
are referred
units, as
too,
The
temperature
Karl Seubert.
to
79
of
Zdtscbr.
p.
433.
VISCOSITY OF GASES
79
193
The
values also
from
temperature.
vapour of mercury.
for
argon, 0-000208,
helium, 0-000165
as the value of the coefficient of viscosity at
tore for which that of air is given.
0, the tempera
we must
Z&itsckr. f. phys.
Chemie 1896,
t
p. 364.
xix.
194
79
= 0-0017810,
at
is
sec.
density
is
p
for
= 0-000191,
From
these
at
is
free
paths,
0-0000099 cm.,
for argon,
for helium, 0-0000240
at
finally,
the collision-
frequency of a molecule
for argon,
for helium,
3840 x 10 6 per
4840
sec.,
The
viscosity of
2
= 0-000642,
28 the value of
while according to the numbers given in
be taken as
must
the molecular speed at this temperature
li
L=
*
Zeitschr, f.
0-30967 p
p, 614.
^1
lix. p.
325
VISCOSITY OF G-ASES
79
19-5
L = 0-0000209
and that
cm.,
of the collision-frequency
stances
= 1245
An
x 10 6 per
great uncertainty,
since the
law
sec.
of
If
we
Koch
we
vi
Ann.
d.
Ghem.
u.
vii.
p. 354.
o 2
196
79
coefficient
the
and a much larger collision-frequency, although it has
and
C0
molecules
the
2
same molecular weight. Further,
NA composed
now we
VISCOSITY OF GASES
79
The
periodicity
is
197
in the
first
groups as the
molecular weight
rises,
the
coefficient
and then
of
fall after
Whether
hold good
these relations
w ith monatomic
T
1-67
54).
numbers
We should
calculated for
coefficient of
viscosity
and
We
mercury vapour
viscosity
not been determined with sufficient certainty. On the other
and
hand, doubts have been often expressed whether argon
and
not, perhaps,
helium are really simple substances,
has really only
mixtures, and whether the molecule of argon
one atom, and not perhaps two or three. In the first two
'
363 ; Journ.
1
Proc. Boy. Soc. 1895 ; Zeitschr. f. phys. Chemie, 1895, xvi. p.
CJiem. Soc. 1895, Ixvii. p. 684 ; Wied. Beibl xix. p. 674.
2
show more than is shown
[The values for three gases only cannot
above.
TB.]
198
tables,
in the molecule.
on the
explanation of the relations described is,
for
formula
theoretical
the
from
to
not
difficult
whole,
give
the free path,
The
L = X3 /7rA/2,
according to which
it
we may
it
increases-
VISCOSITY OF GASES
79
199
still
The
The formula
is
with increase
According to the formula a greater value of the collisionfrequency may result from two different causes which are,
On the one hand this
to a certain extent, antagonistic.
number is increased by an increase in the speed of the
it rises in value with the extent of the
particles, on the other
and
therefore
with the mass of the moleof
action,
sphere
collide the oftenest
molecules
those
cule.
Consequently
the
with
move
which, firstly,
greatest rapidity, that is, the
are in a certain measure
which
and
those, secondly,
lightest
;
200
79
increases.
= 0*30967
y[ T
0-30967 p{tf
p=
the
pressure by means
we can bring
^7r/>O
7?/r
If therefore
T is
collision-interval T,
of
the
and by
formula
form
is
of viscosity.
This law is important in the examination of the behaviour
of vapours, which we have still to investigate ( 86).
But we
must first discuss some other relations which there come
efficients
into consideration,
mixed
gas,
VISCOSITY OF GASES
80
201
The
for
mixed
the formula which expresses the theoretical value of the coIn this reasoning mean values were
dealt
the
with
mean momentum which is carried
always
the
by
moving particles in unit time over unit area in
efficient of viscosity.
;
mean
necessary
mean
We can
values.
"We
may
either
without neglecting anything, or strive to find simpler forwhich are easy to evaluate for numerical calculations.
Maxwell has followed the first way; here we employ an
easier method, which is due to Puluj. 2
The formula for the coefficient of viscosity
mulae,
i)
we
0-30967
where
stand m,
1
is
li,
0-30967 wfl/wA/2,
and
$>
as representing
ii, p. 72.
[4] xxxv. p. 212 ; Scientific Papers,
Wiener Sitswngsler. 1879, Ixxix. Abth. 2, pp. 97, 745 ; Carl's Repert. d.
Experimentallys.
xv. p. 590.
202
80
and
mixture of two gases whose molecular weights are
them
of
and
JV
molecules
contain
respectively in
2
2
unit volume, the mean value of the molecular weight is
N=N
where
whence
for
9?
we may put
and
s2 for
action
is
^=
of
the sphere of
is
to
v*
= 0-30967 m il
0-30967
WjOJw^Va,
/9r^V2,
whence we obtain
we have
for the
mean
section
N
If
we
mean
VISCOSITY OF GASES
80
20S
denominator of the formula for */, we obtain for the cothe mixed gases the expression
efficient of viscosity of
N N = Pi
:
:p
and
N N=
2
_p 2
form
P
(P
slight
amount
*
of hydrogen.
WM. Ann.
PhiL
204
80
the formula,
explain this remarkable phenomenon by
fraction
the
of
may be put
note that the numerator
To
we
,J
p}
w =
2
2 andfl^
= 43*67
becomes
JPV?/
iJP
.P
_ 2 h _ fay f^)*l& v y
3
%
W/
= 1 - 0-52^
p
if
For
of
it
coefficient of
mixture with
some atmospheric
air.
But
if
97
and from
this
follows
it
values of
_p a ,
as
that
amount
of
97
with p^ and
even for small
2
have actually
increases
added
air,
observed. 3
1
[The numerator
of
the fraction
is greater in this
nator so long as the ratio $Jp does not exceed 0-615. TB.]
2
Ber. tib&r d. Verh. d. naturf. Ges. in Freibwg i. B. 1882,
Wied. Arm. 1882, xvii. p. 422.
3
[This behaviour
is
117
viii.
p.
VISCOSITY OF GASES
80
205
Now
0*21 and
oxygen rj r Putting p /p
mixture
of
ratio
the
known
with
accordance
p^/p
31*76 for
of oxygen and nitrogen in air, and also
27-88 for nitrogen, and lastly, Graham's
oxygen and m^
viscosity
to that of
77
== 0-79, in
m =
l
value
i? 2
/^i
nitrogen,
atmospheric
which
is
we
for
the
obtain for
the
0*873
of
transpiration-coefficient
of
air
= 0-898,
in complete
obtained from
81.
transpiration-coefficient
agreement with
Graham's
the
mean
value
A gas
Warburg.
The
= 2-664, which
numerator and denominator are equal only when pjp
does not correspond to any possible combination.~TB.]
1
525. Previously published in abstract in
Fogg. Ann. 1875, civ. pp. S37,
Monatsber. d. Berl. Akad. 1875, p. 160.
for the
206
more the
81
decrease
just
of
diminished as well by
of the external friction.
A measure for the amount of the external friction we
might obtain in the same way as that for internal friction,
viz.
by means
But
slip
viz.
coefficient
of
have
defined
first
slip,
as
it
which we denote by
is
77,
the
r=w*,
appears a really suitable measure of the slip
increased by increase of ^ and diminution of s.
so that
is
which
of the coefficient of
slip is vanish-
The
ingly small, so that f might be put equal to 0.
external friction, therefore, in the circumstances under
which those experiments were carried out, is
infinitely
might be
left
of
the
observations.
Warburg
undertook to test
xl. p. 607.
VISCOSITY OF GASES
81
207
Maxwell's law
We
82.
Kundt
and
Warburg
inversely
paths.
The
external friction
fixed wall
forward velocity by
its collision
the
208
82
sum
But
to
form
this
sum we have
to
make
of the loss
as to the
a hypothetical
which a
single
magnitude
assumption
on impact at the wall. In the case carried
particle suffers
out in 11 we supposed that no kinetic energy was lost on
the collision, but that the particle was reflected from the
wall with the same speed as that with which it struck it.
could not now consistently suppose that the particles of
assert that of the
gas lose any of their speed ; but we might
forward motion, which all the molecules in common possess,
We
How
we have
to
form a
definite
conception.
into them.
and
also,
platinum
Prom
when helped by
foil.
such observations
we
most
upon the
excellently polished, as
Or adsorbing them, according to the modern nomenclature. Compare the
observations of Bun sen and Kayser, Wied. Ann. 1883-5.
2
H. Sainte Claire-Deville and Troost, Comptes Rendm, 1863, Ivii.
965
p.
Pogg. Ann. 1864, cxxiL p. 381.
3
Helmholtz, 'Bericht fiber Versuche des Herrn Dr, Elihu Root,'
if
VISCOSITY OF GASES
82
209
them
gases.
amount
to
double the
particles
%NG,
volume and
is
mean
number
of
impinging particles
is
and for accuracy we must replace it by the number JJVX1 calculated in 37 and 41* of the Mathematical Appendices, which
PHENOMENA DEPENDENT
210
ON.
MOLECULAR PATHS
82
of a molecule as
%fimNlv
= sv
gives the friction of the gas per unit area per unit time on a
solid body.
The
coefficient
viz.
= %8mNfl,
For the
coefficient
of slip
f=*7/
which
fi
= mN.
we have
= 0-30967 i/i/3,
= 1*23868 LJI3,
the molecules
and
it
83*
foregoing
given
by them, and
clusions.
From
VISCOSITY"
83
OF GASES
is
aD/=2
where
a,
whose
relation to
given by
x 760,
numbers given on
In
In the
last
=.
From
this
we
/3,
which was
intro-
where the
=
By
means
1*23868
of this assumption,
4,
1 J.
theoretical requirement
confirm this behaviour, but gave smaller values for the coThe
efficient of slip f, and therefore larger values for /3.
numbers
1
lately
clix. p.
399.
Ixvii. p.
826.
p 2
212
Phenomena
The molecular
free
84
volume
so too,
But we cannot,
in excessively
therefore, believe that the molecular paths
a
of
mercury air-pump,
rarefied gases, as in the vacuum
If we assume, for instance,
attain a considerable length.
were to rarefy the air 100,000 times, or to
that such a
pump
path
450 metres per second. The number of encounters to which
a molecule is exposed remains still very great even in such
a condition of rarefaction it would amount to 46,500 per
;
second.
is
nomi-
remove
paths
longer
be the ranges within which all inequalities will be removed.
This remark remains of force, even when the inequality
This is the
consists in the distribution of electrical tension.
reason
why
YISCOSITY OF GASES
84
213
rebound
is,
it
leaves the
warmer
side.
Consequently,
the mill must so turn that the white side of the vanes
precedes.
If
we think
of the
magnitude
[2]
xxix. p. 57
p. 667.
2
elxiii p. 277.
cli.
p. 96.
zii. p.
337.
214
84
1
law concerning the emisair; according to Kirchhoff s
radiant
heat, they have therefore
sion and absorption of
much
warmer
the
I have
which a warmed
heat
amount
to the
of
re-
pressure.
= 0-0017
units;
degree
millimetres,
we have
= 0-00017,
and
fofmd
value
which,
reduced
to
Kundt
and
degree
C.,
4
is
have
Warburg
JfcjL 0-00015; and, lastly,
The perfect
0*00014 in the same units.
Obtained h
for their correctvouches
numbers
these
between
agreement
ness.
2nd
2
3
4
cix. p.
275
[4]
xxx. p. 314.
VISCOSITY OF G-ASES
84
215
degree.
According to Stewart and Tait's
observations, the emissivity of a surface of bare aluminium
is four times smaller, so that, in accordance with Kirchhoff's law, it seems just to assume that a vane made of
warmer by 1
be capable
if
mill there
to be looked
explain.
We
PMl
216
85
gases
and
it is
temperature
not unreasonable to expect the same also
free path.
the earlier of the
As a fact,
made on the increase
temperature seemed
coefficient of expansion,
drew from
same ratio
and thus proportionally to
Maxwell 1
all gases.
a&,
where
and a
3-
is
is
the
0-30967
plrfl,
dependence of the
coefficient of viscosity
on the temperature
(1
1
oS)*,
ii.
p. 1.
203.
Puluj, Wiener
Stesswngsb&r. Abth. 2, 1874, Ixix. p. 287 ; Ixx. p. 243 ; 1876, Ixxiii. p. 589.
Obermayer, ibid. Abth. 2, 1875, Ixxi. p. 281; 1876, Ixxiii. p. 433 ;
Warburg,
von
GarVs
clix.
VISCOSITY OF GASES
85
217
and
air
this
(1 -f oS)*.
According to the observations of C. Barns , who investigated the flow through capillary tubes within a very
wide range of temperature, n = f both for air and for
to
hydrogen; with this value the formula holds from
inwould
the
free
path
1,300. For both gases, therefore,
1
crease proportionally to
(1
<*)*.
n cannot hold
in
general.
Puluj
0*92 for
obtained by the oscillation method the value n
3
with
observed
and
von
carbonic acid,
capilObermayer
C0 2 ethylene and ethyl chloride,
2
lary tubes values for
N0
by the usual
series of
powers.
0.
of the
Schumann 6
form
p. 465.
*
1899,
5
iii.
Nat. 1876,
Ivl. p.
273.
p. 81
6
et
IxTii. p. 816.
xxi. p. 199.
Ueber die Beibung
1885, xxi. p. 1
PMLMag.
[5]
tionsschrift
von Gasen
u.
Dampf en
xxiii. p. 35S.
u.s.w.'
TttUng&r HaUUta-
318
the
ture, while the quadratic factor is to express
change
85
of
power
L =X
we may assume
the
that
:j
/VsV'2
free
Schumann made
it
range of temperature.
reasons for the assumption that the section of the
The
The
forces
U. 8. Geological
Sum. No.
54, p.
278
VISCOSITY OF GASES
85
The
forces of cohesion
219
an
of
increase
which the
of the gas
when
particles
;
it is
the temperature rises. The consequence of this beis an increase of the molecular free path with the
haviour
temperature.
We
is right,
1
phenomena Sutherland had
According
by the
is
connected with
value
its
L^
relation
coefficient of expansion.
and also the value
If
we
a=
we
= 0-30967
[5]
xxxvi
p. 507.
85
this
formula
Atmospheric Air.
With
G he was able to
Bar us. 2
represent right
Atmospheric Air.
if
*
s
277
is
put.
VISCOSITY OF GASES
221
Carbonic Acid.
calculation,
We
find
hesion which
mayer's
observations.
Nitrogen
O
CO
Oxygen
Carbon monoxide
Nitrous oxide
C 2 H4
Ethylene
86. Viscosity of
84
=127
100
=260
=272
Vapours
To
viscosity of
this
xiii.
2, p.
433
222
86
method
when
from
sufficiently far
its
The
point of condensation.
normal
measured under
such circumstances stand in a simple relation to the chemical
nature of the substances.
have equal, or at
least
it
Alcohols
OnH* +1
Chlorides
Cja^Ol
Esters
C n H2
Bromides
CUEL^+jBr
Iodides.
CJI^I
exceeds 1
77
= 0*000142
150
155
182
210
Lothar Meyer
as
may remark,
Graham
xw.
A
Wwmr
p. 573.
497
1881,
xiii.
p. 1
flp^ Annt
;
VISCOSITY OF GAB T
86
that
223
we must not
Graham's
If,
however,
regular.
were to be affected
Lothar Meyer's
circular nor
observations also
without
effect
number
bound together
in
as this
many
87.
The
first of
the
on
of
mode
in
which
the
viscosity
vapours depends
temperature.
Vapours exhibit a much more marked variation with temperature than gases. Hence Sutherland's formula, which
is in excellent agreement with the behaviour of
gases, is
We
viscosity
Capillary tube
JBT,
the temperature.
PHI. Tram,
224
87
more
larger
easily the vapour
state.
and
is
when G is so large
with it. The length
+*
= 0*30967 pQL,
O that is variable
formula, increase
Sutherland's
in
proportion to
with the
temperature proportionallv to
(1
in
which the
coefficient of
1
a$)
expansion a
is
taken equal to
VISCOSITY OF GASES
g7
225
Sutherland's
We
theory.
with similar difficulties in regard to the ob-
meet
made by O.
servations
Schumann
on the
viscosity of
for the func-
benzol
is
(1
0-00185 S)V(1
0-004
5-),
which each term increases more rapidly with the temcan explain. For many of the
perature than the theory
alteration with temperature that
esters, certainly, the law of
was found lies between the theoretical limits; but the
in
function
(1
Schumann
which
of all his
0-00164 3) V(l
0-004 S),
observations on
mean
Sutherland's
theory.
Let us now examine whether Schumann's observations satisfy this limiting value, and for this let us express
the coefficient of viscosity by
where ^
is its
value at
equal to 0*00367
or
C.,
and a
more simply
taken for
us put
is
let
all
vapours
Schumann
by
mean
and
them.
1
Compare
0.
xxiii. p.
92.
353.
2i>6
From
the
mean
values of these
numbers found
for
87
if
the formula
benzol,
by
which
Schumann
formula.
10*71
VISCOSITY OF GASES
87
The
calculated
numbers
10V
Meyer's
Schumann's
calculation
observation
fit
We
pretty
exactly to reality even in the limiting case now considered,
in which the cohesive forces of the vapour are taken ex-
We
measurement
a 2
of
228
87
the viscosity and partly in the estimation of the temperaBut against this view there is the circumstance that
ture.
first
a fluctuation is shown in the values of the constant
calculated
esters in
almost exactly
similar fashion.
We
are also forced to the opinion that these fluctuaby the consideration that
by
temperature.
88. Dissociation
The case
all of the same size and mass as in a perfect gas.
often occurs, in consequence of the forces of cohesion, in
which two molecules which chance to come near enough to
each other join together to form one molecule of double
and just as frequently the case arises that such a
size
double molecule splits up again into its constituents, either
as a result of collision with another molecule, or simply in
;
The
ratio in
increase of speed, as brought about by a rise of temperature, increases the probability of the division of the larger
The number of simple molecules therefore inmasses.
VISCOSITY OF GASES
88
229
when
the temperature rises, and that of the compound molecules when the temperature falls, until finally,
at a definite temperature, all are broken up and the perAt lower temperatures, on
fectly gaseous state Is attained.
creases
number
of the
compound molecules
is
in
excess.
This behaviour of the molecules explains, as before remarked, the fact that the density of vapours, when compared
with the density of air or of any other gas under the same
pressure but at different temperatures, does not remain constant.
Since, in accordance with Avogadro's law, the
number of molecules in unit volume is the same for all
gases at the same pressure and temperature, we find a
smaller value of the vapour-density if the molecules are
broken up into smaller ones at higher temperatures ; at
lower temperatures, on the contrary, when all the molecules
are joined together, either in pairs or larger aggregates, the
much larger.
of pressure has just the same kind of effect
If the pressure
the degree of dissociation attained.
A change
upon
is
of combining.
views are right, the law of alteration of the
density of a vapour with its pressure and temperature must
be determinable from theoretical considerations that go more
deeply into the nature of the phenomena. The first who
succeeded in finding this law was J. "W. Gribbs.
In
addition to the general laws of gases he employed the two
principal laws of Thermodynamics, viz. the theorem of
energy and the theorem of entropy ; from these he deduced
If these
New Haven
1, p.
234.
Van
der
Under the
'
230
and
similarly,
later,
88
in
starts
discussed
atoms as vortex rings ( 124), J. J. Thomson
intervals
mean
durmg
the
probable
it by an investigation of
combined.
or
remain
separate
which a pair of molecules
of
theoretical
He investigated the
kinetic theory of gases.
ceptions of the
of an encounter,
results
of the different possible
probability
an
on
average, two particles
and thus determined how often,
on an average two
often
on
together, and how
meeting join
combined
5
and also G. Jager, who considerably simplified the
formula by introducing approximate values.
of these
I must omit from this book a reproduction
we have only to investigate whether
theories and formulae
the dissociation has any effect on the value of the coefficient
of viscosity, and, if so, of what kind this effect will be.
larly,
by Dissociation
We
of viscosity
1
*
a
Versl.
v. Wet.
Amst. 1880
[2] xv. p.
when
199.
Wwd. Ann.
1884, xxii*
p*39.
4
TTied.
Ann, 1889,
Wim.
mann*s
xxxviii. p. 288.
c.
p. 1189
il pt. 2,
1895,
p 561.
cii.
p.
671
Winkel*
VISCOSITY OP GASES
89
231
the molecules are broken up. The free path will certainly
become no smaller, but in all probability larger, because the
section of the molecular sphere of action cannot be increased, but only diminished, by the separation of the moleand of the other two factors the density p decreases
cules
increases as the dissociation progresses, but
and the speed
;
their product
pO
2
magnitude pl remains
how
9?
will be altered.
We
however,
find,
surer
way
of
answering
this
w /X = 2,
2
if
We
(NJN
77
7ll
molecular weight
m=
we
obtain
(N1m
232
of the particles,
89
and
putting
in its actual
so that S represents the density of the vapour
the density
and
is
dissociated,
it
state in which
partly
broken
were
molecules
the
all
when
attain
which it would
iV2
and
N,
densities S
Hence
= N,
it
Since,
by the
partially dissociated
is
vapour
r?
of the
represented by
-8
the coefficient of viscosity of the vapour
when completely dissociated into simple molecules; 7? 2? on
the contrary, denotes the value of the coefficient of viscosity
in
which
tj l
is
no
vapours as exhibit
is
no dependence
no material
influence.
hitherto
Wwwr Sitmmgsb&r.
2, p.
279.
VISCOSITY OP GASES
89
those
in
most of
of
Schumann
233
esters
Schumann,
in
however,
For a
sure resolution
ably,
bility
may
help to good
results.
90.
In order
77
= B(A - S)~*
xxiil. p. 394.
B&r.
tib. d.
Wied. Ami.
8
Hence the viscosity
12, I. 3.
Warburg and von Babo, Results,
seems to be connected with the density much more simply than "with the
pressure.
4
FortBchr, d. Phys. 1896, 52. Jahrg. I. p. 442.
234
ness, in the
$, viz.
90
= 2%^)*,
4= .0(2-
/(I
-*), B=5^Z>*(l-*)- s
it
may
pass
What
unit
we
we wish
to
Since
denotes the density of the gas
molecules are decomposed into simple ones, we
might be inclined to assume that the gas then behaves as a
when
all its
VISCOSITY OF GASES
90
23&
op,
==
6pi,
where a and b are independent of the pressure, and therecan be variable with the temperature only.
The assumption that the gas obeys Boyle's law
fore
if we remember that
appears, however, as very improbable
the temperature in the experiments lay between 25 and
41, and was, therefore, on neither side very different from
the critical temperature 31, at which, under certain circumstances, the distinction between liquid and gaseous
carbonic acid ceases. Instead of Boyle's law, therefore, we
might with equal justice assume that the density of the
carbonic acid vapour varies but very little with the pressure,
That this assumption is for
just like that of any liquid.
the most part really satisfied under the circumstances of
Warburg
larly
the pressure.
I have therefore calculated the results of the observa-
and von
Babo
under
both assumptions, supposing in the first place that Boyle's
is
law is obeyed and in the second that the density
tions
communicated by
Warburg
invariable.
The
Table
xii.
clix. p.
575
of their complete
memoir
90
4=
1-208,
B = 0*000378,
is
very satisfactory.
viscosity,
which
are
= 0-000378
(1*203
- 8) -3
shows
When
numbers
I thus obtained
this value for all the series of experiments.
and J5 at the temperatures
the following values for
named
1
I have submitted to calculation only those results of observation in
which the amount of air mixed with the carbonic acid was as small as
possible.
VISCOSITY OF GASES
90
237
shown by the
following table
is
Carbonic Acid.
~L
then calculated
"the
i)
= &p*$i(<zp
8)-*,
=-6
35
40-3
= 0-01332
1374
1142
= 0-0001734
1859
1665
238
90
Carbonic Acid.
If
and
being certainly
VISCOSITY OF GASES
90
239
We
upon as constants
if
be put
the pressure
A=
ap,
is
sufficiently high,
but to
B=
91. "Viscosity
Gas
and of the
I have, on the other hand, sought for a more comprehensive proof of the theoretical formula by returning to the
theoretical meaning of the magnitudes A and B, the values
which I have obtained from the observations, and investigating the conclusions of another kind to which they lead.
of
According to
90,
A = D(2 -
240
91
where
is
is
named above
are not
direct
any experiment,
But from one of the laws of
theoretical chemistry
of a
gas from
we
its
culation of
justified
D to the
formulas in which
A=
op,
B=
we put
6p*.
&==
wherein
represents the
30.
VISCOSITY OF GASES
91
241
A =
we
= 0-0019653
1*5198/773-3
A
where a
If
= 0*OG19653/(1 +
we put
this value of
a$),
of expansion.
we
obtain
of the ratio
coefficients of
rjj^
0*634
637
621
8550
8236
35
40*3
If
we now employ
we
two
way
coefficients of
firstly
that of
for
C 2 H4
thus,
32-6
^ =
0-000142
%=
148
151
35
40-3
0-000224
232
244
C0 2
the temperature-factor
1 + aB,
C2
(1
the temperature-function
aS)
87).
^ =
0-000127
131
132
which exhibit a
fairly satisfactory
^=
0-000189
194
198
agreement.
242
91
the
comparison of these values with those obtained for
is obacid
under
carbonic
of
ordinary
pressures
viscosity
in
79,
vionsly the next thing. As was mentioned before,
Graham's
gave
experiments on carbonic acid at
0-000145; Puluj found the nearly equal value ??
??
0-000143, while von Oberniayer and Schumann agree
in finding the rather smaller value 17 = 0'000138
[and the
s=
mean
of
all
these
is
1?
Since this
is
0-000141],
^ =
mean
value
0-000130
now found
to
put for the density of the intensely comhas become independent of the pressure
which
pressed gas
I therefore
the value
77 x Q'QQ19653/(1
wherein $
is
sion Q"OQ367.
+&) =
016133/(1
o$),
the temperature and a the coefficient of expanOn putting this value in the formula
A =D(2~
1
p. 421.
VISCOSITY OP GASES
91
where
we
to
A =
243
1-203,
be the same in
all
2*(
% /^ =
0-874,
^/ %
0*6465,
so that the ratio in this case only slightly exceeds that resulting from the other calculation [viz. 0*631 in the mean].
We
ij l
from the
S =
B=
lOfy =
378
398
35
409
404
131*9
139*4
143*4
142-3
32-6
25*1
40*3
their
mean
value
T?!
0*000124.
the gas
value
*7 a
0-000192
of the
first
is
improvement that
is
244
91
law of dependence
on the pressure, which embraces both the
A second
liquid and the gaseous states of the substance.
lies in the consideration
the
of
theory
improving
possibility
that, in addition to single molecules C0 2 and double ones
in the gas, there may also be present triple molecules
2
4
named
It
of the density
C3
G,
The want
me
But an experiment
of
Lothar Meyer's, on
1
the vis-
He
cosity of the vapour of benzol, seems to contradict this.
allowed saturated benzol vapour to pass through a capillary
tube Into a space where the pressure was less the vapour was
here condensed by cooling, and the mass of vapour which
had traversed the capillary tube in a given time was determined by weighing the liquid.
Prom this weight the
coefficient of viscosity of the vapour was calculated by
;
Poiseuille's law.
Calculation gave the value of this co(which Is constant for gases) to be the smaller the
higher the back pressure at the exit of the capillary tube.
may therefore also say that the friction seemed to be
the larger the mean pressure was in the tube
smaller
the
is
the
this
exact opposite of the theoretical conclusion,
but
the
that
viscosity of a vapour is the more considerable the
the
nearer
vapour is to the saturated state.
Lothar Meyer has explained this apparent contradicefficient
We
vii. p.
531.
VISCOSITY OF GASES
92
245
high
sufficient exactness
back-pressure,
into the state
it is
a perfect
of
gas.
To make
precipitated
of the tube,
246
92
by Synesius Koch.
vapours mentioned in
87.
1
Compare
also
108.
CHAPTEE
VIII
DIFFUSION OF GASE8
93.
