1221 
action are totally absorbed by those surrounding atoms, and do not 
exert any force whatever outside the molecule. 
This is, indeed, in perfect harmony with what was found lately 
by Erste *) with regard to the influence of the molecules on each 
other as far as their sphere of action is concerned (this would 
namely not extend any further than to the surrounding molecules). 
But now the same thing is found with regard to the atoms in the 
molecule. Only the peripheral atoms take part in the cohesive 
attraction, the central ones — for so far as they are entirely shut 
are perfectly inactive. If, however, 
off by the surrounding atoms 
these central atoms are partly free, as e.g. with the doubly and 
triply bound C-atoms, either half the action or the full action imme- 
diately shows itself again. This is certainly one of the most remark- 
able results to which our researches have led us. 
In what follows I shall confine myself to what is indispensable 
as a support of my contention, reserving the fuller discussion of 
this subject for another occasion. *) 
2. The calculation of the values of a and b. 
As a and 5 are in general functions of v and 7’ (a is probably 
independent of 7), it is desirable to calculate these quantities for 
corresponding states, and for this the critical state is first of all to 
be taken into account. Also when we pay attention to the fact that 
the variability of 46 with the volume is different for every substance, 
and that therefore something special continues to adhere even to 
the critical state, this circumstance can yet be eliminated by the 
introduction of a new parameter. As I have lately demonstrated *), 
in the first place we may take for this the quantity y, Le. the co- 
efficient of direetion of the straight joining line between the densities 
YD, and D, in a D,T-diagram, which quantity is at the same time 
='/, (b-:6,). We can then express all the quantities in y, which 
henceforth represents the only independent parameter; thus the idea 
of correspondence will also have obtained a wider meaning. 
Thus e.g. the quantity s— RT): prvrx is always = 8y: (y + 1), 
and the quantity r=v,:b, will always be =(y+1):y. For 
f= RT: pr (ve-—be) = 1 + (ar: preven’) we find 8y. (f/ is the critical 
coefficient of pressure for the case that @ and 4 may be taken in- 
dependent of 7 at 7;). Further rs=8, /’ (r—1)=8, (f-A)r?-= 
‘) Bemerkung zur dem Gesetz von Eörvös. Ann. d. Ph. (4) 34, 165 (1911). 
*) In the Journ. de Chimie physique of Prof. Guye at Genève. 
8) These Proc. of March 26, April 23, May 29 and Sept. 26, 1914, resp. p. 808, 
924, 1047 and 451. 
79* 
