CRYSTALLIZATION-DIFFERENTIATION IN MAGMAS 399 



it may be regarded as not improbable that two molten magmas 

 might in some cases mix with evolution of heat, but that the 

 solution of solid rock in molten magma is ever accompanied by 

 evolution of heat may be regarded with extreme doubt. It is 

 this reasoned conviction that has led the writer to limit the 

 importance of assimilation. 



LIQUID IMMISCIBILITY AS A FACTOR IN DIFFERENTIATION 



Having in mind then the factors that may be considered to 

 restrict the formation and dilTerentiation of syntectic magmas, we 

 may now return to the question of the separation of phases as the 

 agency of differentiation of magmas, presumably dominantly 

 juvenile and uncontaminated. The possible phases are, as we have 

 seen, the units 2, 3, and 4 of Daly. Units 2 and 3 are, of course, 

 necessarily concomitant, and to them the writer assigns the funda- 

 mental role in differentiation, but Daly is inclined to belittle their 

 importance in favor of unit 4, that is, non-consolute liquid phases. 

 The writer considers these to be non-existent in silicate magmas, 

 this conclusion being based in part on experimental evidence. Not 

 a single example of liquid immiscibility among silicates has been 

 encountered in experimental work, in rjietallurgical practice, or 

 in any of the many industries where molten silicates are treated. 

 The single example of immiscibility of which there is good evidence 

 in natural rocks, that between silicates and sulphides, is promptly 

 revealed experimentally and in metallurgical practice. To many, 

 however, the experimental evidence will never be convincing. 

 When the range of experimentation with silicates shall have been 

 extended to pressures of 20,000 atmospheres it will be possible to 

 state that at pressures of 30,000 atmospheres wholly different 

 phenomena might occur. Or again, an experiment made by human 

 hands will probably never be prolonged to such a degree that its 

 duration might be measured in centuries, so that it will always be 

 possible to say that the time factor was not adequately considered. 

 However, it is desirable to give some consideration to that experi- 

 mental evidence which Daly brings forward, and which seems to 

 him to indicate the probability of immiscibility under high 



