ZOOLOGY AND BOTANY, MICROSCOPY, ETC. 
99 
a wide angled cone at the lower focus aro blotted out. To state the 
combinations more concisely at the upper focus, we have two first order 
spectra and a dioptric beam ; and a first and second order spectrum and 
a dioptric beam. 
At the lower focus you have two second order spectra and a dioptric 
beam ; and two first orders and a third order. 
It may be as well to explain to those not acquainted with optics, 
that these combinations are caused by the spectra T N and H U passing 
through the same zone of the objective. The union of a set of spectra 
such as S D M makes a certain kind of image, and the union of P Q R 
will make a very similar image, not absolutely similar, but so similar 
that it would be difficult to tell the difference between them. So it 
comes to pass that the superposition of a number of very similar images 
strengthens the picture and gives a resultant image very close to the 
original structure. But the image caused by the union of T D N is 
totally dissimilar to the original, and HQK would also be very dis- 
similar and the superposition of a number of these can only make a 
stronger dissimilar picture, or if the pictures, which are superposed, 
differ widely from one another, then the superposition of them will pro- 
duce a fog. By way of illustration, suppose I made a large number 
of photomicrographic lantern slides, using certain spectra, which gave 
an image closely resembling an original known structure, and suppose 
each lantern slide to be a picture, resulting from a different narrow 
dioptric beam, such as D and P in our diagram, and others lying 
between them, we should then have a number of lantern slides, all 
very similar to the original and consequently to one another. Now 
suppose we had a number of lanterns and projected these several images 
at once on the screen, the several images would combine to form a 
strong image closely resembling the original structure. If, however, 
we make other lantern slides, using spectra, such as T N, which double 
the original structure, and if these are projected on the screen in place 
of the others, we shall get a strong image of a structure altogether 
dissimilar to the original. But if we increase the number of our 
lanterns, and project the other images as well, we shall have a confused 
image on the screen, or fog. Another illustration may help to simplify 
the matter. Suppose it were possible in photographing a dog with an 
ordinary camera, by manipulations at the back of the objective, to 
obtain, either an image only very slightly dissimilar to the real dog 
(such as an image slightly out of focus), or with other manipulations to 
obtain a picture of a hayrick. If a number of these slightly dissimilar 
images of the dog were projected on the screen, we should still have the 
image of a dog, and one that we could readily recognize. But if we 
projected the images of the hayrick, we should not have the slightest 
idea that the original object was a dog, and further, if the images of the 
hayrick were projected at the same time as those of the dog, the result 
would be a confused mass of light in which it would be impossible to 
recognize any image. Whether any particular lantern slide turn out a 
dog or a hay-rick, depends on the physical union of various other 
oscillations, but whether the image of either the dog or hayrick be a 
strong one, or a mass of fog, depends on the mechanical combination of 
similar or dissimilar images. 
H 2 