BY
diffusion
liquids
Observations
we understand
or gases
the
slow
mixing
of
two
Such
gas, dealt
248
93
problem before
Passin^
Berthollet,
the
over
3
us. 1
the
older
Graham/
experiments
carried
and
observations
of
Dal ton, 2
out by
Loschmidt 5 on
diffusion of gases.
the
mm.
diameter,
975 mm. long and of
glass tube,
flat
ends
glass plates and
by
which can bo closed at both
of
two
cut into
equal lengths,
parts
glass stopcocks, was
which could at will be shnt off from each other or put into
means of a slider made of a thin sheet of steel
26
connection by
analysed
1
We must not conclude that, because Thos. Graham so judiciously distinhe was the first to employ
guished these differing phenomena by different names,
these names. The term diffmion was used by Da It on (* On the Tendency of
Elastic Huids to Diffuse through Each Other/ Mem. Manck. Soc. 1805, new
Amer. Phil Soc. 1802, v. p. 15; 'Exseries, i. p 244) and Priestley (Trans.
to Various Branches of Natural PhiloObservations
and
relating
periments
M4m.
*
d'Areuett, 1809,
ii,
p. 463.
p.
74
Pogg.
Ann. 1829,
xvii.
p. $755
Experimentaluntersuehungen iiber die Diffusion von Gasen ohne porose
Seheidewande,' Wim. Sitsungsber. Abth. 2, 1870, IxL p. 367 ; hdi. p. 468 ; continned by Wretschko,
p. 575 ; and by Benigar, IxiL p. 687.
6
Wim&r Sitmmg&ber. Abth. 2, 1871, Ixiii p. 63 ; 1872, Ixv. p. 323.
7
PMl Mag. [4] I860, rat. p. 21 1868, xxxv, p. 199. Scientific Papers, L
*
ML
p.
302
ii p. 57.
*
Weitere Studien n.s,w.' Wim. Sitzung&ber* Abth. 2, 1872, Ixvi. p. 324
1878, teriii, p. 733; 1882, Ixxxvi. p. 63 ; 1883, Ixxxviii. p. 835.
*
DIFFUSION OF GASES
93
of the
idea
of
This
resistance
249
resistance
densities
speeds of diffusion.
heat.
This similarity is not only in respect of the mathematical form, but is founded on the nature of the matter.
Just as heat spreads in a conducting body, so in diffusion a
gas spreads from one region to another. The speed with
which heat is transmitted is determined for each substance
by a constant which is termed the conductivity ; in like
is
penetrates
into
determined by
magnitude which we might
but which is more usually termed the
a
or constant, of diffusion.
of these two constants
The meaning
is quite analogous.
obtain the strength of the flow of heat by multiplying
by the conductivity the difference of the temperatures at two
We
places, which are distant from each other by unit length along
the line of flow, i.e. the so-called rate of fall of temperature.
if
we
is
proportional to
it.
We
may
From
it is
easy
250
93
For
second gas into the first depends.
in the total pressure is brought about
by diffusion in an experiment wherein the initial pressures of
the two diffusing gases are equal to each other, the strengths
of the two flows which occur in opposite directions, but
otherwise under the same circumstances, must be equal,
diffusion
of
the
no inequality
since
this coefficient
is
mixing occurs
Loschmidt
it
itself
We
94.
together
which
is
The cause
One layer
is
A transference
of
to a greater distance takes place, therefore, only
the interaction of very many particles in their to-and-fro
momentum
by
motions ;
DIFFUSION OF GASES
94
251
The
cules.
filled
frequent
are driven hither and thither, so that
it
is
impossible for
meaning and
its
95. Theoretical
interval occupied by an
be in such a state of
the
may
diffusing gases
experiment
motion that the equilibrium of the total pressure exerted by
the mixture is nowhere disturbed ; there is then at every
point the same pressure p, and this is the sum of the partial
pressures jp x and
by itself, or
jp 2
sent
by
Pogg. Ann. 1858, cv. p.
Mag. 1858
39
AWtmdl<imgm, Abth.
2,
1867, p. 260
2/52
95
kind, just as those of the other kind are at the opposite end.
The pressure and density of one gas diminish along the
in the
= $p +
$x
N = 91 + nx,
-sp
jp 2
_ $P _
p/P
N =N
2
nx.
3^
it
decrease or
will result, in
has not yet been established along the whole tube, the
foregoing simple formulae can still be employed without
error, if we use them for only a very short portion of the
tube, and therefore, for instance, if by x we understand a
length which is shorter than a molecular free path, as in the
following calculation.
We
cross
described, or,
more
correctly,
number
of particles
cross
in
were
which in unit time meet
If the distribution
DIFFUSION OF GASES
95
253
for the
it,
82 be
would by
two
sorts of molecules,
O2 respectively.
In consequence of the unequal distribution there coine
fewer particles of the first kind from the side of the smaller
x and more in the opposite direction, while for the other
gas the reverse holds good. The particles of both groups
abscissa
larger part of the particles not only reach this section but
from the
layers which, on an average, are distant by x
x
origin of the tube ; the density in these layers is not deter-
mined by
but only by
91
n(x
- Ai) = Nt - wAj
From
are only
particles of the first kind which in unit time meet unit area
of the section ; there similarly cross this unit area in the
of the
same
distance x
side the
number
kind,
+A
sequently an excess of
f>54
ill
95
more than
which
cross the
same surface
in the
is
wA 2 Q 2
where
A2
i7z(A,l I
arises
which
at
once
and thus
- kJlJNilN
and
particles of the
The
same
From
number
in unit time
Gross
we
DIFFUSION OF GASES
05
is
number
which give
of particles
255
rise to
it,
numbers n and
p,
of the distribution.
equality
We
have
still
We
D = $*r(#&0 + N&OJIN,
2
we
PIIEXl-OIEXA
2r,G
95
of
number
We
kind.
culating
can
another kind.
particle
96.
traverse in a crowd of
This calculation
manner
as
it is
in
Kind
may
it
seems unnecessary
it
The
sions, is
probability of a collision, or the frequency of the colliproportional to the section of the sphere of action or
whose radius
is
DIFFUSION OF GASES
90
not be
fore
viscosity.
We
shall there-
considering
the
encounters
learnt
justified,
when
257
by
7TCT
2
.
1
According to a suggestion made by Stefan, which
seems to be confirmed by experiment, the magnitude <r
stands in a simple relation to the tw o magnitudes $ t and
being probably
<T=
^+
s?
this relation
2 ).
ascribe to
^s and ^
l
The
with
as the
explanation
we may
of force surrounded
but
it
admits of a dynamical
by spheres
if
we
The
free
path
of a
molecules.
us first for
molecules do
of space, but
those of the
let
not
make
Wwn,
Compare
2, p.
323.
238
9o
wanders past
during the time in which the molecule
into the condition of
more
comes
therefore
and
easily
it,
The probability is increased in the same
colliding with it.
measure as the relative velocity of the two moving particles
with respect to each other is greater than that of the
In the case supposed the magnitude
particle moving alone.
the absolute
relative
the
of
velocity is easily obtained
itself
If these
relative velocity of each with respect to the other.
and
1
as
hitherto,
absolute velocities are represented by f^
2
is
move only
coming
in
among
all
probability of a collision is
exactly the same measure as by that to-and-fro motion
which we have hitherto assumed.
have, therefore, in
this case too, nothing further to alter than to substitute the
We
medium
at rest.
Maxwell 1 seems
to
have
mean number
of the
Phil Mag.
[4] xix. p.
28
Scientific Papers,
p. 387.
DIFFUSION OF GASES
96
259
of action,
there are
whole path
f! 1
number
of collisions,
L=
for this
assumption gives
fi 1==
2,
<r
s,
N,
= JV = N = X2
3
.
97.
By
different kind.
s
260
97
and 2 particles
once from the
of
number of
whole
the
that
established
have
we
formulae
first
the
of
a
which
collisions
gas undergoes in unit
particle
If there are tf, particles of the first kind
the second in unit volume, we see at
time
is
given by the
r, =s
sum
^WAv^ +
while
= ir*Wf/2
4-
which a molecule
represents the whole number of collisions
of the second gas experiences in unit time in the mixture.
Hence for the mean free path of a particle of the first
kind
we
and similarly
and
are dependent on the numbers
2
and are therefore variable with the ratio of the amounts of
the two gaseous components in the mixture.
Both values
we
in the formula of
we
obtain the
two
free paths 81
and 8 2
95, viz.
which we
it
keep
value in
But
1
all
experiments
this
made with
expectation
is
p.
266; Phil.
DIFFUSION OF GASES
98
For by
substitution
we
^I
of diffusion
D=
261
for it
given before. This value is by no means constant
depends not only on the temperature by reason of fl and O 2
and on the pressure of the mixture by reason of A7 but
;
and
and thus with the ratio in
2
which the two gases have already mixed with each other.
The dependence of the coefficient of diffusion on the
temperature and pressure of the gaseous mixture can raise
no doubt. The coefficient, however, is to be looked on as a
constant in this respect, that it has the same value everywhere throughout the space filled by the gas in an experiment
on diffusion, provided that the temperature is kept constant
it
everywhere
the diffusion
if
if
the coefficient
and
total
too.
From
it
follows
sure.
This too
is
in perfect
agreement with
Loschmidt's
observations.
But our
theoretical formula exhibits still another variathe coefficient of diffusion, which we should not
a priori expect. The value of
is dependent also on the
values of J^ and N& and these numbers, which express how
many molecules of the one or other kind there are in unit
tion of
262
98
each
of
the two
coefficient is variable
of the experiment.
will not be able to exhibit
D
if
of mixture
made
N%.
theoretical calculation of an experiment on diffusion therefore becomes very complicated, and so excessively
The
Another Theory
99.
The theory
of diffusion
made
if
of Diffusion
somewhat
different
fundamental assumption
which the
two gases proceeds. There can be no doubt
but that this cause must be sought in the mutual encounters
of the particles, which
prevent a forward path in a straight
line.
But the question may be raised whether the two kinds
of molecules take part in these
processes in the same way
and in equal measure.
It does not seem
improbable to assert that the encounters
between molecules of the same kind have little influence on
the velocity of the current with which each of the two
gases
is
mixture
of the
loses velocity
flow.
DIFFUSION OF GASES
99
263
phenomena
would not
really represent the actual free paths of the molecules in the mixture of the two gases
but Z l would be the
;
mean
first
kind
values of Z\ and
&!
8' 2
deduced in
and 8 2
In this way we should obtain
.
for the
coefficient
of
all
Scientific Papers,
p. 199
[4]
Wien. Sitsungsber. 1872, !xv. Abth. 2, p. 323.
xxxv.
It
ii.
depends
p. 57.
99
and on the
only on the nature of the two diffusing gases,
are ; its value also
which
at
and
they
pressure
temperature
increases as the temperature rises, and it varies inversely as
the pressure, just as in the theory first developed and in the
Loschmidt
has made.
with
But
experiment does not prove the
agreement
It is still open to the
of
the
simplified theory.
accuracy
offered
the
resistance
by the other kind of gas
objection that
a
molecule molecules of
is not the only one experienced by
a
the same kind also obstruct
particle in its forward motion,
in
its
since they, too, can get
way just like those of the
same medium, and the
of
the
This
resistance
other kind.
observations that
this
free path,
consequently ensuing shortening of the molecular
cannot be left out of account.
The assumption that the current of diffusion of a gas
should be resisted only by molecules of the other kind which
are
to be found in
collide so
it
much
of different kind
their
own
why
particles of
vanishingly small.
th.e
Theories
of
ii.
p. 61.
DIFFUSION OF GASES
100
265
made
by Waitz
simultaneously, and somewhat
this
question,
first
and von
later
to decide
Obermayer
3
by Winkelrnann.
if
cir-
cumstances.
Waitz, who
He
for
0-151,
0-158,
C.
show an
too,
go
whether the observed variations of the
coefficient
conformed
formula
is
incorrect
for
an objection that
is
2
3
xxvii. p. 479.
266
100
of
to the case
which
conclusions
Winkelmann
his
Hausmaniger
Waitz's
calculate
of
coefficient
too,
attempt,
diffusion
has
fall.
taken
unfortunately
as
But
variable.
been
unable
to
this
obtain
agreement between theory and observation for the measurement of the variability of the coefficient, since it is
small, will be made too uncertain by the errors of observation
;
for
exactness.
observers shows that the uncertainty of the observations isFor the diffusion of carbonic
really as large as we say.
acid into air Loschmidt 2 has found the value of the
coefficient, referred
and 760
while
mm.
to
pressure,
C,
D = 0-142,
von Obermayer 3
gives
D = 0-135.
These two numbers show as great a difference as the
numbers observed by Waitz, which, according to von
Obermayer's calculation, assume the values
0-131,
0-133,
0-137,
when they
We
2
3
2, p.
1073.
DIFFUSION OF GASES
100
and
small.
when
267
for shortness
it
when
we put
smaller
the ratio of
first of
numbers
The
of the
itself
changes.
alteration of
magnitudes
D will be very
small
when the
and that
values
will occur
no
fact
known which
contradicts it;
we
D = rv/8AVvW
and on the contrary, when
AT2
P%s.
IV),
Guglielmo,
Bepert. d.
AW
D = V/ 8#
1
0,
+ n 2 2 ),
Ewner's
Iviil p. 599.
268
100
which are
2
or
lt
where
which we
find
on putting
N =N
1O1.
To
test
perature.
and
where
By means
we denote by
then
If
p = ^TrCZ^
takes the simple form
N=N
all
pressure
coefficient of diffusion
where
NJN, +
If
we put
sl
= =9
S9
crV(l
DIFFUSION OF GASES
101
into that
269
particles of the
We need not
D:=
7r
(^^
6icr%
)'
\~
V/fi
observations, and
all later
ex-
perimenters.
Loschmidt
dealing with experiments that are made at constant ternperature and pressure. If the magnitudes 7 were also
independent of the temperature, D would increase proportionally to the f power of the absolute temperature (since
this on the kinetic theory is proportional to IP), and therefore less than is really the case according to Loschmidt's
experiments.
But according to the observations on viscosity the radii
$ 2 of the spheres of action are dependent on the
and
$!
temperature, either actually or apparently ; this is also possible, and, indeed, highly probable, for the newly-introduced
magnitude <r, which has a similar meaning. If we may
assume that
<r
^70
of viscosity
17,
78, is given
which, according to
formula
<ri
101
by the
0*30967 wH/7r$V2,
expansion.
with the
85, increases
has con-
The value
two
of
pairs,
ments on
In
1
1.
fact
later,
by experiments
l\
Abth.
2, p.
367
Ixii,
Abth.
2, p. 468.
DIFFUSION OF GASES
101
on
071
dif-
Air
carbonic acid
Hydrogen
0-968
oxygen
0*755
1-050
Carbonic acid
Carbonic acid
Oxygen
nitrous oxide
....
hydrogen
nitrogen
0*742
0*792
and these values so far agree with those which the same
obtained for the dependence of the coefficient of
viscosity on the temperature, viz.
observer
....
....
....
Air
Hydrogen
Oxygen
Nitrogen
Nitrous oxide
Carbonic acid
Q-76
0'70
Q'80
Q-74
0'93
0*98
viscosity
diffusion,
We
may
magnitudes
further conclude
and
a-
xiii. p.
ISO.
Compare
85.
2, p.
magnimutual
272
the distance
denotes
cr
101
with
rise of
temperature
is
explanation, we
way, that the magnitude
shall
have
to
is
with Sutherland's
assume,
in
the
same
heighten
we
shall therefore
D=D
have to put
(l
a9)t (1
coefficient of expansion,
and
the
absolute temperature
+ a- =
1
9-
between dissimilar
particles,
272-5,
at
C.,
and O
is
does
for
similar
particles.
DIFFUSION OF GASES
101
17Z
of
Carbonic acid
Carbonic acid
Oxygen
Oxygen
Oxygen
hydrogen
nitrous oxide
hydrogen
nitrogen
carbon monoxide
106
-W)
100
136
12 4
for their
experiments
on transpiration
86) served also in these experiments.
Here also the agreement between the calculations from both
(
series of
experiments
is
satisfactory.
from
of the vis-
well
with Loschmidt's
1
experiments.
274
possibility of
at once evident
102
when we remember
is
L =
}
of each of the
are known.
mentioned in
Stefan used
96, viz.
= \( +
&
$l
%),
measured
two gases
DIFFUSION OF GASES
:102
Coefficients of Diffusion.
The agreement
of
is
of
for us to
see in
o-
lf
= ^(^
-v
a- $
2)
.
276
102
each
is
affected
with an
have contributed to
of themselves, defects in the theory may
has not been
diffusion
of
the
since
theory
the discrepancies,
mathematical
strict
such
accuracy as was
carried out with
calculation
accurate
more
a
That
the theory of viscosity.
more
me
the
to
seems
the agreement
prob-
would improve
able as
those of
my
figures
viscosity, to
path.
in so far as the processes of
preferred for this purpose,
friction are more easily and simply treated theoretically.
But when there is a question as to the magnitudes of s and
(r,
something about
cr.
means
Von Obermayer
of teaching us
has employed his
He
observations on diffusion for purposes of this kind.
of the spheres of actions
sections
the
for
obtained
thereby
2
numerical values which are about 1 J times greater than
by me from the viscosity ; but the reason
of this simply resides in his having employed a formula for
his calculation which leads to a constant value of the co-
those calculated
Experiments on Diffusion,
III.'
2,
p. 188.
3
Compare Chapter X.
CHAPTEE IX
CONDUCTION OF HEAT
103.
ONLY
very
Low
little
Conductivity of Gases
of
4
distinguished, as was long ago known to Priestley
5
and Ac hard, by so considerably greater a conductivity that
Magnus thought it might be compared with that of
is
and the
Berzelius
TraiU de
iii.
p. 418.
to
Leiden fr os t, De
Duisburgi 1756.
ac[iu&
communis wmwullis
Oguscula phys.-cJiem.
et
med*
gualitatibiLs tractates 9
iii.,
Duisbtirgi 1797.
51 r
278
10$
on the contrary,
of a
of balls.
If,
for instance,
we allow one
of these
and all the others remain at rest except the last, which
from the rest with the same speed with which the
In this experiment the energy alone
first struck the row.
of the motion has been propagated, and, indeed, with very
great speed, without the mass through which the transmission was effected that is, the row of intermediate balls
being carried on with the energy. We might likewise
conclude that the energy of the heat-motion may be very
size,
flies off
now and
CONDUCTION OF HEAT
103
But
in itself
move
in the direction
points to actuality.
same
279
direction,
and
all
wide stretch in
But
a short time.
energy
molecules.
104. Kinetic
Theory of Conduction
last
Stefan
have
doubts
3
and
before
men-
von Lang
The same
Maxwell's, according
to
of
gases
repel each other with forces that are inversely as the fifth
power of the distance. The theory given in the Mathe-
and
rests
Maxwell's
speeds.
p.
1
Pogg. Ann. 1862, cxv. p. 1 ; AbJiandl. il. Warnietliearie, 1867, Abth. 2,
277 ; ilechanisclie WtirmBtJworie, iii. p. 105.
2
PML Mag. 1860 [4] xx. p. 31 1868, xxxv. p. 214. Sclent. Papers, L
;
p.
403
3
1
li.
p. 74.
xlvli.
Abth.
exlv. p.
290
2, p. 81.
p. 529.
5
Ixvi.
Abth.
2, p.
330
280
1CH
mass
is
obvious.
effected
Diffusion consists
of the
by means
molecular
momentum
motions and viscosity in a transfer of forward
is likewise
heat
of
conduction
the same means
caused
by
a transfer
motion
of energy,
which
of the molecules.
the propagation
of heat
diffusion-phenomenon
particles
diffuse
in
is
effected
as before
by the
may
directly
among each
other.
to a
cules as they pass from an upper and warmer layer
a
till
collision, the
lower and colder one retain their energy
the temperature 1 C.
bution of temperature
a distri-
between the
produced
walls which is independent of the time, and such that at the
is
height x above the lower limiting plane the temperature
is
of itself
CONDUCTION OF HEAT
104
281
apply
denoted the difference of the values which the number of
We
;
particles which
the difference of the values of the
warmer
by
particles in
unit of length.
particles
which
two
number
different layers
Then
it
in unit time
number
cany
of
may
nD,
where
to consider
we have no
number
n,
longer to
we must
We
;^
specific
It
cule.
we
are looking
is
Nmc
in thermal units,
if
= NmcD
number
of
282
104
results
95 by the formula
&(&& + N&flJ/N,
D=
becomes
D=
i-TrflL
our special case in which the two kinds of gas are the
same, for we may neglect the small difference in the values
of fl and 8 that arise from the inequality of temperature,
in
and therefore by
96 and 97 put
o
iij
and by
95
o1,
o
!>
we have
jt
-eg
>
also
N,
The
ft
x,j
conductivity of a
K = N.
homogeneous gas
fore
f
Heat to
the Viscosity
The formula shows a simple relation
which by
of the conductivity
78
is
given by the
formula
n
0-30967
NmOL.
The
=
=
(|7T/0-30967) yc
1-26812
7]c.
greater than
rjC.
if
we had
CONDUCTION OF HEAT
105
28S
Stefan,
viz.
which,
if
seems
for
and
aimed
rj
^=
viz.
77,
calculation of
and attained.
have already met with a similar uncertainty of the
numerical factors that come into formulae it is in all cases
at
We
In an elementary theory
it
is
cedure, not only to deduce the mean result for each part of
the calculation by itself, but also to employ from the first
magnitudes that
come
into
the
we
conductivity
the relation
and
coefficient of viscosity
I
vj
are connected
by
Wim.
Mathematical Appendices,
Compare
48* in the
284
105
This
the value of the coefficient K as exactly as possible.
been
have
taken
with
values
mean
the
which
in
calculation,
due regard to
has given
Maxwell's law
f
If
we compare
distribution
of
1*6027
of
speeds,
TJC.
which has
value 1-6027 of the coefficient K to apply only to the propagation of the molecular energy ; and we might consider it
possible that the remainder of the energy forming the heat
of the gas, and therefore the energy of the motions which
energy.
In
fact,
many
1
2
especially by Stefan and Boltzraann, that the kinetic
energy of the molecular motion is passed on from place to
place with greater speed than the remaining energy, which
in
Chapter V.
1
we have termed
the
atomic
energy.
Ibid.
We
CONDUCTION OF HEAT
105
285
into
= KTjr
the
observations
then
greater atomic
energy.
Theoretical reasons could also be adduced for this hypothesis wr hich at that time seemed to be unavoidably necessary
encounter
become
be altered by the
collision,
while
many
atomic energy.
In the first edition of this book, which was published in
I explained these ideas as possible and admissible
because I knew no better explanation of the facts, although
even at that time I raised several weighty objections to
them. At the same time, however, I pointed to another
1877,
interpretation, in expressing
my
belief that
all
differences
286
105
Glaus i us, be
that a compensation
would at once result by the atomic energy gaming at the
molecular until the proper ratio was again
expense of the
this the untenability of that hypothesis might
restored.
of
ing to this proposition
By
Now,
many
temperature.
law the discussion of considerations respecting the conditions during an encounter has no bearing on the resolution
If a particle has flown from one place to
of our doubt.
another,
it
it
to
its
its
new
place the
energy and
it is an entirely unimportant question whether and how this
that afterwards
energy is transformed by the collisions
or atomic
occur.
Hence
it
kind of energy
= 1-6027
rio
it
CONDUCTION OF HEAT
106
106. Theoretical
Laws
287
of the Conduction of
Heat
The conclusion of the theory which we have found requires
the conductivity of a gas for heat to obey the laws which
hold for the coefficient of viscosity and for the specific heat.
The
kinetic theory of
gases has
led
to the discovery,
specific
heat at
warmed by
Wiedemann's 2
observations.
On
the
kinetic
theory,
of viscosity.
cited
for
the
specific heat
do not,
M6m. de VAcad.
1.
288
106
Pressures
That
independent
ductivity of gases
was given
time in absolute
measure.
had convinced himself by preliminary experiments that the method followed by Magnus is unsuitable
After he
ployed
before
shortly
experi-
2, p. 45.
5
*
this.
olviii. p.
177.
xiii.
p. 164.
CONDUCTION OF HEAT
107
28$
Andrews' method.
bring forward some of Stefan's results for the
experimental proof of the theoretical law that the conducI
first
on
observations
air
= 0-0000555
air to
then, there
mm. he
-0000552.
If,
is
Kundt
Warburg
59-3
G. to 19-6
gases
named
C.
when
filled
Air
19-5
mm.
= 277
4
0-5
Carbonic acid
7-7
1-5
Hydrogen
154
8-8
The constancy
of
Ann. 1877,
xliv. p.
177
66
68
the
gg.
1893,
xlviii. p. 180.
Ann. 1875,
clvi. p.
xiv. p.
232
'
3
Wied Beibl. 1879, iii. p. 701.
Munch. HaUUtatwnsschrift, 1881.
-
1875, p. 160.
5
sees.
277
278
280
349
350
with the
290
107
the
conies out plainly from these numbers, at least within
the
Under
the
pressure
greater
pressure.
given limits of
the
of
currents
gas.
regularity was disturbed by
In the opposite case also, when the pressure was made
Kundt and Warburg observed deviations from
very small,
currents
no question
constancy; there is here, however,
^of
to that
kind
similar
a
of
in the gas but of a phenomenon
This behaviour is quite analogous to that observed in connection with viscosity for in viscosity the internal friction
alone comes into account in dense gases, while in rarefied
the internal
gases the phenomenon depends, not only on
;
friction,
in the
of the
practically depends only on the internal conductivity
it is conditioned by the external
in
rarefied
while
gases
gas,
conductivity as well.
same
law.
Kundt
and
Warburg,
1
Crookes,
CONDUCTION OF HEAT
107
291
essenti-
ally contributed.
him
Air
Apparatus I
= 750 mm.
138
0-000509
501
480
Apparatus II
= 750
43-3
13-1
1
Hydrogen Apparatus
750
0*000277
260
260
259
0-00294
290
258
245
216
91-9
4-7
3
1-92
Winkelmann
from
possibility
itself in
1
Wied. Ann. 1898, Ixiv. p. 101 ; Phil Mag. [5] xlvi. p. 192
Anzeiger, B6. Jakrg. 1899, p. 1, and Siteungsber. 1899.
3
Phil. Mag. 1898 [5] xlv. p. 81.
Wien. AJcad.
tr
292
108
known.
Stefan's experiments
The formula
by
= 1-6027 yc,
We
gether.
case of gases
iv. p.
321.
CONDUCTION OF HEAT
$108
293
By
numbers
'Values of
c.
determined by
Values of
From
this
of
the
von Obermayer:
77.
value of the conductivity t for both temperatures and compare them with the values observed by Winkelmann,
which I sub-
1*53
Values of
f.
294
108
From
these numbers,
Values of
his text-book
still
more com-
0.
ment
of the
to 100
,J!
Wiillner, Lehrb.
p. 574.
d.
Exgerimentalphysik,
ii.
'Warme'
5. Aufl. 1896,
CONDUCTION OF HEAT
108
295
down
their
The
last
0.
1G ,
Winkelmann
observed by
In the case
of
be explained
and
that
the
errors
of
observation,
they cannot
merely by
be charged to a defect in the theory.
InWiillner's
differences
table there
still
is
left
may
of the
law
when we
consider the uncertainty of the basis of the calcuThe specific heat c of the vapour has
lation in this case.
not been directly measured, but has been theoretically calculated by the rule given in
56, from the ratio of the two
specific heats and the density deduced from the atomic
weight. The coefficient of viscosity 77 has been calculated
from the observations of S. Koch, 1 at 273 and 380, and
also
for
the
much
which
"We
f.
may assume
p.
857.
296
108
Prom
we
PART III
MOLECULES
299
CHAPTEE X
ON TEE DIRECT PROPERTIES OF MOLECULES
1O9. Section of
tlie
Molecules
As the
the conclusions
an attempt to deter-
mine the
on,
in
1867, there
2, p.
395;
p. 129.
SM&mlcWs
Zeitschr.f.
300
Maxwell.
