82 PROF. H. E. ROSCOE ON CERTAIN 



The exact determination of the melting-point of the 

 hexachloride is attended with some difficulty, as the lique- 

 faction takes place gradually, and the smallest trace of oxy- 

 chloride depresses the melting-point down to about 180 C, 

 that given by the older observers (180 Forcher, 183 

 Cahours). A mean of several determinations of the pure 

 substance gave the number 275 C. (corrected) as the true 

 melting-point, and 270 (corrected) as the point of solidi- 

 fication. The constant boiling-point of the liquid was found 

 to be 346°* 7 (corrected) under a barometric pressure of 

 759*5 millims. 



The vapour-density of tungsten hexachloride was deter- 

 mined (1) in sulphur- vapour at 440 , and (2) in mercury- 

 vapour at 350 . As the hexachloride always leaves, on 

 distillation, a small quantity of solid residue, the substance 

 was distilled (either in a current of carbonic acid or of 

 chlorine) into the heated bulb from a smaller one attached 

 to it, according to the method employed by me in the 

 determination of vapour-density of vanadium tetrachloride. 

 The narrow neck of the bulb was kept open by inserting a 

 platinum wire ; and, after the sulphur or mercury had been 

 boiling for some minutes, the neck was sealed. 



Determinations of Vapour -density of WCl^ in Sulphur- 

 vapour at 440 . 



I. II. in. 



Weight of bulb and air 227797 14*5806 21*8781 grms. 



Barometer 755*7 755*4 755-4 



Temperature 14 14 12 



Weight of bulb and vapour, "i 



S ealedat 44 o°C ) 23 ' 66 ° 5 , ^ 6 5 a »W5 g™ 



Barometer 761*0 755'4 755'4 



Capacity of bulb I 93*63 84*25 97*36 



Residual air* 0*20 c. c.t 0*420. c| 0*94 c. c. I 



Hence vapour-density (H-=i) 167*8 169*7 168*8 



* Coefficient of expansion of glass, 0*00003. 



t At 1 6° and 633 millims. \ At 14 and 640 millims. 



