736 Mr. J. Chadwick on the Charge on the 



the atomic number, but was equal to N — a where N is the 

 atomic number and a a constant for the series. For the 

 K series a — 1 and for the L series a = 7*4. In order to 

 obtain perfect regularity in the X-ray spectra, it was 

 necessary to leave three places for unknown elements 

 corresponding to atomic numbers 43, 61, and 75, and to 

 adjust the places of the elements A, Co, and Te, where the 

 order of atomic weights clashed with the order of chemical 

 properties. Then, in every case from Al, for which N was 

 assumed to be 13, to Au, N = 79, the X-ray spectra of an 

 element were defined by the number assigned to it. On the 

 nuclear theory of atomic structure this characteristic number 

 must be closely connected with the charge on the nucleus, 

 and Moseley concluded that the number gave in fundamental 

 units the actual value of this charge. He supposed that the 

 constant a represented a disturbing effect due to the elec- 

 trons in the vibrating systems, depending upon their number 

 and position. 



This is one of the most important generalizations in 

 modern physics, and gives a starting-point for the develop- 

 ment of the external structure of the atom. It is therefore 

 necessary to be quite clear as to the nature of the evidence 

 on which it rests. Moseley's discovery that the X-ray 

 spectra of an atom are defined by a characteristic number 

 is an empirical expression of his measurements, and is 

 independent of any theories of atomic structure or origin 

 of spectra. The identification of this number with the 

 atomic number, however, is attained by an arbitrary choice 

 of the constant a, and the statement that the atomic number 

 gives the actual value of the nuclear charge implies a theory 

 of the origin of X-ray spectra. The fact that the charac- 

 teristic number changes by unity from element to element 

 is, of course, very, suggestive on the nuclear theory, but 

 there is no experimental proof of Moseley's conclusion. 

 The good agreement between the observed frequencies of 

 the K lines and the frequencies calculated by Debye * and 

 Kroo t on the Bohr theory afford strong presumptive 

 evidence in favour of the correspondence between atomic 

 number and the value of the nuclear charge, but here the 

 assumption is made that there are no electrons between 

 the nucleus and the K ring. (This corresponds to Moseley's 

 choice of the constant a.) It is true that no X-ray series 

 of higher frequency than the K series has been observed, 



* Debye, Phys. Zeit. xviii. p. 276 (1917). 

 t Kroo, Phys. Zeit. xix. p. 307 (1918). 



