THE GENESIS OF IGNEOUS ROCKS 311 



Still another complication presents itself where the magma 

 comprises two silicates which are capable of forming with each 

 other a combination in the nature of a double salt. This acts 

 to some extent as an independent substance, with the result 

 that there will be a different eutectic mixture according as the 

 original magma was intermediate in composition between the 

 double salt and the one or the other of the simple constituents. 



The influence of the volatile fluxes on crystallisation and 

 other phenomena connected with the consolidation of igneous 

 magmas has not yet been worked out in any detail, but there 

 is little doubt that they must profoundly affect the characters 

 of the rock ultimately formed. It is highly probable, for 

 instance, that they play an important part in determining the 

 late crystallisation of the more acid minerals, to which allusion 

 has already been made. Silica, we know, combines freely with 

 water under pressure at high temperatures, and this appears to 

 be also the case with the constituents of the alkali felspars, for 

 these with quartz make up almost entirely pegmatoid veins, 

 which are undoubtedly formed from magmas containing a large 

 proportion of water. The affinity of the alkali felspars for 

 water is easily explained by the supposition already advanced 

 that they are present in the form of silica and the soluble 

 aluminates of the alkalies. 



These felspars have, as a matter of fact, never been obtained 

 by the crystallisation of an artificial magma without the elements 

 of water or something capable of playing a similar part, and this 

 is usually explained by the fact that a magma consisting largely 

 of the elements of the alkali felspars is too viscous at the 

 temperature of consolidation to allow the molecules the mobility 

 necessary for crystallisation. 1 



Volatile fluxes are also needed for the formation of other 

 minerals, such as quartz, amphibole and mica. These are termed 

 " low temperature minerals " because they are transformed at 

 high temperatures into other minerals, and it is possible that the 

 volatile constituents may operate, in some cases at least, by 



1 Mr. Harker suggests that the abundance of orthoclase (sanidine) in many 

 lavas shows that in some magmas volatile constituents are not necessary for 

 the crystallisation of that mineral. It is, however, doubtful whether there are 

 any liquid lavas which do not still contain an appreciable amount of water. 

 Simultaneously with the loss of this, the whole rock crystallises out, so far as it 

 had not already done so. See, however, note, on p. 312. 