Lord Kelvin,
109
earlier
its
If
is ?.
of the sphere of action ( 44, 63) whose diameter
and
the
we use the relation between the size of this cube
number
of molecules contained
in unit volume,
which
is
given by
the former formula
which shows
that,
may
if
is
known
in absolute
110.
Numerical Values
From
Values of Q.
1871,
Papers,
ii.
[4]
xlvi. p.
453
Ency.
Brit. 9 ed.
iii.
p.
36
Scient.
110
301
of
gases.
that,
if
all
The sum Q
of the sections
may also
of
by reason of the largeness of the number
the molecules, and in this case we should at once be able to
2*: 1
1-26
1.
We
may analogously
in
of their light
which
of
molecules
the
air,
spite
suppose
is given above, to form
area
much
as
cover
so
can
weight
an extremely fine dust,
of enormous number.
302
110
fore seems
to
111.
If
Compound Molecules
Section of
we compare
for hydrochloric
is
follow this rule, since its value for Q, viz. 19200, is greater
than the mean (18000) of those for nitrogen and oxygen,
but in this case the conformity to
viz. 18600 and 17400
;
their determination.
in
C0
and
GO
found for O r
These examples seem to indicate that the section of a
molecule is equal to the sum of the sections of the atoms
111
303
which, form it
for if this law holds for single molecules,
then by Avogadro's law it will hold too for the
magnitudes Q which represent the sum of the sections of all the
molecules in unit volume. As a test of this supposition I
have calculated the most probable values of the sum of the
;
H=
5082,
= 8877,
01 =
N=
9513,
= 9796,
19513.
By
and
'
observed/
Values of Q.
exhibit
an
304
111
llg.
the section
If the hypothesis were general and exact that
is
equal to the
of the molecule of a chemical compound
of but a
allow
sum of the sections of its atoms, it would
peep
We
than
We
path.
value of their section, as found by observation, is not identical with the surface-extension of the plane system of
molecules. But if we consider in our calculations only the
relative
mean and
the greatest.
may
still
be
Q agree
the grouping of the atoms within the molecule to be such
that all the atoms that are bound together in the molecule
do not need thereby to assume that
lie in one plane.
We
least
112
molecule, and
may
at the
same time
305
own
axes.
moves
change
is
recognisable only
another.
Further, when
we must suppose
common
centroid.
it
We
306
chemists,
viz.
atoms
of
112
hydrogen are so
There
is,
been deduced from other observations on viscosity, diffusion, or conduction of heat. I here subjoin some examples
which I have taken from a very compendious collection
1
by Landolt and Bornstein,
methane), in the case of which the
deduced from Graham's obser-
(or
CH
deduced from one of Puluj 's 3 experiments on viscosity, which agrees exactly with that (73400)
And
theoretically calculated from the formula CHC1 3
Graham's number 40100 for methyl chloride CH 3 C1
the calculated number
agrees, at least approximately, with
assign a value 73700,
44600.
From
these examples
also as are
made up
of
2, p.
323.
112
307
loses
its
justification if
at rest.
we suppose
;
we
is
are
being 68800.
We
group
of alcohols
agreement
Vabues of Q.
but a
less satisfactory
Methyl ether
Ether
1
2
3
4
59000
98900
:-
43500 G
89700
80380 P
observed
Winkelmann's
'
those marked P and (?, which are deduced respectively from Puluj's experi.
ments on viscosity, or Graham's on transpiration.
x2
308
we have
112
Here, too, the deviation is the greater the higher the molecular weight and the greater the number of atoms in the
molecule.
diffusion experiments on a series of
from which we may draw an
material
rich
a
esters afford
in
hand.
answer to the question
They give
Winkelmann's
Values of Q.
As
before, the
column
of
numbers marked
Winkelmann's
observed
experiments on
and
'
diffu-
Bornstein's
112
309
'
We
observations on viscosity or
diffusion.
perature
rises.
We can,
113.
Molecular Volumes
310
them, as
spheres and
calling
them
its
113
centroid into
their spheres of
But,
after
spheres of action.
this,
which
also in looking
spheres.
surface conceived to
dif-
The
which we may
well so term
now
that.
113
311
stance
that
is
is
to estimate
With this assumption it is very easily possible to compare the volumes of the spheres of action or the molecular
volumes
Hi.
5.
Abth.
1,
153, 803
2, p. 395.
SnppL-Bd.
p. 129.
1856,
c.
p. 19.
312
=
or,
more
strictly
by
97
is
of service, in
113
formula
\/(3ir/8)
78,
= 0-30967 mfl/7rsV2
w
which, as before,
we have
From
this
=M
formula
spheres of action of
we
fl 2
or
TW^H!
= W *fl
2
two
gases, viz.
V /V =(m /m )"(^ /^
l
In order to be able to compare the values of the molevolume calculated by this formula with those given by
Kopp, the molecular volume F2 for any normal gas, chosen
arbitrarily, with which the others are compared must be
cular
way
he,
many
the calculation.
Such
113
doubts
may
Kopp's
all liquids
that
313
laws, although
investigated with sufficient exactness could be so well calculated by them as to agree excellently with experiment.
For there remains the very grave objection that by these
rules one
to one
of
now
7*8,
now
in
in
12*2
the
ammonia
cyanogen
volume in nitro-compounds.
is
that
the molecular volumes calfurther objection
culated for gaseous bodies cannot be represented by the
different
atomic volumes given by Kopp for the liquid state, for this
seems to contradict the assumption that the atoms are to
be looked upon as invariable. But in order to obtain agreement between calculation and experiment, Lothar Meyer
was obliged to assign, both to nitrogen and hydrogen, a
different atomic volume in gaseous compounds than in
liquid ones.
The variation in the occupation of space by the atoms
which we should have to assume in accordance with these
at
least
314
The
113
Molecular Volumes.
Column
Column
rules, of the
S
01
= 22-6
= 22-8
= 11-0
JE=
5-5
N=
ON =
Kopp's
For the
2-3
28-0,
= 20-0.
Column
III.
which
result
from
113
atoms.
and
= 12*2
also
N = 7-7
H = 3*0
and
and
in nitrous oxide
for
Loschmidt's
= 26,
14,
11,
Cl
are assumed,
one of the atoms
accordance with
315
assumptions.
22-8,
7*8.
as calculated in
In them are
H = 8-5,
N = 13,
but in Column V.
11,
21,
N = 12, ON = 28.
show
But
that
itself if
all
aside.
from
114.
Volume
One
316
114
liquid state
than in
he adduces for
weightiest reason that
gaseous.
our gaseous
of
that
like
a
on
rests
view
this
conception,
the molecules of a
which
motion
of
the
state
of
theory,
of their heat.
liquid take in consequence
atoms in lively motion not only
the
consider
to
have
in the liquid and solid states.
also
but
in the gaseous state,
characterised
to
be
seems
The solid state
by the centroids of
The
the
We
The
space
if it
demanded by a molecule
will
presumably
increase, not only with the kinetic energy but also with the
speed itself, in such a way that, of two different kinds of
for
the atoms.
By
this
consideration
Lothar Meyer
We
is
On
c. p.
360
Abhandl
2.
114
317
corresponding property, of the whole assemblage of molecules contained in unit volume, and we could thus by
experiment decide the question with certainty whether the
difference of the states of aggregation consists only in the
But so
also in other properties of the molecules.
the
a
kinetic
of
are
without
theory
liquid state,
long as we
we cannot in the determination of the extension in space of
motion or
115.
Molecules
We
limiting
values, at least, of the sizes of molecules in absolute measure by comparison of the two fluid states of aggrega-
tion.
already cited.
2
3
318
that
we have
already
115
may be written
in the form
actually
by
the
contained
molecules
unit
in
volume.
we
with
shall
Loschmidt
call
it
We thus
calculation in
absolute
it is
if
possibility of
we may assume
that
when
it
repre-
transformed
maximum condensa-
a gas
is
the
ratio
of
the
gaseous condition. The values of the coefficient of condensation B so obtained will thus presumably be too large, and
this must also be true of the values of the molecular
116
116.
The
319
calculation
actually
carried
vapours,
since
of
out
now
for
may
be
large
the liquid state of the series of gases whose free paths are
well determined from viscosity experiments
and when several determinations are to hand for the same
material I shall always choose the greatest, as this is to be
sufficiently
Pierre 1
ammonia
chloride
is
at
20.
has found 0-866 for the density of cyanogen
in the liquid state, and 0'9 for the approximate density of
condensed sulphuretted hydrogen. 10 Ansdell 11 found for
hydrochloric acid 0-908 at 0.
80 is given as
The density of liquid chlorine at
1-6602 by Knietsch, 12 and Cailletet and Mathias 13 found
that of liquid ethylene at - 21 to be 0-414. Wroblewski 14
Faraday
of liquefied
3
5
Ibid. 1869,
cl.
[3] xxi. p.
[3] xv. p.
oxygen
at
336.
;
362.
Gomptes Rendus, 1878, Ixxxvii. p. 987 Ann. Chim. 1879 [5] xvi.
p. 169
Pogg. Ann. 1848, Erg. Bd. ii. p. 215.
11
Chem. News, 1880, xli. p. 75; Wied. Beibl 1880, iv. p. 310.
13
Journ. de Phys. 1886
Ann. Chem. 1890, cclix. p. 100.
Cowries Eendus, 1886, eii. p. 1010.
14
p. 427.
10
12
clx. p. 214.
p. 221.
;
200, and
[2] v. p.
555.
320
116
l
determined
found the value 1-24 Olszewski has likewise
to 0'905,
0-859
from
as
181
at
the density of nitrogen
0-4148.
as
And,
164
finally,
at
and that of marsh gas
at its
2
of
the
for
liquid hydrogen
density
Dewar obtained
- 238, the sixfold smaller value O07.
;
boiling-point,
in
these values of the density S for the substances
the
numbers
the
combined
have
giving
I
the liquid state
and from their ratio I have
density A in the gaseous state,
With
= A/S.
but
density A of the gases,
calculated from their molecular weight
their values
formula
as
M by
of the
theoretically
means
of the
A = M! (28-88x773*3).
the
I consider this procedure the more correct as it gives
smallest
the
also
therefore
and
for
possible
A,
smallest values
values for the coefficient of condensation t>.
In order to now find from these values of to the values
of the diameter $ I have employed the numbers for the free
116
and
also of
on
Puluj,
Winkelmann's 3
2
3
tables.
For these
given in
The
values
116
we have found
of
them do not differ by so much as one-millionth of a millimetre. Hence it would seem that 'all gaseous molecules and
their spheres of action
But if we recall
sizes.
differ
meter are found just for the simplest chemical compounds, and
especially for those of them whose molecular weight is small.
We easily see that such relations between the weight
and diameter of molecules cannot in general correspond to
actuality if we try to employ the numbers we have here
obtained to calculate that magnitude which
the volume of the molecular path.
terms
Loschmidt
He
under-
116
323
find this constancy when we proceed to calculate the pathvolume for a series of gases from the foregoing numbers.
fied in
s-
for, as we concluded in
112 from a large number of facts, they are more probably,
without exception, flattish discs of very small thickness. We
therefore cannot arrive at true values of their diameter and
volume by looking on them as spheres. At most we may
expect ( 113) to obtain from this calculation an estimate of
be
less
of this sphere.
We may
not,
is
which we have
still
to mention.
117.
tlie
Law
'
transl.
i.
p. 10.
324
117
Begnault's
particles.
radius
value
<?
is
smaller,
and equal
to
L=
From
this
we
(\
single molecule
corrected formula by
which a
-f m )/7rA/2.
is
contained,
is
cube X3 in
,
diminished in the
is,
From
117
325
is, it is
we
obtain
so that
we can
known
values of
7T
dJU
,
/n/0 ,
(v2/o)s,
$ from the
In this calculation we have still
one precaution to take for b and L are both dependent on
the pressure, 5 being proportional to the number
and
therefore to the pressure, and L being inversely proportional
The values given in
79 and 116 for
to these magnitudes.
the free path have reference to the pressure of one atmosphere, while the values of b calculated by van der Waals
and others from his theory presuppose, at least for the
to comgreatest part, the pressure of 1 metre of mercury
pensate for this difference we must multiply the formula for
the calculation of the radius of the sphere of action by the
ratio of the pressures, and thus put
and L.
;
where
.....
...
Air
Carbonic acid
Hydrogen
1
....
pp. 400-2.
5 = 0-0026
30
069
Phys. Soe. TransL
326
117
On combining
in
values
Air
Carbonic acid
Hydrogen
.5 = O80
0*63
x 10~ r cm.
040
we have
started.
We should obtain
We
struck
by another
surface
if
In the
this results
Ixxviii. p. 200),
me
to be theoretically better
established.
3
iii.
p. 102.
3
2, p. 97.
117
327
the formula
ment
is
118.
established.
1
Stefan, in his memoir on the theory of the diffusion of
gases, drew attention, to a simple relation in which the
values
free path
product
(n
- l)L
Now
Maxwell's electromagnetic theory of light requires the refractive index of a substance to be equal to the
square root of its dielectric capacity ; and this law has been
shown byBoltzmann's 2 experiments on seven gases
to
be
The
surprising
fact
that
simple
connection exists
them
are taken to be insulating. According to this assumption the dielectric polarisation must depend on the size and
1
Oefo.
2
civ. p. 421.
328
118
on the same
elements which regulate the molecular free path. This
of the two magnitudes to
explanation makes the connection
"We see
viz.
its
meaning
by
Lo schmidt
for
come
the molecules
if
thing
K,
utmost state
But
of compression.
is
it
possible,
in their
and even
denoted by
If,
$,
and can
therefore,
we
at
most be equal
replace $ in
s
by
g,
we
shall
cular diameter
= 6V2
it.
ai
probably obtain a smaller value for the molethan is given by either Loschmidt's or
s*
Dorn
formula.
to
Loschmidt's formula
is
the
who has
2),
of the dielectric
capacity
xiii. p.
first
by the formula
378.
118
329
determined by
from
calculated
Graham's
experiments on trans-
piration.
are
really
much
smaller than those obtained by the other modes of calculation, as we see from the following comparison of his numbers
Loschmidt (116)
of
Values of 10 7 s in cm.
We
Loschmidt's
magnitude $,
somewhere between the two. I will
not contest the admissibility of this idea, but I must consider
it very uncertain, as so
many kinds of unproved and even
lies
me
observed by
1
xxi
2
Wi&ii.
Dulong.
Sitmngsber. 1885,
For the
xci.
Abth.
p. 446.
et
Nat. 1890
[3] xxiii. p.
197.
330
118
by means
of the
relation
cite
is identical
these
with
figures only
it.
To
avoid
from Exner's
memoir.
Values of 10 7
Ammonia
Water vapour
.1
Nitrogen
Nitric oxide
Oxygen
0*16
0-09
0-17
0-16
0-16
s-
in cm.
Sulphuretted hydrogen
Hydrochloric acid
.
Cyanogen
Sulphurous acid
Chlorine
0-22
0-18
0-19
0-17
0-19
since,
0-2 millionth
of
millimetre.
acid
Sulphurous
is
the
is
= 0'2 millionth
at
we
On
of a millimetre
the contrary
we
find
1 millionth of a millimetre
mechanics.
rely
118
331
smaller than
is
microscopically visible,
we remember
in
as to
molecules.
still a further reason which we may give
the
values calculated from the dielectric
preferring
and
the
refractive
indices.
The equality of the
capacities
the
former
volume
which
values
failed to give
(
69),
pathare
into
view
when
the
latter
comes
(
116),
employed in the
But
there
is
for
calculation, as
is
proved with
sufficient
accuracy by the
Air
Carbonic acid
1-55
....
....
....
.
Hydrogen
Carbon monoxide
Nitrous oxide
Ethylene
Marsh gas
by
.1-42
.
1'84
1*86
.1-50
1-56
2'12
Dor n
33-2
118
part of a millimetre.
Volume
Section and
119.
of Molecules
Although we
may
we have assumed
as a
mean, yet
it
justifiable to
seems
an approximately
suppose that this number may
lost
It is therefore not
trouble, and it is
correct estimate.
more than a play with figures, if we calculate the sectional
area and the volume of a molecule from this estimate of its
serve as
We
diameter.
can
a cube of
120.
a millionth of a millimetre.
is
Now
that
we have
apart of Molecules
an
invisible
the
number
world by determining
at once obtained
area
just
discussed
first
the value of N, or
This
volume.
is
we compare
computed with
109, which we likewise know in absolute
in
on a plane
numerically in
the area covered
110.
it
magnitude Q we understand
all
By this
when we range
we now know,
at
least
approximately, the
size
Since
of
the
120
333
is
easy
find.
In
3xlO~ 16
number
is
N = 18400/ (3 x 10~ = 61 x 10
16
18
N\3 =
i,
whence
X
and
= 2*6 x 10~
this
number,
cm.
too, is
the
same
for
all
gases under
atmospheric pressure.
The
values
in a
remarkable way
paper,
for
the
is
X from
we find
61
If
that
we have now
in round
1
numbers
65.
for air
Compare
p.
250
AbhandL
2.
Abth. 1867,
p. 273.
334
L = 40 X
about.
the relation X
2 x 10~ cm.,
further find, since 9
the volume
for
obtain
we
13
and, finally,
1 cubic
in
contained
molecules
the
We
<?,
by
filled
actually
120
jy.W =
2-5
x 1CT 4 ccm.
i cubic millimetre
eight
diameter, occupy
of the
molecules, and therefore about a 500th part
whole
volume.
is so close
example
appear
it
N = 80000
about
thus remains
still
very
see again that a space containing gas which is
we
and
large,
so extremely rarefied is still very far indeed from being
1
In this case the mean distance apart
completely empty.
molecules is X = 23 millionths of a
of two
or
billions
it
neighbouring
millimetre about.
We can
alter
of 1,000
how
these relations
when
the gas
is
we must
is
Compare
P.
84 and 110.
121
121.
335
Eemembering
this possibility,
it
will not
appear an un-
From
the
number
of
From
cule of
this
air,
we can also
we know
since
x 10~ 23 gram.
the specific gravity of the actual substance of an air molecule referred to water is 5, which is considerably greater
than the
Hydrogen
is
air,
and
Comparison
quadrillion of
witb. other
we
hydrogen atoms
Limiting Values
and atoms.
336
The most
direct
judgment
122
as to the smallness of
the
By
peared.
Annaheim 3
hydrogen must
milligram [which is
weigh
6 x 10
15
really
much
tility
of a
these leaves
must contain
of a
atom
he estimates
wave
at least one
Faraday
of light ; since
layer of atoms, it
of gold is equal to or
This limit corre-
less
3
4
p. 504.
e
jPogg.
Ann. 1857,
ci. p.
318.
'
in
Gehter's
WorterbucJi,
1826,
ii.
122
337
The experiments by which it has been attempted to determine the size of the spheres of action of the molecular
forces in liquids, or the distance to which capillary forces
sensibly extend, also afford help towards the answering of
our question. For this purpose Plateau
measured the
1
smallest thickness at which films of a soap solution containing glycerine could still hold together he took a half of
this thickness as the range of the molecular forces, and thus
obtained the value 0*00005 mm. This value of the range
of the molecular forces was also found by Quincke 2
by measuring the thickness of thin layers on a wall which
exerted capillary force on a liquid through the thin layer.
;
Sohn eke 3
much
millimetre. 6
2
2
Ee
n o 1 d and
ii
cker
1847,
Ibid. 1891,
Johonnott
xliii. p.
had, by measurements
xvL
p. 35.
xl. p.
345.
158.
estimates 6 x 10~ c
mm.
;
f!3S
122
Bunsen, and
1
to be
12 millionths of a millimetre.
Warburg
of the
also
same
But
3
Lord Bayleigh has pointed out, instituted measurements for the determination of the range of the capillary
of an inch, i.e.
forces, for which he found a 250-millionth
that is
calculation
a
but
a
of
millionth
0-1
millimetre,
by
as
really uncertain.
by discussing in
another way the phenomenon examined by Plateau, viz.
a liquid
of
sponding
that inconceivable
tions
if
we
Boltzmann 5
calculated the
work which
is
are held
necessary for a displacement of the particles that
it
the
with
tenacity
together by capillary forces, and compared
this consideration led to
of the substance in the solid state
;
D e H e en
7
Jager obtained
Gr.
for
the values
methyl
and chloroform,
for ether
1
3
5
s
7
p.
O5
345
e.
Abth.
Winkelinann's Handbuch
d.
2, p.
;
2,
ix. p.
255.
Physik, 1896,
ii.
Abth.
2, p.
602.
122
339
is
Such a
possible only
when
The
utilised for
Joiirn. de Pliys.
Chem.
Helmholtz,
Nature, 1870,
p.
55*
p. 557.
340
if
the
122
between adjacent
particles
mining
capacity,
it
makeuse
by
there are
electrical
also
measurements.
of considerations
deter-
other ways of
For
this
we can
of
other kinds of
the electrical forces brought into play with
in respect
mentioned
indeed, as those we have
energy, just,
of the capillary forces.
By
copper,
i.
p 551,
ii.
p. 58.
122
341
as a tolerably accurate
1
particles.
123.
Molecular Forces
of their state
We
know
which
as a fact, directly
come
into
when they
no ground
is
forces,
is
that
is itself
independent
of
1
[An admirable rdsumd of the experimental investigations on molecular
magnitudes is contained in Backer's lecture 'OntheEange of Molecular
Forces,' Jowrn. Chem. Soc. 1888, Hii. p. 222.
TB*]
342
123
Even though
process
Some
upon
and
theorem of
the theorem
if
small distances.
They can
also be fulfilled
bility of
hard bodies.
1
SchlCmilch's
DIIIECT
li>3
clearly
from
vestigates
PROPERTIES OF MOLECULES
the laws
of
343
1
who
in-
massive particles
suppositions than the
the collision of
without
We
conservation of
impact.
oblique
impact, in
magnitudes
is
account.
elastic
encounters of molecules also without thereby ascribing elasticity to the molecules themselves.
answer
an encounter
we
what
that question
to the question,
of
still
cannot pass
it
over in silence.
Schlomilch's
,344
123
becomes
latter may also corne into play when the distance
molecules
the
certainly attract
small enough, though
gaseous
of
distances
mean
separation.
each other at their
On this account Maxwell, too, for a long time defended the assumption of repulsive forces between gaseous
These repulsive forces were supposed to decrease
particles.
more rapidly than the attractive forces as the distance increased, and, indeed, inversely in proportion to the fifth power
of the distance, while the attractive forces were to be taken
as following Newton's law of being inversely proportional
Maxwell
arrived at this
treatwhich allows of
of
the
viscosity of a
ment, because the law of dependence
this hypo#as on the temperature, which is deduced from
But since
thesis, is that which he obtained by experiment.
hypothesis,
contradicted by
is
portional to the absolute temperature,
stable state
Further,
is
attractive forces. 2
Boltzmann
it is
also sufficient
elvii. p.
[3]
49
xi.
p.
193
[4]
xxxv. p. 129
Xinetisohe Theor.
d.
Scient.
Oase, 1889-91,
p. 264.
3
p. 37.
2, p.
714
123
345
We may
made mention.
as
which we have
of
therefore consider
all
attractive
other,
the forces
upon
apart.
each
This
temperature.
The law
of variation of
between the
the energy that comes into play during the action of these
unknown
to estimate the
which
With Boltzmann we
will
first
of
all
calculate
the
amount
mean
on the compressibility
pressed by OOOOQ48 of
atmosphere, so
1
of liquids, viz.
its
Ixvi.
Abth.
2, p.
213.
346
123
is
the pressure the distance apart p of two molecules
is the pressure
An
atmosphere
diminished by OOOOOlGp.
it therefore exerts
of 10334 kilograms per square metre, and
2
a
to
single molecule the force
on the area p corresponding
2
the
wherein
kilogram is taken as
10334 gp absolute units,
of
unit
as
metre
the
length, g being the
unit of mass and
diminishes p by
force
this
Since
acceleration of gravity.
diminish
to
the force required
p by xp is
OOOOOlGp,
10334^/0-000016,
if
we may make
mately
is
doubtless approxi-
is
0-000032.
Boltzmann
m=
3
metre as units,
lOOOp since 1 cubic metre of water
weighs 1000 kilograms further, with our former notation,
wherein G represents the mean value of the molecular speed
as calculated from the energy, the sum of the kinetic energies
,
of the
two
colliding molecules is
24mG = mG\
2
where by
1000 x
614y
whence we obtain
10334 x 9'81p 3
/0-000032,
DIEECT
123
in
OF MOLECULES
1'IlOPEltTIES
347
-x=
According to this calculation the centres of two mole0, if they undergo direct collision
mean kinetic energy, approach to
cules of water-vapour at
when moving with their
when they
Though
may have
encounter of
of 20,000
we
only
learn
on the
two molecules possess a very considerable
intensity.
From
form an estimate
ratio of the
great.
exception is that of the case when the molemonatonnc. In this case we might assume that
there is not interior motion at all in the molecule, since it
But this assumpis simple and consists of but one atom.
tion cannot be right, for monatomic gases and vapours
mercury vapour, for instance can radiate light the spectrum
of which consists of a series of bright lines, and therefore
internal motions must be present in a gaseous molecule that
contains but one atom just as in a polyatomic molecule.
Eilhard Wiedemann 1 has cleared up this apparent
The only
cule
is
made
348
123
compare
it
He
124.
An
Atoms
The
to be expected,
about
We
We
conceive that such large bodies are really indivisible elementary atoms.
shall be so much the less inclined to this belief as
We
many
different reasons
afforded
connection
perties
of
Compare Lot liar Meyer, Die modernen Theorien der Chemie, section
Das Wesen der cliem. Aiome.'
1
9,
124
349
Prout, and
after
him by Thos.
Thomson and
Dumas.
The theory
given by Helmholtz to
be so drawn in the liquid that at
may
is
motion
of rotation,
which
is
of vortex lines.
Helmholtz
atoms the theorem, which this law proves, that the pronew vortices and new vortex filaments would be an
act of creation. He considers the so-called atoms to be vortex
filaments, and represents them by the smoke-rings which
tobacco smokers blow.
However strange this view may appear at first sight, it
will be found by everyone who takes the trouble to master
of
duction of
2
3
PUl
1858,
Iv. p.
25.
SOO
124
it
We
by
spite of,
to the assumption that small ring-shaped or
even thread-
atoms
of these vortex
is
deter-
T
mined alone by the states of the motion in which the w orld
was at their creation. The multiplicity of these states has
given rise to a multiplicity of kinds of vortex atoms, which,
of their multiplicity, were all formed of the same
in
spite
by the properties
of atoms,
by
same kind
ether
is
of elasticity as,
to possess.
Compare
154
Sol
shown
which
are in
tion.
All
shape
this
if
the vortices
and
The same
forces are
act
The
weightless
close rela-
G-alitzin. 6
Each
of
these
between atoms
1
Treatise
to
Cambridge 1879,
1.
p.
Crelle-Borchardt's Journal,
4
Gott. Nachr. 1876, Nr. 11
dynamik, 1881.
p.
18C7,
i.
p.
204;
2nd
ed.
330.
F.Auerbaeh,
Theoret.
5
Schriften d. phys.-ukon. Ges. zu KGnigsberg, 35. Jalirg. 1894
4; 37. Jahrg. 1896, p. 1.
6
Bull, de VAcad. de St. Petersb. 1895 [5] iii. p. 1.
Hydro-
Sitsunysber.
DIPtECT PROPERTIES OF
352
MOLECULES
124
endeavour
electricity
and
light.
MATHEMATICAL APPENDICES
APPENDIX
first
we can
that
same mean
kinetic energy
exist.
no objection can be
1*.
dumber
of Molecules
oo.
The number
of
components
between the limits
an infinitesimal of
of the
from
v, w, so that they lie
u and u + du, v and v + dv,w and w -f dw, is
the order du dv dw it may be expressed by
it,,
v w)du dv dw
9
A A 2
MATHEMATICAL APPENDICES
356
where
are in
Just as the values of the velocity for different particles
different
during which
also will the times be
general different, so
with
direction
unchanged speed.
the molecules move in the same
Let the length of this time that is, of the interval between two
be t, and let the probability of a molecule's
successive collisions
line with unchanged velocity,
moving for the time t in a straight
then
and
are
colliding in the following
u,
v,
w,
whose components
infinitesimal interval dt be
f(t,u,
This function
v,
w)dt.
