102 BRroOMAN. 



transition at low temperatures. I made an attempt at 0°, where the 

 separation of the Hnes would be greater than at higher temperatures, 

 but the transition is so very sluggish that II would not change into I 

 even when carried 3000 kgm. into the region of I. Under the condi- 

 tions it was useless to try to establish the existence of a transition 

 with an abnormally small volume change and equilibrium pressure 

 very close to that already measured, and I abandoned the attempt. 



Acetamide is a substance of unusual interest because at atmospheric 

 pressure it forms, in addition to the ordinary stable form of the solid, 

 an unstable modification of such persistence that Korber ^^ has been 

 able to measure its melting curve up to 3000 kgm. Korber, in fact, 

 says that the unstable form was always the one that crystallized out of 

 the subcooled melt, and that inconvenient manipulation was necessary 

 to force the stable form to appear. I was very anxious to measure 

 the melting curve of the unstable form, because the general relations 

 between melting curves of stable and unstable forms is still unde- 

 termined, in spite of work of Wahl ^^ and Korber. Not once, however, 

 did this unstable from appear in my apparatus, although several times 

 the liquid was forced to freeze by raising pressure across the melting 

 line at constant temperature. Once I thought that I had the unstable 

 form, and measured the melting data for it, but it turned out after- 

 ward that this was merely a point on the melting curve of II prolonged 

 into the region of I. That the new modification II is identical with 

 the previously known unstable form is made impossible by the direc- 

 tion of its melting curve, and the volume relations. II is more dense 

 than I, while the unstable form is less dense, and the melting point 

 of the unstable form is about 10° below that of I, while the melting 

 point of II, if it could be realized at atmospheric pressure, would be at 

 least 30° below that of I. It may be that the failure of the unstable 

 form to appear is connected in some way with the material of the 

 container. For this experiment the acetamide was placed in a nickel 

 steel container, whereas Korber used glass. It will be recalled that 

 an analogous eft'ect of the container was found in the case of ice ^^ ; 

 the unstable ice VI crystallized out of the melt in the usual metal 

 container, and the stable V could be forced to appear only by bringing 

 glass somewhere into contact with the liquid. 



The fact that the line L-II can be carried so far into the region of 

 I is of interest. Ordinarily, when measuring points on the transition 



17 F. Korber, ZS. phys. Chem., 82, 4,5-55 (1913). 



18 W. Wahl, Trans. Roy. Soc, 212, 117 (1912). 



19 A, p. 502. 



