438 Allen, Wright and Clement — Composition MgSiO r 



3. Of the four polymorphic forms of magnesium metasilicate, 

 (I) is the stable form at all temperatures and the others are 

 monotropic toward it, the order of stability being I, II, III, IV. 

 This order is established by changing one form into another at 

 various temperatures and by proving that (II) and (IV) change 

 to (1) with evolution of heat. (Ill) has not been obtained in suf- 

 ficient quantity for this test. Though enstatite and the amphi- 

 boles are not stable, any more than glasses are, on account of 

 the great internal friction of the molecules, they have less ten- 

 dency to change, when once formed, than many glasses. We 

 cannot state definite limits of stability for the various minerals, 

 as it is possible to do where the relation is enantiotropic; it is 

 possible, however, to fix certain temperature limits below which 

 one of these forms may crystallize from a melt of pure magne- 

 sium silicate. Thus enstatite could only form below about 

 1250°, since above that temperature it passes into the monoclinic 

 form; but it must be remembered that in the silicate solutions 

 of nature this limit would probably always be lower on account 

 of the general occurrence in them of such compounds as ferrous 

 silicate. 



4. The specific gravities of the four forms, in the order of 

 their stability, are : (I) monoclinic pyroxene, 3*192; (II) ortho- 

 rhombic pyroxene (enstatite), 3-175; (III) monoclinic amphi- 

 bole, not determined directly, but its relation to (II and (IV) 

 is fixed by its index of refraction; (IY) orthorhombic amphi- 

 bole, 2-857. 



5. While our experiments do not settle completely the mys- 

 teries of the formation of unstable bodies, they do show that 

 temperature and viscosity are two factors of prime importance. 

 Thus, from melts or from silicate solutions, the stable mono- 

 clinic form of magnesium metasilicate crystallizes at the highest 

 temperature, enstatite next, and the amphiboles probably low- 

 est of all. From thin solutions the stable form is obtained at 

 still lower temperatures, 800°-1000°, while from aqueotis solu- 

 tions at 375°-475° an amphibole results. 



6. Our study of the Bishopville meteorite indicates that it 

 must have been cooled very rapidly from a high initial temper- 

 ature, and there is evidence that the same is true of other mete- 

 orites. 



7. The intergrowths of enstatite with the monoclinic pyrox- 

 ene, and of the two amphiboles, which we obtained in close 

 resemblance to those of nature, are cases of falsee quilibrium, 

 and their occurrence establishes the fact that it cannot be as- 

 sumed that all rocks or mineral aggregates are systems in true 

 equilibrium. 



8. In the course of the investigation a useful method has 



been developed for detecting sluggish heat changes. 



Geophysical Laboratory, 



Carnegie Institution of Washington, July, 1906. 



