R. H. Rastall — Differentiation and Ore-deposits. 291 



As we have already stated, much difference of opinion has prevailed 

 and still continues to prevail among petrologists as to the real 

 nature of the processes by which differentiation of rock-magmas 

 has been brought about. From our j)oint of view the elucidation of 

 the exact mechanism of the jorocesses involved is of less importance 

 than a clear recognition of the fact that differentiation has taken 

 place. Here it is necessary to consider for a moment what we really 

 mean by differentiation. As a matter of fact two distinct ideas are 

 here involved, namely, (a) the production of heterogeneity within 

 the limits of a single continuous mass, and (&) the splitting of an 

 originally homogeneous mass into two or more separate and discrete 

 portions.^ It may be regarded as certain that both of these effects 

 have been produced in different ways in different cases. 



Theories of differentiation may be divided into three principal 

 categories, as follows : (1) differentiation by sinking of crystals in 

 a liquid magma, (2) differentiation by diffusion of molecules in 

 solution to the marginal portions of a cooling and crystallizing 

 mass, (3) separation of a solution into two immiscible fractions as 

 temperature falls. It will be observed that (1) and (2) are very similar, 

 while (3) is essentially different, in that its occurrence may be long 

 antecedent to solidification. Hence we see that there are really 

 two types of differentiation, one depending on solidification 

 (crystallization), the other being entirely independent of it. Another 

 point which has led to much discussion is this, whether any form 

 of differentiation is dependent on assimilation of rock from the 

 walls of the magma-basin. This hypothesis of the production of 

 heterogeneous differentiation products by absorption of material 

 acting as a flux has been largely developed by Loewinson-Lessing ^ 

 and by Daly.^ Both of these authors consider that there were two 

 primary magmas, granitic and basaltic respectively, and Daly regards 

 the alkaline rocks, the syenites and trachytes and their congeners, 

 as formed by differentiation from the primary granitic magma after 

 assimilation of limestone. To this process he applies the name 

 syntexis. Neither author has apparently considered the origin of 

 the two primary magmas from a still earlier common source, a 

 possibility which must also be taken into account in a complete 

 theory. 



It will, I think, be generally agreed that in a case of this kind the 

 best method of procedure is to start from a foundation of known 

 facts and work back to the unknown. Let us, therefore, begin with 



^ It has been well pointed out by Daly that the formation of crystals of 

 different minerals, such as quartz, felspar, and mica, from a magmatic solution 

 is in reality a process of differentiation in its simplest form. Each single 

 crystal is in such a case a unit of differentiation. This point has been ignored 

 by most petrologists. It is in reality only the mono-mineralic rocks which 

 can be regarded as strictly homogeneous, and these are very rare. 



"^ Loewinson-Lessing, Geol. Mag., 1911, p. 248. 



* Daly, Igneous Bocks and their Origin, New York, 1914. 



