SANGER AND GIBSON. — DETERMINATION OF ANTIMONY. 723 



of antimony, when introduced into the Marsh apparatus, can be con- 

 verted to the hydride and collected as a metallic mirror. This conclu- 

 sion is reached by a consideration of the averages, for each series, of 

 the ratios of antimony precipitated on the zinc to that deposited in the 

 heated tube. These ratios vary from 7.7 in Series A to 0.8 in'Series J), 

 and the rate of progression is not the result of chance. Plotting these 

 results, we have the following curve (Figure 1), in which the ordinates 

 are the amount of antimony taken and the abscissae the ratios of anti- 

 mony precipitated on the zinc to that deposited in the heated tube. 



From inspection of this curve, it seems to us reasonable to suppose 

 that amounts of antimony under a milligram when introduced into the 

 reduction Hask would be practically converted to hydride, and we have 

 proceeded on this assumption in the development of the method which 

 follows, since in that method the amount of antimony usually taken 

 for a determination in no case exceeds one tenth of a milligram and is 

 usually considerably under that amount. 



With the exception of an abstract of a paper by Rieckher,^ we have 

 found no previous study of the ratio of antimony deposited on the 

 zinc to that evolved as hydride. In this abstract the ratio is said to 

 be between 92 to 8 and 96 to 4, i. e., from 11.5 to 24, but the data for 

 this conclusion are not given. As the original paper of Rieckher is 

 not accessible to us, we can only assume that the concentration of the 

 antimony ions must have been much greater than in our experiments. 



II. The Temperature and Cross-Section of the 

 Deposition Tube. 



The preparation of pure, gaseous antimony hydride has been accom- 

 plished in recent years by Stock and Doht 7 and by Stock and Gutt- 

 mann.^ The pure gas has been shown by these authors to have in 

 many respects (|uite different properties from those of the various 

 mixtures of hydrogen and hydride of antimony which have been in 

 the hands of so many investigators. As to the property which con- 

 cerns us in this investigation, the decomposition of the hydride by heat, 

 Stock and Guttmann have shown that the pure gas decomposes slowly 

 at ordinary temperature, but very readily in the presence of catalytic 

 agents, notably antimony itself In the case, however, of the mixture 

 of hydrogen and antimony hydride which comes from a j\Iarsh appa- 

 ratus containing minimal amounts of antimony, the concentration of 



6 Neucs Jalirbuch. d. Pliarin., 28, 10 (1867 ?) ; ref., Jahresbcr. 1867, 255. 



7 Ber., 34, 2339 (1901) ; 35, 2270 (1902). 



8 Ibid., 37, 885 and 901 (1904). 



