3 i6 SCIENCE PROGRESS 



centage 59J)." " There is throughout a gradual transition from 

 one variety to another, and, for the most part, the constituent 

 minerals are the same, but associated in different proportions." 

 He has shown that the chemical relations of the different 

 types thus linked together may be illustrated by diagrams, in 

 which the silica percentages are represented by the abscissae 

 and the percentages of the other oxides by the ordinates. It 

 is found that with rocks differentiated from the same magma 

 the points corresponding to each of these oxides in the different 

 types form a continuous curve ; which is not the case when 

 unconnected igneous rocks are dealt with in the same way. 



In some igneous masses the rock shows similar local 

 variations, but these are not arranged in the same regular 

 concentric fashion. In other areas a number of separate in- 

 trusive or volcanic rocks show a similar relation in chemical 

 composition. It is supposed that in these cases the magma was 

 differentiated by crystallisation in an intercrustal reservoir, 

 and that subsequently while still in a plastic condition the 

 different types were forced into new positions by pressure 

 due to earth movements. 



In some instances under the stress of such movements the 

 liquid portion of a partially consolidated magma has been 

 strained off from the more basic material that had already 

 crystallised, and the magma separated in this way has crystal- 

 lised out as an acid igneous rock. In other cases the minerals 

 which have first crystallised are believed to have sunk, under 

 the influence of gravity, to the lower portion of the reservoir. 



But though no one can doubt the importance of crystallisation 

 as a factor in rock differentiation, it appears to be insufficient 

 to account for all the cases in which rocks of varying com- 

 position appear to have been derived from the same magma. 

 In some areas we have evidence that both acid and basic 

 products of segregation have crystallised out from the fluid 

 state after separation has been effected, and in others the com- 

 plexity of the relations between the different types is so great 

 that it cannot be explained as the result of a single crystallisation. 



Mr. Harker and others have suggested that igneous masses 

 may be remelted and again crystallised, and that this process 

 may be repeated indefinitely as in the laboratory operation 

 of fractional crystallisation. The temperature of the magma 

 and rock is supposed to remain throughout in close proximity 



