Series St/stem in the Spectrum of Gold. 19 



intensities — or the chance of the particular configurations 

 corresponding to those multiples existing — alter with each 

 addition, rising to maxima under certain as yet unknown 

 conditions. That sometimes such a displacement suddenly 

 confers increased intensity is certainly true. This whole 

 question is a most interesting one and of fundamental im- 

 portance in any theory of spectral origins. Before it can he 

 profitably discussed, however, a greatty increased collection 

 of raw material bearing on this point must be obtained. The 

 spectrum of gold seems especially promising for a more 

 exhaustive investigation under this head. 



In the following list many lines have been included to 

 illustrate the effects of displacement as above indicated. 

 Several of the individual illustrations are doubtless spurious 

 and mere numerical coincidences. This may be specially the 

 case with lines occurring in the region embracing ~W,N 25000 

 to 28000 (say, XX 4000 to 3500), which is exceedingly crowded 

 owing to this cause, containing as it does lines belonging to 

 different sequents and different limits. It is, however, the 

 number of such coincidences which give evidence as to the 

 general existence of the displacement effect. Their con- 

 sideration is, however, of importance from another aspect. 

 A line associated by one of the usual links to an unseen one 

 is evidence in support of the possible existence of the unseen, 

 but it does not give the power in general of exactly de- 

 termining its measure, owing to uncertain linkage modifi- 

 cations. On the other hand, an observed line satisfying a 

 condition of displacement from a suspected one, gives no 

 reliable evidence that such a suspected line can actually 

 exist, but it does give evidence as to its theoretical measure, 

 if it were capable of being emitted. This point is of impor- 

 tance in settling the mean limits as determined from the 

 means of difference and summation lines, i. e. limit = J(F + F), 

 where one of the lines in question is determined by displace- 

 ment from an observed set. 



In the following table sets of lines have been allotted whose 

 mantissse in the different orders (m) are of the same order of 

 magnitude — a condition which normal series lines must 

 conform to. Also a considerable number of other lines have 

 been added where they seem to show the presence of dis- 

 placement, especially where what should be a strong line 

 appears represented by a number of weak ones of similar 

 character. Under each order is inserted the change produced 

 in the sequent by the displacement of one oun. The corre- 

 sponding change on the limit is 9*14. Deduced values are 

 in ( ). 



■ C2 



