( 1259 ) 
of 6= 2,05, and that of n= 9, and then we find for the value of 
the first member: 
0,019, 0,023, 0,019, 0,01, 0,023. 
We find for the factor of x: 
0,193, 0,176, 0,146, 0,12, 0,129 
and for « 0,1 O15, 013, 0.083, 0,18. 
If we consider how many numerical values which are not absolutely 
accurately known, have had to serve for the calculation, and how 
greatly an error in the measurements affects the slight amount of the 
expressions which determine «, the agreement must be called satisfactory. 
[ have undertaken this investigation: 1 to see if in the observations 
a support could be found for the existence of quasi-association, but 
chiefly 2 to try and find out what is the cause of the anormal 
behaviour of the alcohols. For this reason I have tried to determine 
the value of a for ethyl-aleohol. 
The same reasons which made me conclude to the existence of 
quasi-association in general, apply also to alcohol. Very near 7, 
T dp 
p dT 
e.g. at 7' only 0,6° different from 7; = 516,1, — 1 wil! not 
differ appreciably from /f;,—1, and us exceedingly little from 1; 
Pk 
UU 
— is found 
Uk 
hence it may be asked by way of simplication whether 
equal to 1 + (L—m) or equal to 1 + 1—m. Now at 
1 ee == 0,001164 
PET 
we find the value of “1 1,095 for alcohol. Now 0,025 is certainly 
UK 
not equal to 0,001164, but of the order of V1—m == 0,034. We 
need only assume the value v, == 3.63 given by Sypney Youre, to 
be equal to 3.61 to change the value 0,025 to 0.34. So we have 
every reason to expect for alcohol just as for the so-called normal 
substances, that a value for « will be found in the saturate vapour. 
I may mention here that long ago I concluded to the existence of 
molecule-complexes also for water vapour at 100°. Otherwise the 
volume which would be theoretically equal to 1689, would not have 
descended to 1649 — a diminution too great to be ascribed to the 
existence of a and 6 in the equation of state. 
Proceeding in the above way, we find accordingly a value of x 
in the saturate vapour of alcohol, and in the value of # a course 
which does not deviate much from that which occurs for normal 