/, like
more
closely
rf(t,u,v,w)dt
With
NF(u,
At the end
of time
v,
w)f(t, u,
v,
the
l.
number
end,,
of molecules in unit
u, v,
w)dt du dv
new path
in a
new
new velocity.
we assume that the
direction with a
But
the interval
ended
after
time
t,
is
NF(u,
v, to)
f (t, u, v,
2*.
w)
~~
du dv dw.
Siimmation
We
2*
Take the
number
the
axis of
3o7
of those particles
its
condition
r
\)
=u
w,
z
of the point are x
1),
fc.
y
Divide up also this region, from which all the molecules in
question come, into infinitely small oblique parallelepipeds
df dy
by planes drawn
In one
jt,
cZj
dy dz
die
of these
volume-elements
u, v,
it
will,
by
1* happen
,
w) ~du dv dio dp dy dz
u
time
t.
The particles will actually reach their mark, the element dy ds,
and pass through it if the time t is sufficient for the path to be
traversed, and therefore, in the case of particles moving in the
positive direction of x,
if
j> ut.
We
that
1C
all
is,
by integrating
it
MATHEMATICAL APPENDICES
358
the total
number which
mark, we have
2'
to integrate
to
= oo.
Hence
"
di,
J
dw Pf
ch du dv
Hi
dtNF(u,
v,
w)f (t, u
v,
w)
t *
the
particles which pass through
whose
the
with
components
velocity
element clyth in unit time
is
are
the required
11,
number
of
v u\
9
3*. Reaction.
component u is positive.
the particles whicth
If in like manner we obtain the number of
half
of the medium
second
the
from
direction
reverse
in
the
pass
t, which must be
to the first, the value of the relative coordinate
since for this motion the component u is negative, must
such that
its
positive
o
Therefore the
number
ut.
tion is
a **
dy
J ch du dv dw [
JO
In form
sign, but
variables
drNFfa v, w)f(t,
n, v,
w)
v,w}f(t,
u, v,
w).
t Jut
this
if
t,
functions.
by reason of a difference in the signification of these
4*.
1*",
dydsdudvdwNF(u
'v,w)]
f(t,ti,v
w)\ d?
4'"
as the
number
of particles
359
v,
which, in a
unit of time, cross the surf ace -element dy dz in the interior of the
space filled with gas. The upper sign applies to the passage in
NF(u,
v,
w)v du dv dw dz dx,
NF(ii,
v,
w)w du dv dw dx dy.
For u dy dz
from the
number
the
many
velocity,
it
by
it,
state of equilibrium
therefore, goes
moved about
on
just as
in all directions
without hindrance.
The reason
that,
when
simultaneous
collision.
5*.
The
possibility thus
the case of
demonstrated
is
on a
is
1
separated by the surface.
1
See Chap.
12.
MATHEMATICAL APPENDICES
360
come
we have
5*
to deal only
no moment,
for
since, in the condition of equilibrium,
another appears
particle that loses its velocity
same motion by
We
every
collision.
momentum
from the
dy dz in the unit of time in the positive direction of x
first half of the medium into the second by multiplying the abovefound number of particles which in unit of time cross the surface
with the velocities U,
v,
w, viz.
NF(u,
v,
w}u du dv dw dy
momentum
riiu,
and then
dz,
mv, vnw,
The summations
are to
noting that the passage from the first half of the medium to the
second in the positive direction of x can only occur with such
velocities as have a positive component u.
Consequently, the
momenta
carried
positive
direction are
coo
(CO
Q
du\J
/CO
poo
00
dv\J
dw
CO
tPF(u,
V
v>
?),
'
/-CO
pCO
dy dz Nmj du\
rCO
/CO
w),
pOO
dy dz Nm J du J -co dv J
o
dw uvF(u, v,
dv\
dw uwFlu, v, 10}
co
Eor the particles which pass from the first half to the second,
and for the momentum carried with them, the first half is compensated by the passage that occurs in the opposite direction into it
from the second half. In this direction the passage can occur only
with negative values of u and the number of particles which
pass in this direction with velocity-components u v, w being
;
NF(u
v,
w)u du dv dw dy dz,
5y
momenta
momenta
the
it
backward
dz
Jz/
Nm
du
ru
ra
J
03
dw uvF(ti, v, 10}
00
cfe
?w
cZ-y
00
rco
fco
CO
dv
/Q
6*.
00
00
CO
(Lu
J
dw u*F(u, v, w),
dv
J
_co
mtt,
direction are
/-co
r>oo
dy dz Nm J
361
-co
muF(u
v,
w)
CQ
Components of Pressure
by the difference
pCO
|>CO
dv
clu
(CO 00
Y = dy dz Nm
dz
Zx = ^?y ^ Nm
00
_oo
."00
oo
-co
dw uvF(u,
dv
J
CO
00
dw uwF(u,
w\
v,
."CO
pCO
duJ
CO
dw u 2 F(u, v, w),
roo
dv
dio
-CO
>
/-co
/-co
dy dz
cZ?/
v,
w)
the unit of
time, pass over the surface dydz from the first half of the medium
to the second, or, more briefly, are the components of the force
on a surface perpendicular
suffix x,
two
to the a?-axis,
we may
axial
We
/'CO
|*00
(CO 00
du J
Ico co
duJ
Z =:Nm\
s
CO
rco
dv
co
/-CO
CO
du\
*"
duJ
(oo CO
du\
CO
dv\^
CO
dv
oo
CO
dv\J
v,
w)'
dwuPF(u,v,io\
oo
dw mvF(u, v,to),
pco
poo
CO
dwwuF(ii,v
w),
poo
du
dv
J
dw v 2F(u,
rco
poo
Ico
-co
co
/-CO
poo
(oo
-co
CO
*J
roo
/CO
dw u2F(u, v, w\
dv
oo
co
dw uvF(u, v, 10}
MATHEMATICAL APPENDICES
36i>
the positive side, The action of the second half on the first is
with changed sign.
expressed by the same formulae
The first three of these six formulae give the pressures which act
i.e. the normal pressures, while
the
three
the last
magnitude of the tangential pressures
express
whose directions lie along the surface itself.
7'*.
mean
values.
that
is,
mass
the
we may write
X,
Y =Z
= p T?
volume, or
its
density,
== p
vw
where the bar denotes the mean value of the magnitude placed
under it,
These formulae, which were first given in such generality by
1
'
2
for the pressure in a gas in any state of motion that
not
does
depend on time or position, can be much simplified
when there is only the heat-motion of the molecules and not a
ivlaxvfell
_=_=
^
vw
wu
uv
= 0.
of
motion
is
along one
[The author uses the notation H(x) to denote the mean value of %, but
PUl
TB.]
7"
Further,
we
shall
when
363
have
The value
means
is
easily found
for
if
is
= u + r 4?? =^ = 'w = |^
and therefore
and consequently
In
<*r
tc-,
2
,
x = A; = y = i; = ^ = ^, =
y
that
is,
is
as a whole is the
to
0,
mass takes
part, the
as easily found.
as a whole moves in the direction of the
we put
u
= Ui +
component u
then,
if
Ui
hold good,
from that
We have
for the
flow.
if
we now
of the flow
by
a>
1}
where
form
components
On
we have
MATHEMATICAL APPENDICES
364
since u^
for
we
also
this is greater
have
7*
2
,
__
pressure
of flow, viz.
i,
is
it
sp
for in
mean
so that the
coming to rest the gas retains all its energy,
value of this energy remains unaltered, or
a gas at
__
rest,
does not
since the
which
of rest, viz.
is
differ
11-13), viz.
p as proportional to the
expresses Boyle's law.
This deduction of the law is preferable to the former in being
have assumed
as free as possible from unproved hypotheses.
Since this formula gives the pressure
density
p, it
We
move
Boyle'
rectilinear paths,
The
and have
suffices
by
itself
law.
Since the
If
in
sum
%Nm
\
-'
CO
du J[
dv f
00
dw
CO
-f v
iu
volume
}F(u,
V
'
v,
w)J
is
r
we
obtain by this
method
known
also the
relation
365
16)
w w
of the
molecules
m m
i9
2)
formula, therefore, we
calculation, the equation
Dal ton's
law.
8*.
It is,
motions to be rectilinear.
For
Boltzmann
They
if
Z&itschr.
fUrphys. Cliem.
vi.
1890, p. 474
vii.
1891, p. 88.
MATHEMATICAL APPENDICES
S6G
to continually
and
8*
this motion,
as in
gases.
produces a kinetic pressure just
This pressure may be calculated in the same manner as gaseous
pressure,
to exactly the
same
result.
If
the liquid,
continuously
like the particles of gas in vacuum. The only difference is this, that
the alien particles collide very much oftener in the liquid, and that
therefore the free path traversed between successive collisions is
much shorter. But this difference has no influence on the
very
The
sponds
is
which
better corre-
between the
when
alien
rectilinear
which
is
To
we have
of time.
it
when
and
collision only,
8'
367
pressure in liquids
different forces
is
At
1.
(App.) p. 321.
MATHEMATICAL APPENDICES
308
APPENDIX
II
MAXWELL'S LAW
9-'.
On Some
Older Proofs
of gas, as
state of equilibrium of a multitude of molecules
in
their all
consist
not
does
in
shown
22,
has been already
On the contrary, the velocity of any
with
THE
equal speed.
moving
particles
from
this
mean
value
was
among the
first
Max
well
perceived by
molecular velocity are
;
of the
same regularity as we
and processes which are really subject to fixed changes. For the
distribution of the speeds the same law holds good which, accordto Gauss, regulates the distribution of chance errors of
ing
observation
law.
One
such proof was attempted in the first edition of this book, wherein
the law was put forward as the most probable of all conceivable
laws.
Although the mathematical investigation of this idea is
1
of the method
closely connected with the proof given by Gauss
cannot
in
form
be
admitted as
its
first
the
of least squares,
proof
2
and the doubts thrown upon it by Bolt zm aim and von
Kries 3 must be held to be well founded,
4
N". N. Pirogoff, however, showed that nay proof can be
valid,
Principien
tier
ccel.
175-177.
p. 378.
p. 192.
4
Joum. d. Rtiss. pJiys.-chem. Ges. xvli 1885, pp. 114-135, 281-313. Abstracted in Fortschr. d. Physik, 1886, pt. 2, p. 237.
Exner's Repertorium,
MAXWELL'S
9<
LA\V
369
justified
The
maximum
value
when
in the
its
all
these different
be distributed in different
ways possess an equal degree
may
in accordance with
values also the average state of the motion of the group, its
If now a second
average speed and energy, are determined.
of
particles is again picked out by chance, there will be
other values of the components in this second group ; but the
average values of the speed and energy may, in spite of this, be
group
the stricter
way
suggested
munication.
proof in a
sive
I will therefore
way which
will, I
improvement.
Festschrift
Gymnasiums in
zur zweiten
Scicularfeier
Berlin, 1881, p. 295 ; Ueber die
des
FriedricJisiverderscfaen
Bewegung
eines kugelformigen
Atwns.
B B
'
MATHEMATICAL APPENDICES
10
10*.
Maxwell's law
of
which
occurs
it
strictly
paseous
whatever
number
If
of
result,
velocity
of
the
number
be understood otherwise.
of
particles
is
altered in a perceptible
all
sufficiently long period are
these
distributions
different
occurred simultaneously
law would not thereby be changed
particles, the
law must be equally valid in both cases.
but
Maxwell's
On the
of
particles pass through.
different states of a single particle come into
group
NZ
whole, then,
account. Among
will often
it
components
of
for
which expresses
this
law
of
dependence
is
the probability of
MAXWELL'S LAW
10'
371
} ,
in
should move
one particle
lt
NZF
is
particles also.
For the same reason, if the particles of the group are all like
each other, the probability of the event that a second particle m$
has the components u 2 ^ 2 w 2 is determined by the same function
with different arguments, viz.
,
simultaneous
their
product
of the
occurrence
F^ = F(u
2
we
If
i' 3
^7"
product of the
F F<>
l9
v l9
)F(u 2 %,
velocities u$
on
therefore
is
expressed
ws
by the
w2
).
N factors
Fs =
This formula and those which follow later would contain numerical
which would have to be formed according to the rules of combinations
If we sought the probability
if we did not fix a definite series of the particles.
that one of two particles
and 2 had the components 1? v v w^ and the
other the components u Z9 v^ w,2 this would be twice as great as in the
1
factors
case
it
mates
to infinity.
(Encke,
B B 2
MATHEMATICAL APPENDICES
372
the components
UN %,
,
all
2,
WN
among
10*
v2
2,
v lt u\, the
mN having
may have
at
whatever.
F as the first,
arguments,
The two products are equal to each other in value, for each of
the states of distribution is as likely as the other, because, according to our supposition, both form part of the state of equilibrium
which finally ensues. For equilibrium, therefore, it results that
the function
equation
for all
systems of the
1
values of the variables that occur,
This theorem
proved differently by Pirogoff (Joum. d. russ. phys.Pirogoff considers an infinite number of gaseous
particles which are in a state of equilibrium. Prom this infinite multitude
The probability of finding given values of the
particles are taken out.
is
components u,
v,
w among
these
particles is
expressed
by the given
same motion
assumption.
of their centroid.
Maxwell's law
MAXWELL'S JAW
11"
moment
exists
determined.
Before, therefore, the final state, which seems to the observer
one of equilibrium, that does not alter with the time, is arrived at
in a mass of gas left to itself, the function which expresses the
probability of a given value of the speed continually changes its
nature according to the fixed laws of mechanics. But when the
final state is attained the
same
form
of this function
of the
arises
The laws from which this conclusion is the necessary consequence are contained in the theorems which deal with the
mechanics of systems of free particles in motion. Eor our problem, the establishment of the law of distribution of the energy
and speed, only those theorems come into consideration which
contain and determine these magnitudes alone. These theorems
are
1.
2.
The
The
principle of the
centroid.
Further hypotheses are not needed for the present indeed both
these theorems depend on a single common basis, if we may assume
that the action exerted by one particle on another is equal to the
;
reaction
which
at
each moment
it
experiences
itself
of these
MATHEMATICAL APPENDICES
374
IF
In the former
of atoms.
simple massive points or as made up
and
kinetic
kinds
of
two
potential, the sum
energy,
ease, of the
to
the
of which, according
principle, possesses a value that is
constant for
comes
all
into account
for since
all
two molecules
during a
We
collision.
sum
of
We
and
in
21*.
13*.
By
F(u,
v, 10}
Determining Equations
F(u
lt
v l} WI)F(UZ, v 2}
w2
F(uy
zv,
the
function
wA
r)
has the same constant value G for all values of the components
w of the molecular velocities that occur together. According
to the last discussion the values of these components are not
u, v,
theorems
of
mechanics.
then,
of the kinetic energy of one of the
If,
we
must
satisfy the
two named
denote by
the mean value
molecules in question, by
+
+
% + ... -f <U
Nb v + 2 + -"+
No = w + w 3 +
4- iv
.
Na~
MAXWELL'S LAW
12*
375
We
G = F(u
lt
v lt Wi)F(u z v 2
,
w2)
F(UX, V N
wN
)9
coupled with the conditions represented by the above four equadetermine the function F(u, v, w).
tions, to
done by the known processes of the calculus of variarepresent a second system of values which satisfy
the equations by the symbols u + hi v n + $vnt io n 4- hv M where n
may represent any integer between 1 and N, then we have also
This
tions.
is
we
If
tl
C=F(ii
tt ,
\m{( Ul +-^,)
4-
Na =
Nb =
^b
fy) + (W Y + Iwtf 4- ..
2
(U N 4- hixY 4- (V N 4- a%) 42
4- (v, 4-
HI 4t
ciii 4-
4- cvi 4-
??! 4-
&#!
UN
4-
... 4- tv
4-
o^
4- w.v 4-
CW N
4-
CUN
.
and neglecting
terms are
If for
we
all of
shortness
we put
wn
vn
the
first
equation becomes
of all
the functions F,
A __ $F\
Fi
The
^2
^"^ ^N
FN
i
'
we may
MATHEMATICAL APPENDICES
376
In addition
we
to this
12
::
simple conditions
= u^ui +
vjvi
-f
cw
4- ... 4-
= ci^ +
= cv +
= t^!
l
-i-
-h
+
+
uN Miy
V N $V N
+ ^y^%,
CUN
<Xy
cw.v
must be
it
remains perfectly
and what
variations,
therefore,
by means
ratios
of
infinitely small
we
how-
to the
we
If,
eliminate
F,
four of the
3N
variations,
we
+^+T
+
F,,
..'
+ ***
F*
we can
so
BN
its
of
by which the equations of condition are multiplied before being added to the principal equation
these coefficients, which I will take as 2&m,
2kma,
2Jcmj3,
2&wy, are 'then so determined that four variations out of the
whole disappear then, by reason of the 3N
4 other variations
initially
undetermined
coefficients
3N equations
of the
We thus
obtain
form
M AXW ELL
12*
the
same
and
of u, v, w.
value,
which
LAW
577
is
Another Method
13*.
We may
'
F(u l9
*?
two
particles
with
state
its
immediately
any
Of
all the particles, then, only those two which collided have changed
their motion.
may, then, in the product neglect the factors
that have remained unchanged, and thus conclude that
We
F(u l9 v
if
u l9
two
Vi,
} ,
wO-^K
Wi and u 2
= F(Vi>
**> v>*)
^1)^(^2,
7i
2?
TF2 ),
v%,
second
proof,
3
Lorentz, and
and
then in
memoirs
the
of
Boltzmann, 2
It
m
'
which the
depending on the product F(u lt Vi, ?i)-F(%, w 2 ^2)*
w
are
into
is
u,
v,
U",
TF,
F,
changed
components
exactly as great
as the number similarly determined by F(Ult Vl} Wi)F(U2 F2 TF2 )
,
in
it,,
U, F, TF.
The functional equation
4- ^i
^i
-f
p.
Wiener Sitziwgsb&r.
891
&c.
p.
v2
w +
tt? 2
=V +F
= W + TF
157
Iviii.
^2
I'a
4-
vl
is
xevi. 1887,
MATHEMATICAL APPENDICES
378
we employ
makes no
method
alteration in
13"
of the
may
hesitation, as the
it,,
We
result.
= Fu.
2,
or
in which, as before,
cF(u,
v, 10)
and we have
is
put
du
=
By
cwi
_l
hv%.
we then
at
equations
(j tzz
,._
F(u,
v,
F(u,
v,
A
o
JT
in
dio
known
'
dv
T -'^dv-L^
i? ?
dF(ii,v,w)
\
/
_ _
14*.
aw
4.
w)
(^6,
_j_
du
w)
t' 2
w2
^_
^hyfluii
zjm^o
\
o,
a)'
/3)
j
i
%fcm(w
y).
v,
w.
u^v^^
We
MAXWELL'S LAW
14*
379
we may
write with-
du
aw
are
the
F(u,
v,
the velocities u,
By
where
v,
a constant of
the base of Napierian logarithms and
Instead of the latter we may introduce another
integration.
constant A given by the formula
e is
= Adu dv dw.
We
tudes
for instance
velocity of perfectly definite magnitude
properly, a value lying between the limits u and
u+
or,
more
du
can
have a probability that is only infinitely small and of the order
du, we see that G must be an infinitely small magnitude of the
order du dv dw. We may therefore understand by A a constant
of finite
+ (-*;
F(u, v, w) = Ae -*{<>"
This
function,
exhibits
expression
viz.
that
functions, each of
it
may
F(u,
and
<-**} fa fa dw.
have
if
v,
w)
U(u}V(v)W(w}du dv dw,
U(u)=Be~
km(u -
a}
'\
for
we
U*
MATHEMATICAL APPENDICES
380
T3iis
F is
the expression
independent
U(u)du,
the probabilities
occurring velocities v and w of the same particle,
the
correctness of
fact
a
of which are 7(v)dv and W(w)dw;
which
axiom
is so
Maxwell
chose
proof of
it
as an obvious
by
him.
15*.
the conceivably
all
that the
pretation of the functions U, 7,
or
must
be
values
of all possible
1,
B\J
sum of
the probabilities
and that for the two other functions two corresponding equations
must hold. By simple substitutions all three formulae give the
same result
1
= B f dre-**
J
</(*/ km),
CO
of integration
A
9
determined.
We further
arrive at a
of the
knowledge
constants
a, /?,
y by
motion
of the oentroid.
From
duue -**>
00
we
obtain simply
= a, =
&,
ft
c,
since
ln
**r=
dr(r
J_oo
Phil Mag.
[4] six.
1800, p. 22
a)e~
Scientific Papers,
Lmr *
i.
p. 377.
MAXWELL'S LAW
15*
of
which the
first
term
is
dre~ Lm>
381
"~
CO
=/.
We
12*, thus
equation in
00 J
drdsdt{(r + a)* +
= ~? + \m
4/
CO J
(tl
~ ar~
(l
The meaning
~ 6)fl +
(w
<H
(s
6)
(t
+ c y^ e
~ Lm(r ^ s ^ t2)
CO
2
(o? 4- 5
).
of the constant k
equation
first
CO
M
[
__co J
value
by the
du dv dw(tf + v * + w*)e~ lm (
00 J
mean
A = S-^(a
is,
6a
8
),
a; A/
is,
by the energy
16*.
Law
As the values
of its heat-motion.
of the Distribution of
a,
Speeds
we can sum up
way
1),
energy 3/4& when the equilibrium stage has been reached, out of
every
^^
we
= 6 = c = 0,
dv
dw
all directions
MATHEMATICAL APPENDICES
382
for the
and
number
of
16
;r
Thus,
too, the
independently
determined, is
number
of
of
particles
The constant
fc
kinetic energy
u,
is
have the
molecular motion by
44
velocities
mean
the relation
"^F
same
rule
among
among
with
17*. G-as
in notation
when
the gas
is
Notice of
Watson's
MAXWELL'S LAW
17"
383
moments
momenta
the
conservation
of
areas
the
the
of the
we
particle of
u, v,
denote
mass
by
x, y,
m whose velocity is
c2
c3
= 2i.m(yw
= 3(jn(zu
= ^.m(xv
zv)
xw)
yu),
in
position,
of
to a fixed
of obtaining formulas
advantage
about the centroid.
differ
which
refer
only to rotations
the coordinates
as well as the
earlier in
12*
components
of
the centroid.
degree
provided that they satisfy the laws of mechanics,
have to add to the formulae before developed the conditions
of probability,
we
GI
c2
c3
3i.m{y(w
Sw)
= S.m{jsr(w + Su)
= $.m{x(v + Sv]
z(v 4- Bv)}
x(w
y(u
in these formulae only the velocities u, v, w, and not the coordinates x, y, z, are varied, since among the changed values
MATHEMATICAL APPENDICES
384
w+
$u, v 4- $v,
&w those
velocities
are to
17*
be understood
of formulae
it
= ^.m(y to
= S.w(# u
= $.m(x
Si?
z Sv)
x 8w)
y
SM).
We
them by
which we
will denote by
2&, and adding them to the
2&,
%krj,
former formulae, and then the coefficient with which each of the
SN variations 8, 8-y, to appears to be multiplied in the sum of
the equations
In
is to
be put equal to
0.
this
stand at
^/ +
_//
CuU
= lf? +
av
If
= 1^+
Jc
ciw
2km(iD
-y + scq- y),
by the magnitudes
is
the gaseous
mass
sum
if
as
it
which
this is
17
MAXWELL'S LAW
385
is
about the axis of z from the #-axis to the ;z/-axis with the angular
In exactly similar ways are the two other magnitudes
velocity
to be interpreted.
The formulae therefore express motions of the
gaseous mass which correspond to those of a nut which moves on a
.
screw-spindle that itself moves along and turns about a second axis.
These kinds of motion of the gaseous mass are, as our examination shows, to be simply subtracted from the molecular
The
actual velocities of
made up
of three
of a
up
}.
When
always equal values of speed for all the different particles which
belong to the system, but for each particle the values of the
velocity-components which belong to the whole gas at the point
This remark discloses the possibility of
where it actually is.
widening still further the limits of the region wherein Maxwell' s
MATHEMATICAL APPENDICIS
S86
17*
for the propositions are only strictly true for the gas as a whole.
We
not too long. At the same time, this interval may be long
to allow the very rapidly resulting l arrangement in the
distribution of the velocities according to Maxwell's law to
is
enough
occur.
gas
is
In such cases, therefore, the law must also hold good if the
divided into unequally moved layers.
In each of these
layers, then,
Maxwell's law
be subtracted from the value which the particle would have in the
state of rest and equilibrium of the gas as a whole.
Transformation of Coordinates
Maxwell's law it is often of advantage to
18*.
In using
give it
another form by a transformation of coordinates.
Naturally
it is not always
necessary to know how many molecules have a
velocity the components of which are u v, w, that is, a
velocity
of given
as to the
is,
components
u, v,
lute velocity
w= v/(tt -M +
2
Ann.
2
?
),
xssiii. 1886, p.
82;
Natanson,
Wied.
MAXWELL'S LAW
18*
whose
=
= & sin
w = sin
and
387
with reference to
<*>
<*>
cos
sin 0.
0,
= v>
&2
2
).
Then
first
formula
of
n=
16*, the
new one
NJm-*-*"
00
*'*****^ sin
ds
From
all
this
N move
those which
viz.
s
With the
motion,
i.e.
special
sin s e -*<^-^>
formula, found by
>
Maxwell, becomes
an important circumstance
or
= (km
~i
= W.
c c 2
MATHEMATICAL APPENDICES
388
The
18
by
this
on
p. 52.
The ordinate
of
the curve
y dx
so that
it
is
such that
= vjN,
on the contrary
is
of
a value w
the abscissa
= u *J (km) = w/
19*.
From
Mean Values
of the Speed
we
and Energy
easily see
We may compare
with this
E = |mff = 27iM;3w*
2
du a> 4 <?-*ma)a
Jo
mean
value
G of the molecular
==
=
4&
ajto?
We
addition or subtraction of
energy the speeds of all were made equal.
Then the simple relation in which these two mean values
is
= 1-0854 a
MAXWELL'S LAW
19*
389
it
p
as in
of
gas
V at the
= ^G* ^ 1.,^
which
is
pressure
work equal
if
to
each particle
moved with
or the value of
probability.
cases in which a
than
occur
as
often as those in
speed
just
which the speed of the molecule exceeds the mean ralue 0.
Since now the probability of the occurrence of the former case
0, is determined
molecule has a
is
less
given by
and that
while the
it
sum
of
both probabilities
is
value
^.
The mean
therefore be determined
or
by
by the formula
by
Saalsehiitz,
Sitzungsber. p. 45
Sclir. d.
phys.-olxm. Ges.
m KQni^slerg, 1878,
19. Jahrg.,
MATHEMATICAL APPENDICES
390
To determine
of the integral is
the value of
transcendental integral
means
is
by
*=
and
19*
therefore
= 1-0875
1-0875,
"FT
=0-964
Thus the mean probable value is greater than the most probable
W, whereas it is less than the other two mean values, so that the
four special values of the molecular speed arranged in order of
l
magnitude form the series
W<
<O <
G.
To
We
wherein there
20*.
Mixed Gases
brium
of the
form
where
1
TO I;
a,
90
98
_TB.].
MAXWELL'S LAW
20*
391
is still
number
we denote
by
2,
i9
and
if
we
further write J3 l9
energy of a particle
them
of
2,
for the
mean values of
mean value
the
for the
for
so that
all,
then
If
condition
methods
components
u, v,
by the same
12*, there is
The form
For each
of constants only is increased*
obtain differential equations of the form
=4
F
as before, wherein
a, /3 9
du
2fo(ttv
may have
a)'
kinds; the constant k however has the same value for all the
kinds of molecules, since it is the elimination-coefficient by which
the last equation of condition is brought in which takes into
account
all
MATHEMATICAL APPENDICES
392
that there
is
20*
expression
7im(tt2
**
(ir-iJm^e-
^du dv
dw
results for every gas to represent the probability that its comw. In this expression
ponents of velocity have the values u, v,
value of the molecular mass in has to be taken
the
corresponding
same value
for
all.
employed in place
as in
of
thus, that
is
From
this
of a molecule
is
exactly the
or
If,
therefore,
two
which are
of
mean
energy, are mixed together, no change in the distribution of the energy occurs.
