﻿and the 7 lieory of Atomic Structure. 547 



and it is rather difficult to get them in a very pure state. 

 Some preliminary investigations with rather impure speci- 

 mens of these elements showed clearly that from these no 

 certain information could be obtained about the weaker lines, 

 since the great number of " foreign " lines appearing on the 

 plates make the interpretation of the photographs nearly 

 impossible. 



The present investigation has been rendered possible by 

 the kindness of Mr. Auer von Welsbach, who offered to this 

 laboratory a beautiful collection of very pure salts of these 

 elements. On the photographs taken with these salts usually 

 no lines were observed belonging to any other rare- earth 

 metal. As the presence of about O'l per cent, of another 

 element would be sufficient to give the stronger lines, we 

 may conclude that the salts used were extremely pure. In 

 the one case, however, of Yb (70) several lines were observed 

 belonging to Lu (71). 



§ 2. The apparatus was the same as that used for the 

 former work. In this region of wave-lengths the spectro- 

 graph need not be exhausted. The slit of the X-ray tube 

 was covered with an aluminium sheet of 10 ju.. After being- 

 glowed in a Bun sen flame the salts (sulphates) were pressed 

 on a roughened copper or silver plate, which was soldered 

 on the anticathode. Imperial Eclipse plates were used, 

 which appeared to be much more adapted to X-ray work 

 than the technical X-ray plates previously used in this 

 laboratory. These plates are extremely sensitive, especially 

 for wave-lengths o of more than 3A.U., and even for wave- 

 lengths of 1-2 A.U. they give better images than the 

 technical X-ray plates in half the time of exposure. 



For the rather small glancing angles in question (10-20 

 degrees) the apparatus seemed in the beginning to be very 

 inconstant, giving sometimes fairly good and sometimes very 

 bad plates. This phenomenon appeared to be connected 

 with the reflexion on different parts of the crystal ; very 

 good plates were only obtained if the radiation was reflected 

 by the middle part of the crystal. As it seems to be impos- 

 sible that this phenomenon should depend on the geometrical 

 conditions of the apparatus we must assume that the reflect- 

 ing power of the crystal was greatest at the middle part. 

 Once the cause of this phenomenon was known, it was only 

 a matter of time to obtain rather good plates for all the 

 elements, as by a change of the position of the focus spot on 



2 N 2 



