Minerals of the Composition Mg Si 3 . 415 



enstatite changes to the same form above 1260°, and that on 

 cooling neither of the changes is reversible, these facts in 

 themselves do not prove monotropy. In certain cases of 

 enantiotropy, mere cooling of the form stable at high tem- 

 peratures does not suffice to revert it to the other form, even 

 when both forms are in contact and unlimited time is allowed.* 

 This inertia is apparently due to the great internal friction 

 between the molecules at temperatures below the inversion 

 point. The molecular immobility may be overcome by the 

 use of suitable fluxes in which solution of the unstable form 

 and precipitation of the stable go on hand in hand, the unsta- 

 ble form being the more soluble. Thus pseudo-wollastonite 

 is less stable than wollastonite below the inversion point, yet 

 it does not revert to it at lower temperatures except by the 

 aid of proper solvents. In this case, molten calcium vanadate 

 proved well adapted to the purpose. The pseudo-wollastonite 

 dissolved in it and cotemporaneously wollastonite crystallized 

 out. 



Similar tests with solvents were applied to the different forms 

 of magnesium silicate with the result, as we have already 

 recounted in detail (pp. 387 et seq.), that from a considerable 

 number of solvents magnesium silicate was found to pass into 

 solution and gradually to crystallize, invariably as monoclinic 

 pyroxene, whatever crystal form was originally taken. This 

 indicates that at atmospheric pressure the monoclinic pyroxene 

 is the most stable form of magnesium silicate at all tempera- 

 tures between about 800° and the melting point of the mono- 

 clinic form (1521°). 



(2) In order to confirm this evidence of monotropy, we sought 

 to ascertain the direction of the heat changes, whether exother- 

 mic or endothermic, for two of the unstable forms. The method 

 first tried was that of Frankenheim, which consists in observ- 

 ing at regular intervals the rise in temperature of a mass of 

 the substance. Heat was supplied by means of the electric 

 resistance furnace used in this laboratory, and the temperature 

 measured by a thermoelement, f The curves thus obtained by 

 heating the different forms are perfectly smooth, showing 

 neither absorption nor release of heat until the melting point 

 of the monoclinic form is reached. The fact that all forms 

 seem to have the same melting point indicates in itself that 

 the other forms change into the monoclinic pyroxene before melt- 

 ing. The change, however, is so sluggish that it proceeds 

 throughout a temperature interval of several hundred degrees, 

 and the heat is therefore so uniformly distributed as not to be 

 indicated by the curve. 



*See "Wollastonite and Pseudo-Wollastonite," this Journ. ser. 4, xxi, 

 p. 95, 1908. 



fDay and Allen, Pkys. Kev., xix, 184. 



