trichloride, which latter, however, can be removed by wash- 

 ing with either carbon disulphide or ether. Antimony 

 determinations in two different preparations gave 



(1) 75-45 % Sb. (2) 75-88 % Sb; corresponding chlorine 

 determinations gave (1) 12-43 % CI; (2) 12-49 % CI. 



Hence we ha^ve : — 



By acting upon 15 parts by weight of antimony trichlo- 

 ride with one part of trioxide in a sealed tube Schneider 

 (Pogg. Ann. cviii. 407) obtains a crystalline oxychloride to 

 which he assigns the formula 7SbCl3SbOCl. Repeating 

 Schneider's experiments I obtained a pearl grey crystalline 

 mass melting at 72° C, the melting point of the trichloride. 

 When acted upon by absolute alcohol it yields powder of 

 algaroth SbiCljOj, and its composition appears to be even 

 more complicated than that assigned to it by Schneider, 



Antimony determinations in two specimens gave (1) 

 54-24 %Sb; (2) 54-16 %Sb; whilst the corresponding 

 chlorine estim_ations were (1) 45-69; (2) 45-87 instead of 

 55-08 % Sb and 44-02 % CI required by Schnider's formula, 

 but agreeing with the formula SbieCl^eC, which requires 

 54*2 % of antimony and 45-357 of chlorine. 



The differences here found between the substances as 

 prepared by Schneider and myself may arise from the ad- 

 mixture of antimony trioxide with the oxychloride in the 

 former preparation. When the tube in v/hich the substance 

 has been prepared is placed in an upright position and 

 allowed to cool, the undissolved oxide sinks to the bottom 

 of the tube, but on still further cooling when the contents 

 of the tube are about to solidify the oxide rises from the 



