278 Transactions of the Royal Microscopical Society. 
exception, however, that instead of a spot in the cavity of the 
hollow tube which carries the mirror, I have in my case placed 
in an angular slot a rather dense negative photograph of a milli- 
meter scale, divided into tenths, which serves as a micrometer ; the 
lines are therefore produced as spaces, through which the reflected 
rays from a small mirror pass, thence through two double convex 
lenses, which combination is capable of being adjusted for the 
purpose of focussing the image, which is then reflected upon 
the prism in such a position that it falls just above or below the 
spectrum ; this is obtained by moving the eye-piece fitting, and a 
small pin is placed in a slit at the side so as to prevent it from 
going too far. 
I then found that I must either reduce the scale from its pre- 
sent size, or put up with distortion, a difficulty which is quite 
overcome by the means I have adopted, viz. to elongate the tube 
at the back of the scale, to use a larger and longer focus lens, and 
an angular slot instead of a round hole, through which the inci- 
dent rays pass ; by this contrivance aberration and distortion are 
avoided, and the whole of the scale is seen quite clear and distinct, 
which fact appears, at least to me, to be a very important point in 
its favour. 
The adjusting screw, to which is attached the milled head scale, 
is used for setting the micrometer when necessary ; this has seldom 
to be altered, as it will answer for nearly all observations. It may, 
however, at times be useful to have the movement, more particularly 
for the purpose of arriving at the centre of a band. Care should be 
taken always to put it back to its previous position, which is easily 
managed by noticing the milled head, one revolution of which is 
equal to five divisions of the micrometer scale. (See Table 1.) 
So much for the piece of apparatus. We will now proceed to 
consider briefly the manner of using it, previous to commencing 
upon the actual measurement of bands by its means. To ensure 
success, therefore, I must first of all lay down a few rules, which I 
have divided into four heads, as follows : 
Firstly. Before proceeding to make an observation, arrange 
your light, focus the slit carefully, and put on the spectroscope, 
when the spectrum ought to be quite distinct ; then, by means of 
the small mirror, illuminate the micrometer scale, and focus it till 
the lines appear quite sharp and free from blur ; avoid a glare, or 
this is sure to be the case, but try to get as white a light as 
possible. 
Secondly. Care should be taken that the lines of the micro- 
meter are quite parallel with the bands seen in the spectrum. 
Thirdly. As the size of the slit greatly influences the appear- 
ance of the bands in the spectrum, it is necessary to modify it 
