( 41 ) 
negative Ab the complex-formation again greatly decreases in the 
neighbourhood of v = b in consequence of increase of pressure (or 
diminution of volume), the following possibilities will exist: 
I a. a phase with large specific volume (slight density ): sigasphase. 
b. a phase with small specific volume (great density), but internally 
still sufficiently mobile in consequence of comparatively slight complex 
formation: a liquid phase. 
L- c - a phase also with great density , but internally immobile in 
consequence of extensive complex formation: a solid phase. 
For negative A b the density of the solid phase is even less *) than 
that of the liquid phase; it is however not so much the density, 
but the formation of the large complexes, which makes the phase 
immobile — as we have set forth fully above. 
And just as there are all kinds of transitions possible between 
gas and liquid phases, and often no external difference is noticeable 
between the two phases, e.g. in the neighbourhood of the critical 
temperature, somewhat below it, or for a gas phase under high 
pressure above the critical temperature — in the same way there 
will of course be transitions between liquid and solid phases: liquid 
phases with relatively slight mobility (so-called amorphous-glassy 
state) and solid phases with comparatively great mobility (e.g. the 
fluid crystals or anisotropic liquids). 
The real solid crystalline state is always, just as the corresponding 
liquid-crystalline state, connected with extensive complex formation; 
whereas the ordinary liquid state, together with its viscous continu¬ 
ation to the internally immobile amorphous-glassy state will be 
characterized by comparatively slight complex formation. In the last 
named state only the comparatively low temperature in connection 
with the greater density and the comparatively large volume of the 
molecules causes the immobility, the fixedness of the phase. 
h For then Av is generally also negative. Yet we must not forget that Ab 
and av generally differ. On p. 8 of my paper in the Arch. Teyler I derived the 
A b + 
At? = - 
1+0 
l 2 a v—b 
v* 
As we see, only for very small v—b (compared with Ab) Av is approximately 
= — Ab. But if v—b is comparatively large with respect to Ab, Av may even be 
positive, even if Ab is negative. We need only think of the gaseous state, where Ab 
negligible by the side of v—b, and where av is simply = 
1+0 1+0 
