( 200 ) 



Tiionaii in tliis way the two investigators concur in considering 

 the formula H.,S . (^ aq the most probable, neither reasoning seems 

 to me to be conclusive for the given composition. In the tirst place 

 Villard's arguments can at most make the said formula probable, 

 while we cannot attach absolute certainty to de Fokcraxd's calculation 

 in my opinion already for the reason that an analogous calculation 

 applied to the hydrate of S(J.. yielded the value SO., . 8 JiJJ, whereas 

 the analysis of Bakhuis Roozeboo.m already made 7 HJ) probable, 

 and Villard's later determinations, which were carried out with 

 great care ha\ e j)roved with certainty in my opinion that the content 

 of water cannot be higher than 6 molecules. 



For this reason a renewed analysis seemed desirable to me. I have 

 carried it out in a way which the spacial tigure naturally suggests, 

 and which was already applied before for the determination of the 

 composition of gas hydrates by different investigators, among others 

 a single time by Bakihis Roozkboom and rejieatedly by Villakd. 

 Ill the spacial figures of the gas hydrates in general a large region 

 is found where the solid hydrate occurs by the side of a gas phase, 

 which pi'actically consists entirely of the most volatile component. 

 So when we tirst realise the state (jii SLJJ, and then evaporate the 

 lirpud phase Lj , we shall retain the pure hydrate by the side of gas. 



The determinations were made in wide tubes provided with stems 

 with two capillary constrictions; after the tube had been weighed 

 empty and with a certain iiiiantity of water an excess of H„S was 

 condensed, and the tube was fused to at the upper constriction. By 

 first heating the mass to the qnadi-uple point and then cooling it 

 shaking it violently the mass was converted into hydrate ^). After the 

 state SLJt obtained in this way had been preserved for some days, 

 the tube was cooled in carbonic acid and alcohol, and opened ; then 

 we either evacuated the hydrogensulphide at — 80° by means of 

 the water-jet circulation pump, or removed it by boiling about 

 — 13° (ice and salt). In both cases we are in the -S- (9- region, as 

 will be clear from the /-*- 7^-projection of the preceding paper. Three 

 weighings, the last after the tnbe has been sealed at the lower 

 constriction, yield the data required for the calculation. 



1 have carried out some thirty analyses of this kind ; the values 

 obtained thus oscillate round 5,3 mol. watei-; almost all lie between 

 5,3 db 0,2 mol. water (28 observations) so that I think I may conclude 

 that the formula H^S .^ HJ) is the most probable. I think 1 am 

 justified in this because the found oscillations can only be accounted 



1) In some analyses this was jn-omoted by glass rods or small glass spheres, 

 which, hov.'ever, had no influence on the result. 



