POLYMORPHISM 555 



ciently rapid, even in the immediate neighbourhood of the 

 transition point, for this to be determined by observing the 

 growth of one form at the expense of the other. 



When a large crystal of either form is growing, its edges are 

 sharp and distinct, but when it is dissolving they become ill- 

 defined, and the whole crystal takes on a cloudy, etched appear- 

 ance. 



Tubes containing a mixture of crystals, of moderate size, of 

 the two forms, just covered with acetone, were sealed up and 

 heated in a bath of liquid paraffin. Starting from about 54°, 

 the temperature was raised by tenths of a degree, the bath 

 being kept constant for many hours at each temperature, and 

 individual selected crystals from time to time examined. 



At 55 the pale yellow crystals, and at 55*5° the orange 

 crystals, were unmistakably growing, although very slowly. 

 Between these temperatures no definite alteration of either 

 form could be observed. The transition point must therefore 

 be between these temperatures, and lie in the neighbourhood of 



55*25°. 



The difference in density of the two modifications allowed a 

 dilatometric method also to be used. This was found to be very 

 sensitive, although the rate of transformation near the transi- 

 tion point is slow. 



A dilatometer, packed with a mixture of fairly coarse crystals 

 of the two forms, and filled to a suitable height with a saturated 

 solution of the compound in acetone, was placed in a constant- 

 temperature bath adjusted to about 55°, and the capillary tube 

 sealed off. A thermometer wired on to this tube gave the 

 temperature of the bath and provided a scale for reading the 

 height of the column of liquid. 



Experiments extending over a considerable period, during 

 which the limits of temperature where opposite changes in 

 volume occurred were narrowed, showed that the volume 

 diminished at 55 and increased at 55*2°, whilst there was no 

 appreciable alteration at 55*1°, which may therefore be taken 

 as the transition point. 



By means of phthalylphenylhydrazide van't Hoff's deduc- 

 tion that the ratio of the solubilities of enantiomers at a 

 given temperature is independent of the solvent has also been 

 put for the first time to the test. It is a substance very 

 well suited for the purpose, as it is easy to obtain in the 



