292 Mr. S. P. Langley on the Amount 



original heat A would become Ap after passing through one 

 stratum (ES) ; and, according to what has been assumed, it 

 would become (when the sun's zenith distance became ESF) 

 Ap 2 after absorption by the two strata between F and S, Ap 3 

 after absorption by the three strata between Gr and S ; &c. A, 

 the original heat, and p, the coefficient of transmission, are 

 unknown ; but if we make an observation of the heat actually 

 reaching S along FS (let us call this heat m),and again later 

 in the day along KS (calling this second observed quantity 

 n), we have in the particular case supposed 



Ap 2 = m, Ap 3 = n, 



whence A and p both become known. Designating the 

 number of strata by e and the observed value (m or n) by t, 

 we have then Ap e = t ; the exponential formula of Pouillet and 

 later investigators. Its fundamental (and erroneous) assump- 

 tion is, that the coefficient of transmission (p) is a constant. 



It is no doubt true that a very sensible portion of the 

 solar rays is scattered by an analogous process in our atmo- 

 sphere ; but we have in our present knowledge to consider 

 that, by whichever of its effects we note it, this radiation is 

 not simple, but complex. Thus we must remember that heat, 

 like light, is of different kinds. To use Melloni's illustration, 

 radiant light would, to any eye that could see it, appear to be 

 of totally different colours ; and hence, it may be added, we 

 ought no more to attribute to it a single rate of absorption 

 with regard to any absorbing medium than to assume that a 

 blue and a red ray would pass through a red glass with equal 

 facility. The statement of this tact may perhaps seem 

 superfluous to the reader, for it has long been in one sense 

 well known. But in another and most important sense it is 

 not well known ; and it cannot be superfluous to recall its 

 bearing on our present research, since it is the neglect to 

 follow it into its consequences which has led to the error in 

 question. Since, then, the solar energy, whether regarded 

 in whole as " heat " or " chemical action," or in part as 

 " light," is the sum of an infinite number of radiations ; these 

 may be conceivably influenced in an infinite diversity of ways 

 by the different atmospheric constituents. In fact the larger 

 particles rather reflect than diffuse the heat or light, and hence 

 treat all wave-lengths nearly alike, or diminish the direct 

 radiation by a nearly general absorption ; minuter ones begin 

 to act selectively, or, on the whole, more at one end of the 

 spectrum than the other ; smaller particles, whether of dust 

 or faintest mist, and smaller still, form a probably con- 

 tinuous sequence of more and more selective action, down 



