594 JV. L. Bowen — Melting Phenomena of the 



ture is nearly linear with respect to composition and at higher 

 temperatures this relation probably does not change to any 

 great extent.* Of the heat effect on mixing, nothing is known, 

 but it seems that two such similar substances might prove 

 analogous to members of the same homologous series among 

 organic liquids, and that no great effect need be expected. 



Turning to the solids, we find the same facts true. They 

 are miscible in all proportions and again the change of density 

 with composition is linear. It would seem, then, on more 

 mature thought, that we are dealing with an especially favor- 

 able case, and that Raoult's law might be expected to apply to 

 both phases. This is indeed what has been found. 



In the case of solids not completely miscible, it is easy to 

 prove that both saturated solid solutions can not obey Kaoult's 

 law, i. e., if one does the other can not. In the case, there- 

 fore, of two series of mix-crystals with an hiatus, and especially 

 when this hiatus is so large that the two components separate 

 in the pure state, it does not appear possible that any rule of 

 freezing-point lowering, derived from the considerations set 

 forth in the present derivation, would be found to apply.f 



The close agreement found between observation and theory 

 points to other important conclusions relative to these com- 

 ponents. One is that albite and anorthite, both in the liquid 

 and solid state, consist of simple molecules, or else that they 

 are equally associated and that the degree of association is not 

 affected by mixing them. This follows from the fact that, in 

 plotting mol fractions (mol. percentages), molecular weights 

 were assumed equal to formula weights. If they had been 

 assumed to be equal multiples of the formula weights, the 

 plot would have been, of course, exactly the same. It might 

 be assumed that since L (calculated) for anorthite is N found to 

 be equal to the experimentally determined latent heat per 

 gram multiplied by the formula weight, that the molecule is 

 necessarily simple, but L is merely the difference between the 

 heats of vaporization and of sublimation of a mol and refers 

 therefore only to the gas mol. It is a constant characteristic 

 of the substance and in the present instance it is easy to see 

 that, whether simple or equally associated molecules are 

 assumed in its calculation, its value remains the same. The 

 experimental results do not, therefore, throw any light on the 

 actual degree of association of albite and anorthite. They 

 indicate, as pointed out, that each is equally associated in the 

 solid state and also in the liquid state, though the association 

 may be, and probably is, different in the different states. 



* Day and Allen, 1. c, p. 71. 



fCf. E. W. Washburn, J. Am. Chem. Soc, xxxii, No. 5, May, 1910, p. 

 670, for liquids. 



