On changeable Colours and Glories. 295 
to the simple principle of considering each unit as equal 
to about the 45000th of an inch. Hence.it seems that the 
necessity of a correction depends on the different state of 
the lights reflected from one side of a fibre, and diffracted 
round its opposite side, and that, when they proceed in a 
similar manner from two neighbouring parallel lines, the 
necessity no longer exists. What may be the cause of this 
irregularity, will perhaps be understood when we under- 
stand the cause of the singular phenomena of oblique re- 
flection discovered by Mr. Malus, and we haye no reason 
to expect to understand it before. 
Glories. 
*€T have had an opportunity of ascertaining, that the 
clouds which exhibit the white and coloured circles, some- 
times denominated glories, are certainly not composed of 
icy particles; and I have succeeded in deducing an ex- 
planation of these phenomena from the same laws, which 
are capable of being applied to so many other cases of phy- 
sical optics. In the theory of supernumerary rainbows, 
(Young’s Nat. Phil.) I have observed that the breadth of 
each bow must be the greater as the drops which afford it 
are smaller; and by considering the coloured figure, ‘in 
which their production is analysed, it will be obvious, that 
if we suppose the coloured stripes extremely broad, they 
will coincide in such a manner in one part as to forma 
white bow ; the red, which projects beyond the rest, being 
always broadest; so that, if all the stripes be supposed to 
expand, while they preserve their comparative magnitude, 
the middle of the red may coincide with the middle of the 
blue; and it will appear on calculation that a white bow 
will be formed, a few degrees within the usual place of the 
coloured bow, when the drops are about 355 OF saeco Of 
an inch in diameter. It is remarkable that in such cases 
the original rainbow is altogether wanting; and probably 
for a similar reason, we scarcely ever see a rainbow in a 
cloud which does not consist of drops so large as to be ac- 
tually falling, although I have once seen such a rambow 
eriding abruptly at the bottom of a cloud: it may be con- 
jectured that the edge of the light is in such cases so much 
weakened by diffraction, that it is too faint to exhibit the 
effects occasioned by a larger drop. Dr. Smith has made 
a remark somewhat similar, (Opt. r. 50!.) which, if not 
completely satisfactory upon the priuciples which have been 
mentioned, is certainly altogether unintelligible upon his 
owns. 
T4 ‘© The 
