34 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 52 



In the year 1900 the first bolometric observations of the corona 

 were made by Smithsonian observers/ and from these observations 

 certain inferences were drawn by different authors as to the quality 

 of the radiation of the inner corona- 

 All bodies, by virtue of their temperatures, emit radiation ; but it 

 is only when the temperature is fairly high that any considerable 

 part of the radiation is visible. The higher the temperature the 

 larger becomes the proportion of the radiation caused thereby which 

 is visible. 



All bodies exposed to radiation reflect some fraction of it dif- 

 fusely, but thereby generally alter the quality of complex radiation. 

 When the reflecting bodies are particles whose diameters are small 

 compared with the wave-length of light, they reflect the shorter 

 wave-lengths better than the longer ones, and thus tend to render 

 a larger proportion of the radiation visible. Larger particles and 

 gross bodies, like the moon, by reflecting, generally alter the quality 

 of radiation in a way to diminish the proportion visible. 



Msible rays are sometimes emitted by bodies which are apparently 

 far below the temperature of incandescence, as in the cases of 

 electrical discharges and of luminous insects. Such radiation may 

 perhaps be almost wholly visible, without much intensity in the 

 infra-red spectrum. 



In view of these considerations and others, the inferences drawn 

 by the writer from the bolometric study of the corona made in 1900 

 were contrary to the view that the radiation of the inner corona is 

 produced mainly by the incandescence of matter heated to high tem- 

 peratures by reason of its proximity to the sun, and more favorable 

 to supposing the coronal radiation due largely to luminescence, or 

 perhaps to the reflection of solar radiation by small particles. 

 Arrhenius came to a different conclusion ; but, as pointed out in the 

 reference last cited, he misinterpreted the position of the bolometer 

 in the coronal image. 



The bolometric observations at Flint Island were designed to test 

 the inferences above referred to and to measure more definitely the 

 quantity and quality of the coronal radiation- 



^ See Astrophysical Journal, vol. 12, pp. 71-75; also pp. 366-375, 1900. '"The 

 1900 Solar Eclipse Expedition of the Astrophysical Observatory of the Smith- 

 .'ionian Institution," pp. 22-26. Washington, Government Printing Ofifice, 1904. 

 Lick Observatory Bulletin No. 58. Astrophysical Journal, vol. 20, pp. 224-231, 

 1904; Astropliysical Journal, vol. 21, pp. 194-195, 1905. 



