IGNEOUS ROCK-MAGMAS AS SOLUTIONS 251 



variable composition of the ground-mass, but also throws light 

 upon features which would otherwise be anomalous ; notably 

 the occurrence of both felspar and quartz among the porphyritic 

 elements in the same rock, and the evidence sometimes seen of 

 early crystals having been partly redissolved by the magma. 



A mineral does not necessarily begin to crystallise from a 

 magma at the theoretical freezing-point. This freezing-point 

 (for a definite compound the same as the melting-point in the 

 same magma) is the temperature at which crystals of the mineral 

 are in equilibrium with the magma. But equilibrium is not 

 established in the absence of crystals, and the temperature may 

 fall considerably below the freezing-point before any crystallisa- 

 tion is set up. If we may apply to rock-magmas the results of 

 Professor Miers' researches on saline solutions, crystallisation 

 can in no circumstances be initiated, without " inoculation " by 

 solid crystals of the same substance or one isomorphous with 

 it, until a certain definite degree of supersaturation has been 

 attained ; and there are many indications that for the rock- 

 forming minerals the requisite supersaturation is very con- 

 siderable. Crystallisation once started, the latent heat evolved 

 soon raises the temperature to a point at which equilibrium is 

 adjusted. 



Suppose, now, a magma consisting of quartz and felspar, 

 with an excess of the former above eutectic proportions. 

 Cooling having brought about a sufficient degree of super- 

 saturation, quartz begins to crystallise, equilibrium is attained 

 with some rise of temperature, and the separation of quartz 

 proceeds with falling temperature. When all the excess of 

 quartz is separated out, and the magma thus reduced to eutectic 

 composition, felspar should begin to crystallise simultaneously 

 with the quartz ; but, since no felspar-crystals are present to 

 establish equilibrium, this will not yet take place. The quartz- 

 crystals will continue to grow alone, until the magma is 

 sufficiently supersaturated with felspar to start the crystallisa- 

 tion of that constituent. The rapid crystallisation of the felspar 

 will set free enough heat to raise the temperature and also to 

 redissolve part of the quartz, with which mineral the magma is 

 now undersaturated. When the eutectic point is again reached, 

 the presence of crystals of both minerals will prevent any 

 further supersaturation. In this illustrative case it is easy to 

 see that the intervention of supersaturation causes consider- 



