M. Debray on the Chlorides of Tungsten. 255 



In the Giessen laboratory, in which this investigation was 

 made, chloride of rubidium was prepared from a large quantity 

 of ashes from beech wood which had grown on a basaltic soil. 



Debray has investigated the chlorides of tungsten*. When 

 pure and dry chlorine is passed over tungsten heated to dull 

 redness in a tube of hard glass, intensely red vapours are pro- 

 duced which condense to a liquid of a deep grey colour. This 

 is a mixture of perchloride of tungsten, WC1 3 , and of subchlo- 

 ride, W 2 C1 5 (W=92, Cl=35-5). To obtain the perchloride as 

 pure as possible, it must be repeatedly distilled in a current of 

 chlorine ; and though it can never be obtained quite free from 

 subchloride, the quantity of the latter is made so small as to 

 have no appreciable effect on the composition. 



There are also two oxycblorides of tungsten; one, which is 

 red, has the formula WO CI 2 , and the other, which is of a yel- 

 lowish white, the formula WO 2 CI. They are obtained mixed 

 with tungstic acid by passing chlorine over dry oxide of tung- 

 sten, WO 2 , and are separated by their varying volatility, the red 

 being most volatile. 



They may also be obtained by distilling perchloride of tung- 

 sten with oxalic acid in suitable proportions. The red is pure ; 

 but however the yellow chloride is procured it is always mixed 

 with red oxychloride or with tungstic acid, for when distilled 

 it decomposes into these two substances : 



2W0 2 C1=W0C1 2 + W0 3 . 



When perchloride of tungsten is mixed with anhydrous tung- 

 stic acid, the two combine with disengagement of heat, forming 

 the red oxychloride : 



2WC1 3 + W0 3 =3W0C1 2 . 



This reaction is analogous to that observed by Persoz in dis- 

 tilling perchloride of phosphorus and anhydrous phosphoric 

 acid. 



Debray was only able to determine with certainty the vapour- 

 density of the yellow oxychloride and of the red chloride. The 

 theoretical density of the chloride, WC1 3 , corresponding to two 

 volumes, is 13*75, and 6*875 if it correspond to four volumes; and 

 in like manner the density of the oxychloride, WO CI 2 , for a two- 

 volume vapour is 11*86, and for a four-volume 5*93. Now the 

 mean number for the density of WC1 3 is 11*72, and for the oxy- 

 chloride, WO CI 2 , is 10*58 ; which are therefore materially dif- 

 ferent from the numbers required by theory. 



But if the equivalent of tungsten be altered, and it be supposed 

 that the perchloride has the formula WC1 5 , this- would require 

 * Cmnptes Rendus, April 24, 1865. 



