On the X-ray Spectrum of Nickel. 275 



arm. Both the distance of the slit and that of the plate from 

 the spectrometer axis were 19*45 cm., since under these 

 circumstances the point at which a beam reflected at a given 

 angle meets the photographic plate is independent o£ the 

 point at which it meets the crystal face, and therefore of 

 the exact angle at which the crystal is set. 



Special experiments were made to see whether the strong 

 lines called a. and ft could be resolved into other components. 

 For this purpose the platinum slit was made very narrow, 

 and the width of the interference line on the photographic 

 plate was not greater than 0*15 mm., corresponding to a 

 difference of 0° 3' between its edges. Photographs of the 

 spectrum in the second order showed clearly that both a. and 

 ft were close doublets, the values of the angles of reflexion 

 from rock-salt for the first order being 



« = 17° 6', 17° 9'6', 

 ft = 15° 22', 15° 25'. 



The lines were not completely resolved in the first order 

 spectrum. Since it is conceivable that the doubling might 

 be due to crystal imperfections, the spectrum of the same 

 radiation was examined when reflected from the (001) face 

 of heavy-spar. This crystal had previously been examined 

 by Sir Ernest Rutherford and E. N. da C. Andrade *, and 

 used by them in the determination of the frequency of y rays. 

 Here again the lines were found to be doublets, and for both 

 rock-salt and heavy-spar the ratio of the frequencies of the 

 components was a = 1*0033, ft= 1*0032. In addition to the 

 strong lines a and ft, evidence was also obtained of a group 

 of faint lines lying between them. The angles of reflexion 

 of the two main lines were determined as accurately as 

 possible by examining the reflexions on either side of the 

 direct beam and applying a correction for the error of 

 centring of the crystal. 



In the case of elements belonging to the series called the* 

 K series by Barkla, a very large proportion of the radiation 

 seems to be concentrated in the two " lines " a and ft of 

 Moseley. The question naturally arises as to how far the 

 whole radiation from a nickel anti-cathode is concentrated in 

 these two wave-lengths. The problem was attacked by the 

 ionization method. The ordinary method of measuring the 

 leak of a gold-leaf electroscope in a given time is not very 

 satisfactory, since changes in the hardness of the bulb involve 

 alterations in the intensity of the radiation and also in the 



* Phil. Mag. May 1914, p. 854. 

 T 2 



