Royal Society. 147 



under 760 mm of mercury. The second method depends upon a fact 

 already noticed in the preceding communication, that the oxytri- 

 chloride (VO Cl 3 ), prepared, according to the directions of Berzelius, 

 by passing dry chlorine over a mixture of the trioxide and charcoal, 

 possesses a port-wine colour instead of the canary-yellow tint of the 

 pure substance. This dark colour is due to the formation of the 

 tetrachloride of vanadium ; and if the vapours of the oxytrichloride, 

 together with excess of dry chlorine, be passed several times over a 

 column of red-hot charcoal the whole of the oxygen of the oxy chlo- 

 ride can be removed, and at last perfectly pure tetrachloride, boiling 

 constantly at 154°, is obtained. This reaction, it will be remembered, 

 served first to demonstrate the existence of oxygen in the oxytrichlo- 

 ride. In each distillation of the tetrachloride a peachblossom-co- 

 loured solid residue remained in the bulbs ; this substance is vana- 

 dium trichloride, and it slowly burns away in excess of chlorine when 

 heated, forming tetrachloride. 



The composition of the tetrachloride was established by six well- 

 agreeing analyses, made from several different preparations. The 

 mean result is : — 



Calculated. Found. 



V = 51-3 26-54 , 26-87 



Cl 4 = 142-0 73-46 73*02 



193-3 100-00 99-89 



Owing to the facility with which the tetrachloride splits up into 

 trichloride and chlorine, a solid residue was left in the vapour-density 

 bulb, and the density of the vapour (at 21 9°) was found by Dumas's 

 method to be 99-06 (or 6-86) instead of 96-6 (or 6-69). By volati- 

 lizing the liquid in a small bulb, and allowing the vapour to pass 

 into a large bulb already heated above the boiling-point of the liquid, 

 this deposition of trichloride was avoided, and the density was found. 

 to be 96-6 or 6'69 at 205°, and 93*3 or 6*48 at 215°, the last deter- 

 mination indicating that a partial decomposition into VC1 3 and CI 

 had occurred. The specific gravity of the liquid tetrachloride at 0° 

 is 1-8584 ; it does not solidify at —18°, nor does it at this or any 

 higher temperature undergo change of properties on treatment with 

 chlorine. It not only undergoes decomposition on boiling, but at the 

 ordinary atmospheric temperatures it splits up into VC1 3 and CI. 

 Tubes in which the liquid tetrachloride had been sealed up have burst 

 by the pressure of the evolved chlorine. Thrown into water, the 

 tetrachloride is at once decomposed, yielding a blue solution identical 

 in colour with the liquid obtained by the action of sulphurous or 

 sulphydric acid on vanadic acid in solution, and containing a vana- 

 dous salt, derived from the tetroxide V 2 4 . In order to prove that a 

 vanadous salt is formed when the tetrachloride is thrown into water, 

 the solution thus obtained was oxidized to vanadic acid by a standard 

 permanganate solution. The calculated percentage of oxygen thus 

 needed according to the formula 2VCl 4 + + 4H 3 0=V a O s + 8HCl 



L.2 



