165 
The special equation of state deduced from measurements of Kamer- 
LINGH ONNEs and BRAAK gives e.g. 
B, = 0,000893, 
a value higher than might be expected from direct graphical extra- 
polation of the individual values of B from the measurements of 
KAMERLINGH OnneS and Braak. Therefore this special equation of 
state too would also give a too high value for the quadrupole 
moment when the method of calculation of Draijer was used. 
It seems to me preferable to proceed as was done in my preced- 
ing communications viz. to work out the theoretically deduced 
development of. until’) it extends over a sufficiently wide domi- 
nion, to compare then this development with the experimental data 
and to deduce in this way e.g. the value of the quadrupole moment. 
§ 4. Further development of the second virial coefficient for sphe- 
rical polarisable quadrupole molecules. In order to take into con- 
sideration the influence of the polarisability of the molecules in the 
deduction of the quadrupole moment, the development started in 
Comm. N°. 66 § 3 has been extended by a few terms. Thereto 
the formulae given in Leiden Suppl. N°. 39a may be of use. By 
multiplying equation (11) from the paper by 
¥ =sin'.O, +. sin* O, + 40s‘ 0, + 4c08'6,.. . . . (1) 
(Comm. N°. 6a equation (10)), we find, following the notations of 
Leiden Suppl. N°. 39a: 
n 
[yr x] = 22 [Any] + & ) [ A"? BY y] [eos p] + & ra C4] [cost 2p] + 
4 6 ) (|) [.Ar—3 B? Cy][cos?.cp.cos 2p] etc. … meee (2) 
With 
an Oels WEE At ABM vr Ae os (8) 
(see Leiden Suppl. N°. 39a equation (9)) we find: 
[ORs Gade cy in, (APE? OA PPE HE 
sealed Pe ea Se ame ee (A 
1) Of course I acknowledge the objections made by Dene in § 5 of his cited 
paper. They show that the results obtained in the above mentioned way can only 
be accepted with some reserve as has been specially stated in Comm. N°. 6a § 4. 
On account of all this accurate measurements on the second virial coefficient at higher 
temperatures are very necessary. In the meantime however it seems to me of 
some importance that such a good agreement with the experimental data that are 
at our disposition is obtained by using the mentioned simplifying suppositions 
(quadrupolar action, no quantum deviations as yel, collisions as of solid spheres). 
