DIFFERENTIATION OF IGNEOUS ROCKS 419 



low temperature of crystallization. That this is true we have proved 

 above in an empirical way. 



We may note at the same time that a granitic magma, according 

 to its history (as will be more fully explained in a later paper), 

 can never have been overheated above the upper boundary of the interval 

 of the crystallization. 



Any hypothesis based on one hundred degrees or more of 

 overheating of the granitic magma above the temperature at which 

 the solidification began, must be rejected as unmaintainable. 



