OcTOBER 2, 1913]| 
air, are popular presentations of a hackneyed 
theme, almost painfully familiar in this country, 
where liquid air long since descended to the level 
of a music-hall turn. But in parts ii. and iv. the 
author deals in an interesting and original way 
with the theory and practice of actual processes 
for the technical liquefaction of air, and its separa- 
tion into oxygen and nitrogen. Naturally an 
author must be allowed to tell his story in his own 
way, and fight his battles over again, when these 
battles have resulted in success, though almost 
everyone may now be supposed to know that 
working at —200° C. does not confer upon liquids 
any peculiar behaviour or render the separation of 
oxygen and nitrogen from the air a problem essen- 
tially different in its scientific principles from that 
of the separation of alcohol and water in one of the 
oldest of chemical operations. 
The English translation certainly retains to the 
full the racy style of the original, but sadly needs 
careful revision, especially in the mathematical 
expressions. As many as six slips have been 
noted, for example, on pp. 132-4. The units 
employed should be defined to render them in- 
telligible to English readers, and misleading con- 
tractions like calorie: for kilogramcalorie, and 
Kgms. for Kilogram-metres, avoided. In more 
than one instance the real meaning of the author, 
just where it is important, is obscured by some 
slip or looseness of expression, as on p. 184, 
where an improvement is stated to increase the 
yield of liquid air in Claude’s process by 0°85 
litre per H.P. hour, when apparently to 0°85 
litre is intended. In a preface by D’Arsonval 
yields of “finally o95 litres per H.P. hour” in 
Claude’s process are referred to, but in the text, 
apart from the above imperfect statement, we are 
left in doubt as to the best that Claude has so far 
been able practically to achieve. 
Dealing first with the problem of air liquefaction 
Claude’s especial contribution is the solution of 
the problem of expansion with external work, 
following the suggestion made by Lord Rayleigh 
so long ago as 1898. As is well known, Linde’s 
and Hampson’s processes depend only on the 
“internal work,” that is, on the relatively minute 
cooling effect—the Joule-Thomson  effect—pro- 
duced on the expansion of an imperfect gas, like 
air, due to the work done by the molecules in 
increasing their distances apart against their own 
feeble attraction. As in the whole of these pro- 
cesses, the Siemens exchanger of temperature, 
fifty-six years old, is employed, and enables this 
cooling to be used regeneratively until ultimately 
the liquefaction temperature is reached. But at the 
expansion jet, or, at least, inside the exchanger, 
just where it is emphatically not wanted, the 
enormous mechanical energy of the escaping gas 
is quantitatively reconverted into heat. The ex- 
pansion is adiabatic, and temperatures, as in 
Cailletet’s apparatus, far below the liquefaction 
temperature are instantaneously attained, but, in 
distinct inferiority to Cailletet’s simple process, 
are not made use of because the work is quantita- 
tively reconverted into heat inside the system. 
At first sight, but at first sight only, it appears 
NO. 2292, VOL. 92] 
NATURE 
ee 
135 
that an enormous improvement might be effected 
in this direction, increasing the yield of liquid air 
some ten times, and regaining thereby a substan- 
tial proportion of the work employed in com- 
pression. Lord Rayleigh’s suggestion was that 
the air on expansion should drive a turbine, which, 
however inefficient, could not fail both to increase 
the cooling effect and recover some of the power 
employed. 
Claude has successfully employed the energy of 
expansion to do work in a compressed air motor 
capable of working below —100° C. We read 
that ‘while the makers have troubles, which are 
relatively frequent, with the ever well-known but 
still somewhat barbarous ,and brutal appliances 
which air compressors are, they have, so to speak, 
none at all with the new-born appliances, the ex- 
pansion machines for liquid air.” At first petrol 
and even liquid air itself were employed as lubri- 
cants in the cylinders of the compressed air 
machines. Lubrication troubles seem, however, 
now to have been entirely avoided, owing to a 
discovery (1912) of the unique properties of leather, 
«which, after being suitably treated, preserves all 
its good qualities at low temperatures. In the 
present machines the pistons of the expanders are 
provided with stamped leathers instead of metallic 
rings, and do not require any lubrication. The 
chief advantage of the system is that lower 
pressures—4o atmospheres, the critical pressure of 
air—can be employed, whereas the Linde and 
Hampson processes depend on the use of a 
pressure of 200 atmospheres. But the yield, 
spoken of in the preface as finally o°95 litre per 
H.P. hour, is not very greatly superior. In the 
Linde process a practical yield of o°6 litre per 
H.P. hour is realised in large machines, which is 
some three times better than in the Hampson 
laboratory machine. 
The evolution of Claude’s system has many 
points of interest. Exchangers are employed, and 
the gas arrives at the expander at a temperature 
of about —100° C. Now if this process of ex- 
change is carried too far, for example to 
—140° C, “the air which enters the machine is 
not yet a liquid, but it is almost no longer a gas; 
its expansive properties are, so to speak, done 
away with, and the external work of expansion 
becomes detestable.” Even were air a perfect 
gas, it can readily be seen that the more it is 
cooled before expansion the less energy it has to 
lose when expanded, and the smaller the cooling 
effect obtained. Enormously more of it is re- 
quired to fill the cylinder the lower the tempera- 
ture, whilst all the time the external work it can 
do, and the cooling effect it can produce, are 
steadily vanishing. * But actually, at —140° under 
40 At., the volume is already only one-fourth of 
that of a perfect gas. This, of course, though 
Claude does not'say so, is tantamount to admitting 
that the defect in the “internal work” processes 
in not utilising the energy of expansion is more 
apparent than real, and that the advantages of 
utilising the external work are to a large extent 
illusory. The practical solution was found in ad- 
mitting the compressed gas to the expander at a 
