POLYMORPHISM AT HIGH PRESSURES. 135' 



ACp —0.013 to -0.014. The closeness of these limits for ACp is un- 

 usual, and we may accept the value with some confidence. 



It is interesting that the existence of a second modification has beea 

 suspected by a number of observers because of the behavior of the 

 mixed crystals with NH4Br and XH4CI. In fact, Wallace ^'^ made a 

 search down to —16° without result; if he had gone only two degrees 

 farther he would have foimd what he was looking for. 



Ammoxum Bromide. — This was obtained from Eimcr and Amend, 

 U. S. P. Before use it was dried in vacuum at 100°. Whatever 

 impurity there may be present does not form mixed crystals and so 

 affect the transition point, because the transition was always very 

 sharp and rapid, with no rounding of the corners whate\'er. Very 

 little transgression of the transition line is possible before the transi- 

 tion starts; the maximum observed was a superpressure of 30 kgm. 



• "0 2000 



E kgm./cm.' 

 ^ NH,I 



Figure 23. Ammonium Iodide. The calculated latent heat and the 

 change of internal energy. 



The equilibrium pressiu^e was shut within limits which were usually 

 3 kgm. apart, and in the extreme case only 6 kgm. apart. The transi- 

 tion line is very steep; it runs from about 138° at atmospheric pressure 

 to 680 kgm. at 200°. All the measurements had, therefore, to be 

 made with the low pressure apparatus. Two sets of readings were 

 made with the same sample. The repetition was necessary because 

 the first set of Av points was irregular. There seems to be some sort 

 of internal strain set up by the transition which it is necessary to re- 

 move by seasoning. For the second run the transition was carried 

 backward and forward a number of times before any measurements 

 were made. The agreement of the pressure-temperature points of 

 the two runs was very close. The substance was placed in the open 

 steel shell, and pressure transmitted to it by kerosene. 



The observed equilibrium points and the changes of volume are 



27 R. C. Wallace, C. Bl. f . Min. 1910, p. 33. 



