2S8 Corresjiondence — N. L. Boiven. 



CRYSTALLIZATION-DIFFEEENTIATION. 



Sir, — Mr. H. H. Read's acquaintance with modern French 

 literature is evidently thorouoh, and one is surprised that this 

 acquaintance has not rendered him more apt at simile. He compares 

 my dependence upon crystallization as an explanation of many 

 features of igneous rocks with M. d'Astarac's dependence upon 

 sylphs as an explanation of ordinary happenings. Unfortunately 

 for his comparison, we know that magmas do crystallize, and 

 equally unfortunately for his dej)endence upon liquid immiscibility 

 we do not know that silicates unmix. Immiscibility is the petrologic 

 sylph. 



In fact, inimiscibilit}^ is a sort of super-sylph. One sees two 

 related and intimately associated rocks, and one remarks that they 

 were evidently formed from a common magma by unmixing. Since 

 nothing is known about unmixing in silicates, one is therefore 

 relieved from the responsibility of any further thought in the matter. 

 However, Mr. Eead's last paragrajih rouses some hope. He expects 

 that my excessive advocacy of crystallization will lead some one to 

 champion immiscibility. If this some one will analyse immiscibility 

 from a theoretical standpoint, if he will then apply his results to 

 silicate magmas and examine whether associated igneous rocks 

 bear such a relationship to each other in chemical composition, in 

 time and space relations, that they can or cannot be regarded as the 

 result of immiscibility, he will have conferred a boon upon 

 petrologists, Avhether a succeeding generation shall find his con- 

 clusions right or wrong. But no amount of setting up immiscibility 

 as a sylph will advance petrology. 



It is jDossible, of course, that I have overworked crystallization, 

 but no one who has seen the spring break-up in one of our Canadian 

 lakes could doubt that deformation of a crystal mesh must have 

 important consequences in the case of igneous rocks as Avell. There 

 one sees weakened and honeycombed bodies of ice under impact 

 from other masses, locally compacted into a solid mass with the 

 water squeezed out of the comb, elsewhere stretched, fissured, and 

 traversed by streaks of open water. That more or less related 

 features could develop below ground is not to be questioned. 



Though Mr. Read's review is for the nfost part a series of objections 

 they are nearly always generalities, and are only occasionally 

 specific enough to be answerable. My method of deriving some ^ 

 banded rocks by torsion of a crystal mesh would not, as he concludes, 

 give rise to an orientation of the early crystals normal to the bands. 

 It is in the filling of the lenticular spaces that free flow of liquid 

 occurs, and such crystals as would become detached from the Avails 

 of the lens during this action would be carried along by the liquid 



' I have elsewhere suggested that banded rocks may at times be formed as 

 a result of intrusion of heterogeneous liquid (not immiscible liquids). " Later 

 Stages of the Evolution of the Igneous Rocks": Joiirn. Geol. , Suppl. to 

 vol. xxiii, 1915, p. 30. 



