342 PBOCEEDINGS OP THE AMERICAN ACADEMY 



Intestjgatioxs on Light and Heat, made and published wholly or in part with 

 appropriation from the Rumfokd Fuud. 



XVIII. 



THE BOLOMETER AND RADIANT ENERGY. 



By Prof. S. P. Langley. 



rresented Jan. 12, 1881. 



Our knowledge of the distribution of heat in the solar spectrum 

 really begins with this century and the elder Herschel, and, since 

 his time, great numbers of determinations have been made, all with 

 scarcely au exception, by means of the prism, the early ones through 

 the thermometer, the later ones by the thermopile and galvanometer. 

 It was very soon seen that the prism exercised a selective absorption, 

 and that the form of the heat-curve varied with the material of the 

 refracting substance, but a far more important and more subtle error 

 was left almost unnoticed. The elder Draper, I believe, long since 

 pointed out that the prism, contracting as it does the red end, and still 

 more the ultra-red, gives false values for the heat, from this latter 

 cause alone, and displaces the maximum ordinate of the heat-curve 

 toward the lower or ultra-red end. Dr. Miiller (Poggendorff's Ann. 

 CV.), indeed gives a construction showing how we may, from the in- 

 correct curve of the prism-spectrum, obtain such as a grating would 

 give could we use one ; but he despairs of being able to get measurable 

 heat from the grating itself, whose spectra are so much weaker than 

 that from the prism, while even the latter are very hard to measure 

 with any exactness by the pile. 



No one, so far as I know, has hitherto succeeded in measuring the 

 heat from a diffraction grating except in the gross, or by concentrating, 

 for instance, like Draper, the whole upper half and the whole lower 

 half of its spectrum upon the pile, and thus reaching some results, not 

 without value, even as thus obtained, but of quite other value than 

 those which may be expected when we become able to measure with 

 close approximation the separate energy of each wave-length. 



I have tried at intervals for the past four years to do this, and hav- 

 ing long familiarity with the many precautions to be used in delicate 

 measures witli the thermopile, and a variety of specially sensitive piles, 



