248 REPORTS ON INVESTIGATIONS AND PROJECTS. 



between the steel and gold and caused the latter to separate. This was over- 

 come by using a vessel of brass. This vessel was gold-plated, and it was 

 found that the gold remained tightly adhering to the brass during the entire 

 work. With this apparatus the absorption spectra of aqueous solutions of a 

 fairly large number of salts were studied up to nearly 200 . 



It was found that the effect of temperature was in general to widen the 

 absorption bands. The solutions were then cooled gradually from about 

 200 to ordinary temperatures, and the absorption bands were found grad- 

 ually to grow narrower as the temperature fell. This was, then, a perfectly 

 reversible process. This widening of the absorption bands with rise in tem- 

 perature is in complete accord with the solvate theory of solution. Indeed, 

 this result was predicted from the theory before a single spectrum was 

 photographed. Raising the temperature has the same effect as increasing 

 the concentration, i. e., widens the bands. As we increase the concentration 

 we reduce the complexity of the hydrates existing in the solution, and as we 

 raise the temperature we produce the same effect, i. e., the reduction of the 

 complexity of the hydrates ; and both processes widen the absorption bands. 



The question as to whether ions and molecules have the same or different 

 power to absorb light — have the same or different resonance — is an old one. 

 Indeed, it was thought at one time that the ions are the chief absorbers of 

 light in solution. We now think that the absorption of light by solutions is 

 an electronic phenomenon, and this opens up anew the question of ionic as 

 compared with atomic and molecular absorption. 



Our earlier work showed that if ions absorb differently from atoms or 

 molecules the differences were very slight — so slight that it would require 

 special means to detect it. We therefore took a very concentrated solution 

 of the substance in question and passed light through a layer 0.5 cm. deep. 

 This solution was then diluted 100 times, and light was passed through a 

 layer of the diluted solution 50 cm. deep. The second solution was then 

 diluted 5 times and light passed through a layer of the third solution 250 cm. 

 deep, photographing the spectrum in the grating spectroscope in every case. 

 In this way we were increasing the ions present and decreasing the mole- 

 cules, and could see whether there was any change in the absorption spec- 

 trum. In every case the absorption bands were found to be wider in the 

 more concentrated solution, and it is thus highly probable that the ions have 

 slightly different absorption from the atoms or the molecules. In these 

 experiments we were, of course, also changing the complexity of the hy-. 

 drates present ; but a quantitative study of the plates showed that this would 

 not account for the phenomenon. 



The third problem was, in reality, the most important one studied during 

 the year — the measurement of the intensities of the absorption lines and 

 bands by means of the radiomicrometer. This was constructed of the alloys 

 of antimony and bismuth and of bismuth and tin, and a very satisfactory in- 

 strument was built. By means of this instrument we could not only measure 



