SUPPLEMENTARY REPORT ON METEOROLOGY. 139 
Transmission, and the Corona by Reflexion, the same stratum 
of cloud forming both; and he obtains in general a satisfactory 
coincidence* ; (5) by the curious observation of M. Babinett, 
of similar rings formed round the shadow formed on the minute 
particles of gunpowder which had been laid out to dry. 
303. On the other hand, it is not consistent with what we 
know of such phenomena, that the third ring (for instance), 
that having a diameter of 18°, should be seen so vividly, and 
sometimes to the exclusion of any other (for it is hardly likely 
that so large a ring should be one of the first or second order). 
The great rings of 67° to 84° in diameter, can scarcely be thus 
explained. 
304. Dr. Young seems to have felt the full difficulty of the 
subject of reflected glories ; and he has unfortunately expressed 
himself very obscurely upon it. In his earlier writingsf, he 
contented himself with referring them to the colours of thin 
plates ; but afterwards he seems to have hesitated whether to 
consider them as modifications of the corona (by diffraction), or 
of the rainbow. In his article, “‘ Chromatics §,’’ he has given 
a singularly perplexed theory, which we have read over many 
times without clearly understanding ; the reasoning, however, 
appears to be this :—Supposing a cloud to be formed of drops of 
water, light, four times reflected within such drops, will furnish 
rays which will unite in the same phase from both sides of the 
drop, at a point diametrically opposite to the sun, whilst rays 
(also four times reflected) ultimately coinciding in direction, 
and which have undergone reflexion at smaller and greater 
angles, will unite in different phases, the retardation being 
nearly as the angular distance from the central point. It is 
to these supernumerary bows of the fourth order, that he ap- 
pears to attribute the inner circles, whose diameters were, in 
the observation of Bouguer, 6°, 11°, and 17°. 
305. Having found the size of the drops which would give, 
by diffraction, rings of this description ||, he proceeds to find the 
distribution of the colours of a primary rainbow formed by such 
excessively minute drops (less than z,45 inch). He assigns 
to the first supernumerary red bow a radius of 24°, of which, 
however, he inakes nouse. He then observes, that owing to the 
minute size of the drops, the shading off of the primary colours 
is very slow on the concave side, so that the red will sensibly oc- 
cupy a breadth of 73°, the violet of 53°; and as the red and 
* Meteorologie, iii. 111. t Comptes Rendus, iv. 645. 
t Lectures, ii. p. 645. col. i. § Encyc. Britt., seventh edit., p. 638. col. i. 
|| Computed, however, for a radius of 5° for the first ring; whereas, in 
Bouguer’s observation, the diameter was 5° 20!. 
