22 



L. VEGARD. 



M.-N. Kl. 



Values of |^,| found this way are given in Table XI. 



Table XI. 



It appears from the table that both values of ^3 give a very close 

 agreement with observations, and from the mere numerical comparison 

 here given it would be difficult to decide in favour of either of the two 

 possibilities. 



There is, however, one circumstance in connection with the produktion 

 of Ko which might be of great value as a test of the correctness of the 

 various hypotheses, which might come into consideration with respect 

 to the emission process and the number of electrons in the third ring. 



Observations show that Kg always accompanies K^, (perhaps with 

 the exception of ^A'= 11, for which no /vo-line is given in the table og 

 Siegbahn). 



If, now, the homogeneous X-radiation is produced in the way here 

 considered, by recombination between primary rings, it follows that for 

 lower atomic numbers the recombining electron must come from the 

 surface ring; but the number of electrons in this ring will diminish as 

 we pass towards lower values of X. For Js' ^ 12 there will be 2 elec- 

 trons in the third ring, for J\"= 13, 73 = 3, a. s. o. If we assume 8 electrons 

 in the third ring, q^ will keep on increasing until the element .4r(iV= 18); 

 if we put q-i= 12, for the final J7-ring, r/3 will keep on increasing till 

 the element Ti (X= 28). 



From the periodic variation of the properties of the elements we 

 come (as we shall see later on) to the conclusion, that we shall most 

 likely have an increase of the quant number of the surface ring when 

 we pass the element Argon. If, however, for a moment we look away 

 from this possibility, we still have to consider the changes of frequency,* 

 which are due to ihe variation of the number of electrons in the system 



