340 Proceedings of Royal Society of Edinburgh. [sess. 
results show then, that when the eyepiece is correctly focussed no 
trouble will be experienced from diffraction effects. 
In the former note all that was contemplated was a device for 
attachment to an ordinary spectrophotometer to bring the two 
spectra exactly together, that the judging of their relative inten- 
sities might be made more accurate. The author, however, had 
in view the object of measuring the light intensities of various 
liquids, which were to be contained in tubes about a metre long, 
and it was found that for this purpose, in addition to the above 
device, a further modification in the form of spectrophotometer 
was desirable. This new design of instrument also presents 
advantages for general spectrophotometrical work. 
Fig. 4 is intended to give a diagrammatic view of such an 
P f 
Fig. 4. 
[That some parts may be more easily seen, this diagram is not drawn to scale.] 
apparatus. The collimator A is so far distant from the prism R, 
that there is room to insert between the two the long tube B 
containing the liquid. The ends of the tube are made of plane 
parallel glass, so as not to interfere with the parallelism of the rays 
of light passing through it. Before the customary vertical slit of 
the collimator, there is placed a thin piece of opaque metal pierced 
with another slit whose opening is horizontal, so that the effective 
aperture of the two is a very small rectangular hole. This 
arrangement results in the production of a beam of light from 
the collimator lens, which is sensibly parallel, and, the tube B 
being only half filled with liquid, all the upper half of the beam 
of light passes entirely clear of the latter, while all the under half 
of the beam passes through the full length of the liquid. 
Without this arrangement, and using the light as it comes 
