POLYMORPHISM AT HIGH PRESSURES. 153 



seems to be of frequent occurrence in this group, but except for the 

 three ammonium salts, the phase diagrams of the various salts have 

 no relation whatever to each other. The frequency of polymorphism 

 cannot be ascribed to any detailed resemblance in the structure of the 

 crystals of the several salts, but nuist be due to some rather general 

 property of the halogen elements. It may be, for example, that the 

 atoms of the halogens are surrounded by rather complicated fields of 

 force which offer the possibility of being grouped in a variety of stable 

 configurations. 



Sulfides. — K2S*; FeS trigonal; Sl^Ss* rhombic; Sb2S5; BaS 

 cubic; PbS cubic; SrS cubic; CdS trigonal; CaS cubic; CuS hex- 

 agonal; Ag2S cubic*; HgSi cubic, trigonal. 



KnS was found to have a form not previously known. By analogy 

 (NH4)2S would be expected to have a new form, but it is not chemically 

 stable enough to allow the trial. An attempt to remo^•e the moisture 

 from Na2S sufficiently to allow a trial did not succeed. 



Ag2S has been known for a long time to have a second modification 

 above 170°. The transition is accompanied by the evolution of a 

 considerable quantity of heat, as may he shown by taking the cooling 

 curve. There is also a discontinuous change in the electrical resistance 

 on passing through the transition point. No attempt seems to have 

 been made to detect the change of volume, but it has been assumed that 

 in analogy with CU2S the change of volume would probably be very 

 small. I made particularly careful search for any evidence of a transi- 

 tion of Ag2S under pressure. Search was made by the usual methods 

 out to 12000 kgm. at four temperatures, 20°, 120°, 150°, and 200°, 

 and in addition the method of varying temperature from 130° to 200° 

 at approximately 1000 kgm. was used. The change of volume of the 

 transition must certainly be less than 0.0002 cm.^ per gm. and it is 

 probably safe to put the limit at a half or a third of this. This experi- 

 ment answers one question about which there has been some conjec- 

 ture; it has been thought that at higher pressures the difference of 

 volume of the phases might become appreciable, the transition point 

 at atmospheric pressure corresponding to a maximum or minimum of 

 the transition curve. This experiment shows that the smallness of 

 the difference of volume persists at high pressures; there is no reason 

 to suppose that the transition line will show more than the normal 

 amount of curvature. 



CU2S is a substance with a transition point at 103°, the character 

 of the transition being very much like that of x\g2S. The change of 

 volume has in this case, however, been demonstrated to be so small 



