S. P. Langley on the Selective Absorption of Solar Energy. 171 



is approximately double that at meridian ; or else three times 

 daily — at meridian, when the mass of air traversed is approxi- 

 mately 1^ time that at meridian, and again when it is ap- 

 proximately 1\ x 11 = 2^ times that at meridian. 



It will be observed by a reference to the map, that the 

 points chosen for measurement coincide, as a rule, with the 

 summits of the energy-curve; but separate investigations are 

 still in progress on the nature of the absorption in the inter- 

 vals, to determine whether the newly observed bands are of 

 solar or terrestrial origin. The part of the spectrum included 

 extends from \= (T'383 (above H in the violet) to \ = 2 M '28 

 (in the newly observed infra-red region, about two octaves 

 below Fraunhofer's B). 



(As we can, in fact, obtain evidence of heat in ultra-violet 

 waves whose length is little more than O^'S, the length of the 

 solar spectrum as now observable by the bolometer is between 

 b and 4 octaves.) 



The distant slit is separately exposed at each observation, 

 and the extremity of the full swing of the galvanometer-needle 

 is read. In all these measures the galvanometer is used in the 

 same condition of sensitiveness. The slit is opened to a con- 

 stant width of 2 millim. *"(except in measuring the very feeble 

 energy at the most-refrangible end of the spectrum, where the 

 width has been increased without prejudice to accuracy, owing 

 to the corresponding prismatic expansion of the spectrum 

 itself). The same bolometer is used, as a rule having for this 

 purpose 1 millim. effective aperture (except in measurements 

 at the most-refrangible end of the spectrum, where the full 

 aperture of the bolometer is used). 



These observations on the absorption of different air-masses 

 for each spectral ray evidently furnish means for determining 

 the curve outside the atmosphere, by the method already indi- 

 cated. They also, of course, give us the means of making a 

 map of the whole spectrum ; but their use for this latter pur- 

 pose is incidental. 



2nd. The other class of observations is for the special 

 purpose of making a spectral map, extending from the line C 

 to the lower limit of the infra-red. 



This is carried on by means of the linear bolometer, con- 

 sisting of a single strip ^ millim. wide. In this second class 

 of observations, a rough map of the whole infra-red spectrum 

 having been prepared, a very limited part of the spectrum 



* It will be remembered that, the actual distance of the slit being 

 5 metres, this aperture subtends an angle little greater than one minute 

 of arc. 



N2 



