354 CARNEGIE INSTITUTION OF WASHINGTON. 



ceeded in o]:)taining a product suitable for conductivity work. Hitherto 

 the high specific conductivity of the solvent used bj' other investigators 

 has prevented accurate conductivity measurements; but by their 

 method of purification Davis and Putnam have prepared formamid 

 comparable in conductivity with water. 



They have studied the conductivities and viscosities of solutions of 

 salts in formamid as a solvent. This solvent has a higher dielectric 

 constant than water, and would therefore, from the Thomson-Nernst 

 hypothesis, be expected to have greater dissociating power, and such 

 has been found to be the fact. The effect of dissolved salts on the 

 viscosity of formamid is similar to their effect on water. It is a func- 

 tion of the ionic volumes of the dissolved ions. 



Dr. Shaeffer and Dr. Paulus have studied the absorption spectra of 

 indicators bj- means of the new Littrow spectroscope designed by Profes- 

 sor Anderson, and an improved radiomicrometer. A very sensitive 

 thermoelectric junction was built by Dr. Guy, and his radiomicrometer 

 was the most sensitive that had been consti-ucted up to that time. 

 Shaeffer greatly improved the form of junction constructed by Guy, 

 and built an automatically compensating junction which gave far more 

 satisfactory results. 



In the Littrow spectroscope, designed by J. A. Anderson, a plain 4-inch 

 grating was used, the lenses having been made by Brashear. 



By means of the improved radiomicrometer and the large grating 

 spectroscope, the positions and especially the intensities of absorption 

 lines and bands could be measured with a much higher degree of accu- 

 racy than had been hitherto possible. The constants of a number of 

 the more common indicators were worked out with considerable cer- 

 tainty, and about 20 times more accurately than had been previ- 

 ously done. The numerical values of these constants must be known 

 with reasonable accuracy, in order that these indicators may be used 

 scientifically in quantitative analysis. 



The work of Dr. Shaeffer and Dr. Paulus was extended by Dr. Paulus 

 and Mr. Hutchinson to coralUn, using the same apparatus that Shaeffer 

 and Paulus had employed. A satisfactory constant was also obtained 

 for this indicator. 



Jones and Anderson, in Publication 110 of the Carnegie Institution 

 of Washington, showed that when certain salts with sharp absorption 

 lines and bands were dissolved in certain non-absorbing solvents, the 

 resulting absorption spectra were a function of the nature of the 

 solvent in which the salt was dissolved ; for example, neodymium chlor- 

 ide in water had a quite different spectrum from neodj^mium chloride 

 in methyl alcohol. The existence of well-defined ''solvent bands" 

 was thus established for the first time. 



A large number of such "solvent bands" for a large number of 

 solvents, and for a fairly large number of dissolved salts, were dis- 



