486 
PROF. J. W. NICHOLSON AND DR. T. R. MERTON ON THE 
of reproducing in the laboratory the intensity relations which are found in the spectra 
of certain stars, in some of which the line X = 4471 is the most intense of the helium 
lines seen, whilst under ordinary conditions of excitation by uncondensed discharges 
at moderate pressures the D 3 line is the strongest line in the spectrum, and at low 
pressures the line X = 5015 becomes the most conspicuous. 
An investigation of the lines of the diffuse series of lithium with the use of the 
neutral wedge has not yet been made, but qualitative results for these lines, 
photographed with the concave grating spectrograph, confirm the view that the 
Stark effect is mainly responsible for the phenomena observed. It is, of course, 
impossible to control the experimental conditions when spectra are produced in the 
electric arc, and the results may be complicated by reversal. The peculiar character 
of these lines has been noted by numerous investigators. More especially the complex 
appearance of the line X = 4602 has been investigated by Hagenbach # and by 
Kamage. f If the vapour is dense, reversal may be observed in the line X = 6103. 
For less dense vapours the line X = 6103 is not reversed nor the line X = 4132, but 
the line X = 4602, which belongs to the same series and lies between these lines, gives 
the impression of an extremely unsymmetrical reversal. These appearances of the three 
lines have been recorded photographically on the same plate, using as a source of light 
a carbon arc containing a suitable quantity of lithium. It would appear extremely 
improbable that the apparent reversal of the line X = 4602 is real, since the lines 
preceding it and following it in the series do not show the phenomenon. This result, 
however, is precisely what we should predict on the supposition that the appearance 
of the lines depends mainly on the Stark effect. 
According to Stark (Joe. cit.) the electrical resolution of X = 6103 consists of a 
small displacement, but this result is probably incomplete. For the line X = 4602 the 
central undisplaced components, when the line is resolved, are either very weak or 
absent, and the displaced components are unsymmetrically arranged with respect to 
the unresolved line in the same direction as the asymmetry of the line as seen in the 
arc. In the case of the line X = 4132 there are components of considerable intensity 
very slightly displaced from the unresolved line. A minimum unsymmetrically 
localised in the broadened line is therefore precisely what we should expect for the 
line X = 4602, a phenomenon which we should not expect, and do not find, in the 
lines X = 6103 and X = 4132. In cases such as this, it is not improbable that the 
broadening of the lines may be affected by the potential fall between the poles of the 
arc, and indeed such phenomena as the “ pole effect,” or the small changes of wave¬ 
length which have been found to occur in the neighbourhood of the poles of the arc, 
may ultimately be explained in this way. In this case any contribution to the 
phenomenon of the direct action of the potential fall between the poles might be 
detected by polarizing the light. It may also be necessary to consider whether many 
* ‘Ann. d. Rhys.,’ (4), 9, p. 729, 1902. 
f Ramage, ‘Roy. Soc. Proc.,’ vol. 71, p. 164, 1903. 
