NO. 3 RADIATION OF THE ATMOSPHERE — ANGSTROM 73 



the intensity of the scattered skylight as a function of the attenuation 

 of the solar radiation and of the zenith distance of the sun. The 

 theoretical result is not in exact agreement with the few observations 

 that have been made, for instance, by Abbot and Fowle, which may 

 be partly due to the difficulties in this kind of observation; but the 

 theoretical consideration proves that the intensity of the skylight 

 must be a decreasing function of the sun's zenith distance. For the 

 same transmission coefficient of the atmosphere, the skylight must 

 therefore be stronger, on an average, in low latitudes than in high 

 ones. 



Systematic observations on the intensity of skylight in its de- 

 pendence on other conditions are almost entirely lacking. This is one 

 of the most important problems in atmospheric optics, whose conse- 

 quences deeply affect the questions of climate and of the effects of 

 dust and haze and volcanic eruptions upon the temperature condi- 

 tions of the earth. The publications of Nichols, Dorno, and especially 

 those of Abbot and Fowle contain important contributions to the 

 problem. The outlines for further investigations of the subject seem 

 to me to be given by the theoretical considerations of King. 



A question of special interest for the problem I have dealt with in 

 my investigation is this : Is the temperature radiation of the atmos- 

 phere during the day the same as during the night, when temperature 

 and humidity conditions are assumed to be the same, or will the at- 

 mosphere" under the direct influence of the solar radiation assume 

 properties which will result in a deviation from the conditions pre- 

 vailing in the night-time as far as the radiation is concerned ? This 

 question ought to be treated in a general way by methods allowing 

 us to eliminate the short wave radiation and to observe the tempera- 

 ture radiation during different times of the day. Here I will only 

 give a brief account of some observations made during the total 

 eclipse of the sun in 1914 and of conclusions to be drawn from them 

 in regard to the last named question. The observations were carried 

 out at Aviken, a place situated on the Swedish coast, on the central 

 line of the total eclipse, during the two nights preceding and one 

 night following the total eclipse and also during the eclipse itself. 

 As I myself was engaged in other observations I had availed myself 

 of the able assistance of Dr. G. Witt and of Mr. E. Welander of the 

 Institute of Engineering, Stockholm, for carrying out these observa- 

 tions. 



In order to protect the instrument from the direct sunlight, a 

 screen arrangement was used, where the screen, through a simple 



