A REVISION OF THE ATOMIC WEIGHTS OF IODINE AND SILVER. I33 



Two new specimens of silver were employed. A portion of one had already 

 been used in an investigation upon the atomic weight of lead.^ This sample was 

 precipitated once as silver chloride, once as metal by ammonium formate, 

 and was finally electrolyzed. It is designated Sample C. Sample D was 

 employed in an investigation upon the atomic weight of bromine.^ It was 

 first purified in part by precipitation as chloride, in part by precipitation with 

 ammonium formate. The combined material was then subjected to two elec- 

 trolyses. Both specimens were finally fused in a current of pure hydrogen. 



METHOD OF ANALYSES. 



The acid was converted into pentoxide exactly as in the first series. Then, 

 after weighing, it was dissolved and reduced with a slight excess of hydrazine 

 as before, and precipitated with a sUght excess of silver. After filtration and 

 evaporation of the filtrate the excess of silver was determined gravimetrically 

 as silver iodide in a Gooch-Munroe-Neubauer crucible. 



Essentially no change was made in the method of analysis in analyses 8 and 

 9 except that a slight excess of hydrazine was employed. At this point a tube 

 50 cm. in length filled with glass pearls was substituted for the column of bulbs 

 in the reduction apparatus, since it was feared that traces of the spray produced 

 by the effervescence during reduction might have been carried through the 

 bulbs by the current of nitrogen disengaged. The pearls were moistened with 

 sulphurous acid before reduction was commenced. This change in the appa- 

 ratus was without effect, the succeeding five analyses giving results identical 

 with previous ones. In analyses 15 to 17 the method of precipitation was re- 

 versed by pouring the iodide solution into the silver nitrate solution. This 

 change also was without effect, hence it is reasonably certain that the dilution 

 of the solutions during precipitation was sufficient to prevent occlusion of per- 

 ceptible amounts of either iodide or silver nitrate. It is worth pointing out that 

 occlusion of iodide would lower the observed atomic weight of silver, while 

 occlusion of silver nitrate would produce the opposite effect. The occlusion of 

 ammoniimi nitrate by the silver iodide would of course be without influence 

 upon the final result. 



In one analysis, which is not recorded in the table, about one quarter of the 

 iodic acid was reduced by means of sulphurous acid, the ratio of silver to iodine 

 pentoxide in this experiment being somewhat higher than in the other cases, 

 owing doubtless to the occlusion of silver sulphate. In another analysis the 

 attempt was made to precipitate the silver iodide from ammoniacal solution by 

 adding the silver nitrate solution to the slightly ammoniacal solution of the 



* Baxter and Wilson: Proc. Amer. Acad., 43, 365 (1907); Jour. Amer. Chetn. Soc, 30, 187; 

 Zeit. anorg. Chem., 57, 174. (See page 67.) 



* Baxter: Proc. Amer. Acad., 42, 201 (1906) ; Jour. Amer. Chem. Soc, 28, 1322; Zeit. anorg. 

 Chem., 50, 389. (See page 55.) 



