262 
PROFESSOR FORBES ON THE EXTINCTION OF THE SOLAR RAYS 
Thickness, or x in millimetres of mercury. 
2500 
3000 
infinite 
Intensity, or v in degrees of Actinometer B. 2. 
3-1 + 1 5*2 = 18*3 
1*7 + 15*2 = 16-9 
00 4- 152 = 15-2 
Hence, supposing- the approximation to the initial intensity of the solar heat to be 
sufficient, the portion transmitted through an atmosphere balanced by 760 millime- 
tres of mercury will be found to be 
39-03 
— '534 or its whole amount. 
The value of V is by no means given as certain : it may very probably be greater, 
even much greater than has been assigned, but it is very unlikely to be less. 
98. Hence, too, the absolute intensity of the solar ray has been very much under- 
rated by all writers. The portion vertically transmitted probably does not exceed a 
half, instead of being equal to two-thirds or three-quarters, as has generally been 
supposed. Bouguer’s estimate for light approaches nearest to it, but that was 
founded on the logarithmic law, which we have shown not to be applicable, at least 
for heat. 
99. It may be interesting, however, before finally quitting the observations of the 
25th of September, 1832, to inquire what results we should have deduced from them 
upon the old hypothesis of the intensity diminishing in geometrical progression, and 
thus to render the observations directly comparable with those of Bouguer, Lambert, 
Leslie, Kamtz, and Pouillet, that is, so far as I am aware, of every author who has 
published any determination of the opacity of the atmosphere, including Laplace. 
100. Resuming the notation of Art. 7, which we used to describe Bouguer’s me- 
thod, where v x and v 2 are two intensities expressed in actinometric degrees, x x and x 2 
the corresponding atmospheric masses traversed expressed in millimetres of mercury. 
By equation (3.) of that article* we find the value of the coefficient of extinction 
log. 
m = 
0C i 0C o 
And if [u] — the intensity after a vertical transit through the atmosphere, the inten- 
sity beyond the atmosphere being = 1, we have by equation (1.) of that article, 
log M = m x " 60 ‘ 
When more than two values of v and x are used we may divide them into two series, 
and take the arithmetical means of log v and x for each. 
101. Now a good deal depends upon the way in which these series are formed. We 
may combine the observations, so that the observations on the shortest atmospheric 
columns forming one series shall be set against those of the longest columns in another. 
Thus the observations at Brientz alone give the following results. 
* It will be seen that this equation is obtained in exactly the same way as that of Art. 65. 
