CHAPTER VI 



RADIATION TO DIFFERENT PARTS OF THE SKY l 



In the foregoing- chapters an account has been given of observa- 

 tions showing the influence of humidity and temperature conditions 

 upon the effective radiation to the sky. There the total radiation to 

 the sky was considered, independent of the fact that this radiation 

 takes place in different directions. The thing measured represented 

 an integral over the whole hemispherical space. About the different 

 terms constituting the sum this integral gives us no idea. 



In the. historical survey I have referred to the interesting investi- 

 gations of Homen, and mentioned his observations of the nocturnal 

 radiation to different parts of the sky. Homen observed, with a 

 somewhat modified Angstrom pyrheliometer, of type 1905, where 

 two metal disks were exposed to the sky alternately and their tem- 

 perature difference at certain moments read off. In order to measure 

 the radiation in various directions Homen used a screen arrangement, 

 which screened off certain concentric zones of the sky. The chief 

 objection to this method seems to me to be that the radiating power 

 of the soot will be introduced as a variable with the direction, and as 

 this quantity is not very well defined an error will probably be intro- 

 duced, which, however, can scarcely amount to more than about 

 2 per cent. Homen found that the distribution of the radiation upon 

 the different zones of the sky was almost constant for different values 

 of the total radiation. As H omen's measurements have since been 

 employed in extending, to represent the whole sky, 2 observations of 

 the radiation toward a limited part of the sky, and as the question 

 itself seems to be of interest for the knowledge of atmospheric radia- 

 tion in its dependence upon other conditions, I have thought it valu- 

 able to investigate in what degree this distribution of radiation over 

 the sky is subject to variations. For this purpose the arrangement 

 shown schematically in figure 8 was found to be a satisfactory one. 



To the electrical compensation instrument, which has been de- 

 scribed, can be attached a hemispherical screen, abcdef, whose radius 

 is 7.1 cm. From this screen can be removed a spherical cap cd, which 



1 Large parts of this chapter were published in the Astrophysical Journal, 

 Vol. 39, No. 1, January, 1914. 



2 Exner (1903), loc. cit. 



57 



