720 PROCEEDINGS OF THE AMERICAN ACADEMY. 



in the wide part of the tube ; hence a series of standard mirrors under 

 these conditions is impossible. We have found, however, that both of 

 these obstacles are surmountable, and that small amounts of antimony 

 in solution can be reduced with practical completeness to the hydride, 

 which, when suitably heated, deposits the antimony in such a form as 

 to admit of estimation with reasonable accuracy by comparison with 

 standard mirrors. 



In arriving at this result two studies were made: first, as to the influ- 

 ence of the concentration of the antimony ions on the deposition of the 

 antimony upon the zinc ; second, as to the influence of the temperature 

 and cross section of the heated tube on the formation of the mirror. 



I. The Evolution of Antimony Hydride in the 

 Reduction Flask. 



A solution of pure recrystallized tartar emetic was prepared of such 

 a strength that 10 grams contained 0.0996 gram of metallic antimony 

 (average of five determinations of the antimony as the pentasulphide, 

 also confirmed by the volumetric method given below). A definite 

 amount of this solution was introduced in small portions into the re- 

 duction flask (see Figure 2), which contained 0.5 gram of zinc and 20 

 cubic centimeters of dilute sulphuric acid (1 to 12). The evolved 

 hydride was carried through a hot tube by a current of hydrogen 

 according to the method described below, except that the tube was 

 heated, as in the determination of arsenic, at its wide portion. After 

 the deposition of the antimony in the heated tube had evidently reached 

 its maximum, we determined the amount of the deposit, as well as the 

 amount precipitated on the zinc and the residue still in solution. The 

 volume of liquid in the reduction flask was approximately the same in 

 all cases. 



To find the amount of antimony deposited in the heated tube, the 

 portion of the tube containing the deposit was weighed, the mirror 

 dissolved in hydrochloric acid and potassic chlorate, and the tube, 

 after washing with alcohol and ether, reweighed. 



The metallic antimony and zinc .left in the reduction flask were 

 filtered, washed, and dissolved in hydrochloric acid with the aid of a 

 small amount of potassic chlorate. The antimony in this solution was 

 determined by the iodometric method of Gooch and Gruener.s The 

 solution, free from chlorine, was reduced by boiling with potassic iodide 

 in excess of sulphuric and tartaric acids. Any residual iodine was 

 bleached by careful addition of approximately hundredth-normal sul- 



» Amer. Journ. Sci., 42, 213 (1891). 



