246 



REPORTS ON INVESTIGATIONS AND PROJECTS. 



As the matter stands at present the various sorts of material have yielded 

 the following results for the atomic weight of arsenic : 



Ag = 107.870 I = 126.920. 



Baxter and Coffin, by the analysis of silver arsenate, obtained the result 

 74.944 on the same basis (see Year Books Nos. 6 and 7). While it is easy 

 to see how arsenic trioxide sublimed in oxygen, or even in air, might occlude 

 traces of a higher oxide and hence possess less reducing power than material 

 sublimed in an oxygen-free atmosphere, it is more difficult to account for 

 the small but distinct difference between material sublimed in a vacuum and 

 in nitrogen. This difference will be further investigated. It is worth noting, 

 however, that the values for the atomic weight of arsenic obtained by both 

 methods with trioxide sublimed in air, in nitrogen, and in a vacuum agree 

 fairly well with that resulting from the analysis of silver arsenate. 



In previous work upon the atomic weight of neodymium (see Year Books 

 Nos. 8 and 9) a series of fractions of neodymium nitrate was prepared by 

 Dr. H. C. Chapin by crystallization from concentrated nitric acid, the most 

 soluble and the least soluble fractions being occasionally rejected. About 

 2,600 crystallizations were made. These fractions have been converted into 

 chloride and analyzed by Mr. W. H. Whitcomb in exactly the same fashion 

 as in the previous research with pure neodymium material prepared in 

 another way, except that here the chloride, before being weighed, was either 

 fused or heated very nearly to the fusion-point in a current of hydrochloric- 

 acid gas. In this way dry material was, secured, while in the previous 

 research, where the chloride was heated only to 350 , a correction for a 

 trace of residual water was necessary. When the chloride was fused it was 

 necessary to avoid a temperature much above the fusion-point, and to avoid 

 prolonged heating, since otherwise what appears to be an insoluble modifica- 

 tion of the chloride is formed. By observing the above precautions perfectly 

 soluble material was obtained. The following table gives the result of the 

 analyses; the numbers of the fractions increase with increasing solubility. 



