Memoirs presented to the Cambridge philosophical society on the occasion of the jubilee of Sir George Gabriel Stokes, bart., Hon. LL. D., Hon. SC. D., Lucasian professor.

ABSORPTION SPECTRA OF SOLUTIONS OF DIDYMIUM AND ERBIUM SALTS. 307
5811 grams to the litre, and diluting it to twice, four times, and eight times its bulk,
I could find no change in the absorption spectrum produced by it when the thickness
of the absorbent liquid was proportioned to the dilution, either when directly viewed
or when photographed. See plate 4, which however does not include any part of the
spectrum below the green. Nor could I detect any difference between the spectrum of
the sulphate and that of an equivalent solution of the chloride.
Didymium nitrate in four dilutions, beginning with the strongest in thickness of
38 mn., and ending with one-eighth strength in thickness of 305 mm., gave spectra
which could not be distinguished from each other, in the range photographed. See plate
11, where the spectra are those of equivalent solutions of the chloride and nitrate alternately, beginning with 38 mm. of the strongest solution of chloride, next the equivalent
nitrate, then 76 mm. of the solutions of half strength, 152 mm. of one-quarter strength,
and ending with 305 min. of the two solutions of one-eighth strength. This appearance
of identity is brought about, however, by the diffuseness and strength of the absorptions
by which the details of the groups of bands are obliterated. When the spectra of the
same solutions in much less thickness are examined, it is seen that the bands of the
stronger solutions of nitrate are more diffuse, or wider, than the bands produced by
equivalent solutions of the chloride. The weak bands look washed out, the strong are
wider than the corresponding bands of the chloride, and in the strong groups the
component bands are merged together. By increasing dilution the several bands contract
themselves and become better defined, until, with solutions of -3 strength, I am unable
to see any difference between the bands of the nitrate, chloride, and sulphate in
equivalent solutions. In the stronger solutions the weak bands look weaker as well
as broader with nitrate than with chloride, the strong bands are broader but look no
weaker; but I think that when an absorption is very strong the eye does not pereeive,
nor a photographic plate always record, a small difference of intensity. There is no
indication of an increase of intensity of the bands of the nitrate by dilution with corresponding increase of thickness. There are, on the other hand, indications of a shift
of the positions of greatest absorption in the bands in the yellow and green, which
remind me of the much greater shift of these bands by the use of alcohol and other
solvents instead of water.
Comparing small thicknesses (5 mm.) of solutions, the big band in the yellow expands
with the nitrate beyond that produced by the equivalent solution of chloride, especially
on the less refrangible side. Of the four strong components of this band the least
refrangible seems, with the nitrate, to be displaced a little towards the red, and a less
strong diffuse band extends still further beyond the corresponding band of the chloride
on the red side. The less refrangible of the two strong groups in the green, which
for the chloride consists of two nearly equal strong bands separated by a narrow chink
of light, and of a fainter very diffuse absorption extending some way down towards
the red, has for the nitrate the less refrangible strong band widened out by diffusion,
some way beyond its limit for the chloride on the red side, and the more refrangible
is weaker with the nitrate. The more refrangible group in the green appears with the
39 —2