LATER STAGES OF EVOLUTION OF IGNEOUS ROCKS 79 



These differences are to be connected in part with differences in 

 external pressure, but the nature of surrounding rocks and, to some 

 extent, original differences in the magmas are to be considered as 

 well. 



MONOMINERALIC ROCKS 



The formation of " anchi-monomineralic " rocks is possible in 

 the earliest stages (and apparently only then), whatever line of 

 descent is followed, if conditions favor the collection of the early 

 crystals, principally olivine, pyroxene, plagioclase, and iron ores. 

 The more or less perfect sorting of these crystals which results in 

 the formation of dunite, pyroxenite, and anorthosite should be 

 expected normally only in a large body of magma which cools very 

 slowly through the early stages of crystallization. There must be 

 a considerable period during which olivine is separating alone and 

 also a considerable period during which pyroxene and plagioclase, 

 though separating together, for the most part, form only a small 

 fraction of the total mass (liquid and crystals) , so that the pyroxene 

 crystals may assert their greater density without much interference 

 from the hghter plagioclase crystals. It appears, indeed, that 

 plagioclase crystals may at the earlier stages of crystalHzation be 

 only very sHghtly heavier than the liquid at most and possibly even 

 somewhat Hghter. This seems to be the promising period for the 

 collection of plagioclase crystals nearly free from pyroxene. At a 

 later period, provided there is sufficiently slow cooling, these 

 plagioclase crystals will sink and the normal effect on the 

 composition of the liquid will ensue. The formation of syenitic 

 and granitic differentiates is therefore not precluded by the 

 formation of anorthosite, i.e., anorthosite is not the result of an 

 alternative Kne of descent. This fact is shown by the Duluth 

 gabbro laccohth, probably by the complex-gabbro, anorthosite, 

 syenite, etc., in the Long Lake quadrangle. New York, described 

 by Gushing ;'^ by the norite, anorthosite, banatite, granite, etc., of 

 the Ekersund-Soggendal region of Norway f and by a rather similar 

 association in the Nordingra region of Sweden.^ 



' N.Y. State Museum, Bull. 115, 1907. 



^ Kolderup, Bergens Museums Aarbog, 1896, No. V. 



3 Jose M. Sobral, Contributions to the Geology of the Nordingra Region, Upsala, 1913. 



