778 



SCIENCE. 



[N. S. Vol. XVII. No. 437. 



] 



tion bands in the blue portion of tlie 

 spectrum. Schottlander, working with the 

 same object in view, came by no means to 

 the same conclusion. He held it probable 

 that praseodidymium consists of a mixture 

 of two elements, whose oxides burned in 

 air give R^Oj and RO, {i. e., R^Oj + 2RO2 

 = R.j07, the accepted peroxide). His con- 

 clusion was founded upon the small per 

 cent, of oxygen present in the peroxide. 

 Forshing confirmed Bettendorff's results 

 spectroscopically and reported Pr a char- 

 acterized by the yellow bands, and Pr /?, 

 which has the three bands in the blue, 

 indigo and violet. Boudouard arrived at 

 the same conclusions. Brauner concluded, 

 in his work on the oxides, 'from the tend- 

 ency of them both (praseodidymium and 

 neodidymium) to become more highly 

 oxidized than would correspond to the 

 formulee ProO^ and NdaOj, that praseo- 

 didymium and neodidymium may be fur- 

 ther split up.' 



Von Scheele maintains that none of the 

 savants has proved Kriiss and Nilson's 

 theory that there are four elements pres- 

 ent in praseodidymium. By long repeti- 

 tion of the Welsbach process he failed to 

 bring about any variation in the yellow 

 bands, which Bettendorff maintained could 

 be fractioned away, and demonstrated to 

 his own satisfaction the elementary com- 

 position of praseodidymium in a paper 

 reciting much careful and patient experi- 

 mentation. Unfortunately, he ignored the 

 work of Crookes and Dennis, which is de- 

 pendent upon variation of the finer details 

 of manipulation. In our laboratory, fol- 

 lo'^^dng entirely novel lines of research for 

 this element, we have apparently verified 

 the conclusion of the complexity of the 

 element in question. 



The material used was generously pre- 

 sented by Mr. H. S. Miner, the long-time 

 associate and successor of the lamented 



Shapleigh. It was quite free from neo- 

 didymium, but contained a notable amount 

 of lanthanum. The presence of lanthanum 

 facilitates the fractioning of didymium by 

 the Auer method, as pointed out by Den- 

 nis. A pure praseodidymium compound is 

 readily had by using the method of Basker- 

 ville and Turrentine, namely, fractioning 

 a citric acid solution saturated with the 

 hydroxide, and heating. The citrate ob- 

 tained was converted into the oxide. This 

 oxide was free from the other elements giv- 

 ing absorption bands in the visible spec- 

 trum. So far we have not been able to 

 examine the ultra-violet, but shortly ex- 

 pect the arrival of one of Wood's nitroso- 

 dimethylaniline screens, which I am having 

 made for my spectroscope. The instru- 

 ment is a Steinheil plane grating (made by 

 Brashear) with 14,438 lines to the inch, 

 essentially the same as that described by 

 Dennis in his .work with Dales on yttrium, 

 except a size larger. It was purchased by 

 a grant from the Bache Fund of the Na- 

 tional Academy. The oxide was proved 

 to be free from elements which give no 

 absorption band, especially lanthanum, by 

 means of photographs of the arc spectrum 

 obtained with a Rowland concave grating 

 (15,000 lines to the inch and twenty-one- 

 foot diameter). The spectrograph work 

 was done by my friend Dr. "VV. J. Hum- 

 phreys, of the Department of Physics, Uni- 

 versity of Virginia, and will be published 

 by us in full at the proper time. The 

 oxide was very carefully treated with 

 hydrochloric acid, bringing about partial 

 solution, whereby a distinct brown oxide 

 was obtained different from the normal 

 , black peroxide ; further, a separation has 

 been secured by fusion with sodium per- 

 oxide. 



Other methods of attack have been fol- 

 lowed, as, for example, fractional crystal- 

 lization from a concentrated chloride solu- 



