378 
Transactions of the Society. 
words, all the coarser structure which does not happen to refract the 
light strongly from its original path is blotted out. 
Now the dioptric beams of classes B and C of division I. are 
intercepted by the stop also; in the former case, as the objective 
admits at least two, but generally several consecutive diffraction- 
spectra from the same incident pencil of light, the image does not 
appreciably suffer ; but in the latter case, i.e. in class C, where only 
diffraction spectra of first order are admitted by the objective, the 
stopping out of the dioptric beam causes a duplication of structure, as 
in the well-known experiment of Prof. Abbe with the diffraction-plate. 
With respect to division II., the structure corresponding to each 
of our classes A, B, and C, will be to all intents and purposes faith- 
fully depicted as reference to the diagram shows, because in this 
instance we always have the dioptric pencil together with at least 
one, but mostly several consecutive diffraction-spectra admitted to the 
objective. 
The total result then of our black central stop is to give us on a 
dark ground an image of the object, in which the coarser structure is 
to a great extent obliterated, and in which a certain part of the very 
fine structure is depicted somewhat erroneously.* Now it is evident 
that in dealing with the generality of objects, it is of little use to see 
the finer structure clearly, without being able to discern its relation 
to the coarser parts, and the want of being able to do this is in my 
opinion the great defect of dark-ground illumination by the method in 
question. As for the duplicated appearance of any structure so fine, 
relatively to the objective, that only diffraction -spectra of the first 
order are admitted by it, with the knowledge that this occurs, we are 
able to make allowance for it in our interpretation of the structure. 
Moreover, the separation of the lines is so small, that it is only just 
on the border of visibility, and therefore often only gives rise to a 
somewhat thickened appearance. To this defect, then, I think less 
importance need be attached in dealing with the generality of objects. 
The nature of the formation of the image by higli-power dark-ground 
illumination helps us to understand the image seen when using colour- 
discs, under similar conditions, to a consideration of which we will 
now pass. 
To begin with, however, we must take cognisance of a factor which 
has special importance in dealing with colour-discs above the objective. 
It is that in a diffraction-fan produced with light possessing wave- 
lengths of various magnitudes, such as white light, which diffractive 
fan will consist of an infinitely narrow dioptric beam, and a number 
of diffraction-spectra, only the rays of uniform wave-length can co- 
operate — in other words, the light of any particular colour acts quite 
independently of the light of other colours. 
* Certain differences may occur in the image according to whether the light ti a- 
versing the structure falls under division I. or division II., which it is unnecessary 
to discuss in detail. 