The importance of this theoretical
equal
[For
it is
of the
"~dr, if
We
are therefore
experiment, in con&mo> 2
-Taj
MAXWELL'S LAW
20*
393
and
in taking
it
both gases are also under the same pressure, and have thereamounts of kinetic energy in unit volume, the further
conclusion of Avogadro's law, discussed in 31, holds good, viz.
If
fore equal
that two different gases at the same temperature and pressure contain in equal volumes equal numbers of molecules.
Polyatomic Molecules
31*.
The
strictly
speaking,
claim
For systems of polyatomic molecules the investigation is certainly somewhat more complicated ; but for these media too the
distribution of the motion among the individual molecules may be
found by the same method used before, and also the law of
bution of speeds
among the atoms.
among
distri-
i.e.
Let a molecule
consist of a number of similar or different
atoms nti, tn 2
which, in addition to the molecular velocities
u v, Wj execute special atomic motions with the velocities
^ ne magnitudes of these last veloUi, t>i, tt>i, U 2 t> 2 tt> 2
.
cities
must
=
when
S.mu,
= S.mt>, = S.mw,
forming a molecule.
is
extended over
all
the atoms
The equations
molecules.
an
MATHEMATICAL APPENDICES
394
21*
new motion
N(E +
= is. {m(u* 4- v* + w +
2
g)
S.m(tt
2
t>
tt )}
+ 2S.f
J7 of
magnitude
of the
is
introduced by a function
work which
<j>
is
atom
named by Helmholtz
we may
if
easily express by
these forces belong to
central forces.
it a force
Thus,
f(r),
if
is
we
dis-
find
one
the work
of motion,
way
and
as the others.
simpler ones,
may
therefore
But in
this
MAXWELL'S LAW
21*
395
the distances
r/>,
N(E +
where
H stands
we
shall
= H,
@)
53 was thus
which in
denoted.
The
== ^.m(u&u
The
energy
rise in
Clausius
sponding
vtiv
who)
28. {m(uSu
4-
t>St>
-f
= S.mc^, = S.w30, = S.
= S.mSu, = S.m2a, = S.
variation
variations.
<ty
calls
disgregation,
measure, till at last
of
increases in
it
continually correThe
leads to dissociation.
phenomena
is to
be introduced
molecule moves
all
by
MATHEMATICAL APPENDICES
396
= SS.~
-^u
medium
for
is
^H
-[
"jPV&i
21*
^w
c^w
cZ'y
"J~
y^t)
dto
dti
must
4"
<
c&tx)
satisfy the
above
Here
we
Jc,
a, ft,
We
as functions of
is,
however, that
easily see,
contradict each other
if a, b, c
v,
the
w.
changed with
to molecule.
By
differentiating, for
above equations,
we
differential coefficient
du\Fdu
tt,
which
MAXWELL'S LAW
21*
397
and similarly with the others whence we see that the magnitudes
a, b, c must in general be constant in respect to u, v, w also.
;
The
we
obtain
where
atom.
= m{(^-a) + (v
-f m{(u - a) +
2
/3)
This function
Aer ^
1
which the
where by
all
first
0>
(w
c)
4-
may
into the
up
we may put
U(u) V(v
B may be understood
g(u,
y)
a constant which
where
(w
- b) +
of
F = Ae~^du dv dw du dv dvo
2
and
now
is
=m
(( u
t>,tt>)
^)
the
same
for
<r*<
2
0>
is
is
&)
(w
c) }
2^
kind of atoms.
The
same meaning
as before in
14*, so that, for instance, U(u)du denotes the probability that the
atoms of a molecule move parallel to the #-axis with a velocity
du. The determination of the constants can
between u and u
we
have
= B\*
J
We
B = (Tra,
ft
^=
bj
Hence
follow
c.
same equations as
before,
MATHEMATICAL APPENDICES
398
whence
as
it is
proved that
Maxwell's law
21*
weU
Since
Maxwell's law
ft
to determine
as in the
on the
it
the
just as for isolated atoms,
value
for the
mean
Atomic Motions
But another
(u,
t>,
which
is
motions among
determined,
we may
of
one
components
u,
tt, tt>,
must
all
= 0,
= 0,
whole
= 0;
<
X = m(u
2
tf -f rc )
20,
x double the total energy, i.e. double the sum of the kinetic
and potential energies of the atom-complex of a molecule.
or
is
MAXWELL'S LAW
$22*
In order,
further, to
determination
the velocities
is first
u,
t>,
399
magnitude
are variable.
23,
within which
co and
when
their value
If
we
is to
denote by
the
maximum
value of the
= S.J"f(r)<Zr,
v
is
far
energy
to keep the
atom
in the molecule.
The
condition, then,
or,
more
shortly,
is
which
therefore
of %,
This equation expresses that the sum of the kinetic and potential
energies at any moment within the molecule, or the sum of the
heat and chemical affinity within
it,
MATHEMATICAL APPENDICES
400
22*
when
equation
= ^(U + tf +
2
tt
2
)
ty
of the function
as to express 95
integration only so far
by the
2*.
<
we can
carry the
by the transcendental
equation
where
x
and the
limiting value
<J
is
and
being the
(j>
2^,
given by
mg
(f>
m<?
mean
+ 2^ = 2$,
values of $ corresponding to the
and g.
We meet the same difficulty in attempting to find that part
of the energy which we have called the atomic energy ; for, in
taking the sum of the Mnetic and potential energies within the
molecule we ob'tain on the average
velocities #
where the
ix
by
be
for 35.
Molecules
we
rotation
MAXWELL'S LAW
23*
401
suspicion, as in the deduction of the formulae for composite molecules the theorem of the conservation of areas, which was applied
I will
to the case of simple molecules in
17*, was not regarded.
by way of supplement.
theorem the following equations hold for
molecules whose mass m is made up of atoms of mass m, viz.
According to this
d
c2
cs
where
over
4- *>)}
t>)
>)}
)}
all
the atoms of
all
= 0,
S.m)t
and, since
we have
S.mt)
= 0,
S.m&
S.m
first
0,
= 0, S.mw == 0,
= m,
given reduce to
= %.m(yw
GI
S.ms
also
S.mtt
the equations
= 0,
Hence we
zv)
2S.m(mt>
yo)
sow)
yu)
into
independent of
remained double summations, depend only on the position
and motion of the atoms inside the molecule, and cannot alter
still
u were
$.m(yw
zv)
3S.m(gn>
S .m(zu
xw)
^.m(xv
yu]
S S.m(jtt
2 S (x"0
.171
jt>)
pro)
pu)
D D
MATHEMATICAL APPENDICES
402
in
23*
= $.m(ySw
= 2i.m(z%u
xBiu)
S.w(a?8fl
ySu)
zSv)
form
= ss.
d\t
dt>
= :s.m
dm
dw
= 2S.
du
u,
t>, ft)
of its velocity.
In order
to take
we have
again to
number
already in
of formulae.
17"* -for
- + yt-
in
which
equations
Je,
a,
(3,
y,
rj,
are constants,
?)
MAXWELL'S LAW
28*
in
403
of the
form
may
33
is
<f>'
now ^ ne meaning
where
^- as
- b) + (a (D
X = m{(u - a) +
2
c) }
2/,
On
gas
free
which
= 0,
I)i
=0,
b
i) 2
is
exact opposite.
its
= 0,
= 0,
it
c=0,
&3
= 0,
= m(u
4-
2
1>
n>
2
)
The
Hence
theorem
atom m.
of the conservation
of areas alters
may
by
electric currents.
D D 2
MATHEMATICAL APPENDICES
404
24*.
With
24*
of
<f>
We
the
may, however, approximately evaluate
triple integral
Ax
IT,
t>,
tx>
by the
are limited
finite
magnitude
<b
in
ix
We may therefore
= im(u +
t>
=*
k*'
will give a value
2
tt>
which
is
less
1*
than E.
If
this
conjecture
is
we may
correct,
sign S, that
if
we
of
we may
an atom
= @/n,
the molecule, or
e
<
MAXWELL'S LAW
24*
I
405
law
from
energy of an atom
e must be
equal in magnitude to the molecular
The comparison with experiment, which was instituted
energy E.
in
my
cases.
25*.
Mixed
G-ases
The considerations
we have only
We
equations.
each kind
number
of
of molecule,
such
and
differ
from them
= b = c = 0,
has the same value for each kind of molecules. It thus follows
that the mean value of the energy of molecular motion is the same
for each kind of molecules, or
then,
for
20* hold
equalities in
monatomic, but also for polyatomic
good,
only
molecules, so that we are justified in laying
down
the following
compound
gases.
1
Wiener Sitzungsber.
Ixiii.
Abth.
2,
1871, p. 397.
406
MATHEMATICAL APPENDICES
25*
Two different gases are at the same temperature when the mean
energy of the motion of the centroids of the molecules is the same in
"both gases.
is of different magnitude
compounds, the value of the energy ( of the atomic
motions inside a molecule is by no means the same for all molecules, but each kind of molecules has a different mean value of
in different
407
APPENDIX
III
Path of
Given Magnitude
THE first calculation of the mean length of the free path of a
molecule was given by Clausius, in that memoir which we
may rightly consider as the foundation-stone of the theory of
1
methods
that
elementary
Clausius*
No
of transcendental
demonstration
fashion.
gas and
medium
We may therefore
must be regarded
as constant within
We
through a pipe.
length
1
x without a
path of
collision.
Pogg. Ann, cv. 1858, p. 239 transl. Phil Mag. [4] xvii. 1859, p. 81 ;
die mech. Warmetheorie, 2. Abth. 1867, p. 260 ; 2nd ed, in.
;
Abhandl uber
1889-91,
p. 46.
MATHEMATICAL APPENDICES
408
26*
molecule will
If we denote by a the probability that
traverse a path equal to 1 unhindered, a is a proper fraction
which, from the assumption made, is so far of constant magnitude
this
that for every position of the starting-point it has one and the
same value. If the gas as a whole has no motion of translation,
the value of a is also the same for every direction in which the
equal to 2, that
the probability
times as long is a?;
bability of
and we thus see that in general the proan unhindered passage through a length x is given by
the function
a',
where
e is
=^JUa
ZE
log
so that, as a
is
negative,
is
positive.
in
of the
66), viz.
mean
free path.
number
m -*n
ne~
pass over the length x
n [e
dx
&_
<te)/l
= m-
xll
dX
I
26*
409
the path #.
amounts
to
ne~ %llX.dx,
L
This
mean
we assumed
velocity, since
it
is
therefore denoted
used in
particle,
direction,
by
the
Z,
same motion
particles ;
symbol
65.
may
if
magnitudes.
27*. Probability of
Before
we
an Encounter
problem.
Into a space filled with molecules at rest, of which n are
contained in each unit of volume, let a molecule enter with the
What is the probability that this molecule may in a
velocity w.
given interval
t,
In the time
of action therefore
is
therefore
MATHEMATICAL APPENDICES
410
27*
To
this simple
be reduced.
may
reality,
Suppose a multitude
velocity in the
of particles in motion,
same
and
all
with the
same
the same velocity-components
particles
fill
there are
of these particles
which move
another, or even a number of other particles, enter,
with a different velocity in a different direction let the velocity
of this second group when resolved in the same three directions
have the components it,, v, w. We have to find the probability
of an encounter, and the probable time that elapses before an
let
encounter occurs.
The probability
of
an encounter
move
same
as
if,
two different
instead
the
one
swarm was
that
assumed
more
we
had,
directions,
simply,
at rest and the other moved relatively to it with the relative
of
velocity
=V
~~
{(
U)
(v
to
7)
(w
in
F)
}.
of the one
probability, therefore, that a given particle
in
the unit of
of
the
other
with
collide
should
any particle
system
time is to be represented by the same formula as before when for
The
28*.
From
this
Number
simple formula
is
substituted.
Thus
of Encounters
we
case of a real gas by simply finding the mean value of the relative
In this calculation we first of all
velocity of two of its molecules.
assume that all the particles are moving with the same speed.
This assumption is certainly not quite true, as we know from our
former investigations ; since, however, it has shown itself very
serviceable
in the
as
before,
we
denote by
Gr the
velocity
which
all
the
28*
411
particle
U = G COS
V = G sin s cos
W = G sin s sin 0,
</>
which
and
of the motion.
By
molecule.
and the
taken
(?,
0,
w=
0,
is
On
GV(2
we
substitution, then,
-2
cos
s)
= 20 sin
$.
get
sin
and
this
unit time
particle suffers in the unit of time we have to take the sum of the
values of the above expression for all values of the angles s. It
is
particles for
from
s,
s,
or, better expressed, a
so as to lie between s and 5
value differing
ds as its limits.
heat-motion,
goes on in the same way in all directions without distinction,
so that equal numbers of particles move in every direction.
that
it
Consider
so displaced
all
-the latter
MATHEMATICAL APPENDICES
412
28*
of its motion.
therefore,
the
sin s ds
4?r.
n=
where
It is
number
is
the whole
now
s ds,
number contained
volume.
in unit
of the collisions.
to
%N sin
of this
sum
is
"
77
sin
^ cos is ds
Jo
Jo
Compare
which holds
number with
that
first
found
a particle when
it
number
of collisions
denoted by
is
first
all
at rest.
Inversely,
the
particle traverses
1
2.
Phil Mag.
mean
[4] xix.
1860, p. 434
p. 265.
Abhandl uber
28"
the case
when
The value
413
all
of this free
We thus
time.
obtain
7ms 2 o>
for the value of the free
7m<,
swarm
of a particle in a
path
of particles
at rest, but
G"
A
for its value in a
swarm
This calculation
of particles in motion.
Mean
29*.
67 for
of the
Collision-frequency according to
Maxwell's Law
The assumption
is
= N(Jm/irY*e-* *v + r+ ^dUdVdW.
r
ims 2 r
= TOSV {(u -
7)
(v
F)
27*)
(w
"FT) }
which gives the frequency of collision of a particle, whose velocitycomponents are u, v, w, with n others which move about in unit
volume with velocity-components U, F, W, and integrating we
obtain
for the
number
of collisions
which a
x/ {(u
- Uy +
- F) +
2
(o
all
(w
the
molecules
TF)
},
29
MATHEMATICAL APPENDICES
414
Let us
first of all
v,
is
Fo]
com
expressed by
B by multiplying il
to
with
u,
v, w between the
respect
by v and then integrating
the
of
value
oo
The
oo.
and
limits
frequency of colliaverage
sion of any particle whatever is therefore, according to Max-
We then
-vf-N (km!*)* {*
J
where r and
du{*
J
dvT
^
oo
dWreT
dw
f
co
co
co
oo
<
by
the relations
*/{(u-
C7)
(v
F)
much
(w
- W)
2
}
substitutions
for then
of a product of
p= r
J
triple integrals
-co
oo
-co
two
oo
oo
for
of the
form
29*
415
U
<9
a)
cos
t>
s,
sin 5 cos
to
ft
i/r,
co
sin s sin
f27T
which leads
fTT
Too
as sin
rt^r
Jo
Jo
aw
<o
q
d
,,.,_!!
1
~^"
,
Jo
to the value
or, if
mean
the
19*
in
is
velocity
of the
introduced,
This formula,
first
given by
Maxwell,
therefore,
may
which
is
be ascribed to
= 1*41
of collisions
instead of
from that of
differs
28*, only in
= 1*33.
the slightly
The assumption,
all
30*. Collision-frequency in
Mixed
G-ases
different
made
numbers
of them.
values of m,
1
Phil.
p. 387.
s,
Mag.
and
We
MATHEMATICAL APPENDICES
416
ft
for
30*
of gases, since
both kinds
they are at
The
calculation
is,
gases.
but will be
for heat.
of
The
In
} ,
T?J
of a
molecule m^
also
relations
<r
be the radii of the spheres of action or
finally, let s l3 $ 2 and
the distances within which two molecules
2
l9 two molecules
and a molecule 2 approach each other during
or a molecule
>
is
first
29*
number
of collisions
given by
l
y=(7r-Uvni.ir- VWa)
where
r
in which the
formula of
case the
collision.
a molecule
molecule
r oo
J_03
roo
00
r co
foo
dw,
J
00
du a
CO
co
poo
dvA
CO
tfw a
CO
;v"*
30'-
To
417
case,
"k^n*
4"
and by evaluation
of the
sextuple integral
r
_ipoo
oo
rco
roa
r oo
poo
yssCw-^^^.ir-^Wj)"
we
CX3
J-00
rdm\
J-00
d]l\
J~00
d$$
obtain
y
on substitution
or,
do
c2u
J-OO
+ ^V?M
--7
V vKi'niiKtfnz J
__ ^
-----9 v//AiWi
of the
mean
I j
v^WA
two
4-
the
number
98 to the
97 and
These formulae have been applied in
104
into account in
taken
been
also
have
and
of
diffusion,
theory
in the investigation of heat-conductivity.
1
Phil.
p.
386.
349.
[This inter-
E B
MATHEMATICAL APPENDICES
418
dumber
31*.
31*
of Molecular Collisions in a
Current of Gas
friction
influence
which a forward
molecules.
of a gas exerts on
characterised
is
of
the
If this motion at all points
by the
gas
collisions
of
the
and
same
the
same
direction,
frequency
the collision-frequency of its
motion
speed
is
state of things
same
differ in the
apply
itself
it
as heat,
Then
first
the
number
kind, of
and that
of a particle of the
is
r2
wherein y
is
of
one
second kind,
of
= 7rsV2^ a + ^ l7
2
),
given by
is
CO
div^re"
dvt["
J
CO
81*
where
a\
419
of flow of the
two groups in
the direction of u.
By
= U + 4u
= S3 +
= SB + iw
a\ = A + ^a
=
iu
= 85 - &>
w = SB i>
& =^
ui
*> 2
t>
<?,
>!
y assumes
its
,4)
If
we now
=V
2
(tt
TJ&,
tt
2B
+
2
}
2
tt
4-
become
<
2
tt?
4fcm{(u
- a)
4-
2
1>
2
tt)
}.
integrations with
= 1 /^ V
CL
2?r
"
f
J o
of
a gives
If
we
obtain
= W^/TT&m) (1 4-
foa 2
powers
of
a we
...)>
negligible.
of
Here, too,
it
introduction of the
we may
therefore
MATHEMATICAL APPENDICES
420
to a gas
collision-frequency of its molecules as
no other than its molecular heat-motion.
which possesses
The
collision-frequency of
Molecule
may
given speed
31
where
components are
variables
U + u, V +
v,
/^rV
where
for shortness is
-f-
axial
consequently
dU\J _
OQ
put
)* -f
(7 +
v)*
+ (W
Since there
our
for
its
in general of finite
is
coordinate
direction,
the speed
as that of one of
coordiDates ^,
the axes.
7r$
If
we
further
5,
raw
substitute polar
TFwe have
I-TT
j-oo
N(km/7r)
d(j>\
JQ
ds sin
s\
JQ
JQ
where
w2
On
+ 2^ cos $).
and 9
this
becomes
32*
421
form
into another
fWftttO
a,)
du.
v
&-
tion,
Maxwell's.
integration,
same value
which we deduced in
30*
magnitude
this represents
the
number
of
number B with
at least
last
This expectation
few values
with those
given, to represent
the
approximate accuracy.
of the
approximate expression
V(u*
Dissertatio de
Gasorum
+ &2
by the comparison
of a
Also in the
first
MATHEMATICAL
422
The following
series of figures
APPEiNTDICES
show
two expressions
that the
w,
while
absolute accuracy
is
not desired.
33*.
With each
collision a
The number
of its
zigzag path.
as that of the straight bits of the path traversed.
Consefind
mean
of
of
the
we
one
these
bits
length
quently
by dividing
the length of path traversed per unit time (which is measured
same
by
thought
is
to
divide
mean speed by
the
the
mean
L =G
r
collision-
In this
way
gave,
PMl
Abth.
2
Maxwell 2
2,
Mag.
4, xix. p. 434,
note to p. 265
p.
29**,
423
by employing
his
law
of distribution of speeds.
The length
of
s,
L
is
sphere of action.
In the case of central collisions the diameter of a molecule or
its sphere of action would have to be subtracted, and
a smaller amount in every other case. The mean value of the
correction may be found by a simple integration over the hemi-
the radius of
sphere.
made by
impact
is
of this section
mean value
2s f
its
value
is
2 cos
s sin s ds,
and conse-
of the correction is
cos 2
sin s ds
5.
Jo
Van
of this
in a later memoir, 2 He
^9, has given a more exact calculation
pointed out that the correction thus found does not refer to the
it
is
p. 48.
2
34*
MATHEMATICAL APPENDICES
424
that has to be
the colliding particle relatively to that encountered
falls on the striking
of
this
shortening, part
shortened by f s, and,
If the former moves with a
that struck.
particle and part on
with a
u
are
v, w, and the latter
whose
components
velocity
the
of
striking
whose components are U, 7, W, the path
9
velocity
particle is diminished
by
u*
v2
w2
4-
^TTT^
and that
by
^
__
_
__
+
(7)2
7)2
T7
(w
To
number by the
r)*d
where
r
and
If
integrate
now we
=
:
N/
{(-
UdVdWre "
2
27)
the result
collision-frequency
(v
7)
29*)
4-
(w
is
of the velocity
= \/(^
-f
2
-y
+^
2
),
viz.
18*),
we
According
to this calculation,
of the molecules,
we have
on taking account
of the
magnitude
mean
free
path
If instead of
N, the number
volume,
we
34"
introduce
X,
425
m=
3
we
i,
obtain
X3
Advantageous use
of this
f 7TS
3
'
in
improved formula has been made
by molecules
H7).
j
G.
theory
these
considerations
also
to
the
of viscosity.
35*. Influence of
W. Sutherland 2
mula which
of the for-
by calculating in what
two
between
particles is increased
We
now proceed to give his calculation in order to put on a better footing what has been said
in
71 and 85 respecting this action of the forces of cohesion,
we need not calculate the absolute motion of both
but
particles,
only their relative motion with respect to each other,
we may take one to be fixed, while we ascribe to the other a
velocity which is equal to the relative velocity with which they
Since
move
which
The path
one
lies in
of the
moving
particle
by the
of
differential equations
This constant
1
Ji
to
p. 89.
[5]
xxxvi. p. 507.
MATHEMATICAL APPENDICES
426
35*
where b
paths, and the surface h is then equal to the product rb,
fixed
the
from
is the length of the perpendicular
particle on the
the
of
of
the
rectilineal part
moving particle.
path
The angular
velocity is therefore
f
= brp~
2
,
on introduction
differential equation,
of this value,
= - P ry 2
ip*
is
= h/p =
We thus
finally obtain
=r _
2
collision ensues
sphere
of action,
if
and
this occurs
which
satisfies
is less
if
of the
the equation
we may
put as
and
is
equal to
F(p)
0,
is
never reached.
35*
If these gases
427
for
particle.
The
of the collisions
sists in
7r$
minator
given in
71,
and
its
value
was estimated
on internal
in
friction.
36*.
In the case
two
of a
different kinds,
first
we
of
of a particle of the
second kind
30*, viz,
+ TT^X/OL +
& we may
kM/Jcjins)}!*!
write
=1
first
established
2
by Maxwell.
is
87.
MATHEMATICAL APPENDICES
428
37*.
molecular free paths calculated in the prenature of average values, since they are
ceding
deduced from the mean value of the speed Q and from that of
the collision-frequency JT. We may therefore, for instance, look
The values
of the
on the value
found
mean
mean length of free path which any one single particle, moving
with a particular speed, passes over without a collision.
To find the probability that any particle moving with speed w
traverses a path of length x (or rather of a length between x and
x
4-
of
we go back
to the
formula
which give
being the
mean
BZ
free
o>.
is
= oi/B,
particle's
traversing a path
which
e-**i*.
So as
to
show more
clearly
how
mean
free
path
<*>,
37*
The
first, third,
429
1
:
ratios of the
actual speed w to the mean speed &, of the probable collisionfrequency B to its mean value F, and of the probable free path I
to the
mean
together,
as
free
is
plotted from the given numbers, the first representing the collisionfrequencies, the second the free paths, as functions of the speed.
and
=L
speed.
38*. Different
Mean Values
of the
Free Path
path
already calculated in 33*, we can, by a slightly different calculation founded on these last considerations, deduce another mean
value which is important for the development of our theory.
>
[Here
fcrofl*
4/7r
1-27824.
TB.].
MATHEMATICAL APPENDICES
430
Of the
38*
gives
as the
infinitely
little
from
w.
According
B/r
.Ctf/flt
O
FIG. 5
2
PIG. 6
particles attain
is
number
on the average a
free
path
sum
= w/B,
If therefore
which
38*
431
question
The sum,
From
i.e.
which
the
all
between successive
mean
the arithmetic
for the
unit volume.
this
formula
is
We may thus
we
obtain
by M(t), by dividing
thus expressible as
whole number
particles
collisions, is
moment
in
is
path
at
any
particles as starting
at a given
mean
start.
We
we have
to consider not
whole
which
is at
of the
is
therefore given
The sum
by
of all the
the integral
= Ml.
But, according to
number
32*, the
Y Haw
A
"
of these paths is
w 2 Be- towa
= NF.
paths
38
MATHEMATICAL APPENDICES
432
we
value which
mean
Consequently the
L = o/r.
mean
This
value
summation of the
I
is
free
come more
one
each particle
path for
the speed
is
into account
considered, as
w.
We
simple calculation
as given in
32*.
if
we
If,
substitute for
then,
its
approximate value
we put
AW s ^
M(l)
=4
which reduces
to
and from
we may
tables
this
of this integral
calculate the
we
Jf(Z)
This value
is
mean
value.
By
the help of
find
= 0-937 L.
we must
still
remember
that
it is
0-937
is
too small
by
this
is
there-
39*. Interval
collisions of a particle
1
Bessel, Fundaments Astronomies, 1818,
Jahrbuch filr 1834.
p. 36.
Enoke, Astronomisches
39*
433
We
might
first
consider all particles to start new paths at the same moment, and
then ask how long the interval is, on the average, before one of the
Since a particle whose speed is w
particles undergoes collision.
collides on the average B times in unit time, the time that passes
between successive collisions of this particle is on the average
the
mean
cj
0)
= oo.
The
calculation
mean
the
interval
We
number
the
of particles,
else
The mean
is
then nothing
mean number
of
This second
in its
for
each
larger interval
more seldom.
simple results.
first of
We
FF
we
MATHEMATICAL APPENDICES
434
suppose
is
40*
by a gas
We
y,
is
perpendicular to
df.
at a distance r from df, and let the line r make an angle s with
In the element dx dy dz there are at
the negative direction of x.
each moment
V ^V^
C
dx dy dz
an average B
particles with speed w, of which each collides on
times per unit time with other particles and starts a new path. In
unit time, therefore, there proceed from the element dx dy dz
>
N(Tm]irf ~Be-
move
dx dy dz
Km<
w 2 du r~ 2 cos
dfdx dy dz
The number
without collision
of
N(km/w^Be~^eif /3 is
To deduce from
this
by them,
Kmu
mean
/3
force exerted
the particles
is
2
d^ r~ 2 cos
free
number
we must
df is part
I,
dfdx dy dz
or
of the particles
meeting df the
multiply by
2mo> cos
if
path
= 1/Z = B W
s,
of the gas.
6*
passes
through df in unit time in the direction of increasing x, and subtract from this the momentum
simultaneously carried over in the
reverse direction.
40*
momentum
the
435
Qidf
= dfNmHmM^
from
00
Jo
this the
du wV^'Bf^dre-^f^ds sins
Jo
Jo
momentum
Jo
viz.
f
du
Qzdf=dfNm(km/7r)i JO
wV^'B
dr e'*
JO
C ds sin s cos
[%Z0,
JO
J'TT
formula
= 2JVW&m/7r)l[
du a)V" 7lWa)2 B
Jo
which
is
ftt
2
r{
[ "d$ sin s cos
Jo
Jo
d(I>,
On
p
for
dre'
Jo
which may be
= N/2k,
written, in
19*,
Air-iNdGntye-^uPdMBe-^dt dx dy dz
dx dy dz with speed
w,
and
/lTOW
47r-'JVpm>}<r
V^
e~**Bt-
dtdxdy
t,
is
dz.
