468 



DE. W. M. HICKS : A CKITICAL STUDY OF SPECTRAL SERIES. 



= or y = 2'4 '34p. In other words y = 2, p = 1. This 

 makes the limit of the second series less than that of the first by '17, in agreement 

 with the relative values as actually found by direct calculation from the observed 

 lines. Although therefore the limits suggest equality, there is good evidence that the 

 observed difference, though small, is real, and is due to a displacement of two ouns. 

 The sequence depends on an origin of 3667 + 49 = 3716 A 2 . 



Rossi's third series is (6) 28788'08, (2) 3408711, (l) 36536'62, (l) 37869'0, the last 

 being observed by himself. The first three give 



n = 4089673-N/{m+-024118-'043614/m}' 



with C = '36 for the last, probably excessive. The first line is the line adopted 

 as one of the S, (2) group above, the third is F (4). I feel some doubt, therefore, as 

 to these forming a real series, although the sequence has all the appearance of the 

 s (m) type for these gases. The limit does not seem to have any relation to the 

 doublet set. Its denominator is about 1'63. 



TABLE of Constants. 



