48 JOURNAL OF GEOLOGY— SUPPLEMENT 



border phases is therefore substantially correct. It is not, how- 

 ever, necessary that the border phase is the undifferentiated original 

 magma. It is better stated in the more general form that the 

 border phase represents a less advanced stage in the differentiation 

 of the magma. A satelHtic stock of, say, diorite occurring in con- 

 nection with a granitic batholith is to be explained similarly. 



The occurrence of a basic marginal phase at the top of a batho- 

 lith is a reversal of the general arrangement according to density 

 which results from the free working of the processes of differentia- 

 tion. This is not due to the action of any positive process tending 

 to produce movement of material of early separation in the oppo- 

 site direction, but to the constant tendency of cooling to put an end 

 to the gravitative process. It is to be noted that the basic border 

 phase is a completely frozen, solid rock during the period of con- 

 tinuance of differentiation in the liquid immediately below it. 

 There is no violation of the law of gravitation. If a piece of the 

 solid rock becomes broken off, it will, of course, sink in the liquid, 

 but only then. It is in this manner that many basic cognate 

 xenoliths are formed. It is clear also that the salic phase may be 

 injected into fractures in the frozen border types formed from the 

 same magma. The observance of this phenomenon is not to be 

 considered as indicating any considerable difference in age. 



FIELD EVIDENCES OP THE CONTROL OF CRYSTALLIZATION IN 

 DIFFERENTIATION 



The foregoing deductions relative to the course of crystalliza- 

 tion and therefore of differentiation are fully supported by field 

 evidence. Daly's experience has led him to the same conclusion 

 with regard to the results, though the attainment of the results is 

 differently explained by him. Speaking of his theory he states: 

 "It considers the final acid pole of splitting in large post-Keewatin 

 bathoHths as granite." When differentiation is "arrested midway" 

 "rocks of intermediate composition result."^ He rightly con- 

 siders that an intermediate rock such as diorite is the result of 

 arrested differentiation and not an anchi-eutectic rock formed by 

 the crystallization of a final eutectic liquid beyond which differ- 



' Igneous Rocks and Their Origin, p. 361. 



