determining the percentage of antimony and chlorine in pure 

 antimony trichloride, the results agreeing closely with each 

 other and with the theoretical composition. The objection 

 to Schaeffer's method of decomposing the oxy chloride by 

 boiling with a solution of sodium carbonate is that the pre- 

 cipitated oxide of antimony being in a very finely divided 

 state a portion of it is very apt to pass through the filter on 

 washing. 



The simplest formula which agrees with the analytical 

 results is SbgCliaO or three molecules of pentachloride in 

 which two of chlorine are replaced by one of oxygen. 



Calculated. Found. 



Sbs 43-39 43-46 



Cli3 54-71 54-75 • 



O 1-90 — 



100-00 



That this is a definite compound and not a mere mixture 

 of pentoxide and pentachloride (Sb2 05+ 14SbCl5)is evident 

 from the fact that the latter substance is not dissolved out 

 by washing with carbon disulphide. The calculated per- 

 centage of pentoxide contained in this compound is 7*68 ; 

 on heating 2-517 grams of the oxychloride in a tube retort 

 a residue of 0*1799 grams of pentoxide remained, corres- 

 ponding to a percentage of 7*14. 



The second oxychloride formed by heating the mixture of 

 one molecule of pentoxide and three of pentachloride is 

 produced only in small quantities as yellowish crystals. To 

 obtain it in the pure state, that portion of the tube in which 

 the substance is found is cut off" and after the tube has been 

 re-sealed it is placed in a slanting direction in a vessel con- 

 taining water heated from 85° to 90". The SbgCljgO melts 

 and runs down, leaving the other less fusible oxychloride 

 behind ; this is then dried on a porous plate in vacuo over 

 solid caustic potash. Two determinations showed that the 

 melting point of this substance is 97°'5 C. 



