Reports and Proceedings — British Association. 523 



magmas pass from the metastable to the labile condition we may use 

 this knowledge to determine the exact temperature at which certain 

 of their minerals crystallised. 



Ordinary petrographical descriptions supply numerous examples 

 of the difference between the metastable and labile conditions to 

 anyone who will read them in the light of the suggestion which 

 I have made ; others are to be found in such experiments as those of 

 Vogt or Doelter. 



My own hope is that when more experiments have been made 

 upon mixed supersaturated solutions it will be found that most, if 

 not all, of the features of rock development are paralleled by the 

 ordinary process of crystallisation, but that motion, supersaturation, 

 and supercooling are most important factors. 



The very similarity between the differentiation of the alumo- 

 alkaline and ferro-raagnesian minerals on a small scale in the rock, 

 and that of the alumo-alkaline (or salic) and ferro-magnesian (or 

 femic) rocks themselves on a large scale, points to some similarity 

 of origin. 



In order to avoid burdening this address with detail I have merely 

 chosen the researches of van 't Hoff, Vogt, Doelter, and Heycock and 

 Neville as illustrations of experimental work conducted on the lines 

 of modern physical chemistry, and have omitted much that might 

 liave been mentioned ; the valuable researches of Pelouze, Lagorio, 

 Morozewicz, and Loevinson - Lessing, and the melting - point 

 determinations of Joly I have not quoted, because they belong for 

 the most part to an earlier period than that which I am considering, 

 and have been discussed by Teall and other writers. 



Many very interesting speculations I have passed over entirely, 

 because my object has been to focus attention upon experimental 

 evidence. I cannot help thinking that these speculations are often 

 based upon chemical actions and equilibria that may be impossible; 

 but we cannot criticise them for lack of evidence, and I return to 

 my original statement that geology is only beginning to enter the 

 experimental stage. 



An earnest beginning is, however, being made. The researches 

 on mineral and rock synthesis which I have already quoted are 

 laying a solid foundation ; and I see no reason why something of 

 the sort which has been done by van t' Hoff and his collaborators 

 for the aqueous deposits of Stassfurt should not ultimately be worked 

 out for an igneous complex, though it may involve tenfold the labour 

 and tenfold the time. We have already to welcome the establish- 

 ment by the United States Geological Survey of a laboratory for the 

 express purpose of applying to minerals and rocks the exact methods 

 of modern physics and physical chemistry. The very suggestive 

 research of Day and Allen upon the thermal properties of the 

 felspars is a promise of the sort of work that may be expected from 

 such laboratories. 



I fear it will be only too evident to those who have given me their 

 patience during this address that I approach the problems considered 

 in it from the point of view, not of the geologist or the chemist, but 



