PAPERS OX CHEMISTRY AND PHYSICS 263 



as solvent aud the slope of the log X vs. l/T curve of 

 the system is determined, then the ideal slope for A is 

 obtained by dividing this experimental slope by the ap- 

 propriate factor obtained from the chart, Fig. 1. 



From the ideal slope so obtained the latent heat of 

 fusion of the solute may be calculated. This is done by 

 multiplying this ideal slope by the constant 4.5S which 

 gives the latent heat in calories per gram mole. In Table 

 2 are given some values of latent heats calculated in this 

 way. It will be observed that these values compare fav- 

 orably with those determined caloiimetrically. 



A second method for determining the ideal slope and 

 the position in the table may be called the "cut and try" 

 method. This method may be used in those cases in 

 which the solubility of a solute has been determined in a 

 series of solvents whose positions in the chart have been 

 determined. It is evident that there is only one position 

 in the chart which will satisfy the demands of the factors 

 of more than one solvent when the solute has been given 

 any value for the ideal slope. The object is to find that 

 set of values for the ideal slope and for the position in 

 the chart which comes the most nearly to fitting all of the 

 solvents involved. Evidently only two such solvents are 

 needed, but if more have been investigated greater confi- 

 dence may be placed in the results obtained. 



A third method for determining the ideal slope of the 

 log X vs. 1 T curves may be us^d in those cases in which 

 L, the latent heat of fusion, is accurately known. This is 

 seldom the case, however. It is regretable that such an 

 important physical property has been so long neglected. 

 In cases where it has been determined the results are 

 often so discordant that doubt is thrown upon much of 

 the published data. This variation is due, partially at 

 least, to the fact that many of the values have been cal- 

 culated from Van't Hoff's equation and hence the values 

 obtained will depend upon the nature of the solvent used. 



In table 2, column 5, have been tabulated the latent 

 heats of fusion of a representative number of organic 

 compounds. These values have been taken largely from 

 Landolt and Boernstein, Tabellen, and only those results 

 which have been calorimetricallv determined are includ- 



