PRESSURE EFFECT UPON ABSORPTION. 561 



such comparison can be made with the apparatus so constructed as 

 to bring rubber washers in contact with oil at high pressures. If this 

 source of trouble could be removed, it would then be desirable to 

 substitute for the arc a more constant source of light. Nitrogen 

 filled tungsten lamps, the "pointolite" and a tungsten arc kindly 

 furnished by the General Electric Company were tried for this work 

 but were found unsatisfactory because of the increase in the time of 

 exposure required with a less intense source of light. This made the 

 difficulty due to the gradual dissolving of the washers more marked. 



Selection of Material. 



The work was begun upon substances having well-marked absorp- 

 tion bands, a number of which were known to show some change in 

 the bands upon variation in temperature or concentration, since it 

 seemed desirable to make a comparison of any possible pressure effect 

 with variation in absorption already known to be due to other agen- 

 cies. After a preliminary study of various absorbing materials, the 

 following were selected for study: 



Aqueous solutions of salts of the rare earths neodymium, praseody- 

 mium, erbium and "didymium." 



Aqueous solutions of the uranyl salts. 



Aqueous solutions of cobalt chloride. 



A synthetic ruby. 



Samples of colored glass. 



Neodymium Solutions. 



The absorption spectra of aqueous solutions of neodymium salts are 

 known to be rich in bands of unusual sharpness which are much alike 

 for dilute solutions of different salts. Neodymium ammonium nitrate, 

 obtained through the kindness of the Welsbach Company, was the 

 salt used for most of this work but a number of observations were also 

 made upon the nitrate, chloride, and acetate formed from pure neo- 

 dymium oxalate obtained from Professor James of New Hampshire 

 University. 



As is well known, the bands are not all visible under the same con- 

 ditions but different ones may be brought out by varying the thick- 

 ness and concentration of the absorbing layer. In these experiments 

 the thickness of the absorbing specimen was limited by the size of the 



