THE REACTION PRINCIPLE IN PETROGENESIS 



189 



table given by Harker showing the mineral relations in the sequence 

 at Garabal Hill' 



TABLE I 



Mineral and Rock Sequence at Garabal Hill (After Harker) 



It will be noted that the minerals appear in a certain order, as 

 they might in a system where simple eutectic relations prevailed, 

 but they also disappear in a similar order, a feature that is altogether 

 foreign to a eutectic system. In a eutectic system no mineral ever 

 disappears.^ The first-formed mineral is simply joined by another, 

 the pair by a third, and so on until all the minerals appear together 

 in a final eutectic product. Very different from this is the condi- 

 tion actually found, namely, the disappearance of minerals in the 

 order in which they appear which is of the very essence of the reac- 

 tion series. 



Upon examination in detail it is foimd that 2,3, and 4 disappear 

 in B, 5 disappears in D, and 6 in E. From this we conclude that 

 these phases bear a reaction (not a mere subtraction) relation to 

 the Kquid and that, as a result of the reactions, phases appearing 

 later are formed. We arrive at the definite conclusion that 4, 

 5, and 6 constitute a reaction series and at the same time note 

 indications that they are but a part of a series containing more 

 members. 



' The Natural History of Igneous Rocks, 1909, p. 131. 



^ A case of simple inversion without change of composition would require to be 

 excluded from this statement, but it has no importance in rocks at any rate. 



