Ixxii PROCEEDINGS OF THE GEOLOGICAL SOCIETY. [May I9OI, 



of assuming that a solution ceases to be a solution at 200°, 300°, 400°, 500°, 

 or even when heated to a temperature at which it becomes self-luminous ; or 

 will suppose that a crystalline aggregate of ice and calcium-chloride which 

 has become jfiuid is a solution, but that a mixture of quartz and felspar which 

 has been fused is not.' 



He then proceeds to point out that the laws whicli govern the 

 solidification of aqueous solutions must hold good also for igneous 

 solutions ; that by the addition of a certain amount of calcium- 

 chloride to water the temperature may be lowered to —10° C. 

 without the separation of any solid substance ; that by the addition 

 of further amounts the temperature of consolidation of water may 

 be lowered as much, as 59°, and that of calcium-chloride 100"^. 

 Other salts, such as the sulphates and nitrates of potassium, may 

 be made to separate from aqueous solutions at temperatures from 

 600° to 800° below their freezing-points ; moreover, the order of 

 consolidation is determined by the relative amounts of the two 

 substances present ; thus water may be made to consolidate before 

 or after a dissolved salt, by varying the concentration of the 

 solution. 



I have given a somewhat full abstract of this important letter, 

 because I believe that the expansion of the idea which it contains 

 will be the characteristic feature of the next great advance in 

 petrological science, an advance which will come about, not so much 

 by adding to our already large store of facts, as by dint of experi- 

 ment controlled by the modern theory of solutions, and carried out 

 for the express purpose of testing the consequences of that theory 

 and discovering the modifications which may be necessary to adapt 

 it to igneous magmas. 



Almost all recent writers on theoretical questions relating to the 

 igneous rocks have accepted the solution- theory, and the condition 

 of formation of minerals has been discussed from this point of view. 

 Crystals tend to form in a homogeneous liquid mass when the liquid 

 becomes supersaturated with any definite compound. As soon as 

 crystals arc developed the liquid in their immediate neighbourhood 

 ceases to be supersaturated, and there is thus established an 

 osmotic force producing molecular flow from the supersaturated 

 j)ortions towards the growing crystals. 



Erom a consideration of the work of Pelouze on glasses, com- 

 bined with his own work on igneous rocks, Lagorio arrived at the 

 conclusion that the ordinary rock-forming compounds tend to 

 separate out in the following order: oxides, pure iron-silicates, 

 magnesian and ferromagnesian silicates (olivine and rhombic 



