ORIGIN OF AUGITE ANDES ITE 413 



temperature and therefore during a less viscous stage of the magma's 

 history. In an under-cooled silicate melt crystals must sink but 

 slowly. 



The crystallization of most plutonic magmas and of lavas at the 

 earth's surface are thus believed to take place under strongly con- 

 trasted conditions. At the surface of a lava-column olivine, mag- 

 netite, augite, and plagioclase begin to form at high temperatures, 

 when the viscosity is relatively low; the time-interval of crystallization 

 for the whole magma is relatively long. Under plutonic conditions 

 the same minerals crystallize at lower temperatures, when high 

 viscosity is established and (because of under-cooling) the time- 

 interval of crystallization is probably short. 



Nevertheless, in some plutonic bodies of large size a certain amount 

 of fractional crystallization and settlement of basic minerals may 

 have actually taken place. Kemp has suggested that the anorthosites 

 of the Adirondacks have been derived from more normal gabbro 

 as a result of the sinking of the heavy constituents, either crystallized 

 or still molten. He considers that the titaniferous magnetites of the 

 region were erupted into the anorthosite after the anorthosite had 

 consolidated, and implies that the molten ore has been derived 

 from a couche which was formed from settled-out magnetite.^ The 

 relation of anorthosite and magnetite would, thus, be analogous to 

 that postulated between augite andesite and certain olivine-rocks. 

 Secondly, it is further possible that many basic contact-zones represent 

 those parts of the respective intrusive masses where, on account of 

 the relatively rapid cooling, the settling of basic minerals has been 

 specially restrained, while the process has notably affected the com- 

 position of the slower-cooling interior of each mass. According to 

 Kemp and Gushing^ the huge anorthosite masses of the Adirondacks 

 are characterized by wide contact-zones of rock relatively rich in 

 pyroxene and magnetite (gabbro and anorthosite-gabbro) . This 

 gabbroid phase is usually finer grained and was presumably chilled 

 more rapidly than the main body of the anorthosite. 



1 Nineteenth Annual Report, U. S. Geological Survey, pt. 3, 1899, p. 417. 



2 H. P. Gushing, Geology of Franklin County, i8th Report of the State Geologist 

 (Albany, 1900), p. loi; New York State Museum Bull. No. gj (1905), p. 305 and 

 Bull. II j (1907), p. 471. • 



