I02 RESEARCHES UPON ATOMIC WEIGHTS. 



RATIO OF SILVER IODIDE TO SILVER CHLORIDE. 



In any atomic weight investigation it is extremely desirable to obtain the 

 value sought by reference to as many different well-known atomic weights as 

 possible. The method of heating silver iodide in a current of chlorine, which has 

 already been used by Berzelius and Dumas as well as by Ladenburg, furnishes 

 the ratio between silver iodide and silver chloride, and seemed capable of yield- 

 ing trustworthy results; for since the silver halides fuse at a comparatively low 

 temperature, there is no possibility of inclusion of silver iodide by the silver 

 chlorides formed in the reaction, and hence there is certainty that the reaction 

 will be complete. 



First, silver iodide was prepared by precipitating an ammoniacal solution of 

 the purest ammonium iodide (Sample I) with a solution of recrystallized silver 

 nitrate. In this operation a slight excess of ammonium iodide was used. The 

 precipitate was well washed with i per cent nitric acid, rinsed with water, and 

 was collected on a Gooch crucible with the use of a disk of filter paper instead 

 of an asbestos mat. In this way contamination of the precipitate with asbestos 

 shreds was avoided. The silver iodide was finally dried in an air-bath at about 

 100° C. for at least 8 hours. After removal of the filter paper, those portions of 

 the precipitate which had come in contact with the filter paper were cut away 

 with a clean knife. Next the substance was fused in a weighed crucible protected 

 from the flame by a very large crucible. While the silver iodide was fused, a 

 small quantity of the purest iodine was placed upon the lower side of a second 

 crucible cover, and this cover was substituted for the one which had been weighed 

 with the crucible. The iodine immediately vaporized and the silver iodide was 

 thus fused in an atmosphere containing iodine vapor. Finally, the cover was 

 removed, so that the uncombined iodine escaped from the crucible, and the salt 

 was kept fused, covered with the original crucible cover, until it was certain that 

 all excess of iodine had been ehminated. That no excess of iodine was retained 

 by the silver iodide was readily shown in one experiment with i8 gm. of salt by 

 reheating the salt to its fusing point and reweighing. A loss in weight of only 

 0.00003 gm. resulted. 



After the silver iodide had been weighed it was heated in a current of chlorine. 

 This gas was generated by dropping concentrated hydrochloric acid upon man- 

 ganese dioxide, and it was purified and dried by bubbling through water and 

 passing through a 3-foot tube filled with beads moistened with concentrated 

 sulphuric acid. Traces of bromine or iodine in the chlorine would have been no 

 disadvantage, and it is inconceivable that it should have contained fluorine. 

 In order to prevent spattering of the fused salt from the crucible, a perforated 

 porcelain disk, which fitted the crucible half-way between the bottom and the 

 top, was placed in the crucible. This disk was always weighed with the crucible. 



The chlorine was conducted into the crucible by means of a small hard-glass 

 tube, which passed just through the perforated cover of a Rose crucible, while 



