CunnincHam— Crystallization of Minerals in Igneous Rocks. 411 
affords evidence of all three types of crystallization. Nor is it 
wonderful that the above, as well as Mr. Stromeyer’s examples, 
are only the exceptions, for certainly the majority of igneous 
rocks must have crystallized at great depths, where the pressure 
would be enormous (greater than OB). 
The exception proves the rule. It will be evident that Dr. 
Joly’s, as well as any of the other theories alone, breaks down 
in face of these exceptions. ‘Thus, if, according to Dr. Joly, the 
softening point of quartz is actually below that of felspar, and 
hence the felspar crystals are able to develop their proper form at 
the expense of the quartz crystals, then why was this not always 
the case? In the same way (as Bunsen saw long ago) the theories 
depending on the different proportions of the chemical elements 
could not possibly explain differences of crystallization in rocks of 
the same chemical composition. 
As an appendix, I shall now take the example of quartz, and 
examine how its latent heat, and similarly the latent heats of the 
other minerals, may be deduced from simple chemical and physical 
measurements (Plate XXIII). 
In the case of quartz, then,’ we know the specific heat, up to 
1185° C. (slope of OB), and we know that the specific heat of 
amorphous silica is somewhat greater. Dr. Joly? measured the 
specific heat of opal. ‘The density of this substance varies from 
2:1 to 2°22, and it contains a variable quantity (13 to 2 per cent.) of 
water, which would be a factor tending to increase its specific heat. 
The values obtained varied from 0°2379 to 0:2033. Ihave adopted 
the value 0:20 as a fairly safe minimum for anhydrous SiO,. This 
affords us the slope of H F. In order to err on the side of the mini- 
mum I have further assumed the specific heat constant up to the 
melting point. For most solids and liquids the specific heat increases 
with rise of temperature.* Further, it is highly probable that the 
specific heat of an amorphous substance is greater than that of the 
erystalline modification.* If we assume the same mean specific 
heat for silica as has quartz, we get a curve like HF’ to represent 
the heat absorption of amorphous silica. But we do not yet know 
how far apart to place these two lines, or, in other words, where to 
1 Pionchon, Joc. cit. 
2J. Joly, ‘‘ On the Specific Heats of Minerals,’’ Proc. Roy. Soc., xli. (1886) 
p- 250. 
3 Preston, ‘‘ Theory of Heat,’’ p. 212. 4 J. Joly, loc. cit., p. 254, &e. 
