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. 311
With erbium nitrate the addition of acid produces more marked effects:` see plate 17.
All the bands which are more diffuse with the neutral nitrate than with the equivalent
chloride solution, are still more diffuse with the acid nitrate; and the effect regularly
diminishes as the solution is made more dilute. There is however no indication that
there is any weakening of the intensity of the bands by the presence of acid, but rather
a strengthening of them.
With the chloride, on the other hand, there seems to be no more difference
between the absorptions of the neutral and acid solutions than there is between the
corresponding solutions of didymium chloride. Comparing the spectra by eye, I can see
no appreciable difference between the acid and neutral solutions of equal thickness and
equal erbium concentration. Plate 16 gives a reproduction of photographs of the
absorptions of two pairs of equivalent neutral and acid solutions of erbium chloride, the
upper pair being those of the strongest solution. The creeping down of the continuous
absorption with the acid solution is visible in both pairs of spectra, but more evident
with the stronger solution, where it sensibly affects the apparent intensity and breadth
of the broad band at about X 451. The second pair of spectra on this plate were
taken with solutions made by diluting those used for the first pair of spectra until their
volumes were three times as great as before, and they were put into tubes four times as
long as those used for the first pair. There is no indication of any weakening of the
absorptions by the addition of acid.
The absence of any diminution of intensity either of the didymium or erbium bands
by the addition of acid, taken in conjunction with the fact that rise of temperature
does not increase their intensity, go a long way to negative the supposition that these
bands are produced by the metallic ions; and the facts recorded in the preceding pages
rather suggest that the metallic bands are the outcome of chemical interactions between
molecules of the salt with each other and with those of the solvent, while the general
absorption at the most refrangible end, which is evidently of a different class and
resembles the absorptions of glass and many other substances which absorb the more
rapid vibrations but are transparent to waves of less oscillation-frequency, may perhaps
be due to encounters of molecules without chemical change. The effects on the spectrum
when different solvents are used may throw some light on this question. Accordingly
I made some experiments with didymium salts in various solvents.
EFFECTS OF DIFFERENT SOLVENTS.
Didymium chloride solution evaporated at 100~ retains some water, and seems to have
the composition of the crystalline salt. Dried at a higher temperature it may be had
anhydrous, but in that state appears to be quite insoluble in alcohol. Dried at 100~
it dissolves with tolerable facility in absolute ethyl-alcohol, and in glycerol, but will not
dissolve in benzene. The alcoholic solution deposits beautiful pink crystals on evaporation.
The absorption spectrum of this solution shews the same bands as an aqueous solution,
but they are somewhat modified. They are more diffuse so that the weaker bands look
as if they were washed out, and the positions of maximum absorption are all moved