158 S. P. Langley on the Selective Absorption of Solar Energy. 



It seemed at first, then, improbable that the heat below the 

 red should materially exceed or even equal that above it ; for 

 this would demand (since the heat shown by the last ordinate 

 at X = 1' A, is very small) an extension of the curve of heat, as 

 obtained from the grating, to a distance enormously beyond 

 the furthest limit then assigned to the normal spectrum by 

 experiment. The writer's further investigations, however, 

 led him to believe that this immense and unverified extension 

 has really existed, and to thus confirm by independent means 

 the statements of Tyndall and others as to the great heat in 

 this region. He was unable to determine its exact limit with 

 the grating as then used, on account of the overlapping spectra, 

 but was, some two years since, led, from experiments not here 

 detailed, to suspect the existence of solar heat at a distance of 

 nearly four times the wave-length of the lowest visible line, 

 A(\=(r-76)orat\=r-0. 



We receive all the solar radiations through an absorbing 

 atmosphere; and it is of the first consequence to determine the 

 rate of this selective absorption for each separate ray. This 

 has (owing to the difficulties before alluded to) never been, so 

 far as I know, yet attempted. It forms a prominent part of 

 the present design. 



The great difficulty in this investigation, after the provision 

 of a sufficiently delicate heat-measurer, lies in the varying 

 amount of radiant energy which our atmosphere transmits, 

 even for equal air-masses. The solar radiation is itself sen- 

 sibly constant ; but the variations in the radiant heat actually 

 transmitted are notable, even from one minute to another 

 under an apparently clear sky. The bolometer, in fact, con- 

 stantly sees (if I may use the expression) clouds which the eye 

 does not. That these incessant variations are in fact due to 

 extraneous causes and not to the instrument itself, has been 

 abundantly demonstrated by measurements on a constant 

 source of heat. 



Those taken, for instance, on a petroleum-lamp, so placed 

 as to give nearly the same galvanometer-deflection as the sun 



■wave-lengths in the dark-heat region had been estimated by hazardous 

 extrapolations from contradictory formulae — formulae which profess a 

 theoretical basis, but contradict each other. Thus Miiller finds by Red- 

 tenhacher's formula a wave-length of nearly 5^*0 for the extreme solar 

 beat-rays, Draper (as we have just seen) a wave-length of but 1 M, for the 

 same rays, &c. All these formulae (Briot's, Cauchy's, &c.) agree well 

 with the observations in the visible spectrum, which they have in fact 

 been originally deduced from. They contradict each other thus grossly 

 when used for extrapolating the place of the extreme infra-red rays, 

 whose real place we give later from actual measures. 



