450 
DR W. M. HICKS; A CRITICAL STUDY OF SPECTRAL SERIES. 
Unfortunately L.D.’s measures are only given to lA, and thereby their value is 
greatly diminished as they become merely indicatory and cannot serve as quantitative 
data. The accuracy of Baly and of Watson is good and probably about the same. 
Priest claims an accuracy with probable error <‘0005 A, Meissner with error not 
>’0015, but the accuracy of an interferometer measure depends very largely on the 
nature of the individual line. Meggers’ results are exceptionally valuable in that he 
gives interferometer measures of a number of lines of small wave-length 3701 to 3370 
where the S (2) and some of the higher order F lines occur. Rossi has succeeded in 
allotting lines to series. 
Neon affords an apparent exception to the rule amongst the rare gases of different 
spectra, according as they are developed with or without condenser in the tube 
discharge. On the other hand its spectrum would appear 'to be a composite one of 
the typical “ red ” and “ blue ” spectra. It undoubtedly has a portion analogous to 
the “ blue ” as will be seen by the results obtained below, completely analogous to those 
found in this communication for the other gases, which refer to their “blue” spectra. 
On the other hand in some remarkable sets of accurately equal separations discovered 
by Watson* it shows a relation to the analogous well-known constant separations 
observed by Rydberg in Argon. Further, it is specially rich in lines in the red region. 
In the list of lines observed at the Bureau of Standards referred to above there are 
225 between 8783 and 5689. Since in each periodic group of elements the number 
of lines as a whole increases very rapidly with the atomic weight, the excess of red 
lines in Ne is even comparatively greater than the actual number shows. The majority 
of these lines are weak, but they almost all fall into a few definite linkages in which 
the links are the constant separations discovered by Watson. Some of these special 
linkages again are connected together by the p and s links, especially the e.u.v. 
Tliey belong to the F type of order 77r = 2, and should afford most valuable 
information as to the way in which parallel and displaced lines are related. I hope 
to return to this question on a later occasion, and oidy refer to them in the present 
discussion incidentally as they afford some evidence for the determination of the 
value of the oun. 
The wave-numbers of obse^ed lines published stretch from 13251 to 36536. 
From analogy with the spectra of the other gases we must not therefore expect to 
find more than one order in each of the S and D series. Nor, with its small atomic 
weight will the e.u.v links be large enough to act as efficient sounders. On the other 
hand the whole of any F series (m = 1 excepted) shoidd lie within the above limits. 
It is therefore clear that the attack on the problem must be made first on 
this series. One datum at least is at our disposal in the magnitude of the oun. 
Taking the atomic weight at 20‘0 ± '01 the calculated value of d is 14’47 ± ‘01. 
This value of 14'47 may therefore be treated as exact to one or two units in the last 
digit. 
* ‘ Proc. Camb. Phil. Soc.,’ vol. 16, p. 130 (1911); ‘Astro. Journ.,’ vol. 33, p. 399 (1911). 