Nfymlrfer^uPdu
reach the element df
if
e-**B1r l dt r~ 2 cos s
the time
is
dfdx dy dz
F F 2
MATHEMATICAL APPENDICES
436
be traversed, that
pressure
40*
is, if
wtf>r.
again,
integrating.
p=
41*. NuinTber
Colliding Particles
of particles, by the collision of which against the
walls of the vessel this pressure is set up, is easily calculated by
The number
We
p = ^irNmti
2
,
the
by %N& particles
give rise to the pressure
value of the impulse of a single particle
struck
is
mean
is
mean
it is
values in the
42*.
The number
volume is
way
mean
of the
W(fo) JQ
("Bg-^Vdw f "V^/Jr
of these integrals
respectively are
2>/(27r)JVV/fc7
and
is
and
Z^/
their ratio
the
mean
paths traversed
collide is
JQ
The values
Jo
for $w,
free path, as
to
free path.
dr.
43"
437
Motion
The same calculations may also be made in the case when the
mass of gas is in a state of flow, and we shall estimate the pressure
exerted by a gas moving with constant velocity a upon a surface
in the mass of gas which we take at right angles to the direction
of flow.
A
C
Two
Maxwell's
law,
by the more
s)
s
o
for the
momentum
positive direction,
/ sins cos 2 *
e-*^
and
"
4
ds w
si
for that carried over in the negative direction ; and the pressure
exerted by the streaming gas in the direction of its flow is, accord-
can,
on introduction
or,
we
5=
%Nm[(bir&
Similarly
of the
mean value O
19*),
we have
a2)
and thus as
result
we
==
NmfarQ*
a 2 ),
we
obtain
MATHEMATICAL APPENDICES
438
p
or
its
ds sin
much
43*
simpler calculation,
lM(ca
2
cos s Q~
"
equivalent
duF doF
J-oo
We have obtained
and from
and force
this
J-o
35
formula in
(p.
72) in the
form
it
of reaction of
resistance of air.
The magnitude
of the
resistance
is
calculated in another
in some memoirs. 1
formula
way
for
while
face.
area,
This
mode
to the third,
and
power
of the
this reduces to
speed
was
a.
first
3
theory, which were, however, answered by Clausius.
It is
Q and Q 2
for
W.
B.
Smith,
Toepler, Zwr
Fheorie,
Zur
Gottingen 1879; E.
Lufiwiderstandes tiacli der Mnetisclien
Sussloff Jowrn. russ. phys.-chein* Oes. xviii. p 71),
JErinittlung
Wien 1886
G-.
Molecular-Kinematic
des
1887.
2
Him,
Tempera-
4e Belgfyue
1886.
I'Acad.
439
APPENDIX IV
VISCOSITY OF GASES
44*. General
THE
Remarks on Viscosity
friction of gases
viscosity or internal
else
is,
according to this
nothing
another of the momentum of translation or flow of the medium by
means of the heat-motions of its particles. In order to find the
of the force exerted by one layer upon another by
theory,
magnitude
that
appeared
here
first
in
1865,
reproduce.
theory
of
However, since
I then
tion of speeds.
45*.
'Ueber
'Ueber die innere Reibnng der Oase,' erste Abhandhmg.
den Einfluss der Luft auf Pendelsohwingungen,' Pogg. Ann. 1865, exxv.
i
p. 586.
MATHEMATICAL APPENDICES
440
45'
importance in comparison.
z such
Consider a system of rectangular coordinates as, y,
that the 2/-axis is parallel to the direction of the forward motion,
which the
friction
in
plane pass through any arbitrary point
an
take
this
In
z.
plane
the medium with coordinates x, y,
with edges dy and dz, and find the
infinitely small rectangle
which pass through it and the amounts of
number of
2/0-plane,
and
let this
particles
momentum, which
I will denote
by Q and Q 2
l
carried over
it
in
point
dx'dy'dz'
the number of particles which, starting from it in ti straight
If
is
course, meet the surface dy dz and pass through it.
are
there
unit
in
contained
volume,
the number of molecules
,
T denotes the
number
of particles
and
commenced
is
67 of the
text, or
we have
-$7rs'
\~
33* of Appendix
III.
and A denoting
VISCOSITY OF GASES
45*
441
dis-
f
,
O,
or
r2
we may
as the
= (x - xY
r
interpret the
number
-f (y
y)
- z)*,
(z
magnitude
of particles
which
start
s,
JNT-V^Trr
)-
cos s
dydzdx
dy'dz
How
Let
this occur
(x,
z),
and
let
is
dQ = (m/tyNr-We-fir-* cos
From
dy dz
dx'dy'dz'.
r
this,
of increasing
is
~ l
f dx
Ja,
dz e-^v'r'* cos
f
J~co
dy' Jf
co
5,
MATHEMATICAL APPENDICES
442
"
-00
The
difference
sum
or the
"
CO
45*
is
CQ
of the gain
and
and so
given by
db (x f
f
(y
= r cos s
=
r s n 5 cos
y]
=
r sin 5 sin
z)
x)
~~
(z
</>,
= dy dztynJ^NT^r
*
[
V*V cos
sin s dr ds
d<t>
'
Jo Jo Jo
Q2
where
= dy d^m/^NT"
for shortness
V]f
=v+
%'
-y
j j"
JJ
d<j>,
we put
cos 5
cos
^J?
may
Then we obtain
5+
-~ sin s cos
^
J? sin
cos
+
+
~? sin
ssin0
j?
sins sin
V+
0V +
are independent
VISCOSITY OF GASES
45"
we
get,
443
tions,
of
of
the
friction exerted
F = dyd,mNT-*(l
\3/3
The
series obtained
of the
may
if
rapidly that we
for the friction
5/3
and
+.
% + J^
dx
dx
by rising powers
square for by substitution
its
or,
1/L
more
for
correctly, of
/3 it
becomes
is
may
F = %mNT ~
L*(du/dx)dy dz.
dv
ITT
in
which
T?
is
viscosity.
for
75)
or
77
of this co-
MATHEMATICAL APPENDICES
444
where
and
Gr,
45*
of the
formula, which was first deduced by Maxwell / we put the value of the mean free path given by C 1 a u s i u s
theory, we ohtain
If in this
molecules.
'
If the
which is sufficient as an approximation.
so
it
flows
as
is
the
is
assumed
small,
gas
always
speed with which
simplification
in the theory
of
we can
viscosity,
consider
it
as negligible in
is
manner Boltzmann
and Tait
We
momentum
way and
dyds in the same
In unit volume there
carried this
manner
are
and therefore
47T~ W(&m)
/lTOCO
V<$w r 2 dr sin
dy dzN(km/7r)%e~
Bach
particle begins
hme
B new paths
in unit time,
TT
wii
where
Jo
ds d$.
=
V
p. 390.
46 X
VISCOSITY OF GASES
445
e- B >/<".
(or,
y, z) is
Jtmeo
7rYie~
*Be-'
Br/ClJ
<o
d<j>.
d(^
w 2 B0
~ lm<0 "
(co
for the
momentum which
^ds
roo
dr e-* rl ~
j
%v
sin s cos
s]
d$ v
we
put for
friction
of
is
series, or
2
dr re~* Tla
or, as it
may
The value
and
also
*^,
to
be written,
of this integral
byBoltzmann
= 0-088942636 w/
T ait's
by Tait,
this
11
and with
find finally
77
See
48*,
446
MATHEMATICAL APPENDICES
The numerical
factor,
materially differ
assumption
46*
of equality of
47*. Calculation,
speed for
all
molecules.
Law
claim to perfect
improved calculation cannot lay
is valid for the
which
distribution
of
law
accuracy. Instead of the
for a gas in a
holds
that
law
the
state of rest, must be employed
Even
this
state of flow.
of course be extended without difficulty
of gas has a forward motion, as has
mass
the
wherein
to the case
But this generalisation has
18*.
and
16*
in
shown
been
to the particular case wherein the forward motion
reference
only
has the same direction and speed at
may
paths.
compared
And
we may
flow as uniform.
This assumption
hypotheses
may
be
justified quite
independently of the
of the
difference in the
ingly small.
On these grounds
we
16* and
VISCOSITY OF GASES
47*
447
We
now
to put sin
cos
if,
</> t
as hitherto,
we
The formulae
of the preceding
paragraphs are thereby changed so far that the exponent
which occurs in Maxwell's formula, is to be replaced by
meaning
it
The number
on the average
= mw +
B of
wv
sn
cos 0.
in unit time,
We put therefore
dy dzN(km/7r)se~
'Be~
"^ du dr
Rrlt
cos
s sin s
ds
d<f>
<f>
poo
roo
dy dz
fin
in unit time is
r%ir
dr
ds
d&Bs
dydzmN(kmJTry\ dw
JO
JO
JO
JO
\
Br/'
e
o
q
ws
6;
and
in
45*,
is
f
substituted for v in q,
and
also the
momentum Q 2
47 "
MATHEMATICAL APPENDICES
448
which
for
is
when
t/ 2 is
put
a/.
In
of a
this
collision,
velocity
is,
o>,
We
at all points.
may,
gas is not in the same state of motion
B
to have the same value
take
and
this
variation,
however, neglect
v and v'
everywhere if we retain the assumption that the velocities
forward motion
of the
may
In
this
in comparison with
e~*
o>
2
,
= 0~
of t/
and put
toa>2
(l
%kmwv'
sin $ cos
<j>),
Q = (yy -f
$ = (yv
l
and
also
i?;
dv/dx}dy
dz,
wherein
y
37
28
15:r
The
last
the gas.
48*.
even thus
from
first
it
it is
is
as
not
little
integrable as
difficult to
its
grasp
Boltzmann's;
but
obtained.
We
first
of
all
by
is
For
VISCOSITY OF GASES
48*
we put
and
the value of
given in
449
32*, viz.
new
under the
sign of integration disappears, and r\ becomes inversely proportional to \/(km), i.e. directly proportional to the mean molecular
replace the free path I of the particles which move with the speed
w by the mean free path
given by the equation
2
),
which
results
in
is
independent of
m and is
is
also
of
the
thus the
law
of probability,
Maxwell's
by
then
By means
value
we
K must lie.
The mean value in the formula is to be formed exactly as the
mean value M(fy, which is introduced in 38* with the like notation. The arithmetical mean, therefore, of all values of the product
G a
the numeric
MATHEMATICAL APPENDICES
450
&H
for the
48*
whole
of the
found, wherein I
Thus for each particle
the speed with which it is traversed.
01
has been begun.
which
we are concerned only with the single path
value
mean
another
by considering
38* we also formed
In
a lengthened
for each particle all its paths traversed during
traversed
the
all
account
into
take
paths
now
If we also
period.
B
instead
calculation
the
into
to
paths
have
we
bring
in unit time,
mean
a
in
results
value,
This
larger
of a single path.
process
37 *) the larger
since the number B and the free path I are (by
is
defined by the
formula
If
first
if
given by Stefan,
we
On the other hand, we get too small a value for the coefficient
we substitute for the variable free path I its mean value L, that
a length which is too great for small values of w, and too small
for large values of w, so that in the calculation of the mean the
have
smaller values come more into account than the larger.
is
We
therefore
or
of the
coefficient
therefore,
K,
we
B<?
or
0-27<K<0-39,
1
Ixv.
Abtiu
2, p. 363.
obtain the
VISCOSITY OF GASES
48*
is
r/\/2,
it
closer.
451
r<0-37.
= 0-30967,
which
I have
is said to be correct to all five places of decimals.
tested this calculation in different ways, firstly by repeating
former calculation with a greater number of calculated values of Z,
my
from
by breaking up the
to oo, into a
number
of parts,
integral, which
and putting Z in
lies
gases
= 0-30967 mNQL
G G 2
MATHEMATICAL APPENDICES
452
48*
employed
indeed,
on
As has
assumptions which are not exact, but only approximate.
can
in
47*, objection
especially be
already been mentioned
has
been
of
distribution
raised that the law of
employed
speeds
in a form
which
is
and
that the
But
hitherto
49*. Friction
much from
the truth.
In addition to the kind of internal friction that has been considered in the foregoing investigations, and which takes place in
like fashion in both liquids and elastic fluids, there can be exhibited in the latter class of fluids a second kind of friction
which
depend on a
different coefficient of
For as the elastiviscosity.
extension and tension require two different values of the
constant of elasticity, so also the friction that comes into
play
cities of
with dilatation
of
friction
may
from
that concerned
in
the Motional
phenomena
Sitmngsber. Abth.
2, 1880,
Ixxxi.
VISCOSITY OF GASES
49*
453
roll
friction
of
We
it.
momentum
make than
to
and
to
&m(w
dy dz in unit time
2w%
m sin s cos
momentum normal
un
-f
cos
for
s)
cos
which
Thus the
s.
is
carried over
is
Q = dydz mN(kml^
cZw
Jo
dr ["* ds
Jo
Jo
sin
cos s
Jo
==
-f-
- f -- cos
1 \dx
which corresponds
to a
e -q
with
sufficient
.__
-^
dy
Q%=
<
sin s sin
dz
4>
}r
...
-jcm<*
_j,
we
approximation,
2&mom' cos s)
^p
-f
f
yu
\p
where
p
is
sin 5 cos
\n dujdx
\rf du/dx,
~ ^itmNQ?
whose meaning
is
The
latter is the
;
g2
Q ssp
l
if
du/dx
=p
49*
MATHEMATICAL APPENDICES
454
Our theory
have
Q=
y<v
= %mN&v
from
same
is,
is given
intensity, therefore,
/3Q
where
/3
is
by the expression
= %/3mNlv,
a numerical coefficient.
viii, p.
287),
Cauohy
(Hxerc. de
VISCOSITY OF GASES
455
many
/3
of the particles
as certainly in like
oppositely directed to their initial motion,
manner occurs at the edge of flowing water; and in this case
we should have to assume ft > 1. According to Kundt and
83, it
Warburg's observations, which were discussed in
appears that
ft
whence
=f
would have
to be put
= /3y = ^mNQ.
lose their
whole velocity
of translation
MATHEMATICAL APPENDICES
466
61*
APPENDIX V
DIFFUSION
61*.
IT
is
Maxwell's
As
law.
make only
the
P-Pl + PSP
possesses everywhere the same constant value P, and therefore
keeps this same value always and also that, corresponding to it,
;
same number
N,+N,
of molecules of the
two kinds
in unit volume.
As
time pass in the direction of increasing x through a surfaceelement dS of a section of the diffusion tube at a distance x from
the beginning of the tube.
of the validity of
We form this
sum with
the assumption
Maxwell's law
This
of distribution of speeds.
assumption is not strictly admissible, since the deduction of this
law presupposes the state of motion of the whole gas to be every-
DIFFUSION
51*
within a
large
tolerably
region,
be considered to hold.
457
Er!ea
e~
rM<
**u
we have
is
du dS cos
s sin s
ds dr
<ty
is
s)
position x.
are not concerned with the whole
We
s,
number
of particles that
pass through dS, but only with the difference between the numbers
which pass from the right and from the left ; this difference does
is conditioned only
Hence, on introducing into the above formula
the expression we have developed for N, we neglect the first term
and investigate only the second
by
its variation.
%e- B^e~*m*V& dS
cos 2
which we have
to integrate between
and oo in respect to &> and
unit
over
as
area
from
to \v in respect to 5, and
r,
regards dS,
from to STT in respect to 0.
thus obtain as expression for
the number which pass through unit area in the direction of
We
2
3\/7r
dx
same
MATHEMATICAL APPENDICES
458
52*. Signification
52*
of the Integral
and Value
The
in
mean perpendicular
distance, represented
diffuse
by A, by which the original layers of the molecules that
section.
that
from
distant
are
section
through any
In the preceding paragraph we first found the number of the
to this was added the
particles meeting the section considered
;
factor r cos
s,
Consequently the
layers,
would be nothing
else
If we
for all the particles that meet the unit area in unit time.
divide this sum by the number of particles that meet the area,
which by
41*
is
^N&, we obtain
the
mean value we
are seeking,
= fr-iQ
M to denote
the
mean value
of
Maxwell's
law,
we may
A
where
is
the
= fQ-W^/B) =
mean
gQ-WfwZ),
speed w.
Since the integration cannot be carried out, we are obliged to
estimate its value by approximations. It seems sufficient to put
for
its
mean
value,
then get
A
and hence,
since
by
= fa~
r~W(o>
= f TrO
M(<**) = G
2
we
),
2
,
obtain
As^O/r^iirfc
where &
is
the
mean
is
in
29*.
We
DIFFUSION
53*
459
the
of increasing
more
If
now we
gases,
and
particles of the second kind pass through unit area in unit time
in the positive direction.
By repeating the procedure employed
in
95,
JN^+tfsStf
is
a constant independent of
x,
we
We have to
where
value of
D were constant,
dt
ancl
we
substitute in this
then obtain
put
54*. Differential
If this
dx*'
Equations
dt
MATHEMATICAL APPENDICES
460
54*
alone.
But
if
consequently with
i
#,
_
~
Afr)dNi\
dt
dx\
v~--' =zjjdx J
But now
_i
2,
and
dx
+ $2
^^2
5#! &?
&?
dx
dx*
dx J
do;\,
^-
'
da?
in
which
JN"
x and
tt
of
The
dD_
fdD _dD\dNi _
dx
\dN
dNj
dD
\dN2
dx
dt
dx*
f dD
^i
dN2 _ n d?Nj
+ rdP
dt
dNj
'
dx
form
dN \dx
dP\fdN<j\*
'
dNJ^dx)
\dN2
dx*
dD\dN^
___
dN
^Ii^D^L
dt
dx*
experiments that
alter with time
dt
nd*N2
dx*
'
'
and
The
and place
If
>
461
APPENDIX VI
CONDUCTION OF HEAT
56*.
Transference of Energy
As
momentum
forward
heat-conduction
is
is
concerned,
we
which denotes the kinetic energy or heat carried across unit area
of a surface in unit time in the direction of increasing x,
Maxwell's
strictly applicable
if
only
19*, p. 388,
is
55*
MATHEMATICAL APPENDICES
462
ft.
Further, the
number
of mole-
must
also vary
us
magnitude, I shall again neglect
of
viscosity
variableness with place, as in our former investigations
that in
and diffusion for the approximation will be permissible
value
mean
constant
a
the short length of a molecule's free path
value.
variable
a
in
reality
of B may be substituted for what is
and k must, however, be both
The two other magnitudes
reason of their continuity we can use
As regards the
last
By
treated as variable.
theorem to the
we can
first
limit
two terms,
may
put
^ dx
instead of
their
dx
ft
to
lies nearer
according as the position (r, s, <)
or further from the origin of the ^-coordinates than the point
to be taken
fo^'*)If
we
of
we then
r,
obtain for
the kinetic energy which passes through unit area in unit time at
the point (x, y, z) in the direction of x the amount
27r
'"
d</>
wherein
s
si
('"ds sin
cos
Jo
last
Jo
and
defined,
is
given by
2=
a/Be-^-**
fl =F (
I
L
+
Lf^
N
dx
(f
^
ft-
w<* 2
cos
)fHr
dx
J
si,
the upper sign corresponding to a flow in the direction of increasing # and the lower to one in the direction of decreasing x.
7
CONDUCTION OF HEAT
56*
463
56*.
N and
For
together, either can be considered as a function of the other.
the complete solution of this problem the determination of their
is necessary.
This can, as Olausius
has
from
the
be
obtained
condition
that the transfer of heat is
taught,
not bound up with a simultaneous transfer of mass.
mutual relation
we make
If
according to Olausius:
(1) The mass of gas which passes in unit time through the
7/-surface in the direction of increasing x must be equal to that
which passes through it in the reverse direction, i.e. of decreasing
x for otherwise the density would alter with the time.
;
(2)
momentum,
the
every value of
direction
x,
increasing x
of
in
the
opposite
direction.
The
just given.
factor
and
\mu
We
is
satisfy
absent
first
the
mw cos s
and
assumed
as,
'*
cxv. p. 32
2.
MATHEMATICAL APPENDICES
464
But we can
ference of heat is
56*
by
of
of
mean
we
is
integral
arrive at a relation
between the
of the form
o
here
= l^- k-^dx*
N dx
"h
which
differential coefficients
and according
to the
latter
The
true value of
integral,
li
must
lie
between these.
It
would then be
practically sufficient
we
if,
differential coefficient of
We should obtain
-1
But
ness,
if
~ W6>
~ rcoss
also possible to calculate the value of h with exactare not afraid of the tedious work of calculating by
it is
we
communicated
to
me
h = 0-71066,
which
gives for
the value
if (2*21066
k~ 1
- wu
2
)
rcos
CONDUCTION OF HEAT
.07"
Conductivity
57-".
The
4i\5
integration, so far as
it
can be carried
du
is
(2-2107 &-
may
constant,
heat
of
in
transfers, which,
the
direction
positive
from
resulting
in accordance
both
is
put
- f d ^ = - $wiN(kmM :dJc-r<lu o
Jx
dx) o
Now,
as
we found
km
19 *,
before in
= 4*-
0- 2
~
47T
C)
m.dk
^
A
T
l
of
expansion
We therefore
-m>
-a
fi
/i
+
i
a(l
To reduce
mass
"J
.
)e"
p. 388,
-
dN-->
air)
^w
aS)j
it
/-' H
/(l
$)
0.
^=
-
aa;
a^
--
f
1
4-
d&
-
5
a$a&
= $7r-W(Jfcw)'(l +
of unit
consequently
Ja?
pressed in
B ~*(2-2107
it
B-^ww
- 2-2107 7r
Awwt
)c"
this expression to
in mechanical units
thermal units
we
is
i.frQ'^fcwO-
1
,
0. /or absolute
11
MATHEMATICAL APPENDICES
466
57"
Hence the
~l
mu*
where
speed
w.
is
the
mean
free
- 2-2107 Ar )*-*"*"*
(mu*
1
2-2107 i- )*-***,
with
path of the particles which move
We find
this integral by
upper and lower limits of the value of
the approximate methods employed in former paragraphs, in which
In this calculation I have
for B or I their mean values.
we
put
not taken the exact value of h, but its approximate value f and
therefore 2-25 instead of 2-2107 and I have obtained
,
<
<
mNClLc c
<
or
0-333
If
we compare
0-818
this determination
of
is
given by
we have
$*W <
< HirV,
1-047 ijc<
<
or
2-570
170,
so that the upper limit coincides nearly exactly with that calculated by Boltzmann.
Since these limiting values are rather far apart, an exact
evaluation of the integral by mechanical quadrature, as in the
case of viscosity,
at
my
left
his kindness.
The
tof,
f= 1-53716
if
is
taken equal
i?c,
was assumed in the first edition of this book but with the
more correct value 0*71066 for h calculated by Conrau we have
as*
1-6027
rjc.
Maxwell,
lies
viz, ft?c
INDEX
The numbers
ABSORPTION, 39
Berthollet, 248
Bjerknes, 351
Boedeker, 59, 139
ether, 129
Aceton, 58, 129, 307, 321,
338
with
prefixed.
Boiling-point, absolute, 97
108, 139-41,
59,
144, 175, 230, 248, 279, 284, 327,
329, 338, 344-6, 351, 365, 367-9,
377, 405, 444-5, 448, 452, 454, 166,
Borda, 69
Acetylene, 126
Boltzmann, 46-9,
335
Aldehyde, 126
Allyl bromide, 126
chloride, 126
Amagat, 14, 17, 108
Ammonia,
192,
15,
198,
57,
300,
319-20, 330
alcohol, 307, 321
isobutyrate, 308
propionate, 308
Andrews, 32, 97-8, 235, 242, 277, 289
Amyl
Annaheim, 336
Ansdell, 319
viii.
Boyle, 12, 14
Boyle's law, passim, esp. 13, 24, 27,
99-103, 364
deviations from. See Gaseous laws
Breitenbaoh, 203, 211, 217
Broch, 25
Brash, 291
Brusotti, 75
Bryan, 26
Buff, 139
Bull, 26
CADMIUM, 119
Gagniard de la Tour, 98
Cailletet, 14, 17, 32, 325
and Colardeau, 68
Mathias, 319
Calderoni, 74
Capstick, 125
Carbon, 303, 306-7, 814
dioxide, 15, 58, 84, 86, 108, 125
129, 138-4, 185, 187, 192-3, 196
203-4, 211, 217, 220-1, 233-44
265-6, 270-1, 273, 275, 287, 289
293-5, 300, 302-3, 314, 319-20
S25-6, 329, 331
disulphide, 15, 126, 129, 821, 338
INDEX
168
Carbon monoxide,
331
tetrachloride, 126
Carnot, 127. 133
Darling, 319
Depretz, 191
Descartes, 349
Cazin, 125
Charles, 65
395
Dixon, 132
Cauchy, 454
330
iodide,
125,138
338
Clausius, passim, esp. 34, 60-1, 70,
107, 118, 131, 151, 158, 161, 279,
326, 328, 333-4, 344, 393, 407, 409,
412, 438, 463
law, 118, 136-7
Coefficient of condensation, 318, 320-1,
210
M^
8rf,
231,
Dobereiner, 40
Dorn, 328-31, 451
Drude, 337
du Bois-Eeymond, 10
Dulong, 123-4, 329
and
214,288,311),
406
Dumas, 349
EJ>DY, 139, 144
Effusion, 77, 84, 92, 114, 247
thermal, 87
Eichhorn, 289
Electrical atoms, 351
Energy, atomic and molecular, 115--15,
285, B94, 400, 404-5
molecular kinetic, 80-1, 117,281,
364-6, 381-2, 385, 392, 398, 405
potential, 3, 167, 374, 394
Esters, 225, 233
338
bromide, 129
butyrate, 308, 322
chloride, 58, 126, 129, 192, 217,
222, 300, 303, 314, 319-20
ether, 58, 126, 129, 232, 307, 321,
338
Ethyhdene
Exner, 329-30
DALTOST, 27, 32, 65, 248
pressures,
149,203,365
31,
92,
05,
Formic
acid,
308 321
}
Fourier, 219, 4 do
Frankonheim, 97
Hoffmann,
Im'.edom, degrees
Presnel, 1213
of,
S3,
50*
'from change of density,
internal.
See Viscosity
49
>
Graham, passim,
esp.
195, 203, 215, 264
Graham's
P.,
IS?
Hofmann, A. W.,
60
#.'5f>
Hydrogen, 15-7,
lOti,
137,
217,
289,
320,
arsenide, 08
bromide, 125, 138
chloride, 15, 58, 125, 129, 138, 192,
phosphide,
f>8
law, 84
Grassi, 345
Greely, 77
Gross, 254
Gugliclmo, 267
Guye, 329
propionate, 308
valerianate, 308
Isovalerianic acid, 308, 321
of,
430
and Lord Kelvin, 7,
IKDEX
4TO
KOE
319-20
chloroform, 126
120, 13fi, 102, 1%, 198,
300, 303, 307, 314
formate, 226-7, 233, 308, 321
ether, 58,
IiALA, 32
Lament, 33
Landolt and Bdrnstein, 306-9, 321
von Lang, 19, 175, 279
Lange, 319
Laplace, 75-6, 102
Lasswitz, 11
Leibniz, 3
Leidenfrost, 277
Light-mill, 213
Linnexnann, 319
Liveing, 26, 145
Lorentz, H. A., 46, 75, 377
Lorenz, L., 340
Loschmidt, 49, 162, 248, 250, 261,
264, 266, 269-70, 273-4, 299, 311-2,
315, 317-8, 322, 328-9
de Lucchi, 124
Ludwig, 94
Liibeck, 343, 345, 369
Lummer and Pringsheim, 125
Lussana, 134
MACHB, 108
Magnus, 28, 93, 277, 288
Maneuvrier and Fournier, 124
161-2, 109,
188-200, 212, 256-60, 324, 407 9,
413, 422-32, 436 \Uu;.u'i.<:c. h ,'
mass, 335
motion, 5
path-volume, 162, 322, 331
sections, 300-4, 332
separation, 333-4, 340
speed, calculation of, 24, 27, 29,
Sec Maxwell's
law
Masson, 123-4
Maxwell, passim, esp. 45, 59,
88, 139,
151, 162, 178, 185, 188, 216, 258,
263, 273, 279, 287, 344, 362, 415,
417, 422, 427, 444
sphere, 102
structure, 304-9
Monatomic molecules,
of speed-distribution,
passim,
44-52, 60, 66, 370, 381-2,
esp.
law
of-
H>.%
collision-impulse, 436
collision-number, 83, 436
19*
56,
distribution of.
