396 N. L. BOWEN 



was inserted in the definition to make it expressive of the general 

 agreement among petrologists to exclude from the category of 

 differentiation those variations of igneous rocks that result purely 

 and solely from assimilation of foreign material. That syntectic 

 magma so formed could not suffer differentiation was no part of 

 the intent of the definition. 



In that part of the paper in which he discusses the differentia- 

 tion of a syntectic magma the writer does, however, state his 

 belief that such differentiation must usually be of limited scope, 

 giving his reasons and describing the limiting factors. Contrasted 

 with this is the assertion of Loewinson-Lessing, to which Daly 

 offers adherence, that "a small degree of contamination with foreign 

 material may change equilibrium in the magma, which, therefore, 

 separates into strongly contrasted parts."' Such a generalized 

 statement leaves one wondering as to the nature of the equilibrium 

 referred to. Between what is the equilibrium? What is the nature 

 of the strongly contrasted parts ? Are they solid, liquid, or gase- 

 ous ? Do rocks show the evidences of such processes ? Have 

 analogous processes been encountered in experimental science, 

 and is the analogy of real significance ? 



In going on to a discussion of differentiation Daly proceeds to 

 enumerate what may be called the units of differentiation. Some 

 of these are presumably the strongly contrasted parts formed as 

 a result of the disturbance of equilibrium mentioned above, but 

 no particular mention is made of those that are to be so regarded, 

 nor is any discussion offered of the manner in which solution of 

 solid rock causes the separation of the units. 



The units enumerated by Daly are the following: 



1 . Contrasted fluid phases of an initially heterogeneous magma, 

 including parts particularly rich in volatile constituents.^ 



2. Solid crystals (fractional crystallization). 



3. Mother-liquor left after partial crystallization. 



' This is Daly's form of expression of the Loewinson-Lessing idea. 



^ Units of type i need not be considered, for according to Daly's own statements 

 on the page immediately preceding, "rock phases resulting from originally different 

 parts of the magma can be called differentiates only by destroying the useful definition 

 of differentiation already adopted by most petrologists, " and in this the writer concurs. 



