LATER STAGES OF EVOLUTION OF IGNEOUS ROCKS 63 



stituents of the magma, it is to be expected that the minerals 

 precipitated would give evidence of richness in those volatile con- 

 stituents which are known to be constituents of magmas. The 

 minerals are, in the present case, sodalite and cancrinite, in other 

 cases noselite and hauynite as well. Moreover, under this theory, 

 this light residual magma tends to accumulate, during the course of 

 differentiation by crystallization, in the upper and outer parts of 

 the batholithic chamber, and in a limestone terrane it is especially 

 likely to border on the limestone. This theory, then, assigns to the 

 limestone no essential action in the formation of the alkaline types 

 as is obviously true in other localities where limestone is absent 

 in the country rock or present only in relatively small quantity. 

 It is, for example, difficult to imagine how an occasional bed of lime- 

 stone in a terrane consisting principally of siliceous gneisses could 

 be considered a possible agent of desilication, for its absorption 

 would entail simultaneous absorption of a much greater quantity 

 of relatively siliceous material. 



To return to the idea of the desilication of the magma, either 

 localized or otherwise, through interaction with limestone in the 

 Bancroft area, it may be pointed out that there are in this area 

 many bodies of diorite of no inconsiderable size which one would 

 expect to have suffered similarly, for they also have caused amphi- 

 bolitization, etc., of the limestone. Indeed, the task set the lime- 

 stone would seem to be much less arduous in this case than in the 

 case of granitic magma. The reduction to nephelite of the poly- 

 silicate molecules represented in the alkaline feldspars would need 

 to be preceded in the granitic magma by the binding of the quartz, 

 whereas this would be unnecessary in the diorites. Nevertheless 

 the diorite bodies apparently fail to show borders of nephelite- 

 bearing rocks. They belong, not with the diorites, but with the 

 quartz-rich rocks, the granites, as they should under the present 

 conception of their origin. 



Neither is the Bancroft area the only one that exhibits this inti- 

 mate connection of the nephelite-bearing rocks with granites. 

 Several cases have already been referred to in discussing the 

 sequence of types and several more could be added. At this point 

 only three others will be mentioned. The nephelite syenites 



