DR. W. M. HICKS : A CRITICAL STUDY OF SPECTRAL SERIES. 443 



Argon. There are a large number of strong lines in the visible spectrum 

 evidently connected with D(l) and F(2) lines. The oun is so small that the 

 qualifying test for D lines is not so definite as in the other gases, although this is to 

 some extent remedied by the fact that the measures on the whole are good with a 

 possible maximum error of '02A. We shall therefore make no attempt to discuss the 

 F and D series with the same fullness as in the other cases. The groups selected are 

 certainly not the only ones and possibly may not be the most important ones, but 

 they will be sufficient to give data for the determination of the oun to about the 

 same degree of accuracy as for the other gases of this family. Take for the first 

 group. 



m = 1. m = 3 



(1) 23782-51 (K.) 179-92 (6) 23962*43 (E.V.) (l)44827'37) 175-51 (l)45002'88 



(2) 23899-08 (K.) (3)44835'41 



m = 2 

 (3)39357-06 178-95 (4)39536-01 



(3)39420-68 



39^ 



(5)39439-33 



The lines for m = 2, 3 are by EDER and VALENTA. Again the low intensity for 

 D n (l) is to be noticed, but we have here some indication of the source of this 

 peculiarity. EDER and VALENTA give the line at 23899'83 of intensity 9, and state 

 that it appears only with a very strong condenser discharge. The difference in the 

 two measures (d\ = "13 A) may possibly be due to observation although it is greater 

 than is the rule between the measures of K. and of E. V. As an oun displacement in 

 the sequent produces a change dn = "4, the difference may be due to 2^ displacement. 

 If so, no error will be introduced in the succeeding considerations by overlooking this, 

 treating it as due to observation error and using KAYSER'S measurements, subject to 

 a smaller possible error. For in either case the dependence of the sequent mantissa 

 on the multiple law is not affected and although the multiple itself is different, at the 

 same time the difference in the two readings will not modify the formula constants to 

 an extent to appreciably affect the calculated lines for m>2. 



Using the limit D ( oo) = S ( oo) = 5173T03 and the first two D n lines the formula 

 found is 



n = 5173r03-N 



m 



