( 197 ) 



the aid of soaled liilx^s. A tubo providcMl wiih a cjipilltiry was 

 wciglied first ciiiptv, then with water. At'tci- coiideiisatioM of hydrogen 

 sulphide the Inhe was sealed at the eoustriclion, and the weight of 

 the eoudensed gas was easily found by weighing of the tube with 

 the melted oil' |)ai1. Then the tube was slowly heated in a water- 

 bath, and the temperature was determined at whicli the last crystals 

 disappeared. It will be (dear that this observed temperature will lie 

 on the liquid branch of the three-phase line only wiien liie vapour 

 phase disappears at the same moment as the solid substance. This 

 can, of course, no! be carried out in j)ractice; it is, however, possible 

 to choose the (piaidity of vapour small, and then the observed tem- 

 perature differs in general oidy very little from the leal one. This 

 latter precaution, which will have little, if any, influence for the 

 determination of the li([uid points on SLJJ, must, however, be 

 carefully observed for the determination of the discussed points on 

 SL^G. We can easily see this in the following way. If the last 

 crystals disappear, the observed temperature agrees entii-ely with the 

 liquid [)oint of the mixture, the quantity of vapour being neglected. 

 If we now are on SLJr, the vapour consists for the greater pai't 

 of hydrogensulphid?, the substance which is present in the mixture 

 in great excess; so a neglect of the vapour s])ace is undoubtedly 

 justified, when it is small. For the line /S'L./j^, however, the situation 

 is just the i-everse ; the vapour which remains at the observed 

 vanishing point of the solid substance, consists for the main part of 

 hydrogensulphide, the substance which is present in a small quantity, 

 and with appreciable vapour space tlie errors can i-eaeh here a 

 high value. Accordingly in the first place the vapour space should 

 be (diosen small; to eidiance the accuracy, however, I have determined 

 the vapour space by calibration with water at the end of the 

 determination. The quaidity of hydrogensulphide in the vapour 

 could then be roughly calculated by the aid of the three-phase 

 pressures IVoui llic preceding communication, and the law of Royi-k, 

 which though probably not holding sti-ictly here, can yet be used 

 in the deternunation of the correction which is already snuill. 

 Moreover it ap|)eared that after the said correction the points deter- 

 miiic<l with great gas xoliime, ngreed satisfactoi'ily with the ob.sei- 

 valioiis with small gas volume. The observed li(piid points have 

 been collected in the followinu' table: 



