380 
Transactions of the Society . 
in a different colour to the image composed chiefly of the finer struc- 
tures, but the two blend together, and the total result is that the finer 
structure appears more distinct than by ordinary illumination, whilst 
we retain the image of the coarser structure just as distinctly as 
usual. 
When we use a disc with the periphery coloured only, e.g. red, we 
get rather less contrast, but the image is almost monochromatic. The 
reason is that the image of the whole of the object, excepting part of 
the coarser structure (division I., class A) is formed by means of the 
red rays, whereas we get more light in the formation of the back- 
ground than in the last case, thereby lessening the contrast. On this 
account it is sometimes possible to get better results with a red than 
with a blue disc, because the former colour is more luminous. 
If we use a disc with central portion and the remainder both 
coloured (plate IX. fig. 6), we can obtain very good contrast, but we then 
introduce the one error attaching to the use of the dark-ground stop, 
viz. the duplication of structure of which only diffraction-spectra of 
first order are admitted by the objective, when their dioptric pencil 
falls within the central zone.* Also, in this case, w r e should bear in 
mind the relative intensity of the lights we are using, and modify them 
with a view to the desired effect, which is usually to emphasise the 
finer structure. Thus, in general, the central spot should be more 
deeply stained or coloured than the peripheral part of the disc. 
It will now be seen that in the last two cases we are using a 
species of illumination offering almost all the advantage of ordinary 
monochromatic illumination, vrith the additional advantage of great 
contrast of background. 
. It may be objected that the effect of chromatic aberration of the 
objective has not been eliminated. But, in reality, it has, for by 
hearing in mind how the differently coloured parts of the objective act 
independently of one another, it will be understood that we are 
practically working at the same time with two objectives, one of which 
is of very much smaller numerical aperture than the other. But 
since the penetrating power varies inversely as the N.A., it follows 
that the central zone, when used independently, has far greater 
penetrating power than the "whole objective or the whole objective 
minus that zone. This then allows the image from the central zone 
to accommodate itself to that of the remainder of the objective and to 
be in focus at the same time, and corrects for chromatic aberration. 
A very interesting feature about the colour-disc method is, as 
you will already have noticed from the foregoing, the additional evi- 
dence it affords on the Abbe theory; and at this point I must 
* Theoretically, since the two colours of the disc are never truly monochromatic, 
a certain range of wave-lengths is transmitted by both, and this tends towards 
reducing the error. Owing to their comparatively feeble luminosity, however, when 
we use a disc with strongly contrasting colours, the effect is not noticeable^ visually. 
Photographically it is possible to detect it. 
