SANGER AND GIBSON. — DETERMINATION OF ANTIMONY. 729 



amount which is inappreciable in our work on antimony.^'* By u.se of 

 hydrochloric acid and purification of the hydrogen from the generator, 

 there is no necessity for purification of the hydrogen from the reduc- 

 tion flask, provided sulphur compounds reducible to hydrogen sulphide 

 are absent. Whenever the hydrogen contains hydrogen sulphide, the 

 deposit in the heated tube is more or less reddish-yellow in color, due 

 presumably to the presence of antimonious sulphide. 



The fused sticks of calcic chloride used for drying (Merck) dissolved 

 clear in water, and the solution showed an alkalinity of not over 0.8 

 per cent. We have seen no reason to believe that hydride of anti- 

 mony is held back by this preparation, nor has there been any loss of 

 antimony observed from an unduly moist condition of the chloride. 

 AVe have, however, kept the chloride as active as possible by refilling 

 the rear half of the tube with fresh chloride after every half dozen runs, 

 as explained, or when the accumulation of moisture becomes noticeable. 

 When not in use, the tubes are kept stoppered. 



The selection of hard glass tubing is a matter of the highest im- 

 portance for the success of the method. We used at first a German 

 glass, source not known to us, which had been used in arsenic work 

 without apparent disadvantage. This glass had a slight brown color 

 after long ignition. Antimony mirrors deposited on this, under the 

 conditions about to be described, were often white and not clearly 

 defined. We next tried a sample of American glass, but this gave the 

 brown color on ignition still more, and the deposits of antimony were 

 entirely white. Next a Jena glass, which on ignition gave the well- 

 known opaque appearance, and on which the mirrors were also entirely 

 white. Finally we resorted to a glass of Kavalier, which did not give 

 any color or opacity on long heating. The deposits of antimony on 

 the capillaries drawn from this glass were satisfactory, and it was used 

 in the prepai'ation of the standard mirrors. ^^ 



^* Tliis would mean not over 0.00004 mg. of arsenic in the amount of hydro- 

 chloric acid used, which is beyond the limit of the delicacy of the process as 

 applied to arsenic. 



^5 A cursory qualitative examination of the different samples of tubing showed 

 no marked points of difference except tliat the fourth tubing contained no barium, 

 while the others did. The amount of barium in the others was proportional to the 

 degree of change in appearance of the mirror produced on the samples. Lack 

 of time prevents an investigation on this point, and an opinion as to the influence 

 of the barium, if any, would be mere conjecture ; as, for example, whether the 

 barium oxide could act catalytically in causing an oxidation of the antimony, 

 since it seems probable that the white deposit is due to an oxide of antimony. In 

 the absence of any definite knowledge on the matter one can only determine the 

 availability of a sample of glass by actual trial. 



