DIFFERENTIATION IN INTERCRUSTAL MAGMA BASINS 



ALFRED HARKER 

 Cambridge University 



Dr. N. L. Bowen's comprehensive article on "The Later Stages 

 of the Evolution of the Igneous Rocks," issued as a supplement to 

 the last volume of this Journal, will be hailed with satisfaction 

 by all petrologists, and indeed with gratitude by those who have 

 the misfortune not to be chemists. It contains the first serious 

 attempt to deal with the problem of magmatic differentiation 

 directly from the standpoint of experimental knowledge. In 

 demonstrating how the course of crystallization may be changed 

 by the sinking of crystals, or the straining away of liquid from 

 crystals, or the formation of zoned crystals in isomorphous groups 

 of minerals, the author scarcely goes beyond actual laboratory 

 experience, and his conclusions accordingly carry a great weight 

 of authority. When he proceeds to construct on this basis a general 

 theory of differentiation, the element of hypothesis is necessarily 

 introduced, and, as the author recognizes, his argument can no 

 longer command unquestioning acceptance. It is a very interest- 

 ing contribution to a discussion which is not likely soon to be closed. 



I wish to make a few remarks upon one of the subsidiary issues, 

 which, however, touches the main theory at numerous points, viz., 

 Bowen's predilection for differentiation in situ as opposed to differ- 

 entiation prior to intrusion. That an appreciable settling down 

 of crystals may take place after intrusion is not to be denied, but 

 I think that the experience of any field geologist goes to show that 

 it is a rare and exceptional incident. Daly has given a list of about 

 thirty stratified sills and laccolites in which such "gravitative" dif- 

 ferentiation is believed or conjectured to have occurred, but proba- 

 bly a critical examination would dispose of many of the examples 

 cited. In some, such as the Loch Bordan mass in Sutherland, there 

 is not a gradual transition but a sharp boundary between the several 



554 



