POLYMORPHISM AT HIGH PRESSURES. 159 



FeSOiTHijO content to pure ]MgS047HoO, a greater density is found 

 than that of the natural form of MgS047H20. One is therefore led to 

 expect a second denser form, and to look for it at higher pressures. 

 This matter is discussed in greater detail in Groth's Chemical Crystal- 

 lography, p. 92. It was a matter of great surprise to me to find that 

 there is no other form at high pressin-es, for I had regarded this evi- 

 dence as establishing a greater a priori prol)al)ility for the existence of 

 a new form than for anv other substance within mv knowledge. Two 

 runs were made on ]MgS047H.20. The first was with the pure salt 

 at 20°. For the second run the salt was inoculated with FeS047H2(). 

 and runs made at 20° and 100°. Higher temperatures were not tried 

 for fear of running into a decomposition of the hydrate. The expected 

 new form should haxe been found at low temperatures rather than high. 

 Runs were made with FeS047H.,0 at 20° and 100°. At 20° nothing 

 was found out to 12000 kgm., but at 100° a deconiposition point of the 

 hydrate was found. This will l)e described in detail under the group 

 of substances with water of crystallization. 



Substances with Water of Crystallization. — Every one of the 

 following list of substances has been investigated both with and ^^'ith- 

 out its water of crystallization, except LiXOs, which was tried only in 

 the anhydrous state. For brevity, both the hydrous and the anhy- 

 drous salts are specified in the following list only when the crystalline 

 forms are known; in many cases the crystalline form of the anhydrous 

 salt is not known. 



Oxalic Acid, +2H2O — monoclinic, anhydrous — rhombic; Po- 

 tassium Oxalate, +H2O — monoclinic; Potassium Acid Oxalate, 

 + H2O*, anhydrous-monoclinic; Potassium Tetroxalate; Potassium 

 Tartrate, +I/2H2O — monoclinic; Ammonium Potassium Phosphate, 

 (NH4)2KP04*,+4'H20; FeS047H20 — monoclinic; MgS047H20 — 

 rhombic; CUSO45H2O — triclinic ; Hg2(N03)22H20*; LiXOsSH.O. 



All of these substances were tested at the two temperatures 20° and 

 100°, instead of at 20° and 200°, as for most other substances. All of 

 them were hammered into an open steel shell and pressure transmitted 

 directly by kerosene. ^Yith one exception, the water of crystalliza- 

 tion was removed by heating to 100° in vacuum for several hours. 

 Potassium Tartrate was dehydrated in vacuum at 160°, since 100° was 

 not high enough. 



It is a curious freak of chance that the first substance with water of 

 crystallization that I tried, Hg2(N03)2, showed a transition in the 

 hydrous condition, but not when anhydrous. I was prepared to expect 

 a number of such cases, but the tetroxalate and FeS047H20 were the 