Mariotte, 12
law
;J21
Lehnebach, 213-4
Marsh
207
NARK, 288
Natanson, 34, 90, 230, 386
Natterer, 17
Naumann, 111, 139, 299
Navier, 454
138,
471
192,
129,
273,
320,
330
Nitrous oxide, 15, 58, 84, 126, 129,
133-4, 192, 196, 217, 221, 271, 273,
275, 293-5, 300, 303, 314-5, 319-20,
329, 331
isobutyrate, 308
prop ion ate, 308, 322
valerianate, 308
Prout, 349
Noyes, 194
QUINCKE, 337
OBEBBECK, 340
von Obermayer, 189, 192, 216-7, 221,
242, 265-6, 270, 273, 276, 298
Oil, 337
Olszewsld, 320
Osmose, 247, 367
Osmotic pressure, 367
15, 25, 5B-8, 84, 86, 125, 129,
131, 137, 184, 191-2, 196, 205, 221,
270-1, 273, 275, 294-5, 300, 302-3,
Oxygen,
PECLET, 277
Pfaundler, 19
Phosphorus, 126
trichloride, 129
368
9, 37'2
Plank, 288
Plateau, 337-8
mixed
365
75
Priestley, 248, 277
Probability of a particular free path,
157, 159, 408 411
of
Rankine,
3, 99,
349
140
Rontgen, 120, 123-4, 337
Root, 208
Roth, 100, 104, 325
Rubenson, 327
Eucker, 337, 341
Rumford, 4
Roiti, 75,
139
Preston, S.
Ritter, 103
Pierre, 339
Pilling,
RADIOMETER, 213
gases, 31,
T.,
208
14
INDEX
472
Slip, 205,
Smith,
Torricclli, 78,
210
W.
B.,
Smithells, 61
76-7, 123
Sound, propagation
on pressure,
Specific heats, dependence
of, 73,
134
on temperature, 132-7
ratio
of,
120-6,
129,
137-8,
140-4
of action, 102, 154-5, 165,195,
218, 257, 272, 275-6, 310, 312, 314,
322, 337-9
Sphere
Staigmuller, 139
Stefan, 19, 74, 152, 166, 175, 184,248,
257, 263, 273-6, 279, 283-4, 288-9,
292, 306, 327, 417, 450, 454
Steudel, 222, 273
dioxide,
329-30
Sulphuric ether, 129
Sussloff, 438
Sutherland, 166-8, 218-28,272-3, 345,
425
129
VAPORISATION, 37
Vapour, 98, 228
density, 05
Vincent and Dolachanal, 31 J
Vinyl bromide, 126
Violi, 140
Vu-ial, 344
51
Viscosity, 171-246, 271, 43ilVoigt, 34, 98
Vortex atoms, 230, 3i9-51
(
4,
100
8,
Winkelmann,
291, 293,
128,
2!)r>,
Wretschko, 248
Wroblewsld, 319
YOUNG,
S.,
Thos.,
107, 320
3,
338
ZERO, absolute, 28
Tomlinson, 190
Topler, 267, 438
&
0,
240
and Ihmory, 338
Water, 57, 126, 129, 11K5, 190, 233/240,
BOO, 303, 305, 314,321, 330, 338 40,
345-7
Waterston, 10
Watson, 139
Weber, F,, 128
Wiechert, 351
Wiedemann, E., 121, 128, 133, 217,
287, 293, 347, 350
Wilson, 339
'
230,
42,'J
Waitz, 265-6
Wanklyn, 111
Wantzel, 85
145, 230
Thos., 349
See Lord Kelvin
W.
oil of,
Thermo-diffuslon, 88
J. J.,
Traube, 59
Turpentine,
Thomson,
212
438
Square y Londoit
28'J,
CATALOGUE
CLASSIFIED
OF
WORKS
SCIENTIFIC
PUBLISHED BY
MESSRS,
32
HORNBY ROAD.
CONTENTS.
PAOL
30
AGRICULTURE
27
AsTRONoivn
14
BACTERIOLOGY
BIOLOGY
25
25
BOTANY
CHEMISTRY
DYNAMICS
10
-13
8
PHYSICS
...
-
SERIES
SOUND
18
STATICS
WORKS
A,)
HYDROSTATICS
8 TECHNOLOGY
LIGHT
.
LONDON SCIENCE CLASS- BOOKS - 32 TELEGRAPHY TELEPHONE
LONGMANS' CIVIL ENGINEERING
SERIES
13 TEXT- BOOKS OF SCIENCE
MACHINE DRAWING AND DESIGN - 13 THERMODYNAMICS
-
19
METALLURGY
17
14
15
8
WORKSHOP APPLIANCES
-
iz
17
-
-12
iz
ZOOLOGY
MAGNETISM
MANUFACTURES
MECHANICS
17
32
17
PHYSIOLOGY
.25
PRACTICAL ELEMENTARY SCIENCE
GEOLOGY
HEALTH AND HAGIENE
8
5
PHYSIOGRAPHY
PROCTOR'S (R.
PHOTOGRAPHY
12
14
ENGINEERING
HEAT
NAVIGATION
OPTICS
ELECTRICITY
14
18
26
BUILDING CONSTRUCTION
MINERALOGY
NATURAL HISTORY
NAVAL ARCHITECTURE
29
28
24
14
25
Scientific
Works published
by Longmans,
Green,
Co.
CHEMISTRY.
CROOKES.
ANALYSIS,
By
chiefly Inorganic.
Edition, Rewritten
and Enlarged,
Sir
Organic.
By
School Board.
With 65
Illustrations
Lecturer
Third
etc,
net. 8
Inorganic and
London
Ctown 8vo. 2?. 6d.
on Chemistiy,
L-.- i-
By
F, C.
GARRETT, M.Sc.
JAGO.
Works by W. JAGO,
INORGANIC
THEORETICAL
CHEMISTRY,
PRACTICAL. With an
F.I.C
F.C.S.,
AND
crown
INORGANIC
svo.,
is.
6d.
THEORETICAL
CHEMISTRY,
PRACTICAL
AND
4J-.
6d.
KOIBE.K
SHORT
CHEMISTRY.
By
HUMPIDGE, Ph.D.
Dr.
TEXT-BOOK
HERMANN KOLBE.
With 66
Illustrations.
OF
INORGANIC
Crown
Illustrations.
vols. 8vo.
36^.
71 Woodcuts.
INWith
Scientific
CH E ISTR Y
Continued.
IV!
MUIR.& COURSE
M,
P.
OF PRACTICAL CHEMISTRY.
College, Cambridge.
Part
I,
in
Chemistry
of
By M.
(3 Parts.)
Crown
Elementary,
Part II,
Co.
i3
Greeti^
Crown
Intermediate.
Part III.
\Inprtparation.
NEWTK
Works by G.
strator
Examiner
the
in
in
CHEMICAL
LECTURE
Cro\\n Svo.,
Diagrams.
EXPERIMENTS.
With
224
icw. 6d.
Cro\\n 8vo.
A TEXT-BOOK: OF INORGANIC
Crown
Illustrations.
6s
6d
CHEMISTRY
SOLUTIONS.
Chemistry
in the
additions,
of
with 146
Crown
Svo.,
With
108
ax 6d
By W. OSTWALD,
Professor
of
University of Leipzig.
Being the Fourth Book, with some
the Second Edition of "Oswald's
Lehrbuch der allgememen
'
Chemie'.
in
RE YNOLDS. EXPERIMENTAL
JUNIOR STUDENTS.
By
J.
Part
I.
Part II.
Part IV.
Fcp. Svo
i*f.
6d.
6cf.
Carbon Compounds.
by
Chemistry
W.
Fcp. 8vo.
A.
45-.
SHENSTONE,
Lecturer on
in Clifton College.
FOR
F.R.S., Pro-
for Acids.
Part III.
Fcp. Svo.,
Introductory.
CHEMISTRY
PRACTICAL
With 42
Illustrations.
8vo.,
is.
6d.
INTRODUCTION TO CHEMISTRY.
and
Scientific
Works published
by Longmans,
CH E M ISTR Y- Continued.
WW
ON VOLUMETRIC
THORPE.
THORPE,
Vols.
I.
and
Vol
each
42^
II.,
THORPE
ALYSIS
Fcp. 8vo.,
57
Woodcuts
TILDEN.
vols.
With 88 Wood-
6d.
4-y
D.Sc., F.R.S.,
III, 63^.
and M.
Fcp. Svo
Works by
PATTISON
6d
3.9.
Mum,
WILLIAM
A.
MA
CHEMICAL ANT.
E THORPE, Ph
I).,
Plate of Spectra
.md
With
TILDEN,
D.Sc. London,
South Kensington.
PHILOSOPHY.
With
Woodcuts.
PRACTICAL CHEMISTRY.
Analysis.
HINTS
Fcp. Svo.
is
THE
ON
The
principles
of Qualitative
6d
TEACHING
ELEMENTARY
OF
\Vorte
LLOYD WHITELEY,
by R.
CHEMICAL CALCULATIONS.
Problems and Answeis, specially adapted
Schools
With a Preface by Professor F.
Cro\vn Svo.,
With
foi
use
K.I.C,
West Bromwich.
Explanatory
CUMKS,
Notes,
2s,
ORGANIC CHEMISTRY:
Illustrations.
With
Crown
Svo.
3,5-.
the Fatty
6d.
Compounds.
With 45
Works published
Scientific
Green,
by Longmans,
Co.
PHYSICS, ETC.
Christ's College,
Cambridge.
Crown 8vo
Questions.
6J
,4*
GANOT.Vfwte
Translated and
by PROFESSOR GANOT.
ATKINSON, Ph.D., F.C.S.
ELEMENTARY TREATISE ON PHYSICS, Experimental
Edited by E.
of Questions
GLAZEBROOK
T.
R,
.u D SI1A
Kcp. 8vo.,7s
7 Plates, 624
6i/.
Jr.
FRS,
GLA/RKROOK, M.A.,
Woodcuts.
With
j\
PRACTICAL PHYSICS.
and
N.
\\
With
SHUV, M.A.
With
GuiHRiK, Ph.D.
HELMHOLTZ.
SUBJECTS.
By
134
6</.
Fop. 8\o
Diagrams.
By
6</.
Translated by E. ATKINSON,
Ph. D., F C.S formeily Professor of Experimental Science, Staff College. With
2 vols croun 8va, 3*. 6J. each.
68 Illustrations.
By
CONTLNTS
ff
ELEMENT AR\
tfjDJSJS SON.
HENDERSON,
33.Sc.
Polytechnic.
Crown
I.
(Eclin ),
Svo., 2*. 6d,
ME.,
PHYSICS.
Treatise, \vitli
MKYKR, IVoiessoi of
tal)
Edition.
Tree's
E B
\YNES.
Re.ider in Physics.
M.A
Student
JOHN
Moiough Road
CASES.
By Dr
Appendices.
Translated In ROISKRT
O\fuid, and DI
By
Physics Department,
ElemenOSKAK EMU,
Second
Revis,ctl
of Christ
Chmch,
8\o.
By
J.
IN
a Preface
University of Leipzig
&
Scientific
Co.
PHYSICS,
WATS
N.
W. WATSON,
Laboratory Manual for Use in Organised Science Schools. By
B.Sc. Assistant Professor in Physics in the Royal College of Science, London ;
With 120 IllusAssistant Examiner in Physics, Science and Art Department.
trations
Crown
WORTHINGTON.K
6d.
2s.
OF PHYSICAL
FIRST COURSE
LABORATORY PRACTICE.
WORTHINGTON, M.A., F
8vo.
R.S.
WRIGHT. ELEMENTARY
PHYSICS.
MARK
By
By A. M.
45. 6cf.
R.
College of Science.
BALL.K
CLASS-BOOK OF MECHANICS.
BALL, LL.D.
By
M.
by T.
Sir
R.
S.
C. GELDVRI>,
Crown
GOODEVE.Wo$
By
6d.
AND DYNAMICS.
STATICS
\s.
Fcp. 8vo.,
89 Diagrams.
GOODEVE,
Svo., 5?.
M.A.,
formerly
Crown
Svo.
With 357
Wood-
6^.
PRINCIPLES OF MECHANICS.
numerous Examples.
Crown
Svo.
6s.
A MANUAL OF MECHANICS
for Students of Applied Mechanics.
an Elementary Text-Book
With 138
LESSONS
By W. H, GRIEVE,
School Board,
Stage
i.
is.
Stage
Stage
2.
3.
With
Professoi of
Kngmm ing
With 620
Illusitaiions
Crown
IN
ELEMENTARY MECHANICS.
late Engineer,
London
etc.
165 Illustrations
6d.
With 122
With 103
Illustrations.
Illustrations.
is.
is.
6d.
Fcp, Svo.,
Scientific
Co.
MAGNUS.
B.Sc., B.A.
MAGNUS,
LESSONS IN ELEMENTARY MECHANICS.
Introductory
Key
Teachers only,
pi ice 55.
3^.
with Answers,
ROBINSON.
The Worked
ss.
J.
ELEMENTS OF DYNAMICS
A
numerous Exercises
Fcp, 8vo.,
L.
is.
ROBINSON, M.A.
(Kinetics
and
Statics).
Crown
Cicmn
Sold separately
Statics, 2<
D}namics,
Svo,
With
Svo., 6s.
6d.\
2S.
With
3^. 6tf.
ss.
6*
By FRANCIS A. TARUETON,
TAYLOR.
Cioun
Svo.,
icxs.
MA,
Works by
Head Master
6d.
B.Sc. (LondL),
E. TAYLOR,
J.
of the Central Higher Grade and Science School,
Sheffield.
THEORETICAL MECHANICS,
Pneumatics.
With
175
including
Diagrams and
Crown
8vo.
Illustrations,
ai. 6d.
Papers, etc.
and
With 103
Illus-
THEORETICAL MECHANICS.
trations,
Hydrostatics
FLUIDS.
With 122
Illus-
amination Papers,
etc.
Crown
Scientific
Co.
TWISDEN.WQs
PRACTICAL MECHANICS;
their Study.
8vo IQS 6d.
JOHN
by the Rev.
With 855
F.
TWISDEN, M.A.
an Elementary Introduction
and 184 Figures and Dugiams
Exercises,
to-
t'io\\n
MECHANICS,
THEORETICAL
With
Examples
172
Cro\\n 8vo.,
8i.
6J
WILLIAMSON. INTRODUCTION TO THE MATHEMATICAL THEORY OF THE STRESS AND STRAIN OF ELASTH
By BENJAMIN WILLIAMSON,
SOLIDS.
F.R S
D.Sc.,
Cro\\n Svo.,
'
5*.
\\ith
K.R.S., and
Sc.,
Crown
F. R.S.
Svo.
an
Ele-
A M. WORTHINGTON, M.A
By
45. 6d.
ABNEY.h
TREATISE ON PHOTOGRAPHY.
With
115 Woodcuts.
By R. T.
MA., F.R
Woodcuts
Q^ICKL PROJECTION
WRIGHT.
ami
GLAZJSOOjS:.~PUYSICAL OPTICS,
BROOK,
By Captain
LnerpooL
C^LA/KWith uS^
Crown
Svo.
232 Illustiations.
6s.
With
192 Illustrations
DA tf NUMERICAL EXAMPLES
DAY, M.A.
Fcp. Svo., y.
6d
Crown
IN HEAT.
Svo., 41.
By
Dv
6</.
R,
E.
Scientific
Co.
<S
EMTAGE.LIGWT.
By W. T. A.
Crown
232 Illustrations.
HELMHOLTZm
Svo.,
EMTAGE, ALA,
Royal Svo., 28
MADAN.S& ELEMENTARY
For the Use of Schools. By H G. M
late Assistant Mastei
College, Oxfoid
;
By HER-
i.
TEXT-BOOK ON HEAT,
UHN, M.A
at
Crown Svo
Eton College.
With
6x
By
J.
9^.
CLERK MAXWELL,
F R SS., L. and E. \Yith Corrections and Additions by Lord RAYLEIGH. With 38 Illustrations, Fcp, 8\o,.p 6</.
,
SMITH.
Cams
TYNDALL.\!W\&
By
College, Cambridge.
by
J. HAMBLIN SMITH,
down Svo.,
3-*.
JOHN TYNDALL,
F.R.S.
D.C.L.,
See p. 28,
WRIGHT.
VI v&s> by
SOUND, LIGHT,
Illustrations.
B.iSc.,
M.A.
Fcp. Svo.,
is. 6</.
Ciown
ADVANCED HEAT.
With
numerous
and
Rules
Engine Drivers and Boiler Attendants, with Notes on Steam Engine and
Boiler
M.I.M.E., A.M.I.C.E.
GAS AND
C'LEKK, Associate
Chemical Society,
Patent Agents,
Member of
Member of
With 228
JfOLJfJS.S.Tti.E
Boiler Explosions.
as. 6d.
By M. Povvis BALE,
Fcp. 8vo.,
OIL
ENGINE.
By
DUGALD
Illustrations.
Svo., 15$.
STEAM ENGINE.
By
GEORGE
of the Institution of
Naval
C.
V.
Architects.
io
Scientific
Co.
AND CAS
NORRIS.^ PRACTICAL TREATISE ON THE 'OTTO'
With
STEAM
OIL,
RIPPEJt.
207
Illustrations.
Works by
WILLIAM RIPPER,
Professor of Mechani-
STEAM.
With 142
Illustrations.
Crown
Svo., 2S.
6(f.
BUILDING CONSTRUCTION.
'
Rivingtons'
Crown
Svo.
4^.
'.
the Author
By
With 385
Illustrations.
6d.
By EDWARD
J.
BUILDER'S
TRADES.
Medium
By
Col.
8vo., i6s
Arranged
to
meet the requirements of the syllabus of the Science and Art Department of the
Committee of Council on Education, South Kensington. Medium Svo.
Part
I.
Elementary Stage.
Part II.
Advanced
Part III.
Stage.
Materials.
Course
for
Honours.
los. 6d.
los. 6d.
cuts, 2IS.
Part
IV.
Honours.
Structures.
Course
for
Scientific
Green^
Co.
1 1
Cm icnt
Motor.
DA Y. EXERCISES
MEASUREMENTS,
DU
IN ELECTRICAL
with Answers.
By
R.
E DAY.
i2mo.,
AND PRACTICE.
Berlin.
AND MAGNETIC
3*.
6d.
THEORY
IN
EBERT.
With 93
By H. EBKRT,
Illustrations.
Svo.,
l>y
JfjEJffjOJSJKSOJViPTSiACTlCAL ELECTRICITY
NETISM. By JOHN HKNDKRSON, D.^SC. (Edm,), A.LE.E,
trations
and Diagrams.
Cro\\n Svo
including
With
Illus-
6s, 6</.
TY AND MAGNETISM.
JKNKIN,F.R.S.S.,L andE.,M.I.C E. With 177
Illustrations.
By FLEEMING
Fcp
8\o.,s?.6tf.
AND MAGNETISM.
Founded on JOUHEKT'S Trait6 itlementaire d'ElectiiBy G. C. FOSTER, F.R.S., and E. ATKINSON, Ph.D. With 381 Illustrations. Crown Svo
js 6d.
(
cit(5'.
LAjRDJSN.
Crown
Svo.,
y>
2s.
Cr. Svo.,
6s,
MEIJ?IJSLZ>.
2s, 6tt.
235 Illustrations.
6d.
Crown
Svo.
With
45. 6<f.
Cr. Svo.
as.
12
Scientific
Co.
With 7
etc.
Properties,
1-
6d
35,
PREECE
H. PKEKCE, K.C B
F R.S V.P
hist.,
Sir J
trician,
Sir
W.
With 267
By
E., etc.,
Fcp. Svo.
Illustrations
6s.
a Description of Details
of Railway Construction
subsequent to the completion of the Earthworks and
Miucmreb.
B\ Sir JOHN WOLFE BARRY, K.C B., F R.S,,
I.C.E
With
218 Woodcuts.
Fcp Svo
6t/.
4^
Wh
Son
A MI C E
M.I
K,
Piofessoi of
Engmeeiing
With 620
Illustrations
W. ~~
POCKET-BOOK
GIXEERS.
Professor
specially
prepaied
Illustiations.
'Fcp, 8\o
gilt
edges,
T.
GRAPHICS,
or the
Art of Calculation
by Drawing
H. SMITH,
I.
With
with
Strength and other Properties of Materials
F.R
S., M
I.C.E.
With
5 Plates
and 143
Practical
IN GIRDERS
Observations
By BINDON B STONEY,
Illust. in
the Text.
Royal
on
the
LL
D.
Svo., 36^.
oi
STEEL, AN D TIMBER.
ioj. net.
CONSTRUCTION IN IRON,
Lectures
of Sydney.
Challis Professor
With
iq Folding
Scientific
LONGMANS'
TIDAL RIVERS:
'
their
i6j
13
'.
Improvement,
Navigation
Medium 8vo
Co.
ENGINEERING SERIES.
CIVIL
&
With 75
(3)
Illustrations.
net.
Deputy
Medium
Illustrations.
Svo.,
Civil
21.1.
Engmeer-m-Clnef, Admiralty.
net
PRINCIPLES
STRUCTION
C E. and Executive
With 97 Illustrations.
M.Inst
Medium
C.
With 365
Svo.
155. net.
HYDRAULIC ENGINEERING:
CALCULATIONS IN
Practical
T.
Dundee
Fluid Pressure and the Calculation of
College,
Part
I.
With numerous
gineering Structures
its Effects in
and Examples.
Illustiations
En8vo.
6d net
6s.
RAILWAY CONSTRUCTION.
By W. H. MILLS, M.I.C.E.,
With 516
Illus-
NAVAL ARCHITECTURE.
Manual
for Students of Science Clashes and Draughtsand Naval Architects' Drawing Offices. By
EDWARD LEWIS AIT WOOD, Assistant Constructor, Royal Navy, Member of
the Institution of Naval Architects.
With 114 Diagrams Gxwn Svo., yjr. 6</.
men Engaged
in Shipbuilders'
Naval Architecture
With numerous
at the
Durham
Illustrations.
Royal 8\o.,
155. net.
Works by DAVID ALLAN LOW, Professor of EngineerEast London Technical College (People's Palace).
LOW AND
AND
With 153
Illustrations
W. CAWTHORNE UNWIN,
General Principles,
With 304 Diagrams,
II.
Fcp. Svo.
Chiefly
,
4J. 6d.
OF MACHINE DESIGN. By
F.R.S.
Machinery.
Part
Svo, QS 6d.
UNWIN.TEE ELEMENTS
I.
Crown
DESIGN.
M.I.Mech.E.
Part
and Diagrams.
on Engine
Fastenings,
etc.
and
Transmissive
Fcp. 8vo., 6s
Details.
14
&
Scientific
Co.
ETC.
WORKSHOP APPLIANCES,
By W. H. NORTHCOTT.
and Ornamental.
338
8vo,, iBs.
tions of
DESCRIPTIVE MINERALOGY.
Diagrams.
Woodcuts and
With 373
6s.
Fcp. 8\o.,
Fcp. 8vo.,
6s.
By
all
G.
including
GORE, LL.D., F.R.S.
their Properties
and Treatment. By A. K. HUNTINGTON, Professor of Metallurgy in King's
Lecturer on Metallurgy in Mason's
College, London, and W, G. M'MiLLAN,
With 122 Illustiations. Fcp. 8vo., ?s. 6d.
College, Birmingham.
:
An Elementary Text-Book. By
^JKB^(J3. METALLURGY.
E. C RHEAD, Lecturer on Metallurgy at the Municipal Technical School,
Manchestei
With 94
Illustrations,
Fcp. 8vo.
y.
By
F.
RUTLEY, F.G.S.
With
6d.
and 88 Woodcuts.
grams.
Fcp. Svo.
6s.
A CLASS-BOOK OF ASTRONOMY.
Fcp. 8vo.,
is.
6d.
Dia~
6d.
With 41 Diagrams.
Scientific
ASTRONOMY, NAVIGATION,
CLJEJUZJS.THE SYSTEM OF THE STARS.
CLERKE.
With
6 Plates,
and numerous
DE CAMPIGNEULLE&
DUMRAON, EEHAR,
Illustrations
By AGNES M.
OBSERVATIONS TAKEN AT
INDIA, during
Svo
IQJ
SVO.
Bart.,
H.,
etc.
AND NAUTICAL
Master of the Liverpool
6d.
15
8vo., zis.
F.
Co.
With
9 Plates
By
Sir
JOHN
12.9.
LOWELL. MARS.
Asiatic Society,
etc.
6d.
R.
MARTIN, R.N.
Royal
8\o., iSs.
For
Use of Students
By J. MERRIFIELD, LL.D., F.R.A.S., F.M.S.
Charts and Diagrams. Crcwn Svo. 5^
the
With
WEBB.
By GEORGE
With 84 Diagrams.
Svo.
SCOPES.
By
the
Rev
W. WEBB,
T.
With Numerous
PARKER, M.A.,
of
55. net.
COMMON
M.A., F.R.A.S.
Fifth
TELEEdition,
tions.
Crown
Svo.
6s. 6d.
WORKS BY RICHARD
A.
PROCTOR.
THE MOON:
the
Plurality
Map,
Charts, etc.
Crown
Abode
Svo.
Worlds
Illustrations
3$. 6d.
INFINITIES
in
of
With 14
a Series of Essays conSpace and Time with the Infinities around us.
:
[OVER.
Scientific
&
Co.
8vo.,
3 s &d
Scientific Subjects.
I.
Natural Phenomena,
Crown
Cheap Edition
etc.
3 vols
Familiar Essays
,
US
Moon and
Essays on the
5^,
each.
Firmament
crown 8vo
Essays on the
Crown
Planets,
8vo.
6d.
3*.
Wonders of the
6d
35-
means
of
a Popular Guide
Amusement and Instruction. With
LIBRARY,
and
the School,
the Obsenatory,
Twelve Circulai Maps (with Two Index-Plates). With an
i ntroduction on the
Study of the Stars. Illustrated by 9 Diagrams. Cr. 8vo. 5 ?,
,
THE SOUTHERN
SKIES:
a Plain
a Plain
the position of
the pimcipal Star-Groups night after night throughout the year
With Introduction and a separate Explanation of each Map.
True for every Year,
410.,
Knowledge
3..?.
of the Constellations.
Showing
in 12
Maps
6d.
With
2 Index-Plates
THE STARS
IN
THEIR SEASONS:
Knowledge
Crown
8vo.,
STUDIES.
Imperial 8vo.
Familiar
to
5.?.
Essays
on
y. 6d.
NATURE
an Easy Guide
In 12 Large Maps.
etc.
Folio, 15?.
Crown
8vo.,
y.
6d.
LEISURE READINGS.
WILSON, T. FOSTER, and A.
By R.
C.
Crown
A. PROCTOR, E. CLODD,
RANYASD. Crown 8vo., jr. M.
A.
Scientific
Co.
17
By
F. T.
'
8vo., 12^
net
LUPTON. MINING.
MORRIS
AND WILKINSON. Y}I& ELEMENTS OF COTTON SPINNING B> JOHN MORRIS and F. WILKINSON. "With a Preface
by Sir B. A. DOBSON, C.E., M.I.M
Crown
Svo.
SHARP.
Illustrations.
js 6d. net.
AND TRICYCLES:
BICYCLES
an
Elementary
TA YLOR.
JOHN T. TAYLOR.
WATTS.
Crown Svo
js.
By
6d net
GROWERS.
Crown Svo
By FRANCIS WATTS,
F.I.C.
F.C.S.,
With 20
Illustrations.
6s.
BIRD.
