272 
CAPTAIN W. DE W. ABNEY ON THE TRANSMISSION 
It should be remarked that the methods adopted in determining the absorption 
by astronomers practically eliminated any variation due to haze, since the stars were 
compared inter se, which is what the method employed in the foregoing observations 
also does, though in a more complete manner. 
The above plan of obtaining the coefficient of k leads to another very important result, 
which is that on any day, by taking an optical measure of the total light with any 
standard, such as a candle, and also a photographic measure of the same, we can fix the 
coefficient k to be applied to each wave-length, and also the loss by general absorption, 
with very great exactness, as two equations are formed from which each can be deduced. 
I would point out that even a simpler method is to expose a sensitive surface to 
rays coming through an orange medium and also through a blue medium, and measure 
the relative densities or blackness of the resulting photographs. The exposure 
through the orange medium will be equivalent to an optical measure. 
It must be recollected, however, that each different sensitive photographic compound 
will give results for different parts of the spectrum. 
§ XIX. Application of the above Law for the Comparison of the Actual Variations in 
Sunlight. 
Now on various days exposures were made on sensitive paper during the time of 
observation, and the depth of density measured, and from these measures were 
derived the intensity of the light acting. If the loss of fight is aloue caused by the 
scattering of small particles, and from no other cause, then the values of the ordinates 
of the different curves belonging to the days on which simultaneous observations were 
taken ought, at scale number 56, to be proportional to the values obtained from the 
photographs, the times being those given in the Tables. At the Eiffel the photographic 
value on September 15th was 256; at South Kensington on July 21st, 172; on 
October 29th, 72 ; on November 4tli, 72. The values of scale number 56, taken from 
the curves, are as follows:—Eiffel, 2'45 ; 21st July, 1*74; October 29th, ’73; 
November 4th, ‘7 ; which are fairly in accord. On the other hand, there are values, 
as for the 8th November and 23rd December, which are not concordant, showing 
that there was a general absorption of fight as well as a scattering in its passage 
through the atmosphere. 
§ XX. Considerations as to the Amount of Scattering. 
We are bound to ask ourselves what is the cause of the different coefficients for the 
scattering of fight in its passage through the atmosphere. It must be recollected 
that small particles of any kind will suffice for the purpose, and that it is merely a 
question of quantity which determines the coefficient. Greater or less amount of 
dust will affect the question, and dry weather is not the weather in which, from this 
cause, the scattering should be least. Again, small particles of water must always be 
more or less present, and it is believed by the writer that these are the most active 
