New Way of producing Colour-Contrast. By J. Bheinberg. 379 
This fact had not yet, to my knowledge, been experimentally 
demonstrated, and I therefore made a series of test experiments with 
discs coloured in various manners, using an a a Zeiss objective, and 
the Abbe diffraction-plate as the object, and also with use of a 1/5 in. 
objective and diatoms as the object, which have undoubtedly established 
the above as far as visual perception is concerned.* 
It will be well to consider separately the action of three kinds of 
colour-discs, viz. those in which the central zone (say quarter the 
effective diameter of the disc) is coloured only, those with the 
peripheral portion coloured only, and those with central and peripheral 
portion both coloured. 
In all three cases, of course, we obtain a background of the colour 
possessed by the central zone of the disc, because the whole of the 
direct light passes through that part only and not through the other 
(plate IX. fig. 5). 
Therefore if we use a disc with the centre coloured only, e.g. red, 
the image of the background is red, and the object appears almost 
white similarly, as was the case with the high-power dark-ground 
illumination. But whereas in that case the image of the object was 
determined by the light passing outside the central zone of the disc 
only, in this case it is determined by some further factors. For now 
part of the white light, viz. the red, can pass through the central as 
well as the peripheral portion of the disc, whereas light of other colour 
can only traverse the peripheral portion. The result is, therefore, that 
as far as red light is concerned we obtain a perfectly true image of 
the object, just as if we were using a monochromatic screen. Prac- 
tically coincident in position with this is an image of the object formed 
by the other components of the white light, which is incorrect to the 
small extent in which the image which we got with the dark-ground 
stop was incorrect. But the red correct image is so much more 
luminous than the other one, because none of the component rays of 
the diffraction-fans have been stopped out, that it is quite impossible 
to distinguish any duplication of structure. The image of the other 
than red light, however, being superposed as it were on that of the 
red light, the colour of the resulting image is practically as if it 
proceeded from white light. 
You will note that the above refers to all except division I., class A 
(plate IX. fig. 2). With the dark-ground stop we lost the image of 
part of the coarser structure falling under this category completely ; 
with the red spot stop, however, we do not do so, it is simply depicted 
in red. 
We have thus got rid of both the defects of the high-power dark- 
ground illumination, whilst we retain the advantage of the contrast it 
afforded. True, we have now the image of part of the coarser structure 
* Since this paper was written I have learnt from Dr. Czapski that some different 
experiments with the same end in view were undertaken by Prof. Abbe some years 
