Cunnincuam—Orystallization of Minerals in Igneous Rocks. 409 
much less any relative magnitude of the latent heat. ‘Thus if we 
try to compare oligoclase and augite, we find the available data 
insufficient for our purpose. For we cannot safely compare their 
latent heats until we have actual numerical data over and above 
Von Kobell’s scale. Much less would it be safe to compare the 
values of the ratio he 
— V1 s 
ara in such cases. But let us confine our 
attention to the examples of augite and orthoclase as typical, and for 
which the above comparison is, I believe, reliable. It is apparently 
the general opinion that the latent heats of the minerals are small.’ 
I shall show directly that the latent heat of quartz is very large, 
certainly much greater than the latent heat of ice (the greatest 
previously on record). Hence we might be led to expect that the 
less silica a mineral contains, the smaller was its latent heat.’ 
This is certainly not contradictory to my statement of the order 
d AG 5 
of the values. of : for the minerals; 7.e. its value for quartz 
is least, for orthoclase rather greater, for augite the greatest of 
the three. The diagram (fig. 9) on p. 407 shows this relation. 
We have the curve for augite, crossing that for quartz at a 
pressure corresponding to OA; and the orthoclase line crossing 
the quartz line at a pressure corresponding to OB. Then in a rock 
formed under any pressure greater than OB, augite and orthoclase 
will crystallize before quartz; at any pressure greater than OA, 
augite will crystallize before quartz; but at pressures less than 
OA, quartz must crystallize before both augite and orthoclase. 
We may call OA and OB the “ eutectic pressures” for quartz and 
augite, and quartz and felspar respectively. It is unfortunate 
that we do not know the melting-point of any of the micas, 
although their fusibilities are known to be low. However, in 
the meantime we may class them with augite as typical ferro- 
magnesian minerals. And so the crystallization of an ordinary 
granite is explained at once. 
Let us conceive of a rock magma kept liquid by reason ofits high 
temperature, in spite of great pressure. Now, if the pressure is 
maintained while the temperature gradually falls, the rock will 
* Cf. C. HE. Stromeyer, Joc. cit., p. 10. 
2 This agrees with the observation that the amount of thermal metamorphism is in 
general greater round an acid than round a basic magma. 
