136 
DR. WALTER WAHL: PHYSICO-CHEMICAL DETERMINATIONS 
These inelting-poiDts are plotted in the diagram, fig. 6, and connected by the 
upper curve Pj of that diagram. The curve Pg is not shown in the diagram as it 
runs very close and parallel to Pj. 
The difference between the pressures Pj and Pg is about 20 kg./cm.^, and the 
amount of super-cooling necessary to reach the part C of the crystallization-velocity 
curve as shown in fig. 5, where the velocity of crystallization attains large values and 
growth-structures also appear, is thus only about one degree in the case of CBr^—one 
degree on the melting-point curve corresponding to 20 kg./cm.^. From the fact that 
the difference between the values Pj and Pg does not change at higher pressure, it 
may be concluded that pressure does not infiuence the degree of super-cooling which 
is necessary in order that growth-structures may be produced. The curves showing 
the velocity of crystallization in its dependence on the degree of super-cooling 
(diagram, fig. 5) will, therefore, be similar in shape at all pressures, altlrough the 
absolute value of the velocity at corresponding degrees of super-cooling may vary 
with the pressure. 
(2) Transition-point Determinations .—When the molten carbon tetrabromide is 
cooled between glass plates it first crystallizes in isotropic crystals, and on further 
cooling a crystallization of the anisotropic crystal modification suddenly starts. The 
anisotropic crystallization begins usually from one or from a few points at the border 
line of the isotropic crystal mass, and the anisotropic crystals grow at the expense of 
the isotropic just as if they were growing in the liquid phase. The growth is rapid, 
and long, blade-like crystals are formed which cover the space earlier occupied by 
several of the isotropic crystals, ami no connection whatever seems to exist between 
the border lines of the isotropic crystals and the growing anisotropic crystals. On 
the other hand, when tlie transition of the anisotropic modification to the isotropic 
occurs, on raising the temperature of the preparation, the isotropic crystals are not 
seen to grow at all, but it looks rather as if the anisotropic crystals were simply 
melting. Dark isotropic spots of quite irregular shape appear both at the border 
lines and in the interior of the crystals, and continue to be formed and to grow larger 
until the whole has changed into an isotropic mass. 
Pressure raises this transition-point rapidly, as seen from Table II., and the same 
general difference which exists between the transition from isotropic to anisotropic, 
and from anisotropic to isotropic at ordinary pressure, exists also when transition 
takes place at high pressures. But, in addition, it is found that the transition from 
isotropic (fie., regular) form takes place extremely slowly at the true transition-point 
pressure. When the transition-point curve is passed at constant temperature by 
lowering the pressure slowly, the anisotropic phase at first appears in the form of 
small, bright, coloured spots on the dark field. These are a kind of crystal globulm, 
crystal germs, which then grow to Ijigger crystals. This, as is well known, is tire 
general case when crystals begin to form in a liquid, and the related phenomena have 
been described at length by Vogelsang, who especially investigated the growth of 
