( 399 ) 
Fig. 2. 
This will be noticed by comparing the lines AB, CD and Et. 
Such a behaviour is highly improbable, for it may be expected that 
at 35°, 25°, and 10° the effect on the solubility of the d-mandelic ester 
by the presence of the /-mandelic ester in the solution will manifest itself 
in an equal sense and not cause an increase of solubility at the one 
temperature and a decrease at the other. In the diagram of Findlay 
and Hickmans, however, the course of the curves AB, CD, and EF, 
1S ’ e °ntrary to my investigations, determined exclusively by the 
situation of the points B, D, and F, whose position is deduced from 
an observed angle of rotation of the solution of the residues obtained 
> evaporation. It is not at all improbable that if these points were 
ftermined by another method their situation would be quite a 
liferent one. In this system the verification would have been a simple 
°ne, because the melting point curve of the binary system formed by 
e two biactive esters is known. The melting point for each content 
111 P’ 01 ^ natures of D -J- L is therefore known. 
now, Findlay and Hickmans had only determined the melting 
Point of their residues each time after evaporating their saturated 
utions (R _|_ _£,) and (R + D ), they would have obtained from the 
Situation of this melting point in connection with the melting point 
