POLYMORPHISM AT HIGH PRESSURES. 125 



that of II. The direct measurements with the low pressure apparatus 

 showed that the expansion of I is of the order of 0.0005 cm.' per gm, 

 greater than that of II, and that of II 0.00004 greater than III. On 

 the III-IV Hne, the indications are that III is considerably more 

 expansible than IV. 



It is of interest to notice that the triple point II-III-IV is the first 

 case of its kind that has been found. .Ml three transition lines meeting 

 at this point are abnormal in that the form at the higher temperatures 

 has the smaller volimie. 



KHSO4 was suggested as a subject for investigation by a remark on 

 page 95 of Groth's Chemical Crystallography. KHSO4 forms with 

 XH4HSO4 three series of mi.\ed crystals, of different systems, only one 

 of these being known as a possible form for the pure KHSO4. It 

 suggested itself that KHSO4 might be really trimorphic. We have 

 here, at higher pressures, other modifications as suspected, but they 

 are four in number instead of only three. It would be most interesting 

 to find whether either of the new modifications stable at high tempera- 

 tures at atmospheric pressure is really of the same crystalline system 

 as the mixed crystals of KHSO4 and NH4HSO4. 



Ammonium Acid Sulf.\te. — The reason for investigating this sub- 

 stance was the same as for KHSO4, namely that the existence of other 

 forms is suggested by the fact that under ordinary conditions these two 

 substances form a series of niLxed crystals belonging to three different 

 systems. And as a matter of fact, just as in the case of KHSO4, it 

 is found that there are new forms at high pressures. However, there 

 are, instead of the three forms suggested by the mixed crystal relations, 

 demonstrablv four forms, and I am morallv certain that there is still 

 another. Furthermore, the phase diagrams of KHSO4 and NH4HSO4 

 show no obvious relation to each other. The relationship of isopoly- 

 morphism between the two substances must evidently be more com- 

 plicated than is indicated by the mixed-crystals system at atmospheric 

 pressure. 



The substance used for this investigation was obtained from Eimer 

 and Amend, and was of the " tested purity" grade. Two separate lots 

 were used, which were practically identical in behavior. The analysis 

 on the bottle showed only a minute trace of impurity. There was, 

 however, considerable absorbed moisture. I removed this as far as 

 possible immediately before use by keeping it in vacuum for six hours 

 in the melted condition (at 160°). In spite, however, of the excellence 

 of the analysis and the precautions to remove moisture, there was still a 

 large amount of impurity left, judging by the lack of sharpness of the 



