Series System in the Spectrum of Gold. IM 



Ctiange per oun= - 448. 



(2/-) 25922-38 0-C= -66 



3818-90 



(1)29741-28 



m = 6. 



(-^)(l)26'J10-aG =(26921-42) 0-C=--24 



381338 381552 

 ( - d x ) (2r) 30724-04 =(30736*94) 



This case is specially interesting as the observed lines 

 treated as (&i)S give also the separation proper to the limit 

 <^)S(oo). 



Amongst the results obtained in the foregoino- discussion 

 of most importance to the general theory of spectral analysis 

 may be mentioned : — 



(1) The existence of a D set corresponding to order m = l 

 far up in the ultra red whose satellite mantissa conforms to 

 the general rule of being a multiple of A. Gold therefore 

 does not form an exception, as formerly appeared to be the 

 ■case. 



(2) The indication given of the existence of summation 

 lines for P and S series. The evidence for these lines in [V.] 

 is drawn from D and F series. In general, such lines for S 

 and P will lie far down in the ultra violet, and this is, no 

 doubt, one reason why they have not previously been recog- 

 nized. The evidence here given must be sustained by further 

 numerical coincidences obtained in other spectra before it is 

 regarded as conclusive, but it is sufficient to render it ex- 

 tremely probable. 



(3) The more accurate determination of the oun depending 

 on 81 S = A = 113951 — 6p, where 'lp angstroms is an ob- 

 servational error. This result gives 



S = 1406'802--074/?. 



Taking the ratio of 8 to (W/100) 2 , where W is the atomic 

 weight, to be 361*75 + '05^, the resulting value of the atomic 

 weight of gold is 



197-2024--0052p--013(K/. 



The University, 



Sheffield, 

 April 12, 1919. 



