METAMORPHIC STUDIES 605 



analyses of common schists have been platted graphically in order 

 that they may be compared with both known and possible parent 

 rocks. An examination of these plates brings out the fact that the 

 composition of the schist tends to approach the distinctive chemical 

 characteristics of the dominant platy or columnar mineral in the 

 schist. This is especially well shown in the sericite schist, the com- 

 position of which lies between that of the primary rock and that of 

 the mineral sericite. It is indicated, not only by the position of 

 the analyses on the diagram, but by the shapes of the flags showing 

 the relative amounts of constituents. If we were to include in 

 these plates only those schists in which the processes of anamor- 

 phism and rock ilowage have gone to an extreme, the tendency 

 would be still more obvious, because rocks are included in these 

 diagrams, described as hornblende, chlorite, or mica schist, which 

 have but a small proportion of these minerals. 



Our inference from the available facts is that, while recrystal- 

 lization of substances present has of course played an important 

 part in the production of schists, for some rocks important changes 

 in composition have also occurred; that these changes in compo- 

 sition have been toward the composition of the characteristic 

 end-products — mica, hornblende, or chlorite; that these changes 

 are known both in sedimentary and in igneous rocks, of both acid 

 and basic composition, and that the changes have been sufficiently 

 important to make it impossible, along with other reasons, to use 

 chemical composition as a conclusive criterion for the identification 

 of origin of schists and gneisses. 



Significance of convergence. — If this idea of convergence be 

 correct, our attention is directed to the physical and chemical char- 

 acteristics of a few minerals like mica, chlorite, and hornblende as 

 important factors in anamorphism by rock flowage. Obviously 

 they are adapted to the conditions of rock flowage ; otherwise they 

 would not develop at the expense of other minerals. It is not so 

 clear whether they are adapted by their crystal habits, by their 

 cleavages, by their composition, or by a combination of these char- 

 acters. The fact that they are always arranged according to their 

 dimensions, their greatest and least mean being respectively parallel, 

 and the fact that their dimensions in the schist are clearly those 



