CRYSTALLIZATION-DIFFERENTIATION IN MAGMAS 397 



4. Non-consolute liquid fractions (liquid immiscibility) . 



5. Material of fused country rock, not diffused into the original 

 magma (ultra-metamorphism in part) . 



6. Original magma locally charged with material dissolved 

 from the country rock but slowly diffusing from the source of 

 supply (syntexis). 



Units 5 and 6 are plainly produced by the acquisition of foreign 

 material but cannot be regarded as a separation of the original 

 magma into contrasted parts as a result of this acquisition. The 

 contrast is rather between the original magma and the included 

 foreign material. No doubt such units are real factors in differ- 

 entiation, but one may reasonably question whether they are of 

 great importance. Differentiates formed from such material 

 would have no necessary consanguinity with other differentiates 

 formed from the magma itself, since their composition would 

 depend principally on the nature of the included material, that is, 

 on a purely accidental circumstance. It is, however, the fact that 

 differentiates do normally show consanguinity that has led to the 

 whole conception of differentiation in petrology. A differentiate 

 formed by the accumulation of these clots of foreign material would 

 necessarily show a very simple chemical relationship with the 

 available foreign material. It is the fact that this simple relation 

 does not exist that led Daly himself to reject purely marginal 

 assimilation, and that should furnish an equally cogent reason for 

 rejecting accumulation of xenolith material as an important 

 factor. 



Daly's case is much stronger in this respect when he assumes 

 complete assimilation (usually abyssal) of the foreign material 

 and subsequent differentiation of the syntectic magma, for then 

 all the differentiates would show the requisite consanguinity. 

 We are thus brought back to a consideration of the separation of 

 phases (in the more definite physicochemical sense of the term) 

 as the fundamental factor in differentiation, and the possible 

 units then become those enumerated by Daly as 2, 3, and 4. 

 The processes involved, namely the separation of a crystalline 

 phase or of a non-consolute liquid • phase, must be regarded as 

 possible factors in the differentiation not merely of Daly's syntectic 



