STUDY OF IGNEOUS HOCKS. 277 



The order of separation of several compounds in solution in one 

 another depends on the degree of saturation of each, that with the 

 highest degree of saturation, or that one whose saturation point is 

 reached first upon the cooling of the solution separates first. The 

 relation between saturation, molecular concentration, and the melting 

 point of each compound has been established in general terms for 

 different sets of cases by Meyerhoffer, and further elaborated by 

 Roozeboom, for the cases of crystals of isomorphous compoimds. In 

 all cases where the mixed compounds do not unite chemically to form 

 new compounds, or physically as mixed ciystals, there is one minimum 

 point of temperature for a mixture of two compounds, and more than 

 one in more complex mixtures, at which a certain proportioned mixture 

 remains liquid. At this temperature the two components of a binary 

 mixture will crystallise simultaneously. This minimum temperature 

 and particular mixture are called eutectic. 



Miers has shown that when supersaturation sets in and the labile 

 condition is taken into account, as the state of the liquid in which 

 spontaneous crystallisation takes place, the minimum temperature of 

 separation and corresponding proportions of the mixture do not coin- 

 cide with those already described as eutectic. These he has called 

 hypertectic. 



A study of these pi'inciples shows that there can be no invariable 

 order of separation, or ciystallisation, of the constituent compounds 

 in a series of mixed solutions composed of like compounds. And that 

 simultaneous crystallisation of pairs, or of more than two kinds of 

 separating compounds, may take place in solutions of whatever com- 

 position. Eutectic mixtures may consist of more than two com- 

 ponents. Moreover, the supersaturation of a solution by one com- 

 ponent may affect the proportion between two or more components at 

 the moment of s\Tichronous crystallisation. Synchronously crystal- 

 lised mixtures of certain kinds of components, therefore, are not 

 necessarily similarly proportioned. The bearing of these principles on 

 the ciystallisation and texture of igneous rocks is manifold. A few 

 ilustrations will suffice. Quartz may be the first mineral to separate 

 from a molten magma when the solution is so rich in silica that upon 

 cooling it becomes saturated with silica before being saturated with 

 feldspar or some ferromagnesian compound, or even iron oxide. Quartz 

 may be the last mineral to separate from magmas so rich in feldspar 

 or ferromagnesian compounds as to become saturated by these upon 

 cooling before being saturated with quartz. 



Either labradorite or augite may separate first from a mixture of 

 the two according to which saturates the solution first upon cooling, 

 and this depends upon their relative amounts in the solution, and 

 their order of crystallisation is further modified by the possibility of 

 one or the other producing supersaturation in the liquid. This will 

 account for the differences of texture often noted in certain gabbros, or 

 basalts, of almost the same composition. 



Eutectic mixtui'es, or those whose components crystallise 

 simultaneously, often yield aggregates of intergrown crystals, the most 

 familiar examples of which are found in graphic granite, and certain 

 alloys. But Miers has called attention to the fact that the simul- 

 taneous crystallisation of two compounds in eutectic proportions does 

 not invariably produce intergrown individual crystals, or grapliic 