British Isles,
and 247
Illustrations.
ADVANCED GEOLOGY. A
Crown Svo
Manual
for
of
the
2s. 6d.
Students in Advanced
Crown
>GJtEJEJ\r.PR\$lCA.L
GENERAL READERS.
tiations.
MORGAN.
8VO.,
By
M.A F
H. GREEN,
With 236
G.S.
Illus-
2I.J.
ELEMENTARY PHYSIOGRAPHY.
By ALEX. MORGAN, M.A.,
Treated
F.R.S.E., Lecturer in
Mathematics and Science, Church of bcotland Training College, Edinburgh.
With 4 Maps and 243 Diagrams. Crown Svo., 2r. 6d.
Experimentally.
D.Sc.,
of the
trations.
New
(for Section
With 215 Illus-
Crown
Svo.
2s. 6d.
ELEMENTARY PHYSIOGRAPHY
an Introduction to the
Study of Nature. With 12 Maps and 247 Illustrations. With Appendix on
Astronomical Instruments and Measurements. Crown Svo 2s. 6d,
:
ADVANCED PHYSIOGRAPHY.
Illustiations
Crown
With 6 Maps
and
203
Scientific
&
Co,
ASHJBY.
Lecturer
the
Diseases
Crown 8\o
Illustrations
6J
31.
HEKLTm
BUCKTON.
THE HOUSE;
IN
By Mrs.
Crown
COJVELD.THE LAWS
MA
FIELD,
M.D.
Fcp 8\o
C.
With
41
8vo., 2*.
By W. H. COR-
OF HEALTH.
,
Twenty- five
M. BUCKTON.
6d.
is.
HOTTER
6d,
8vo., 2S
POORE.WV&S
ESSAVS
THE DWELLING-HOUSE.
Svo
Crown,
Illustrations.
X MANUAL OF HEALTH-SCIENCE:
WILSON.
With 74
With 36
6d,
31.
and Colleges
Illustrations.
Crown
adapted
S., etc.
6d
NATURAL HISTORY.
STRUCTURE AND CLASSIFICATION
OF BIRDS.
By FRANK E. BEDDARD, M,A,, F.R.S Prosector and ViceSecretary of the Zoological Society of London. With 252 Illustrations.
Svo.,
,
2is. net.
book
Text.
;
or,
of which are coloured,
LIFE IN PONDS
and 331
AND STREAMS.
Crown
and 549
Illustrations in the
and 241
Illustrations in the
(British).
Text
BRITISH BIRDS.
*js.
With 12 Coloured
6d.
By W. H. HUDSON, C.M.Z.S.
With 160
Illustrations.
By E.
Crown.
Scientific
&
Co.
19
ASHJ8 Y.
By HENRY ASHBY,
DISEASES OF CHILDREN,
D. Lond. F. R. C. P.
By HENRY ASHBY,
Physician to the Geneial Hospital for Sick Children, Manchester, and G. A,
Oxon F R.C.S Eng., Assistant-Surgeon to the ManWRIGHT, B A.,
chester Royal Infirmary, and Surgeon to the Children's Hospital.
Enlarged
and Improved Edition.' With 192 Illustrations. 8vo., 2^.
MB
WILLIAM
JBJSJVNETT.Woik.s by
H.
BENMETT,
F.R.C.S.,
8vo,, $s
CLINICAL LECTURES
chiefly
ON
grams
ON ABDOMINAL HERNIA:
m relation
With
12 Dia-
an Address delivered
Inaugural Meeting of the Nottingham Medico-Chirurgical Society,
Session 1898-99.
Systematically,
M. R. C. S. Eng.
CA30T.
St.
the Text.
Works by
Crown 8vo
RICHARD
6d.
dr.
CABOT,
C.
M.D., Physician
and 9
With
3 Coloured Plates
and
8vo., i6f.
Illustrations.
DISEASE.
Royal
With 31 Tempera-
Part
I.
The Treatment
(Lond),
etc.
Inflammation, Suppuration, Ulceration, Gangrene, Wounds and their Complithe Administration of Anaesthetics,
cations, Infective Diseases and Tumours
With 66 Illustrations. Royal 8vo. ics. 6d.
;
20
Scientific
CLARKE.
Works by
Assistant
F.R.C.S.,
M.B. Lond.,
London and
JACKSON CLARKE,
J.
the North-west
at
Surgeon
Illustrations
POST-MORTEM EXAMINATIONS
AND ORDINARY CASES
and on
of Post-mortems,
IN
MEDICO-LEGAL
With
M,D
With 194
6d
10^.
By JOSEPH
Third
Reused throughout.
Edition
COOXE.
TABLETS OF ANATOMY.
tions given
the Westminster Hospital Medical School.
Eleventh Edition
,n Three Parts, thoroughly brought up to date, and with over 700 Illustrations from all the best Sources, Batish and Foreign.
Post 40.
Part
I.
Part II.
Part
Brain.
6d
IQS.
net.
etc
DISSECTION GUIDES.
such Practical work
an
Aiming
at
Anatomy
A HANDBOOK OF MIDWIFERY.
RADFORD DAKIN, M.D
Midwifery
at
St.
F.RC.P.,
George's Hospital,
connection with
By WILLIAM
and Lecturer on
With 394 Illustrations.
Large
Obstetric Ph>sician
etc.
DICKINSON.
Wo&*
by
W.
HOWSHIP DICKINSON,
to,
M.D.
at,
St.
THE TONGUE
Three Parts
Svo
^3
49.
With
12
6d.
AN INDICATION OF DISEASE;
AS
March, 1888.
Svo., 74.
&
8VO.,
i8<q~
JJ
Scientific
Green, a* Co.
21
DUCKWORTH.
DUCKWORTH,
etc.
JUDGMENT
IN
the Haneian Ouition, delivered before
the Royal College of Physicians, iSth Octobei, 1898
Post 4to. 2s. 6d.
,
a Treatise on Surgical
Injuries, Diseases,
ERICHSEN, Bart., F
Surgeon Extraordinary
H.M
the Queen.
Engravings on Wood.
Royal
8vo.
late
FOWLER
Illustrated
by nearly 1000
48^.
GODLEE.
AND
LUNGb
to
2 vols.
GARROD.
Works
by
SIR
M.D.,
F.R.S.,
etc.,
(RHEUMATOID ARTHRITIS).
Enlarged.
With 6
Plates,
The Thirteenth Edition, Revised and Edited, under the superby NESTOR TIRARD, M.D. Lond., F.R.C.P., Professor
Matena Medica and Therapeutics in King's College, London, etc.
Crown 8vo., 12? 6d.
of
By
HENRY GRAY,
shown
22
Scientific
Co.
HALLIBURTON.
HALLIBURTON,
^Gt\^ by W. D.
M.D.,
London.
8vo., 28^.
104 Illustrations.
Illustrations
With 77
5.?.
L-4WG.
EYE
Being Part
Guide
of a
I.
to the Practice of
Practitioners
HANDBOOK ON
L/7ELVG.
tor Students
Ophthalmology
the Royal
to
Illustrations.
THE
DISEASES OF
SKIN.
foi
Fcp. S\o.,
5.?.
ZZ7SK
TOXICOLOGY,
in
33 Illustrations
m the Text.
2 vols.
Crown
8vo,, 24^.
NEWMAN.
KIDNEY
Practitioners.
SURGERY
PICK.
By
Fcap. 8vo.,
By W. H. and
2s. 6d.
Including
Women
and Children.
etc.
2 Vols.
Medium
&
Scientific
Co.
23
7:
SPECTIVE SUBJECTS.
VOL
PART
I.,
EMBRYOLOGY,
I.
By E A. SCIIAKER, F.R
200 Illustrations
VOL. L, PART
Royal
Svo.
With
S.
,
GENERAL ANA-
II
VOL.
TOMY OR HISTOLOGY.
SCHAFER, F
A.
Illustrations.
By E.
With 291
S.
Royal
Royal Svo.
PART
By E. A.
With 178 Illus-
SCHAFER, F.R.S
trations.
Royal
Svo.,
VOL.
Royal 8vo
PART
gs.
SPLANCH-
IV.
ByE. A. SCHAFER,
F R.S ,and JOHNSON SYMINGTON,
M.D. With 337
Ill
NOLOGY.
With
gs.
THE ORGANS
III.
OF THE SENSES.
II
PART
OSTEOLOGY
ARTHROLOGY ByG D THANE
With 224 Illus Royal 8vo us
VOL IT., PART II
MYOLOGY
VOL
Ill
With 102
THANE.
Illustrations
9$-.
THE NERVES.
II.
By G.
Illustrations.
Royal
8vo., i6j
SUPERFICIAL AND
SURGICAL ANATOMY
By
Professor G. D THANE and Pro-
SCHAFER, F R.S.
fessor R. J.
APPENDIX.
i6s.
VOL.
PART
III.,
trations.
THE SPINAL
I.
Royal
By E. A.
With 139 Illus-
8vo.,
i2s. 6d.
With
GODLEE, M.S.
29 Illustrations.
6s. 6d.
Royal Svo.,
Descrip-
and
tive
Practical.
'
'.
SCHENK.UA'NUAL OF BACTERIOLOGY.
tioners
S. L.
and Students.
SCHENK,
Univ. Dub.
late
Illustrations,
SMALE
some
Practi-
By
Dr..
Trans-
For
With
AND COLYER.
AND INJURIES OF
DISEASES
THE TEETH,
a Manual of Practical
including Pathology and Treatment
By MORTON SMALE, M.R.C.S.,
Dentistry for Students and Practitioners.
L.S.A., L.D.S., Dental Surgeon to St. Mary's Hospital, Dean of the School,
Dental Hospital of London, etc. and J. F. COLYER, L.R.C.P., M.R.C.S.,
L.D.S., Assistant Dental Surgeon to Charing Cross Hospital, and Assistant
Dental Surgeon to the Dental Hospital of London. With 334 Illustrations.
Large crown
SMITH
'(JST.
8vo., 15^.
F.}.
MIDVVIVES.
41 Woodcuts.
the
By
Mechanism of
their
TAPPEINER.
INTRODUCTION
TO
CHEMICAL
By Dr. H. TAPPEINER,
Professor of Pharmacology and Principal of the Pharmacological Institute of
the University of Munich. Translated by EDMOND J. McWEENEY, M. A, M. D.
(Royal Univ. of Ireland), L.R.C.P.I., etc. Crown 8vo., 3* 6d.
,
24
Scientific
TIRARD.
MD
Co
By NESTOR
TIKARD.
King'b
King's College
to the
of
at
E\elma Hospital
\VALLEK.~ Works
AUGUSTUS
by
WALLER,
D.
M.D.,
Reused
Kd'tion,
Illustrations.
LECTURES ON PHYSIOLOGY.
8\o.,
Electricity
Third
8\o., i8r.
On Animal
First Series.
59. net.
of the
Nenous System
By AUGUSTUS
II. in
the press.
8\o
Electro-Physiology
2s.
Part
and
D WALLER
I.
6d
net.
With
ANTON WEICHSELKAUM,
U7LKS
A.YD
^O^OM
ANATOMY.
By
Sir
Sir
SAMUEL WILKS,
sicians,
WALTER Moxox,
on Pathology
at,
LECTURES ON PATHOLOGICAL
Bait., M.D., F.R.S., President of the
and Physician Extraordinary to H. M. the Queen, and
R
F
C
P.,
D.,
Physician to, and some time Lecturer
Guy's Hospital
SAMUEL WILKS,
Bart.,
M.D.,
LL
D.,
F.R
S.
By
8\o., iSs,
late Professor of
Bombay
and Students
Manual
of
Bovine Pathology.
and Students.
being
With
being a
With
bein^
a Manual of Ovine
Pathology for the use of Veterinary Practitioners
With Coloured Plate and 99 Woodcuts. 8vc i2s
and
Students.
etermary
WTZWYGHAM
General Sir F.
2s.
Room.
Manual
Crown
FITZWYGRAM,
Bart.
With 56 pages of
Svo., js
6tt.
By Major-
Illustrations.
Svo
Works published
Scientific
by Longmans,
Gteen,
Co.
25
VETERINARY MEDICINE,
^ ANGORA GOAT
1
auspices of the South African Angora Goat Breedeis' Association), and a Paper
on the Ostrich (leprmted from the Zoologist foi March, 1897). By S. C. CRON-
\\RIGHTSCHREINER.
With 26
EASE
By STONEHENGE
'
'.
8\o
Illustrations.
'
With 78
Wood
8vo
Engravings.
THE DOG.
6d
109.
HEALTH AND
IN
W.
by
DIS6d.
?s.
WATSON,
6d.
Wood
With 33
En-
8vo., 6s
gravings
ZOOLOGY.
ASHBY.
M.D
With
Fcp 8vo
141 Illustrations.
55.
By Mrs.
S.
A. BARNETT.
a Children's
With 113
Illus-
by
FRANK
E.
BEDDARD,
M.A. Oxon,
Crown
SV
With 93
Illustra-
2s. 6d.
8vo
Illustrations
BIRDS.
sir, net.
8vo.
By JOHN BIDGOOD,
,
B.Sc.,
With 226
F.L.S.
Illustrations,
4$. 6d.
for Schools.
ODI.~THE
ESSENTIALS
Fcp. 8vo
in
is.
OF EXPERIMENTAL
PHYSIOLOGY.
To-
Diagrams.
N
R G.S
With 218
PHYSIOLOGY.
Illustrations.
Crown
8vo.,
2s.
By W. FURNEAUX,
6d.
26
Works published
Scientific
Green,
by Longmans^
&
Co.
OS.
HUDSON
6d. each.
4*.
By Surgeon -Captain
8vo.
MACALISTER.
B.
H.
S.
LEUMANN, M.B.,
Indian
3-f.
ALEXANDER MACALISTER,
Write by
M.D.
With 41
Diagrams
Fcp
8vo.
6a
is.
Diagrams.
is,
8vo.,
With 77
6d.
J/0Oi ELEMENTARY
MA
With
PHYSIOLOGY.
By
BENJAMIN
Mo<jRE,
8\o
late
6i/
MORGAN ANIMAL
BIOLOGY
8s.
SCHENK.
an Elementary Text-Book.
With
6d
MANUAL OF BACTERIOLOGY,
for Practitioners
and Students, \uth Especial Reference to Practical Methods. By Dr. S. L.
SCHENK. With 100 Illustrations, some Coloured. 8vo., IQJ, net.
some Coloured.
By JOHN THORNTON,
Crown
8vo., 6s.
BOTANY.
ELEMENTARY TEXT-BOOK OF BOTANY.
AITKEN.
By EDITH AITKEN,
Crown 8\o. 4.?. 6d.
late
BENNETT
AND MURRAY
GAMIC BOTANY. By ALFRED W.
on Botany at
St.
of Botany, British
CROSS
AXJ>
Museum.
.F^K
With 378
Illustrations,
CELLULOSE:
8vo., i6s.
an
Outline
of
the
CURTIS.--A TEXT-BOOK OF
DE
GENERAL BOTANY.
FLORA
By
Columbia University
T.43ZJ5Y.-THE
OF CHESHIRE. By the late
LORD DE TABLEY (Hon. J. B\ RNE LEICESTER WARREN,
M.A.). Edited by
SPEXCKR MOORE With a Biographical Notice of the Author
by Sir MOUNTSTUART GRANT DUFF. With a Map of Cheshire and a
Photogravure Portrait
Crown 8vo., IQS. 6ef. net.
T
Scientific
Green,
Co.
27
6d,
is.
FARMER.
A.
STUDY OF BOTANY
Flowering Plants.
FRANCES
A.
KITCHENER. With
mE
Adapted
CLASS-BOOK OF BOTANY,
McNAB.
AND
MORPHOLOGY
OLOGY.
With
With
Diagrams
42
B.Sc
With 33
F.L.S.
AND
B.Sc., F.L.S.
TUBE UF.
Steel
By W. R, McNAB.
OF PLANTS.
118
Diagrams.
Fcp. 8vo.,
Illustrations.
.A2\rarr.
LOGICAL BOTANY.
BENNETT,
With 20
is.
THOME
CLASSIFICATION
PHYSI-
6d.
is.
Fcp. Svo.,
and
TREASURY OF BOTANY.
Edited by
Plates
Home
to
19 q Illustrations.
DISEASES
PLANTS
OF
INDUCED
BY
CRYPTOGAMIC PARASITES.
WATTS.
K SCHOOL FLORA.
Botanical Classes.
AGRICULTURE.
ADDYMAN. AGRICULTURAL
ANALYSIS.
Manual of
COLEMAN
AND ADDYMAN.
PRACTICAL AGRICULTURAL CHEMISTRY. By J. BERNARD COLEMAN, A.R.C.Sc., F.I.C., and
FRANK T. ADDYMAN,
Svo.
COOKE.
is
With 24
Illustrations.
Crown
6d. net.
AGRI-
CULTURE.
0^-8.5.
adapted to the requirements of the Science and Art Department, the Junior
Examination of the Royal Agricultural Society, and other Elementary Examinations.
With 34
AGRICULTURE.
With 100
Illustrations.
Illustrations.
Crown
Manual
Crown
for
28
Works published
Scientific
Green,
by Longmans,
Co.
FRAGMENTS OF SCIENCE:
Vol
The Constitution
2 vols.
NEW FRAGMENTS.
Crown Svo
los. 6d.
Goethe's Farbenlehre
Atoms, Molecules, and Ether Waves
Count Rumford Louis Pasteur, his Life and Labours The Rainbow and its CongenersAddress delnered at the Birkbeck Institution on October 22, 1884 Thomas Young Life in the
Alps About Common Water Personal Recollections of Thomas Carlyle On Unveiling the
Statue of Thomas Carlyle On the Origin, Propagation, and Prevention of Phthisis Old
Ct^
ri
xi
The Sabbath
A Morning
Alpine Jottingb
'
on Alp Lusgen
LECTURES ON SOUND.
Diagrams.
Svo.
IQS
6d
125 Woodcuts
With
and
125
AND
1872
Crown Svo
Diagrams.
1873
Plate,
ESSAYS ON
and 59
$$
Crown 8vo M
js.
24
Wood-
6d
Crown
8VO., I2j.
is.
6d.
Institution of
Crown Svo
is.
6d.
1875-1876
With
58
ALPS:
Crown
Narrative of Excurbeing
sion? and Ascents
An Account of the Origin and Phenomena of Glaciers, and
an Exposition of the Physical Principles to \\hich
With
they are elated.
numerous Illustiations. Crown Svo dr. 6d. net.
a
Svo.,
6c.
ALPS.
With
7 Illustrations.
6/. net.
FARADAY AS A DISCOVERER.
Crown
Svo., 3 j. 6d.
Works published
Scientific
Green,
by Longmans,
Co.
29
TEXT-BOOKS OF SCIENCE.
PHOTOGRAPHY
By Captain \V DE
WIVELESLIE ABNEY, C.B F.R.S
With 105 Illustrations. Fcp, 8vo
,
y.
6d.
ANDERSON,. C.E.
Illus-
Fcp. 8vo., y, 6d
trations.
RAILWAY APPLIANCES
WOLFE BARRY,
MICE
By Sir JOHN
K.C
F.R S
B.,
Illustiations
With 218
Fcp. Svo
6d
4.9.
6d.
3-r
LL.D F R
With 72
S.
Illustrations.
35. 6d.
PhD.
With
88 Illustiations.
Fcp.
Fcp Svo
Illustrations.
6s.
&
Svo
6d
45.
6d
With
F.R.S.E.
and
Plate
of Spectra
Fcp. Svo.
57 Illustrations.
3^. 6d.
Fcp. Svo.,
4-s .
BALL, LL. D. F
,
130 Illustrations.
SYSTEMATIC
Fcp. Svo
By
With
S
,
G.S.
Fcp. Svo.
6d
6s.
MINERALOGY.
HILARY BAUERMAN, F
373 Illustrations.
By
With
6s,
DESCRIPTIVE MINERALOGY.
By
Illustrations.
Fcp. Svo
etc
,
6.1
METALS
THEIR PROPERTIES
AND TREATMENT. By A K.
HUNTINGTON and W. G. McMiLLAN.
With
7.4
122
Illustrations.
Fcp.
Svo.,
THEORY OF HEAT.
MAXWELL,
tions.
M.A.,
&
E.
Fcp, Svo.,
CLERK
By
LL.D., Edm.,
With 38 IllustraJ.
4J. 6d.
PRACTICAL PHYSICS.
K.C.M
G.
6s.
,
'
With
T
By R.
With
GLAZEBROOK, M.A., FR.S.
OPTICS
Fcp. Svo., 6s
183 Illustrations.
MA
RUTLEY, F G S
88 Illustrations
T.
By FRANK
With 6
Plates
Fcp. Svo.,
WORKSHOP APPLIANCES,
some
Descriptions of
and
44. 6d.
including
of the
Machine
Fcp. Svo., y.
F.R.S.,
M.I. C.E.
PART I. General Principles, Fastenings and Transmissive Machinery.
With 304 Illustrations.
6s.
B.Sc.,
II.
tails
With
Fcp. Svo.
4J. 6d.
By R.
J.
Fcp, Svo
267 Illustiations
PHYSICAL
PART
6d.
F.R.SS., L.
MA
ELEMENTS OF ASTRONOMY.
Sir R. S.
SIVE WRIGHT,
B. Sc. F. L S.
,
Fcp. Svo.,
W. BENNETT, M.A,,
With 600 Illustrations.
A.
6s.
By
30
&
Scientific
Co.
By
Rivmgton's Notes on
With 385
Building Construction '.
Illustrations and an Appendix of
Examination
Crown
Questions.
the Author of
8\o.
'
4$. 6d,
and Kinetics
By A THORNTON,
MA., F.R.A.S With 220 Illustrations, 130 Worked Examples, and
over goo Examples fiom Examination
HEAT
etc.
By
Inter
With 136
Sc (Lond.)
and numerous
Examination Papers.
trations
S.
Map
trations,
Geological
British Isles (coloured),
Illus-
of
and a
set
the
of
Crown
Examination.
HUMAN PHYSIOLOGY
a Manual for
Students m advanced Classes of the
Science and Art Department.
By
TON, M.A.
Svo.,
By W. J. A. EMTAGE, M.A.
With 232 Illustrations Cr. 8\o 6s
M A.
With 317
Svo.,
trations,
and
Spectra,
Crown
Colouied
Plate
of
Ph.D., B.Sc
PHYSIOGRAPHY.
and
LIGHT.
Illustrations
Readers
(Lond.),
4?. 6tt.
Illus-
E v amples
Crown
Questions for
Svo., 7.? 6d,
THEORETICAL MECHANICS
Papers,
GEOLOGY
Crown
With 100
Illustrations.
HYGIENE. ByJ.LANENoTTER.M.A.,
M.D Protessor of Hygiene in the
,
6rf.
Army
Woodcuts
Crown
Medical
School,
Netley,
Medical
Army
and R. H. FIRTH, F.R.C.S.,
Assistant Professor of Hygiene in
Surgeon-Colonel,
Staff,
Written specially
to
ELEMENTARY STAGE
GEOMETRY,
Arithmetic.
By
with
Illustrated
Svo.
ai.
including
I.
Graphic
H. MORRIS.
Drawings.
Fully
Crown
6d.
Art Department.
Crown
8vo.
i s. 6d.
By L H. MCRRIS.
ON
of the Direc-
PRACTICAL,
(Whitworth Scholar).
Svo
25.
Part
II.
Parti. Crown,
Crown
Svo.
%s.
AN INTRODUCTION TO MACHINE,
DRAWING AND DESIGN. By
DAVID ALLAN Low. With
trations.
Crown
153 Illus-
Works published
Scientific
by Longmans,
Green,
ELEMENTARY SCIENCE
Co,
inued.
2S. 6d.
6d.
AN ELEMENTARY COURSE OF
MATHEMATICS Containing Arith- AN INTRODUCTION TO
metic
Euclid (Book
and
tions
Crown
Exercises)
8vo.
6d.
2s.
I.
;
In-
Crown
By
TAYLOR, M.A.,
E.
J.
Crown 8vo
etc.
2s. 6d.
MECHANICS-
THEORETICAL
J. E. TAYLOR, M.A
With 122 Illustrations,
(Lond
numeious Worked Examples, and
FLUIDS
By
B.Sc.
F.C.S., F.I.C.
CHEMISTRY
PRACTICAL
the
WILLIAM
is.
8vo.,
A. TILDEN, D.Sc.
By
Fcp.
6d.
THEORETICAL MECHANICS-SOLIDS.
By WILLIAM JAGO,
Crown
THEORETICAL MECHANICS
PRACTI-
).
F.R.G.S.
ELEMENTARY
GEOLOGY.
Geological
F.G
By
With
S.
Map
of
the
HUMAN PHYSIOLOGY.
WILLIAM
With 218
FURNEAUX,
Illostrations.
By
F.RG.S.
Crown 8vo.
,
A MANUAL OF MECHANICS. With A COURSE OF PRACTICAL ELE138 Illustrations and Diagrams, and
188 Examples taken from ExaminaM.
tion Papers, with Answers
By
Crown 8vo 2s. 6d.
A.
GOODEVE,
Diagrams and
2.?.
By
Crown
Illustrations.
PHYSICS.
an Elementary Text-
RHEAD.
Crown 8vo.
With 94
E. L.
Illustrations.
Alternative
3^.
6d.
Course
MARK R WRIGHT,
Illustrations.
A.
W.
M.A.
Crown 8vo.
With 242
,
2s. 6d.
With 235
POYSER, M.A.
Illustrations.
Crown
8vo.
ORGANIC CHEMISTRY:
R.
LEY, F.I.C., F.C.S.
Compounds.
trations.
Crown
J.
342 Illustrations.
With
2s.
6d.
By
Crown
2s. 6d.
the
the
Institution
Engineers.
Crown
By
tions.
By
Illustra-
6d.
METALLURGY:
Book.
With 226
MENTARY BIOLOGY.
BIDGOOD, B.Sc.
With
of
142
Mechanical
Illustrations.
ELEMENTARY
PHYSIOGRAPHY.
Instru-
Crown
Fatty
Illus-
Ph.D.
With 34
Illustrations.
Crown
32
Scientific
Co.
ASTRONOMY.
WELL
Sir
By
LL D
BALJL,
Fcp. 8vo
41 Diagrams
ir.
6d,
Fcp. 8vo.
>
is.
With 22
Illustiations.
6d.
By
F.R C
P.
8vo.,
Fcp.
is. 6d.
8vo,, is.
AND PNEUMA-
HYDROSTATICS
TICS,
B.Sc
By
8vo., is
Fcp
MAGNUS,
With 79
B.A.
PHILIP
Sir
Diagrams
(To be had also
6d.
The
2? )
Solutions of the Problems.
Worked
Answers,
with,
2>
AND
MOLECULAR
PHYSICS
SOUND By FREDERICK GUTHRIE, BOTANY
With 91 Diagrams.
F.R.S.
8vo.,
Fcp.
6d.
is.
With
8\o.,
6c.
1.5
Diagrams
141
Fcp.
With 42
M.D.
Diagrams
grams.
is.
6d.
Fcp.
6d,
is.
8vo.,
Fcp
Diagrams.
By RICHARD
D.Sc.
8vo.,
With
41
is. 6rf.
(Section
THORNTON,
JOHN
By
I.)
Head
M.A.,
Higher Grade
Bolton
With 215 Illustraa Coloured Spectrum.
School,
and
tions
Crown
Svo.
ay.
I.C.,
6d
J.
With
SICS.
WATSON, B.Sc
Assistant Professor
the
Physics
Royal College of Science, London,
etc.
Exercises,
zs. 6d.
PRACTICAL INTRODUCTION
TO THE STUDY OF BOTANY ELEMENTARY PRACTICAL ZOOBRETLAND
Flowering Plants, By
LOGY, By FRANK E BEDDARD,
FARMER, M.A., Professor of Botany
M.A. Oxon, F.R.S, Prosector to
m the Royal College of Science,
the
of
J.
London
Zoological Society
Lecturer on Biology at Guy's
Hospi;
With 93
tal.
Svo.
2s.
Illustrations.
6d.
Crown
Dint,
533.7
M61
c
CD
38 332