4 REPORT—1840. 
the spectrum is pure. 8. Bands are frequently visible when the spec- 
trum is pure; but in that case they can be seen equally well, whether 
the mica be placed on the violet or on the red half of the pupil. Mr. 
Airy then gave a rapid sketch of the leading features of the undulatory 
theory,—showed how a series of rays or a wave in passing through a 
convex lens, being more retarded in passing through the middle of the 
lens than through the edges, was bent so as to be convex to the lens 
after passing through it, and thus was made to converge to a focus. 
Next, if a thin plate of a transparent substance, like mica, were made 
to cover half the lens, one half the waves were retarded; and thus, 
when the lens represented the eye, and the place of the focus the 
retina, interference was produced, which, when the distance of the 
retardation bore a certain relation to the distance of two waves (or the 
wave length), might obliterate the light altogether. Next, if a lumi- 
nous point be looked at, it never is seen as a point, but as a small circle 
of light; and if a retarding plate be interposed, under some circum- 
stances bands or fringes parallel to the edge of the retarding plate are 
generated in the circular image of the luminous point; and those bands 
are not symmetrically arranged from the centre of the circular image; 
and the amount of their deviation from symmetrical position depends 
on the retardation of the ray in passing through the plate of mica, 
&c. and therefore depends upon the colour of the ray, or upon its posi- 
tion ina spectrum. Thus, if we examine the bands formed by different 
kinds of light, similar to those from successive points in a spectrum, 
the bands formed by these different kinds of light will be shifted suc- 
cessively more and more to the right (or to the left as the case may be). 
Now, if different luminous points be superimposed (as suppose points 
of the different coloured lights contained in white or solar light), in 
general it would be easily understood that the bands belonging to one 
colour would fall unsymmetrically between the bands in the circular 
image of another colour ; and thus if a multitude of them were super- 
imposed, they would tend to obliterate each other. But if the luminous 
origins of the different streams of coloured light did not coincide, but 
were arranged side by side in the same order as the order of their suc- 
cessive retardations by the mica, (which supposition is exactly repre- 
sented by supposing the origins to be the successive points of a spec- 
trum, formed either by refraction or by diffraction,) then it might 
happen that the shift of the centre of each image, in proceeding from 
one to the next, was exactly equal to the shift of the bands in relation 
to the centre of the image; and if this shift took place in the proper 
direction, the bands formed by the light from the different sources 
would unite and would strengthen each other; but if the shift took 
place in the opposite direction, the bands would be more widely sepa- 
rated than before, and would obliterate each other. It resulted from a 
mass of calculation to which he had subjected these conditions, that the 
relation of the wave lengths of the different colours, in passing from 
the violet to the red end, was such, that, under favourable circum- 
stances, a retarding plate being made to cover the half of the eye next 
the violet end, the bands came together, and so strengthened each 
