218 N. L. BOW EN 



could have had not more than 10 per cent liquid. For natural 

 melts the figures would not be the same, and the probability is that 

 the amount of liquid would be relatively somewhat larger on account 

 of the presence of orthoclase in the liquid. Assuming the figures 

 to be approximately the same, it seems necessary to believe that a 

 rock containing 95 per cent or more plagioclase, if it is true that it 

 is formed by the method outlined, should exhibit certain charac- 

 teristics that set it apart from such a rock as a granite, which, as 

 we know well enough, often occurs in the completely molten 

 condition. When the plagioclase rock is formed in situ, it need 

 exhibit no features differentiating it from other igneous rocks 

 except perhaps a marked coarseness of grain. Such a body, while 

 still containing nearly its maximum of about 20 per cent inter- 

 stitial liquid, might be moved en masse, though probably not far, 

 from the position of its original formation, but- this movement 

 would be accompanied by the development of protoclastic structure, 

 especially about the margins. Since all crystalline igneous rocks 

 pass through a stage at which they are 80 per cent crystalline, all 

 are subject to the possibility of the development of similar structures 

 under parallel conditions. The plagioclase rock differs only in 

 that it cannot be moved without developing this structure, since if 

 moved when containing more than 20 per cent liquid the mass 

 moved has not yet attained the requisite degree of concentration of 

 plagioclase crystals. Protoclastic structure and granulation should 

 therefore be perfectly general features of all moved anorthosite 

 masses and very common features of all anorthosites. 



When we come down to the movement of such material in small 

 masses, it seems impossible that it would be capable of being injected 

 into small openings in cold country rock — in other words, that it 

 would form no small dikes in such rocks, though it might occur as 

 dikelike masses in consanguineous igneous types, being injected 

 into them at a time when they themselves were not completely 

 crystalline. Such material should, moreover, be incapable of 

 occurring as effusive flows. 



For the purposes of the foregoing discussion a mass containing 

 20 per cent interstitial liquid has been arbitrarily chosen. It is a 

 matter of opinion how much liquid a mass must have in order to 



