io8 



RESEARCHES UPON ATOMIC WEIGHTS. 



of iodine was employed, the fact that the solutions were always very nearly 

 as strong as tenth normal and that the ammonic iodide was poured into the 

 silver nitrate made it probable that occlusion had taken place at least to some 

 extent even here. It is noteworthy that three of these syntheses yielded results 

 as high as 126.926. Similar conclusions are to be drawn concerning the results 

 of the titration of silver and iodine, although in this case precipitation took place 

 in the reverse fashion, i. e., by adding the silver nitrate to the iodide. 



Accordingly experiments were carried out for the redetermination of the 

 ratios of silver to silver iodide and to iodine, as well as the new one of iodine to 

 silver iodide, with especial precautions to avoid the error of occlusion by using 

 very dilute solutions and no excess of silver. After the iodine had been weighed, 

 very nearly the exact amount of silver to combine with the iodine was weighed 

 out and dissolved in nitric acid with the usual precautions to avoid spattering. 

 The quantity of iodine used in each analysis was between 3 and 4 gm., and the 

 amount of silver consequently about 3 gm. 



The solutions of both silver and iodine were diluted to a volume of i liter each, 

 so that the solutions were about thirtieth normal, and then the silver nitrate 

 was added very slowly to the solution of hydriodic acid with constant stirring. 

 The flask in which precipitation was carried out was then shaken for some time, 

 and allowed to stand until the supernatant liquid was clear. This liquid was 

 tested for an excess of iodine or silver in a nephelometer, and if a deficiency of 

 either was found, it was made up and the solution again shaken, until the point 

 was reached where the extremely faint opalescence produced by both hydriodic 

 acid and silver nitrate was equal in both nephelometer tubes. The excess or 

 deficiency of silver was in no case over o.i mg., and in most cases much less. 



The Atomic Weight of Iodine. Series VII. Ag:I. 



Several samples of iodine were employed in these analyses, each one of which 

 had been three times distilled from an iodide as previously described. Three 

 different specimens of silver, purified by different methods, were used. Sample 

 A was a portion of the material used in the early part of this investigation. 

 Sample B was prepared from silver nitrate which had been recrystallized seven 

 times from nitric acid, five times from water, and finally precipitated by am- 



