THE 
MONTHLY MICROSCOPICAL JOURNAL. 
DECEMBER 1, 1875. 
I . — On a New Method of Measuring the Position of the Bands 
in Spectra. By H. C. Sorby, F.R.S., &c., Pres. B.M.S. 
( Read before the Royal Microscopical Society, November 3 , 1875 .) 
Plate CXXIII. 
In various papers published during the lost few years I have 
described a small piece of apparatus by means of which we can 
obtain a spectrum with a number of dark bands that can be used 
as a standard scale for measuring the position of those seen in any 
spectra compared with it side by side. For ordinary purposes, 
when very great accuracy is unnecessary, nothing could be more 
convenient ; since the position of the bands can be read off at once. 
The only serious objection is that it is difficult to accurately appre- 
ciate the fractional parts of the intervals between the different 
bands of the scale. Thus, for example, it may be difficult to say 
whether the centre of a band were situated at of or 3f. When 
there is any such difficulty I take the intermediate value, ; but 
even with every care, and after having acquired all the skill that 
comes from long practice, it may be looked upon as impossible to 
measure the position of the centre of any absorption band to less 
than one-eighth of the interval between the bands of the scale. In 
many cases this is of very little importance, but when it is necessary 
to measure the position with sufficient accuracy to enable us to 
ascertain the true laws of the relative wave-lengths of a number of 
bands, such a source of uncertainty is very objectionable, and it is 
most desirable to adopt some method that may give the wave- 
lengths true to a millionth of a millimeter. 
For some years past I have been fully alive to the importance 
of some plan which would enable us to obtain a spectrum in which 
a well-marked band could be made to travel up and down, and ad- 
justed exactly in the line of the centre of any hand in the spectrum 
under investigation. There is no difficulty whatever in seeing 
when one band is in exactly the same line as the other. It is a 
totally different thing to estimate by the eye alone the exact value 
of a fractional interval. This will be more easily understood by 
VOL. xiv. u 
