2IO 



REPORTS ON INVESTIGATIONS AND PROJECTS. 



amounts of pentoxide, after reduction to hydriodic acid by means of hydra- 

 zine, against weighed nearly equivalent amounts of the purest silver. A few 

 tenths of a milligram excess of silver were always used, and the excess was 

 determined gravimetrically in the filtrate after evaporation. 



In the determination of the ratio of silver to iodine, the method was ex- 

 actly like the one just outlined, except that, instead of iodine pentoxide, 

 iodine was weighed and reduced, with elaborate precautions to prevent loss 

 by spattering or evaporation, with an excess of hydrazine. The titration 

 against weighed equivalent amounts of silver proceeded as above. 



The iodine was prepared in part by decomposing very carefully crystallized 

 iodic acid at high temperatures in a current of pure dry air. One specimen 

 of iodic acid, which had been made by the action of pure fuming nitric acid 

 on pure iodine, remained from the previous research. A second specimen 

 was prepared from recrystallized potassium iodate through barium iodate. 

 Still another specimen of iodine was purified by double distillation from an 

 iodide, with intermediate boiling in the form of hydriodic-acid solution to 

 eliminate hydrocyanic acid. 



The preparation of the iodine for weighing consisted in sublimation in a 

 current of pure dry air into a weighed tube, of glass in most of the experi- 

 ments, of quartz in the three final experiments. At first the sublimation 

 was conducted in a hard-glass tube, but in most of the experiments a quartz 

 tube was used for the purpose. Several samples of silver of known purity 

 were employed. In order to prevent or detect occlusion by the silver iodide, 

 the method of analysis was varied by diluting the solutions, more in some 

 cases than in others, before precipitation, but in no case was either the 

 hydriodic-acid solution or the silver-nitrate solution more concentrated than 

 thirtieth normal. Furthermore, although in most of the experiments the 

 silver-nitrate solution was poured into the hydriodic-acid solution, in two 

 experiments precipitation was effected in the reverse fashion. All variations 

 in material and method were without perceptible effect. 



In the following table are given the results of all the analyses except one 

 which met with an accident: 



