1920: No. 2. 



ON THE X-RAY SPECTRA. 



31 



This is no longer necessary when we assume recombination from 

 secondaries; because in that case there is altoays one and only one elec- 

 tron in the system of depart. 



As the simplest hvpotesis we shall assume one .V-ring, and calculate 

 the frequency for M^ on the assumption of 73 =- 12. 



The values of B and (' are found from (5 c) by putting: 



i = 3, Ä; = 4, "3 = 3. n^ = T = 4. />:i = 10, /^4 =10 + 73 



B 



^ (qz) -t- 10 10 -t- q-i — I 



Q 16 



= -,093(73= 12) 



^_g2ji0.73) «^2(10 + 73. 1) — 2(l0-|-73-0>5) _ i;,8(7,— 12) 

 9 16 



Further : 



54 /X— 14,98V /iV— 14,42V /^ — 2\\^u 



In Table XV^ are given calculated and observed values of ^ for some 

 atomic numbers. 



Table XV 



The agreement is seen to be very satisfactory. It should be remarked 

 that if we give up the assumptions of two L-rings the residts given in 

 the table would also have folloived as a consequence of the assiimjMon of 

 recombination from secondaries in tlie original form of tJiis hypothesis. 



Quite generally we might say that if we stick to the hypothesis 

 that there is only one system with a given quantnumber, the modified 

 form of the hypothesis of recombination from seconderies will give, for 

 the aline of each series, the same frequency as it did in its original form, 

 because there is no electronic system between the secondary circle and 

 the broken ring. 



