214 REPORTS ON INVESTIGATIONS AND PROJECTS. 



material sublimed in oxygen yields a somewhat higher value, 75.04. Mate- 

 rial sublimed in nitrogen nearly free from air yields a minimum value, 74.95, 

 which is identical with that found by Baxter and Coffin through the analysis 

 of silver arsenate. The meaning of these variations will be further investi- 

 gated and the titration of arsenic trioxide against iodine pentoxide also will 

 be undertaken. 



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

 No. 624. Investigations on the absorption spectra of solutions. (For 

 previous reports see Year Books Nos. 2-8.) $1,200 



The work during the year has been on three problems : first, the nature of 

 the solvent as affecting the absorption ; second, the effect of adding free acids 

 to salts on the absorption ; third, the effect of temperature on the absorbing 

 power of solutions. 



The first problem was studied rather fully, on account of its bearing on the 

 theory of solvation or combination of solvent with the dissolved substance. 

 If solvents in general combine with substances dissolved in them, then the 

 different solvates formed by the different solvents ought to absorb different 

 wave-lengths of light. That this is true will be seen at once if we recall that 

 the absorption of light is a resonance phenomenon. Those wave-lengths of 

 light that are absorbed throw something in solution into resonance with 

 themselves and are consequently stopped. The different solvates formed in 

 the different solvents would have different compositions, and the vibrators 

 in the molecules, whatever they are, would, consequently, be expected to have 

 different resonance. In a word, the different non-absorbing solvents would 

 be expected to affect the absorption of a salt dissolved in them. 



One case was found by Jones and Anderson. Neodymium chloride, when 

 dissolved in water, has a very different absorption spectrum than when dis- 

 solved in alcohol. In a mixture of water and alcohol the water-bands and 

 the alcohol-bands coexist on the spectrogram; showing that the one set of 

 bands was not the other set shifted in position, but that there were two sepa- 

 rate sets of bands corresponding to the two solvents. A fairly large number 

 of cases illustrating this same point have been found. Uranyl and uranous 

 salts have fine absorption lines and bands, which can be carefully studied 

 and their wave-lengths accurately measured. When a given uranyl or uran- 

 ous salt was dissolved in different solvents, the resulting absorption spectra 

 were very different. The solvents used were water, methyl alcohol, ethyl 

 alcohol, acetone, and glycerol. The existence of water-bands, alcohol-bands, 

 acetone-bands, and glycerol-bands has been proved. 



This result is specially interesting, as it furnished another line of evidence, 

 added to the many already discovered, for the correctness of the solvate 

 theory. 



The study of the second problem has led to results that seem to have a 

 bearing on the nature of chemical reactions in general. It is usually sup- 



