IGNEOUS ROCK-MAGMAS AS SOLUTIONS 



By ALFRED HARKER, M.A., F.R.S., 



St. John's College, Cambridge . 



Recent years have witnessed a very marked revival of interest 

 in the rational, as distinguished from the merely descriptive, 

 side of petrology. There is sufficient geological evidence to 

 show that the great diversity of igneous rocks actually observed 

 must be explained. by some sort of evolution from more primitive 

 and generalised types of rock-magmas; and the physical processes 

 by which such " differentiation " may be brought about have 

 accordingly been the subject of much speculation. It is clear 

 that one prerequisite for the intelligent discussion of this im- 

 portant problem is a well-grounded conception of the real 

 constitution of igneous rock-magmas with a knowledge of the 

 laws which control their crystallisation. On the same basis 

 must rest also our interpretation of the textures and structures 

 of the rocks themselves, and in some measure their mineralogical 

 composition. The question of the constitution of rock-magmas 

 thus comes to occupy a very fundamental place with reference 

 to petrological inquiry. 



The chemical analysis of an igneous rock is commonly 

 presented as a column of percentages of silica, alumina, and 

 other oxides. An alternative plan, convenient for certain special 

 purposes, is to state the composition in terms of the elements, 

 oxygen, silicon, aluminium, etc. Again, it may be possible, 

 either by calculation from the chemist's data or by a direct 

 mechanical analysis, to express the composition of the rock in 

 percentages of the several component minerals, orthoclase, quartz, 

 biotite, etc. For a crystalline rock this last manner of repre- 

 sentation exhibits not only the composition but the actual 

 constitution, since it is in the form of these definite compounds 

 that the elements are in fact combined in the solid rock. The 

 first question to be considered is, which of the above ways of 

 expressing the composition will likewise represent the true 

 constitution in the case of a molten rock-magma. Do the con- 

 stituents exist in the magma in combination as silicates, etc., or 



239 16 



