364 



ON SEGREGATION IN IGNEOUS ROCKS. 



COSMO JOHNS, M.I.Mech.E., F.G.S. 



The sugfgestion that cooling fluid masses of the igneous rocks 

 may be comparable to saHne solutions has often been brought 

 forward. The writer proposes, by discussing certain obser- 

 vations that have been made on the occurrence of segregation 

 in granites, to see if the theory of solutions would apply. It is 

 hardly necessary to state that the great progress made in the 

 study of alloys and fused mixed silicates is due in great measure 

 to investigators having treated them as cases of reciprocal 

 solutions of their several constituents, and by utilising physical 

 and chemical methods of research to supplement the data ob- 

 tained by the use of the microscope. That the theory of solutions 

 should have been found to hold good for alloys and fused silicates 

 rather strongly suggests that no difficulty should be experienced 

 in applying it to the igneous rocks. That difficulties have arisen 

 is only too evident, and it can but be suggested that those who 

 have experienced the difficulties either were not sufficiently 

 well informed as to what has been done in applying the solu- 

 tion theory to alloys and fused silicates, or suff'ered from 

 lack of data relating to the rock they were discussing. 

 This brings in the interesting question as to what constitutes 

 sufficient data. Complete chemical analyses would come first : 

 not only the average composition of the mass, but separate 

 determinations of any differentiated portions. If, in addition, 

 we know the thermal history it is very probable that by a strict 

 application of the solution theory we should be able to predict 

 the microstructure and the order in which the various minerals 

 separated out, and all that would be left for the microscope 

 would be to confirm the prediction. Unfortunately, the physical 

 methods that have thrown such a brilliant light on the structure 

 of alloys and slags have not been applied to the igneous rocks, 

 so we are left to deduce from limited data the thermal changes they 

 have undergone. A careful determination of the microstructure, 

 with special reference to the order in which the minerals formed, 

 is of great importance. That this insistence on complete chemical 

 analyses is not unjustified may be proved by looking at the 

 numerous partial analyses that are appended to so many papers. 



When large masses of alloys are very slowly cooled and 

 examined evidences of segregation are not uncommon, so that 

 the appearance of similar features in certain igneous rocks is in 

 no way astonishing. It is further noticed in alloys that these 

 local segregation areas are distributed through the mass and 



Naturalist, 



