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.

300 PROF. LIVEING, EFFECTS OF DILUTION, TEMPERATURE, ETC. ON THE
affected the photograph. A similar effect is produced by convection currents of unequal
density. These were pretty well avoided within the absorbent liquid, but could not be
completely avoided in the water of the bath. The difference of temperature, and consequent difference of density, of the currents in the water was, however, small, and
the thickness of water between the end of the tube and the quartz window of the
trough also small, so that the currents were not of much consequence. Attempts to
use temperatures below that of the room were abandoned because of the dew which
settled on the quartz windows. Wetting the quartz with glycerol was no remedy, because
the glycerol gravitated, destroyed the plane figure of the window, and dispersed some
of the light. Very fair observations by eye of the effect of heat on a solution, not
too dilute, were made by fixing two similar test tubes containing the solution, one in
front of the slit and the other in front of the reflecting prison, and after adjusting
their positions until the two spectra, seen simultaneously, were identical, heating up one
of the test tubes by placing a lamp under it. For dilute solutions, requiring a greater
thickness to give absorption bands of sufficient intensity, two of the tubes used for the
photographs were employed, one of them being heated up in its water bath.
As a source of light a Welsbach incandescent gas lamp without chimney was chiefly
used. This was placed 5 or 6 inches from the screen so that the network of the
mantle was quite out of focus at the slit. It gave a good light up to a wave-length
of X 370, but beyond this point it would not produce a good photograph without an
exposure too prolonged for the less refrangible part of the spectrum. For the region
above X 360 a lime-light was used.
Inasmuch as the bands observed are all more or less diffuse, and fade away gradually
on either hand, any variations of the intensity of the source of light, of the sensitiveness of the photographie plates, or of the development of the image, tend to mask
the effects of varying the composition, or the temperature, of the solutions; so that
two photographs can be fairly compared, for the sake of determining these effects, only
when they have been taken with the saine light, on the same plate, with equal times
of exposure, and have been developed together. This has been attended to throughout.
The photographs to be compared with each other have always been taken in succession
on the same plate, with no other change than the necessary shift of the plate and the
substitution of one tube of liquid and its bath for another. The photographs taken
thus in succession do very well for comparison of the intensities and other characters
of the absorption bands, but cannot be depended on for the detection of a very small
shift in the position of a band. That could be done if the two spectra to be compared were in the field at the same time, one of them reflected in, but I have not
attempted to photograph two spectra in this way, and have been content to detect
alterations of wave-length, in the bands most easily visible, by the eye without
photography,