2l6 REPORTS ON INVESTIGATIONS AND PROJECTS. 



contact with metallic silver. Furthermore, the averages of the three series 

 agree as well as could be expected. If all the results are given equal weight 

 the mean value of the atomic weight of phosphorus is 31.027, a result essen- 

 tially that obtained previously by Baxter and Jones by the analysis of silver 

 phosphate, 31.03. This research will soon be published in the Proceedings of 

 the American Academy of Arts and Sciences. Dr. Moore has also begun the 

 preparation and analysis of phosphorus trichloride. 



In order to explain the slight discrepancy between the value for the atomic 

 weight of iron previously found by Richards and Baxter by the reduction of 

 ferric oxide, 55.88, and that found by Baxter, Thorvaldson, and Cobb by the 

 analysis of ferrous bromide, 55.84 (see Year Book No. 9), ferric oxide was 

 subjected to new analyses by reduction in hydrogen, in the hope that by the 

 use of modern quartz apparatus more satisfactory results could be obtained. 

 This hope was fully realized. Very carefully prepared ferric oxide, upon 

 reduction for a prolonged period at about 1100° in a quartz tube, has yielded 

 the results 55-839, 55-844, 55-851, 55-850, 55-846, 55-843, 55-848; average 

 55.846. This investigation is being continued to include the examination of 

 meteoric material. 



The investigation upon the atomic weight of arsenic has been continued in 

 two ways. The comparison of arsenic trioxide with iodine, commenced last 

 year by Mr. G. W. Harris (see Year Book No. 9), has been extended by him 

 to include the examination of arsenic trioxide which has been sublimed in a 

 vacuum; while Mr. P. C. Voter has begun the comparison of arsenic trioxide 

 with iodine pentoxide. Although no final results have been obtained in 

 either case, preliminary experiments indicate that both methods yield the 

 same value for arsenic that was obtained by Baxter and Coffin by the analysis 

 of silver arsenate, 74.96. 



In addition, the atomic weights of silver, iodine, and neodymium are still 

 under investigation and a new research upon the atomic weight of praseo- 

 dymium has been begun. 



Jones, Harry C, Johns Hopkins University, Baltimore, Maryland. Grant 

 No. 688. Investigations on the absorption spectra of solutions. (For 

 previous reports see Year Books Nos. 2-9.) $1,200 



The work on the absorption spectra of solutions, carried out during the 

 past year, had to do with three general problems : 



First, the absorption spectra of comparatively rare substances have been 

 mapped. These include salts of neodymium in a number of non-aqueous sol- 

 vents, such as propyl, isopropyl, butyl, and isobutyl alcohols ; ether, acetone, 

 esters, and formamide. Certain uranyl and uranous salts were also investi- 

 gated in these non-aqueous solvents. It was always found that the absorp- 

 tion bands widen and become more intense as the amount of salt in the path 

 of the beam of light is increased; the more diffuse a band the greater the 

 widening under these conditions. 



