218 



PROCEEDINGS OF THE AMERICAN ACADEMY. 



sulphide of carbon, alcohol, or ligroin ; soluble in hot glacial acetic 

 acid, but on cooling a heavy white precipitate of oxichloride of anti- 

 mony settles out. It is readily soluble in hydrochloric acid and a 

 solution of tartaric acid. It will crystallize from a solution containing 

 free hydrochloric acid. Since a determination of antimony and chlo- 

 rine gave results which showed an exact ratio of one of antimony to 

 three of chlorine, it was obvious that there must be something com- 

 bined with the potassium in the salt. A determination of potassium by 

 Smith's method left about three per cent of the salt to be accounted 

 for. The use of dried salt eliminated water from the list of probable 

 constituents, and chemically pure materials guarded against impurities. 

 The fact that Professor Cooke invariably found a certain amount of 

 oxichloride of antimony in the chloride led to a search for oxygen 

 (combined) as the fourth element after the described method. Ob- 

 viously heating the salt would drive off all the antimonious chloride, 

 and whatever oxygen there was present was combined as the oxi- 

 chloride, which upon heat left a residue of Sb 2 3 . Calculating the 

 per cent of oxygen from the weighed Sb„0 3 gave the required per cent. 

 After a few preliminary experiments were tried, a series of ignitions 

 were carefully made, with the following results. 



Columns I., III., and V. contain the direct percentages of each ele- 

 ment according to the total estimation. 



* Probably some undeoomposed oxichloride. 



t This sample was undoubtedly heated too high when drying. 



