146 CARNEGIE INSTITUTION OF WASHINGTON. 



(15) The problem of the anorthosites. N. L. Bowen. J. Geol., 25, 209-243 (1917). 



Anorthosites are made up almost exclusively of the single mineral plagio- 

 clase and in virtue of this fact they present a very special problem in petro- 

 genesis. The conception of the mutual solution of minerals in the magma 

 and the lowering of melting temperature consequent thereon is no longer 

 applicable. Yet anorthosites give no evidence of being abnormal in the mat- 

 ter of the temperature to which they have been raised ; in other words, they 

 give no evidence of having been raised to the temperature requisite to melt 

 plagioclase. A possible alternative is that they may never have been molten 

 as such and are formed simply by the collection of crystals from a complex 

 melt, probably gabbroic magma. This possibility is in harmony with the 

 expectations that grow out of experimental studies and for this reason a con- 

 sideration of the likelihood that anorthosites have originated in the stated 

 manner becomes imperative. 



A consideration of the method whereby accumulation of plagioclase crystals 

 might take place leads to the conclusion that the most promising is the separa- 

 tion by gravity of the femic constituents from gabbroid magma while the 

 plagioclase crystals, which are basic bytownite, remain practically suspended. 

 Then, at a later stage, when the liquid has become distinctly lighter, having 

 attained diorite-syenite composition, the plagioclase crystals, which are now 

 labradorite, accumulate by sinking and give masses of anorthosite, at the same 

 time leaving the liquid out of which they settle of a syenitic or granitic com- 

 position. 



Some of the consequences of this manner of origin of anorthosite are as 

 follows : Typical anorthosite very poor in bisilicates should not occur as small 

 dikes, for a mass of accumulated crystals should have little invading power. 

 A proportion of about 15 or 20 per cent bisilicates or other foreign material, 

 such as orthoclase and quartz, should be necessary for the formation of small 

 dikes. Typical anorthosite should for like reasons not occur as an effusive 

 rock, a rather large proportion of minerals other than plagioclase being neces- 

 sary before such an occurrence would become possible. Anorthosite should 

 be intimately associated with gabbro, but perhaps as intimately with syenite 

 or granite. Anorthosites should commonly be labradorite rocks rather than 

 bytownite or anorthite rocks. 



A consideration of anorthosites with special reference to the Adirondack and 

 Morin areas gives some reason for believing that anorthosites do show the 

 requisite characters. For the Adirondack area especially, evidence is adduced 

 favoring the possibility that there anorthosite and syenite may still occupy 

 the relative positions in which they were generated by the process outlined, 

 the Adirondack complex being interpreted as a sheet-like mass with syenite 

 above and anorthosite below. 



Other monomineralic rocks present essentially the same problem and are 

 restricted in their occurrence in substantially the same manner if we consider 

 especially those that approach most closely to the strictly one-mineral charac- 

 ter. All of the monomineralic rocks do occur, however, as dikes and dike-like 

 masses in essentially contemporaneous, congeneric igneous rocks, a fact which 

 may be interpreted as due to the intrusion of a heterogeneous, partly crystal- 

 line mass. 



On the whole the inquiry gives considerable support to the belief that the 

 monomineralic rocks, of which the anorthosites are perhaps the most important 

 representatives, are generated by the process of collection of crystals under 

 the action of gravity. 



