36 ART. 12. — H. HIROBE : ON THE FUSION SURFACES OF THE 



The terms (3 9î-iV)'' and (iV— 9() are liighly sensitive to ex- 

 perimental errors, and the effects are moreover magnified in the 

 calculation according to (/). Hence exact determination of this 

 magnitude is impossible particularly when the concentration of 

 phenol is small. At a constant temperature ^ ought to be 

 independent of the concentration of phenol. This holds true only 

 for curves 1 and 2. For all others Ü diminishes with the 

 decreasing amount of phenol. The deviation is too regular and 

 one-sided to be accounted for by accidental experimental errors. 

 Therefore we must conclude that either this simple hypothesis 

 does not correspond closely enough to the actual state of things ; 

 or that the conditions of the ideal solution are not fulfilled with 

 sufficient approximation ; or that there is some constant source 

 of error in our measurements. 



But in order to test the equation 



the values of i^ as determined above are not accurate enough. 

 Hence it would be more rational to represent .*R as a function 

 of temperature, and then calculate x by means of i\. The value 

 of X so found can then be compared with observation. 

 The constants in the equation 



have been calculated from the values of I02; ivt for fusion curve 

 No. 1 and found to be : 



q = 20940 calories, T. = 373.2.i> 



1) These values do not agree very well with tliose found by Yamamoto ; for he has 

 calculated j? under tlie assumption that tlie heat of fusion of naphthalene Q is constant 

 for all temperatures. Moreover a very slight discrepancy in tlie measurement of the fusion 

 curve may give rise to such deviations. 



