THE PROBLEM OF THE ANORTHOSITES 211 



THE PROPOSED SOLUTION 



For this reason in part, but also because careful consideration 

 of all the possibilities and much experimental work to test these 

 possibilities seem to indicate the inadequacy of any process other 

 than crystallization, it is believed that the gabbroid magma must 

 proceed to crystallization, and that anorthosite masses are simply 

 collected plagioclase crystals. It is believed, then, that anorthosites 

 were never liquid as such, but that their material when liquid was 

 part of a solution probably of a gabbroid nature. Only in virtue 

 of the sorting of solid, crystalline units from this solution does 

 anorthosite come into being. 



Having arrived at this belief, we may examine the anorthosites 

 to see in how far they agree with its consequences, but before this 

 can be done it is necessary to discuss in detail the process of sorting 

 of crystals. 



THE PROCESS OF ACCUMULATION OF CRYSTALS 



General relations involved. — It must be admitted that the 

 problem of the method of accumulation of plagioclase crystals for 

 the formation of a mass of anorthosite is not a simple one. If the 

 plagioclase crystals were much heavier or much lighter than gab- 

 broic magma, all would be plain sailing. It could then be assumed 

 that the crystals sank in the magma or floated in it, and were there- 

 fore accumulated at the bottom or at the top. But laboratory 

 determinations of the densities of calcic plagioclase crystals and of 

 molten gabbro place them very close together, with the crystals a 

 very little lighter, and while the difference would not be the same 

 and might even be in the opposite direction, there is, nevertheless, 

 every reason to believe that it would still be small under natural 

 conditions. As a matter of fact, we actually find this similarity of 

 density expressed in the composition of anorthosites. If plagio- 

 clase were much lighter or much heavier, the mere accumulation of 

 crystals would be, as we have seen, a simple matter, but the 

 anorthosite masses formed would not be such as we find them, a 

 fact which will become obvious from the following considerations. 



During the crystallization of a magma involving the precipita- 

 tion of mix-crystals the crystals first deposited are rich in the higher- 



