230 PROCEEDIXGS OF THE A.MEKICAN ACADEMY. 



these could have been present in quantity, their absorption bands obvi- 

 ously could be found only in concentrated solutions. The marked 

 broadening and running together of the neodymium bands in concen- 

 trated solutions makes the detection of bands due to impurities diffi- 

 cult, except when these occur in regions in the spectrum comparatively 

 free from neodymium bands. Of the three strong erbium bands X 650, 

 X 540, and X 522, the last two are completely masked by the neodymium 

 in concentrated solution. The band X 650, however, occurs in a region 

 favorable for observation, and could not be detected even in the most 

 concentrated solution. In the case of holmium the three strongest 

 bands, X 639, X 535, and X 485, all fall within the limits of neodymium 

 bands at high concentrations. There was no evidence of the strongest, 

 X 535, in solutions of average concentrations. The strongest dyspros- 

 ium band X 451, which falls just within the broad neodymium band 

 X 484-X 453, as it appeared in our most concentrated solutions, could 

 not be detected. The terbium band X 488 practically coincides with 

 the neodymium band 487 which could be seen in our preparations only 

 at the highest concentrations. Since this band could be seen of the 

 same intensity in all our preparations, it is improbable that it repre- 

 sents an impurity of terbium. The sharp thulium band X 700 does 

 not fall within limits of any neodymium band, and we could find no 

 evidence of its presence. Since the absorption of europium solutions 

 in the visible spectrum is faint, and since all the stronger bands are 

 masked completely by the bands of a concentrated neodymium solution, 

 we have no spectroscopic evidence concerning the absence of this ele- 

 ment. It may be added that, even in the extreme fractions of our 

 crystal series, we could not detect the presence of any of these rarer 

 elements. Further study of the absorption spectrum of neodymium is 

 planned with the aid of a large quartz spectrograph, which has recently 

 become available for us. 



The Pbeparation of Neodymium Chloride. 



The neodymium oxide fractions prepared as above were converted to 

 chloride as follows: Fractions 1302 and 1614 were dissolved in the 

 purest nitric acid in platinum vessels, the solutions were filtered, and 

 the oxalate was precipitated from acid solution by means of the purest 

 oxalic acid. After the oxalate had been washed by decantation and 

 drained by suction upon a porcelain Gooeh crucible employing a small 

 disk of filter paper in place of an asbestos mat, the salt was dried 

 and ignited in a platinum dish over a blast lamp. The oxide was dis- 

 solved in a quartz dish in hydrochloric acid which had been distilled 

 with the use of a quartz condenser, and the chloride was three times 



