LATER STAGES OF EVOLUTION OF IGNEOUS ROCKS 49 



entiation cannot proceed. The use of the word "spHtting" indi- 

 cates that Daly considers that the medium of accompHshment of 

 differentiation is liquid immiscibihty, a process of which there is 

 no clear evidence and of which a detailed description applied to 

 rocks has never been offered by any investigator. The facts are, 

 however, in perfect accord with the details of differentiation by 

 crystallization given in the present paper. 



The gravitative arrangement of the various differentiates of the 

 Moyie sills and their individual mineralogy are absolutely that to be 

 expected if differentiation by crystallization as outlined is the sole 

 control.^ They vary from normal diabase through hornblende 

 diorite to biotite granite. 



Somewhat similar sills in the Gowganda Lake District of 

 Ontario, described by the writer, have essentially the same relations. 

 In the original paper it was considered that the surrounding sedi- 

 ments played an important part in the formation of the granophyric 

 bodies at the upper surface of the sills .^ This opinion was arrived 

 at principally because of the difficulty of picturing any process of 

 pure differentiation whereby a quartzose rock could be formed from 

 basaltic magma. With this difficulty removed the writer has no 

 hesitation in concluding that the granophyre and the micro- 

 pegmatite interstices of the diabase were formed after the manner 

 detailed in the present paper and that interchange of material 

 between the granophyre and adinolized sediment was a subsidiary 

 process^ contributing to the soda-rich nature of the border phases. 



The great sheetlike mass at Sudbury with granite in the upper 

 part and norite in the lower shows the same type of crystalHzation 

 differentiation controlled by gravity. 



There are also many cases of this general type with the addi- 

 tional feature that peridotite occurs as a basal differentiate. The 

 Duluth laccolith, closely related in age and in origin to the Sudbury 

 mass, is a case in point.'' 



'Daly, Am. Jour. Set. (4), XX (1905), 187; Schofield, "The Origin of Granite 

 (Micropegmatite) in the Purcell Sills," Geol. Survey Canada, Museum Bull. No. 2, 

 pp. 1-34. 



= N. L, Bowen, Jour. Geol., XVIII (1910), 658. 



3 Cf . CoUins, Geol. Survey Canada, Mem. 33, pp. 59 f. 



4 W. S. Bay ley, Jour. Geol, II (1894), 814. 



