434 E. T. Allen, F. E. Wright and J. E. Clement— 



In explaining the formation of enstatite in the above cases, the 

 influence of temperature is doubtless of importance. When 

 pure magnesium silicate is cooled slowly enough, i. e., when 

 crystallization occurs at a temperature not too far below the 

 melting point, only the stable monoclinic form is obtained; if 

 cooled more rapidly, enstatite begins to form; still more rapidly, 

 and we have amphibole; and finally, when we chill very 

 suddenly, glass is the result. In all these cases, the temperature 

 at which crystallization occurs is the lower the more rapid the 

 cooling. 



The first influence of these solvents is, therefore, no doubt, to 

 lower the temperature at which crystallization takes place, 

 though not all of them are equally effective; the addition of 

 labradorite and ferric oxide makes less difference than albite, 

 orthoclase and the alkaline silicates. 



This explanation is, however, an incomplete one, otherwise 

 why should fluxes like the vanadates of magnesium and cal- 

 cium, tellurous oxide, and magnesium chloride, always give 

 the monoclinic form even at considerably lower temperatures. 

 Comparing these solutions with those which give rise to 

 enstatite, ■ we note at once that the one property which 

 serves to distinguish the two classes is the viscosity. The 

 monoclinic variety is obtained from thinly fluid solutions. At 

 first sight it might appear otherwise with the vanadates, 

 because the whole mass of silicate and vanadate from which 

 we crystallize the silicate seems little more than pasty. The 

 most of the mass is indeed undissolved solid, but the molten 

 vanadates are comparatively thin and so no doubt is the solu- 

 tion which covers the surface of the grains of silicate and colors 

 it yellow. The enstatite, on the other hand, was obtained from 

 solutions which are comparatively viscous at the temperature 

 of crystallization. 



The influence which the viscosity of a solvent exerts on 

 the transformation of unstable crystals which stand in 

 contact with it, has been very well shown by Kastle and 

 Reed.* The substance which they investigated was mercu- 

 ric iodide, an enantiotropic substance with an inversion point 

 at 128°. The yellow form is only stable above this point, yet 

 below that it is always obtained from solutions whatever the 

 solvent may be. The rate at which this form passes over into 

 the red variety, stable at ordinary temperatures, depends on the 

 viscosity of the solvent. Thus with certain mobile solvents like 

 water, the transformation was complete in a very short time, 

 while under a concentrated sugar solution the red appeared 

 only after several days, in glycerine after two weeks, and in 

 vaseline none appeared after a year and a half. In the case. of 

 mercuric iodide, so far as known, the unstable form always 



* Loc. cit. 



