124 BRIDGMAN. 



Because of the smallness of the change and the width of the band of 

 indifference on the hnes I-II and II-IV, several of the early measure- 

 ments of the change of volume were not good and had to be discarded. 

 These comprised three points on II-IV, one on I-II, and one on I-IV. 

 In virtue of the relations at the triple points one can be fairly certain 

 of the values of Av for these lines. 



By a curious grouping of the discrepancies of the early measurements 

 of I-II and II-IV, it looked as if possibly there were still another modi- 

 fication with excessively small volume change. Subsequent careful 

 exploration failed to substantiate this surmise. 



The order of determination of the points was as follows: first one 

 on 1-TV, then several on I-II and II-IV, then the points on III-IV, 

 then II-III, II-IV, I-II, and I-IV again, and finally two points with 

 the low pressure apparatus. The complete run at high pressures 

 with the apparatus set up with one filling and pressure never entirely 

 released extended over seven days; there was no decomposition or 

 change in the KHSO4 in this time as shown by the fact that the last 

 determined I-IV points fall exactly in line with the first. 



The point with the low pressure apparatus on II-III does not require 

 especial comment. To obtain the Av value of I-II special procedure 

 was necessary because of the sluggishness of the transition. On 

 passing over the line from II to I, the temperature was raised to 190°, 

 and then lowered, in order to ensure completion of the transition. 

 The sluggishness was so great that it was not possible to obtain an 

 equilibrium point on I-II at low pressures. At 180° the transition 

 from I to II could not be started by a pressure of 1000 kgm., the limit 

 of the apparatus. 



There are no previous data for the polymorphic transitions for 

 comparison, the existence of the modifications not having been known 

 before. The melting point of KHSO4 at atmospheric pressure is 

 given as 200° by Mitscherlich ^^ and 210° by Schultz-Sellack.^s I 

 found it to be at any rate higher than 200°. 



The accuracy of the measurements does not justify an attempt to 

 compute the difference of compressibility of the four forms. It can 

 be stated, however, that what difference of compressibility there is 

 will be found to be very small. In this connection it should be noted 

 that along the line I-II Av increases with rising pressure, and that 

 along III-IV it falls. Both these effects are unusual. The change 

 along I-II points to a thermal expansion of I abnormally greater than 



24 Mitscherlich, Pogg. Ann., 18, 152 (1830). 



25 Schultz-Sellack, Jahresber., 1871, 217. 



