130 BRIDGMAN. 



volume on I-II was not so consistent as it should be, considering the 

 rapidity and sharpness of the reaction. On the line I-III it was not 

 possible to obtain any consistent measurements at all. Four measure- 

 ments were made, which varied irregularly with temperature from 

 0.0032 to 0.0078. These points are not shown in the figure, and in 

 the table the listed value was obtained from the conditions at the 

 triple point. This value is evidently uncertain because of the lack 

 of agreement of the two points on the lower end of II-III. Besides 

 these four points, on the I-III line, the only other points not shown in 

 the figure are two of the higher pressure points obtained with the 

 first set-up with a patched cylinder, which later developed a perceptible 

 leak. 



The reaction velocity phenomena were as follows. On the line I-II 

 the transition is rapid and easy to measure near the triple point, but 

 becomes so rapidly slower with falling temperature that it did not pay 

 to try for points below 50°. On the I-III line the transition is rapid, 

 in both directions, but is strikingly more rapid with falling pressure. 

 The transition II-III is slow over the entire length of the curve, but 

 III-IV is rapid. It has already been mentioned that on the II-IV 

 fine the transition in the direction from II to IV runs easily, although 

 very slowly, with falling pressure, but that the reverse transition from 

 IV to II cannot be forced A\'ithin the limits of the apparatus at the 

 higher temperatures. About 5000 kgm. superpressure is necessary. 



The unusually large error in the results makes it of no use to try 

 for Aa, A(3, or ACp. 



Since these other forms were not known before, there are no other 

 values for comparison. Even the melting point is not listed; it is 

 without doubt greatly affected by small quantities of moisture. 



Figure 19 suggests the simplest conceivable mixed crystal diagram 

 for KHSO4 and NH4HSO4 at atmospheric pressure that will explain 

 the three known series of crystals at room temperature. Whether this 

 surmise is really correct or not cannot be verified until the crystalline 

 forms of KHSO4 at atmospheric pressure have been determined. But 

 in any event, the possibility of so simple a diagram shows that there 

 is no necessary connection between the new high pressure modifica- 

 tions of NH4HSO4 and those of KHSO4. In fact, the entire dissimilar- 

 ity of the phase diagrams shows that such a connection is not likely. 



Cuprous Iodide. — Two lots of this material were used; the first 

 was 55 gm. from Eimer and Amend, with which the existence of a 

 transition was discovered, and the second lot was of 100 gm. from 

 Hofl'mann and Kropff , AAith which the final measurements were made. 



