22 ART. II. — T. YAMAMOTO : THE FUSION CURVES 



tion by equation (7). It may be argued that the whole treatment 

 has too much of a conjectural nature, and that the equation 

 is nothing more than an empirical one. The justness of this 

 objection must indeed be admitted, particularly because the 

 characteristic constants D and S were determined from the curve 

 itself, and the {igreement between observation and calculation 

 is in general no guarantee for the soundness of a hypothesis. 

 But in the present case the curve is of such peculiar form, that 

 its representation by means of an equation with only two em- 

 pirical constants through so wide a range may be looked upon 

 as a partial vindication of the hypothesis. Moreover there is 

 nothing improbable in the hypothesis itself. Whether it is right 

 or not must be decided by the study of other chemical as well 

 as physical behaviors of phenol under different conditions. 



The discrepancy between the observed and calculated fusion 

 curves of phenol are too large to be accounted for by experi- 

 mental errors. The specific hypothesis employed may not be 

 appropriate ; but it is not impossible that other constants involved 

 in the calculation are inaccurate. 



On the whole a certain degree of probability has been 

 established for the existence of the reversible reaction 



in liquid phases, in which phenol is a component. 



