THE GENESIS OF IGNEOUS ROCKS 317 



to the melting point, so that a slight variation in the conditions 

 may bring about crystallisation or fusion. A serious difficulty, 

 however, presents itself, especially in the case of the more 

 acid rocks. Crystallisation is accompanied by the elimination 

 of the greater part of the volatile fluxes, and without these a 

 much higher temperature would be required to bring the rock 

 again into the liquid state. Many rocks, too, would in the 

 absence of these fluxes, unless the temperature were very 

 elevated, give rise to magmas so viscid that it would be 

 impossible for differentiation to take place. 



The possibility of segregation in the liquid magma must 

 therefore be considered. This has been chiefly discussed in 

 connection with Soret's principle, which declares that a sub- 

 stance with which a solution is nearly saturated tends to 

 accumulate in any portion of the solution which is colder than 

 the rest, or, more exactly, that " equilibrium will be attained 

 only when the concentration at every point is inversely pro- 

 portional to the absolute temperature." The differences of 

 temperature likely to occur in an igneous magma bear, however, 

 so small a ratio to the absolute temperature that the results 

 of the application of this principle may almost be disregarded. 



It is probable that segregation under the action of gravity 

 is of more importance. An ordinary aqueous solution would, 

 we are told, have to be at least 100 metres high before there 

 would be a sensible difference in the degree of concentration 

 between the top and bottom. Many intercrustal reservoirs 

 must be far deeper than this ; there is evidence, too, that the 

 separation occurs more readily in the case of silicate magmas. 



So far it has been assumed that the contents of each subter- 

 ranean reservoir consist of a single magma, whose constituents 

 are miscible in all proportions. Some liquid substances, however, 

 such as aniline and water, are only completely miscible above a 

 certain temperature, below which one cannot absorb the other 

 except to a limited extent. The result is that as the mixture cools 

 it separates into two liquids which arrange themselves in layers 

 according to their density. Each of these liquids consists mainly 

 of one of the constituents with a certain proportion of the other. 



Vogt, however, concluded from his experiments that the 

 constituents of igneous rock magmas are as a rule miscible 

 in all proportions at all temperatures at which they remain 

 liquid. The only exceptions he admitted were certain sulphides, 



