( 40 ) 
For a slight number Of simple molecules could very well form'a 
multiple molecule in some cases, without it being necessary that the 
phase becomes solid. For in all the so-called associating liquids a 
comparatively smaller or larger number of double, at most triple 
molecules are assumed. 
In the theory worked out by us we assumed for the sake of 
simplicity only the formation of double molecules; we observed, 
however, already at the conclusion of our preceding paper that 
qualitatively the theory is not subjected to any change when multiple 
molecule complexes are started from. Still, I intend to revert to this 
later on. 
With regard to the number of simple molecules which contribute 
to the formation of a multiple molecule, it is still worthy of note, 
that of course, this number may be different for one and the same 
substance according to circumstances. This explains too that several 
solid modifications often occur for one substance, which may pass 
into each other under certain circumstances and two or three of 
which — according to the phase rule — can exist by the side of 
each other, both with simultaneous coexistence of liquid and vapour 
That the properties of the solid state depend on the direction , is 
also a consequence of the relatively (with respect to the diameter of 
the molecule) slighter distances, and the being bound to fixed centres 
of equilibrium, in consequence of which also the grouping of the 
atoms in the molecule can come more to the fore. But on the whole 
it will after all be probably the grouping of the simple molecules 
inside the complex crystal molecules, so the structure of the last-named, 
which in perfect accordance with the prevailing crystallographic ideas, 
causes the properties to change with the direction. 
Strictly speaking the equation of state will be different there for 
every direction, because the specific quantities lKt> and l Vb vary . 
with the direction. But yet we may treat v and b, like a, as scalar 
quantities, when only hfere as elsewhere, we have in view middle 
va ues. Then there is no reason why we should not continue to 
speak of a definite specific volume v, or density */„, or of a definite 
molecular volume b in these cases, just as we can continue to speak 
° a definite s P^ific attraction a. For all the so-called optically- 
isotropic solid phases, moreover, the difference of direction disappears 
tor the different properties. 
So we can summarize the above considerations as follows: 
oi a simple substance, the molecules of which can combine to 
arger complexes, and where Lb differs from zero, so that e. g. for 
