LATER STAGES OF EVOLUTION OF IGNEOUS ROCKS 57 



highly femic and highly alkalic types (rich in feldspathoids) may 

 result. The more "basic" types of alkaline rock are not, however, 

 in all cases basic differentiates from nephelite syenite magma. 

 The reactions preliminary to the separation of quartz and biotite 

 begin at an early stage in the crystalHzation of basaltic magma, 

 and the separation of these minerals may take place at an early 

 stage, giving rise to quartz diorites or granodiorites. The possi- 

 bility of the formation of alkalic magma at a stage much earHer than 

 the biotite granite stage is thereby introduced if conditions are 

 favorable. Favorable conditions seem to consist in the oppor- 

 tunity for sinking, not only of the plagioclase crystals and femic 

 minerals, but also of quartz crystals in sufficient amount. Thus 

 may result relatively "basic" alkaline magmas from which such 

 rocks as basanite might be formed, and nephehte syenite itself as a 

 light differentiate. 



In passing from the normal biotite granites direct to the nephe- 

 lite syenites and related rocks, an important stage has been passed 

 over, viz., the stage of the alkaline granites and syenites. Many 

 occurrences, to be referred to later, show that alkaline granites and 

 syenites are intermediate in their time relations between normal 

 biotite granite and feldspathoid-bearing types. These facts find 

 a likely explanation in terms of the general system here outlined. 

 The precipitation from the granite magma of relatively large 

 amounts of potash molecules as orthoclase and of both potash and 

 magnesia as biotite to form normal granite results in a relative 

 concentration of sodic and iron molecules which are precipitated 

 as relatively large amounts of albite and as the alkaline pyroxenes 

 and hornblendes characteristic of alkaline granites, i.e., of this state 

 of precipitation. 



In north-central Wisconsin the sequence described by Weidman 

 — gabbro, diorite, normal granite, soda granite, quartz syenite, 

 nephelite syenite — strictly accords with the outline of differentia- 

 tion offered.^ 



The sequence in Essex County, Massachusetts, studied in 

 great detail by Clapp, might be described substantially by repeat- 

 ing the type names used above in connection with the Wisconsin 



^ Wis. Nat. Hist. Survey, Bull. 16, pp. 349 f. 



