85 



from the tetrachloride by heatmg in a moderately hot zinc 

 bath. 



The Bichloride is a non-volatile loose grey powder, with- 

 out lustre or crystalline structure. It undergoes change on 

 short exposure to air, and is converted by water into brown 

 oxide, with evolution of hydrogen. Analyses of two 



preparations gave as follows : 



Calculated. Found. 



Tungsten W=184 72-15 73-00 



Chlorine Clo= 71 27-85 26-35 



255 100-00 99-35 



Experiments made in the endeavour to prepare the 

 chlorides WCI3 and WCl were unsuccessful. 



5. Tungsten Oxychlorides. 



The Monoxy chloride WO CI4, and the Dioxy chloride 

 WO2CI2, have alread}^ been tolerably fully studied, never- 

 theless we find that Persoz actually doubts the existence of 

 these well characterised compounds, and Debray, obtaining 

 abnormal numbers for the vapour density of the first of 

 these bodies, is unable to explain his results. 



The splendid ruby red needles of the monoxy chloride are 

 best obtained by passing the vapour of a chloride over 

 heated oxide or dioxychloride in a current of chlorine. The 

 crystals melt at 210-4° and solidify at 206-7°; when heated 

 more strongly the liquid boils at 227'5° C. (corrected), form- 

 ing a red vapour rather lighter coloured than that of the 

 hexachloride. On repeated distillation in chlorine over 

 charcoal the hexachloride is formed. On exposure to air 

 the red crystals become at once coated with a yellow ciiist 

 of the dioxychloride. 



Analysis gave : — 



Calculated. Found. 



Tungsten W = 53-80 63-89 



Chlorine Cl4 = 41-52 41-11 



Oxygen 0= 4-68 



100-00 

 Debray found the vapour density of this body in sulphur 



