AT PUGH PEERSUEES BY OPTICAL METHODS. 
138 
pressure is rapidly raised,, the crystals begin at a certain pressure to grow more 
rapidly in a certain direction than in others, and so the cubic structures of growth 
are produced. In this way it is possilde to determine at each temperature two 
crystallization pressures ; one at which the crystals can just he observed to grow— 
this is the pressure closest to the true crystallization pressure we are able to observe— 
and a second higher pressure at which very rapid growth in a certain ciystallographic 
direction takes place. 
This behaviour is not peculiar to carbon tetrabromide, but is common to all substances 
hitherto investigated, and is probably a phenomenon of quite general character. It 
depends upon the following circumstances :— 
Gernez* and Moore! have shown that the velocity of crystallization, when 
measured in thin-walled glass tubes, increases with decreasing temperature, that is, 
when the degree of super-cooling increases. Later, Tammann and his pupils have 
found I that this increase of the velocity of crystallization occurs only up to a certain 
temperature, when a maximum of velocity is reached, which in many cases remains 
constant for a certain interval of temperature, and then very rapidly diminishes and 
assumes quite small values. The general case may he diagrammatically represented 
by fig. 5 : a represents the melting-point temperature, at which the velocity of 
crystallization is very small. With descending temperature the velocity at first 
increases slowly, hut at temperatures h a much more rapid increase takes place, and 
continues until the maximum velocity c is reached. At temperatures d the velocity 
again very rapidly drops to small values, and diminishes then at low temperatures e 
quite slowly. The part a to c of this curve is that which is of interest in this 
particular case, because it gives the explanation of the crystallization phenomena 
observed when pressure is increased. The comparatively short interval of tempera¬ 
ture, where the rapid increase of the velocity of crystallization takes place, evidently 
corresponds to the pressures at which such an amount of super-cooling is reached that 
rapid crystallization ensues and the growth-structures are produced. This view is 
* G. Gernez, ‘Compt. Eend.,’ 95, p. 1278 (1892). 
t B. Moore, ‘Zeitschr. f. Phys. Chem.,’ 12, p. 545 (1893). 
I G. Tammann, ‘Zeitschr. f. Phys. Chem.,’ 27 (1897), p. 152; 29 (1899), p. 58. 
