76 JOURNAL OF GEOLOGY— SUPPLEMENT 



The preceding outline is intended to represent the usual case. 

 The period of crystallization of any stage is somewhat later than 

 that of the stage above it. The crystallization of a mineral placed 

 under any stage is essentially simultaneous with that of the other 

 minerals grouped with it. There is, of course, considerable over- 

 lapping from stage to stage; indeed, the division into stages is 

 purely arbitrary, for the process itself is perfectly continuous. 



The writer has some hesitation in presenting this very definite 

 statement concerning crystallization, for he knows full well that it 

 cannot be absolutely correct. Nevertheless, it is believed that if our 

 knowledge of the subject were complete, the final correction of 

 the statement would be principally an elaboration rather than an 

 alteration. In presenting the foregoing scheme, the purpose is 

 chiefly to emphasize the point of view from which the crystallization 

 of a magma is to be attacked. The great difference should be 

 realized between the simple group of minerals which results when a 

 magma is crystallized quickly and the long series of products which 

 may result from slow crystallization. The fundamental controlHng 

 influence of the plagioclase mix-crystal series must be recognized; 

 also the fact that the minerals precipitated give clues to certain 

 equilibrium reactions in the liquid. Ferromagnesian material in 

 one form or another must appear at all stages of precipitation. If 

 a chemist has a mixture of, say, potassium and sodium salts, he has 

 yet to discover a method of cooling or evaporation whereby all 

 the sodium salts can be removed from the solution so long as any 

 solution is left. 



The great lowering of the temperatures of crystallization of 

 granitic magma as the result of the concentration of volatile con- 

 stituents has often been emphasized, but the converse is worthy of 

 equal emphasis. The relatively great concentration of volatile sub- 

 stances in granitic magma is the result of its low temperature of 

 crystallization and follows as a necessary consequence of the long- 

 preceding period of crystallization in virtue of which alone granitic 

 magma comes into being. 



VARIATIONS FROM THE STATED SCHEME OE CRYSTALLIZATION 



In presenting the schematic statement of the progress of crystal- 

 lization the writer does not intend to imply that this sequence is 



