70 



BRIDGMAN. 



that with rising pressure in some cases being greater, is an interesting 

 feature of these curves, but may be due to experimental error. The 

 sudden increase of the width of the band of indifference for II-IV from 

 zero to a fairly large value at the higher pressures is an interesting 

 effect, but is not unique to this substance; CsXOs affords another 

 example. 



Carbamide. — This has three modifications, but only two transi- 

 tion curves available for measurement.^ On the upper of these two 

 curves above 100°, the transition is extraordinarily rapid, much too 

 rapid to measure, and the width of the band is sensibly zero. In no 

 case was the difference of pressures reached from above and below 

 more than 5 kgm., and in many cases there was no appreciable differ- 



Pressure, kgm./cm.' x 10 

 Ammonium Nitrate 



FiorRE 15. Ammonium Nitrate. The breadth of the band of indifference 

 between the several modifications as a function of pressure or temperature 

 along the transition lines. 



ence. On the lower curve between 0° and 100° for the transition I-III 

 the reaction was slow enough to measure. Three different experi- 

 mental arrangements were used; the carbamide was either rammed 

 tightly into a steel cup under mercury or submerged in loose pieces 

 under mercury, or rammed into a light steel shell and pressure trans- 

 mitted by kerosene. There is a very slight dissolving action of the 

 kerosene, not sufficient to vitiate the pressure-temperature values, 

 but sufficient to slightly round the corners, and so vitiate the measure- 

 ments of the time rate. This was one of the few cases in which a 

 restraining action of the steel shell was noticeable ; the reaction accel- 

 erations were less and the width of the band of indifference greater 



