430 K T. Allen, F. R Wright and J. IT. Clement— 



Although this method is purely qualitative, repeated trials 

 on all these substances have convinced us that the relative posi- 

 tion of the curves is invariable. No point of change, of course, 

 can be found. 



In tlie consideration of these carves the fact must not be 

 overlooked that they include two quantities which may be 

 independently variable, the heat of fusion or inversion and the 

 specific heat. In general it is undoubtedly true, and therefore 

 a justifiable assumption in this case, that latent heats are of a 

 greater order of magnitude than specific heats. Although 

 glasses possess higher specific heats than the solids which crys- 

 tallize from them, polymorphic forms of the same substance 

 in all probability possess similar specific heats. For the pur- 

 poses here described, we are, therefore, perfectly safe in assum- 

 ing that the comparisons made are independent of differences 

 of specific heat in the substances experimented upon. The 

 results obtained for albite by this method are in full accord 

 with well established existing data, although albite, of all the 

 substances examined, might be expected to offer difficulties 

 from this cause. A substance exhibiting a sluggish heat change 

 might perhaps be found in which the relative magnitude of 

 the specific heats of it and its product, when compared with 

 the latent heat of the change, would be such that the latter 

 would have less influence on the form of the curve. In such 

 a case, the method would lead to erroneous conclusions. 



It goes without saying that if the change to be investigated 

 does not take place during the time the substance is in the 

 furnace, the method is useless. We found this true in the 

 case of quartz. When the temperature of the furnace was 

 held at 1560°, only about 1 per cent of the quartz was changed 

 during the twenty minute period of experiment, even though 

 other evidence has shown that tridymite is the stable form 

 above a point at least as low as 800°.* 



Order of Stability. — We have seen (pp. 402, 410 that the 

 orthorhombic amphibole and enstatite both pass over into the 

 monoclinic pyroxene under circumstances which point to the 

 greater stability of the last named form ; .that the heat effect 

 which accompanies this change of state is exothermic in all cases, 

 but quantitatively greater in the change — amphibole > mono- 

 clinic pyroxene. The amphi boles are, therefore, less stable than 

 enstatite, a conclusion which is substantiated not only by the 

 greater difficulty of forming the am pinholes, but by the closer 

 resemblance of all the properties of the two pyroxenes. The 

 fact that the monoclinic amphibole probably forms directly 

 from the orthorhombic amphibole in the presence of water at 

 375°- 475°, though the orthorhombic form is all the while in 

 * Day, Shepard and Wright, loc. cit. 



