PROF. A. FOWLER ON SERIES LINES IN SPARK SPECTRA. 
242 
in the convenient notation adopted by Paschen in his extensive researches on series, 
represent the differences between the wave-numbers of the observed lines and the 
limits of the series to which they belong. Thus, if the Hicks equation were exact, 
the symbol 3D would be an abbreviation for 4N j (^m + D + — J when m has the value 3, 
and so on. Defects of formulae are eliminated in evaluating the variable parts of the 
expressions in this manner, except as regards the determination of the limits. In the 
present case it is probable that the limits are also exact. 
The differences between the two values of 3D, and between those of 3S are to be 
attributed to errors of observation. 
§ 6. A New System of Magnesium Doublets. 
Two conspicuous doublets, a little more refrangible than A. 4481 (Plate 3, fig. 2), 
were recorded in the magnesium arc in vacuo by Fowler and Payn in 1903 A 
Following E. E. Brooks,! who also investigated their occurrence in the magnesium 
spectrum, it will be convenient to designate these the “ F and P ” or “ F.P. ” Doublets 
in order to distinguish them from the “ Wide ” Doublets which have already been 
discussed, and to avoid confusion with the term “ Narrow ” Doublets as applied to 
the F series of Ca, Sr, and Ba. The wave-lengths of these lines were given more 
accurately by FowlerJ in a later paper as 4434'20, 4428'20, 4390'80, and 4384'86 
(Howland), from which the wave-number intervals in the two doublets are 30'5 
and 30'8 respectively. 
The lines evidently form two related pairs, and the experimental conditions for 
their production are identical in every respect with those for A 4481, thus differing 
from the Wide Doublets in the ultra-violet. It appeared, therefore, that the “ F.P.’’ 
Doublets might belong to series, the investigation of which would throw further light 
on the relationships of the “ 4481 ” series. 
Six other doublets with the same interval were eventually traced, and an additional 
one in the infra-red has been recorded by Lorenser. The doublets occur in groups 
of two, the more refrangible being slightly the stronger in each group. They fall 
readily into two series, and as in the other enhanced line series which have been 
considered, it is necessary to employ the term 4N for the numerator in the formulae. 
The stronger doublets have been regarded as belonging to the Diffuse, and the 
weaker ones to the Sharp series. According to Hicks this classification is supported 
by the signs of the last terms of the denominators when his formula is employed. 
Measurements of the lines have been made on photographs taken with the quartz 
spectrograph (Plate 3, figs. 4 and 5), with the exception of the doublets near A 4481, 
* ‘Roy. Soc. Proc.,'’ 1903, vol. 72, p. 255. 
t ‘Roy. Soc. Proc.,’ 1907, vol. 80, p. 220. 
I ‘ Phil. Trans.,’ A, 1909, vol. 209, p. 451. 
