Ferguson and Merwin^-The Ternary System. 183 



that comment here is unnecessary npon this point. Not 

 so well known perhaps are the inversion intervals of solid 

 solutions, the existence of which is demanded by the same 

 theory that indicates the melting intervals. Thus the 

 theory 17 predicts that the inversion temperatures in the 

 case of complete solid solution between two components 

 may pass through a maximum or a minimum or change 

 gradually from the inversion temperature of the one 

 component to that of the other, and that there will be a 

 region represented upon a concentration-temperature 

 diagram in which mixtures of the two solid solutions may 

 exist at equilibrium. We have been unable to detect 

 any inversion interval upon any of the compositions 

 investigated. The segregation of the phases in the solid 

 state is very slow, so that even if the resulting material 

 does show some determinable crystals, still there may be 

 doubt as to the interpretation of the structures seen. 

 Furthermore, inversion is not rapid enough to show a 

 temperature interval by purely thermal methods. Thus 

 the methods are inadequate for the determination upon 

 these charges of whether complete or only partial inver- 

 sion has taken place although such intervals must exist. 

 However, the theory does enable us to say that the inver- 

 sion surface along the calcium silicate-diopside line is 

 convex upwards. 



If now we plot vertically upon figure 2 the inversion 

 and decomposition temperatures, 18 given earlier in this 

 paper, of the solid solutions which bound the area of solid 

 solution and also those upon the side line and then make 

 vertical projections of these we will obtain the curves 

 given in fig. 3 which indicate approximately the effect of 

 solid solution upon these temperatures. No great accu- 

 racy is claimed for these results, but their accuracy is 

 sufficient to establish their general character. The direc- 

 tion and degree of the change in inversion temperature 

 with solid solution in the various cases well illustrates 

 the complexity of the phase relations and is in accord 

 with the theory of such phase relations. These also 

 emphasize the futility of the application to such cases 



17 It should be remembered that this theory assumes that the state of the 

 system is completely determined by pressure, temperature and composition. 

 If further experimental evidence indicates that the crystalline character 

 of the substances concerned enters as an additional factor, some modification 

 of the theory may become necessary. 



18 In reality the temperatures at which such inversions or decompositions 

 begin, i. e., the start of the inversion intervals for the various compositions. 



