BAXTER. 



ATOMIC WEIGHT OF IODINE. 



81 



never more than one tenth of a milligram, was subtracted from the 

 original weight of silver, and no correction was applied to the weight of 

 silver iodide for the amount dissolved in the wash water. 



TABLE IV. 



Series VI. Ag : Agl. 



The close agreement of the averages of the different series makes it 

 certain that no constant error of magnitude still remained undetected in 

 any one of the methods. 



In Series I, II, and III, the three chief possible errors have been con- 

 sidered and shown to have no effect. Occluded silver nitrate in the 

 silver iodide was eliminated by fusion in iodine ; and it was proved both 

 that no volatilization of the halides actually took place and that no 

 halogen was retained by the solidified salt. In Series IV, V, and VI, 

 the chief possible error, that of occlusion of silver nitrate by silver iodide, 

 was avoided by sufficiently diluting the solutions before precipitation. 

 That the dilution was sufficient was evident from the agreement of the 

 results of the experiments with the larger and smaller quantities of 

 material, the total volume of the solutions being the same in all cases. 

 Furthermore, the weight of the silver iodide obtained in the analyses 

 recorded in Series V was in every case very nearly equal to the sum of 

 the weights of silver and iodine employed (Series IV), which could not 

 have been the case if silver nitrate had been retained by the iodide. 

 One is forced, then, to conclude that the average of all six series, 

 126.985, represents the atomic weight of iodine within a very few thou- 

 sandths of a unit, and that the results of the syntheses of silver iodide 

 described in the first paper are slightly too low, owing to occlusion of 

 vol. xli. — 6 



