LATER STAGES OF EVOLUTION OF IGNEOUS ROCKS 77 



rigidly adhered to in every case of slow crystallization of basaltic 

 magma. The plan of presenting the subject — ^that of offering 

 experimental evidence bearing on the questions and then deducing 

 therefrom the crystaUization of a magma — ^has led naturally to 

 the statement of crystallization given above. This may be regarded 

 as the normal sequence, for, considering the volume of rock types 

 produced, it is probably the most important. Any other succession 

 of types may be regarded as a variant from the above. In this con- 

 nection it is intended to discuss what appear to be two principal 

 variants and to suggest that the effects of pressure probably ade- 

 quately explain these variants. The effects are not those of pres- 

 sure per se, for these are probably altogether insignificant, but 

 rather of pressure in so far as it affects the retention of volatile con- 

 stituents, especially water. 



The sequence gabbro-diorite-quartz diorite-granodiorite-granite- 

 alkaline types expressed in the tabular statement given is that 

 normal to the Mesozoic-Tertiary intrusives of the Cordilleran 

 region of North America. The types characteristic of the pre- 

 Cambrian intrusives of the Adirondacks and adjacent regions 

 of Canada are rather gabbro-diorite-syenite-granite-alkaline t5rpes. 

 If we regard the nature of the plagioclase as a measure of the stage 

 of crystallization or the degree of concentration of the liquid residue 

 relative to the original volume of magma, it is clear that the types 

 of the eastern area exhibit a delaying of those reactions, character- 

 istic of the presence of water, which result in the separation of 

 significant amounts of biotite and quartz. It appears, therefore, 

 that at any stage, representing a certain degree of concentration 

 of the residual liquid with respect to the original magma, the active 

 mass of water was less during the development of the eastern types 

 than during that of the Cordilleran types. In any individual case 

 it would be impossible to refer such a difference in the activity of 

 water definitely to a difference of external pressure, but that this is 

 the kind of effect which would result and no doubt has resulted in 

 many cases is hardly to be questioned. 



The production of quartz and biotite in the presence of water 

 results in large part from the breakdown of KAlSiaOg molecules. A 

 low value of the active mass of water, sufficient to reduce this 



