High-Frequency Spectra of the Elements. 789 



In spite of the zincblencle sheet and the long exposure, 

 die /3-lines especially were in many cases so feeble as 

 scarcely to be seen on a copy ; for that reason the copies of 

 Plate XIV. have been retouched in order to be reproducible 

 and do not, I suppose, generally admit of being measured. 

 Naturally the distances of the lines have been measured 

 directly on the plate. Lumiere plates, " Violet Label," were 

 used and were developed with metol-hydrokinone. While 

 exposed, the plates were covered with black paper. 



The calculation of the incident angle was made in the 

 following way. At first the incident angle for the strongest 

 of the silver lines was measured directly with a spectrometer, 

 whose prism had been replaced by the rocksalt crystal. The 

 crystal was turned till the a-line appeared at its sharpest 

 on a barium-platinocyanide screen in place of the plate. 

 Greater accuracy was obtained with a micrometer screw. 

 The crystal plate was then turned till the line appeared on 

 the other side, and the angle turned through, equal to 

 180 — 2 a, gave the incident angle a. The distances between 

 the lines produced by the direct and the reflected X-ray beam 

 for the element in question and silver may be a and b 

 respectively, and the corresponding incident angles A and B ; 

 then A is calculated by the relation : 



tan 2 A a 

 ta~n2B = S' 



Owing to altered arrangements during the research the 

 plates must be compared with two different silver plates, 

 on Plate XIV. marked Ag z and Ag n . With Ag x ought to 

 be compared Mo, Ru, Cd, and Sn, with Ag n all others. 



In the cases of Mo and Ru the incident angles, thanks to 

 the strong reflexion, could also be directly measured for the 

 «-lines. The values thus obtained were 7° 16' and 6° 34' 

 respectively, while the values calculated from comparison 

 with the silver plate were 7° 16' and 6° 34'\5, which shows 

 a good agreement. Generally, however, the errors of 

 measuring must be valued at 3' to 4' owing to want of 

 sharpness of the lines and owing to the incident beam possibly 

 not falling quite on the centre of the rocksalt crystal. The 

 " collimator " had been adjusted by an optical method, having 

 been turned till a pencil of light-rays through the two slits 

 was projected on the centre of the crystal. 



The wave-lengths have been calculated from the well- 

 known relation 



n~k—Zd. sin a (n here = 1), 



where ^ = 2,814 . 10~ 8 cm. 



