\U7 



\alue of" the last column become!? 100 cal. liigliec lluiii the corre- 

 sponding value of table 4. A difference of" one molecule of' water 



TABLE 4a. 



in the composition of the hydrate would require a difference of 

 J 440 cal., and besides the difference lies in the wrong direction. 

 Accordingly the value ?* = 5 does not correspond to the observations. 



It will be clear from the tables 4 — 6 that it is permissible to 

 neglect the values s{)oken about in § 6 in the expressions lO^? and />. 



12. When we now compare the results of this indirect analysis 

 with the determinations carried out before b} a direct way, it api)ears 

 that at the quadruple point the composition of the compound is 

 given by H^S . 6H5O, that, however, at — 80°, when the compound 

 is formed of watei- and an excess of sulphuretted hydrogen, a quan- 

 tum of gas is persistently retained by the solid substance, which 

 does not even escape at a pressure of 2 cm (waterjet pum|)) iii 

 some hours. When the compound is heated, this gas is, however, 

 quickly liberated, and this latter causes two-phase equilibria to be 

 measured in va[)Our pressure determinations at low temperatures fCf. §7). 



The analyses executed before can be explained by this gas-adsorp- 

 tion. The quantity of sulphuretted hydiogen retained was not incon- 

 siderable. When we bear in mind that I he composition of the solid 

 substaiu'e varied between HjS.5.1H,0 and H^S . 5.5HjO, it appears 

 that pel- molecule H^S . 6H3O resp. 0.18 and 0.09 mol. H.,.S have 

 remained adsorbed. That these quantities have not rendered the above 

 three-phase determinations impossible is owing to this that during 

 the filling of the apparatus, it was repeatedly evacuated cooled with 

 liquid air, hence at lowei- pressure, thai therefore the sulphuretted 

 hydrogen could more easily escape, and the time of evacuation was 

 chosen long. 



In the direct analysis exhaustion with liquid air is not permissible, 



