MM. Troost and Deville on Vapour Densities. 459 



the cylinder. Small cylindrically bent pieces of sheet metal, 

 affixed parallel to the sides of the cylinder, form additional layers 

 of vapour surrounding the central space, which exclude all influ- 

 ence from the external source of heat, and also from the sur- 

 rounding air. The upper part is closed by a lid of cast iron 

 pi'ovided with two holes, through one of which passes the finely 

 drawn-out neck of the cylinder, and through the other the tube 

 of an air-thermometer, which does not need to be graduated, as 

 it only serves to show if the temperature in the interior be con- 

 stant. An iron tube, 2 centimetres in diameter, is screwed as 

 high as possible in the side of the cylinder, so that its aperture 

 shall be at least 8 centimetres above the base of the neck of the 

 globe. If sulphur be used, this tube may conveniently terminate 

 in another tube, in which the sulphur may condense and flow 

 out without inflaming. The operation, according to the authors, 

 is easy of execution. In general it is necessary to evaporate 2 

 pounds of sulphur, or 2-4 pounds of mercury in each opera- 

 tion. The experiment is concluded when no more vapour rises 

 from the neck of the globe, which is kept warm by hot charcoal. 



Deville and Troost give some vapour densities determined by 

 this method. 



They found that the vapour density of pure chloride of alu- 

 minium in mercury vapour at 350° C. was 9"35, and in sulphur 

 vapour at 410'' C. 9"34. The density calculated for a conden- 

 sation to two volumes is 9'31. They found for the density of 

 perchloride of iron, which has the same condensation, 11-39 in 

 sulphur vapour. The density calculated for the formula Fe^ CP 

 is 11*25. The vapour density of chloride of mercury calculated 

 for the formula Hg^ CI, and for a condensation to four volumes, 

 is 8'15. Mitscherlich had found the number 8*35 . Deville and 

 Troost find the number 8*2 1. 



An example is given of the use of a knowledge of the vapour 

 densities of substances whose formulas cannot be considered to 

 be fixed. They found for the vapour density of chloride of zir- 

 conium determined in sulphur vapour, the number 8*15. The 

 formula of this substance is generally considered to be Zr Cl^, 

 and its equivalent 174"5. On this assumption its theoretical 

 vapour density would be 12. But assuming for it the formula 

 Zr C1-, its theoretical density would be 8-02, with which the 

 results of experiment closely agree. 



In a subsequent communication, Deville and Troost promise 

 to communicate the results of experiments which now occupy 

 them, in which the vapour of zinc is used as a mode of heating; 

 and for vessels, glolies of porcelain are einploycd, the necks of 

 which terminate in a point fine enough to be closed in a moment 

 by the oxyhydrogcn blowpipe. 



2 IT 2 



