LOTS) 
to reduce these to the same temperature. These corrections could 
not be made with the same degree of certainty as for hydrogen, 
for then the temperature differences were always less than 0°.2, 
while differences of fully O°.4 were experienced in several instances 
in our present argon measurements. 
That the individual coefficients’) for each isotherm may be cal- 
culated without adjustment from other isotherms, it is of special 
importance that points observed at temperatures which lie close 
together should be reduced to the same temperature. 
There is yet another circumstance to cause the accuracy of the 
argon isotherms to be less than that reached in the hydrogen mea- 
surements. It has already been shewn?’) that the desired degree of 
accuracy (from 1 to 0.5 per 1000) cannot be reached with reservoirs 
of content less than 5 cc. Hence the fact that quite a number of 
points have been got with a reservoir volume of 2.5 ce. can also 
account for the smaller degree of accuracy of many points. The 
following table gives for the various series rough values of the 
volumes in ce. of the large reservoir in the compression chamber 
and of the small reservoir that is kept at low temperature. 
Large. | Small 
Series. reservoir. | reservoir. 
XII | 600 Ga 
XII—XVIII 9000 5 
MIX XX | 600 2.5 
XXII—XXII 600 5 
XXIV—XXXV 600 15 
XXXVI—XXXVIII | 600 9.5 
$ 2. The argon used was made according to the methods indicated 
in $ 1—3 Part I of Comm. 115. For the sake of completeness we 
now publish some further diagrams of the apparatus used. 
Fig. 1%) shews the circulating apparatus following that used by 
Fiscrer and Ringe‘). 
1) Proc. June 1961, Comm. No. 71. 
2) Proc, April 1901 Comm. No. 69 § 6, and Proc. March 1903, Comm. No. 84 
§ 19. 
3) Proc. May 1910, Med. 115, Part I § 10. 
4) F. Fiscuer and O. Rixer, Ber. d.d. chem. Ges, 41. 2017. 1908, 
40* 
