470 EEPORT— 1888. 



•wald and V. Meyer for ferric chloride, and makes determinations at tlie 

 temperature of boiling stannous chloride, 606°, which give a mean value 

 3"6. Now Dumas andMitscherlich' had found — Damas for temperatures 

 506° to 524° values 6'512 to 6-o81 nearly constant during interval of 

 temperature 18° ; Mitscherlich 6"9 at 508° (?) ; the very wide difference 

 between these results and 3'6 for vapour-density of sulphur at a tem- 

 perature not very far removed, and the insufficiency of Dumas' and 

 Mitscherlich's proof of Sg, their temperature interval being much too 

 small — these considerations led Biltz to determine, at the temperature 

 518° of boiling phosphoric sulphide, the density by Dumas' method ; this 

 gave, as a mean result of several nearly agreeing determinations, the 

 value 7"0. By the same method he obtained a nearly constant value 

 4' 7 at 606° in vapour of boiling stannous chloride. 



The values by Dumas' method were constant in different deter- 

 minations at the same temperature, and gave higher results than those 

 given by V. Meyer's method of replacement by nitrogen, 



Biltz made a series of experiments by the nitrogen-displacement 

 method at the same temperature 518°, but varying the amount of the 

 substance, and obtained (p. 2016) a series of thirteen results, in which, 

 while the amount of substance varied from "lOG? gram to -0450 gram, 

 the density found varied regularly from "•104 to 4509. These results 

 are due, according to Biltz, to the greater or less dilution of the vapour 

 of the substance with nitrogen, the observed density being less the less 

 the amount of substance, and thereby the more the amount of dilution. 



He brings forward as instances of this effect of dilution Horstmann's 

 experiments - on acetic acid, Meier and Crafts on the densities of vapour 

 of iodine,^ and V. Meyer and C. Langer's'* experiments with bromine 

 vapour. Moreover he himself confirms this explanation by making 

 vapour-density determinations by Dumas' method, but leaving in the 

 vapour-density vessel a mixture of nitrogen with sulphur vapour, and 

 obtaining thus, at the same temperature, ditFerent values for the vapour- 

 density of sulphur, as before. By experiments by Dumas' method in a 

 series of vapour-densities at intervals from 467'y° to 606°, nearly equally 

 divided among ten determinations, results were obtained varying from 

 7*937 to 4' 734; but these gave no constancy for vapour-density cor- 

 responding to Sg through any range of tempei-ature, as may be seen from 

 a graphic representation of the results. 



By varying the method used and by making a series of observations 

 over a considerable range of high temperatures he finds constant results 

 only for S2. 



Biltz continued his investigation into the influence of the size and 

 shape of the vessel on vapour- density determinations ; ^ and in another 

 article ^ he has described a new method of taking the vapour-densities of 

 volatile chlorides. 



The method consists in heating the metal in an atmosphere of chlorine, 

 and thus forming the chloride by absorbing chlorine, the amount of 

 which by volume is either equal to, greater than, or less than, the volume 

 of the product according to the volume of chlorine in the molecular 



' Ann. adm. Ph. [2], 50, 1832, p. 178 ; and 55, 1834, p. 31. 



= Ber. 3, 1870, p. 78. s C. R. 1881, p. 181. 



* Pyrochem. Unteisuch. BrannscJiKeig, 1885. ^ Ber. 21, 1888, p. 2772. 



• Loe. cU. p. 2766. 



