50 
Proceedings of the Royal Society 
objection is reduced to a minimum ; and the method at best is only- 
applicable to small instruments. A spectroscope of sufficient disper- 
sion to split D,- or even h, in diffused light would entirely alter the 
appearance of the lines, and the rainband would appear as a hand 
incapable of comparison with them. With the small spectroscope 
used on a clear sky the D-line appears about equal to E in intensity 
when the rainband is at a minimum. As the band increases the 
D-line appears -wider, becoming in turn equal to &, equal to E, and 
greater than E. . In this way six shades of intensity may be 
easily distinguished, and these may be represented shortly as — 
D -1- 7T = E. 
„ >E<&. 
. .■ „ 
„ >J<F. 
>, =r. 
” 
The words, luide, dark, hlach, and intense, when applied to the 
rainband and the larger spectral lines, mean the same thing, for it is 
difficult to say whether a large rainband makes the D-line apparently 
blacker, or wider, or both, when viewed by the small spectroscope. 
Another point of some importance in the prediction of rain is, as has 
been pointed out by Mr Kand Capron, the visibility of the thin, lines 
in the green. These vary greatly in distinctness, being sometimes 
invisible and at other times very apparent, but it is difficult to avoid 
changing the value of the words used to express the different degrees 
of visibility. A long and intimate acquaiutance with the spectrum 
in its different appearances is necessary in order to overcome this 
difficulty. 
A regular record of the intensity of the rainband has been kept 
since June 1882, and the weather which followed within twelve 
hours of each daily observation at 9 a.m. has also been noted. 
Other meteorological observations were subsequently introduced 
to supplement those of the rainband, but during the first seven 
months the latter alone were considered, and a prediction founded- 
thereon was written down each morning. The degree of intensity 
warranting a prediction of rain was, of course, entirely a matter of 
experience, each success or failure modifying to a. certain extent the 
