324 Alexander Scott — Saturation of Minerals. 



Tschermak's silicate with, the production of olivine and anorthite. 

 Again, minerals such as tourmaline and topaz are usually the result 

 of pneumatolytic or fumarolic action, and as such it is impossible to 

 deduce from their co-existence with silica in rocks that they would 

 form during the normal cooling of an ' acid ' magma. 



From what has been said above it is obvious that a classification of 

 igneous rocks can no longer be made dependent on such a factor 

 as saturation with respect to silica, particularly when applied in this 

 empirical fashion. During the period of cooling, rocks are subjected 

 to many vagaries, owing to the impossibility of the external conditions 

 being uniform, and any classification dependent on the final 

 mineralogical composition without reference to the ' cooling-history ' 

 is not of much use. Any group of igneous rocks must be subdivided 

 according to the method of development of the particular rocks, and 

 it is impossible to compare directly on a purely modal basis a rock 

 which has formed through a noi-mal cooling process and is therefore 

 in a state of equilibrium and one which is not in equilibrium owing 

 to such causes as undercooling, inhibition of transformation either in 

 the liquid or solid states, or the formation of some or all of the 

 constituent minerals under abnormal conditions of pressure and 

 temperature. In rocks of similar composition, the minerals which 

 develop in the latter case are often widely different from those arising 

 in the former one, and hence such minerals as the garnets, topaz, etc., 

 cannot be said to be ' saturated ' or ' unsaturated ' in the sense that 

 albite, nepheline, or leucite can be. Similar objections exist with 

 regard to a classification on a basis of the chemicul composition or of 

 the 'norm' derived from it, and one of the chief advantages of the 

 American system is that it provides an extremely good method of 

 indexing the increasing volume of rock analyses. 



These two factors, however, the chemical and mineralogical com- 

 position, are the only ones we have for a basis in classification, but 

 they must be correlated in such a way that as much information as 

 possible is obtained regarding the ' cooling-history ' of the rocks. 

 The crystallization of rocks can no longer be considered either as 

 wholly fortuitous, or as wholly the result of a normal cooling process, 

 but as the result of the operations of the laws of physical chemistry. 

 Hence the objects of a rock classification must be to correlate rocks 

 not only of similar chemical composition but also with similar 

 ' cooling-histories'. 



The data for this can be derived from (a) the determination of 

 chemical composition, (b) of mineralogical composition, (c) the field 

 relations of the rocks, {d) the evidence of synthetic work in mineralogy. 

 The defects of most of the current classifications can be traced to the 

 position of paramount importance given to some one of these bases 

 and to the subordinate positions assigned to the others. The occurrence 

 of minerals in rocks, as shown by the 'observed' facts of distribution, 

 must be correlated with the other evidence — particularly that of field 

 relations and of synthetic work — concerning the conditions of 

 equilibrium both during the cooling process and in the solid phases 

 observed in the rock. It is only in this way that we can acquire 

 a knowledge of the genetic relationships and the general history of 

 igneous rocks. 



