62 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 65 



top of Mount Whitney, 4,420 m. above sea level. In every instance 

 the sky was perfectly clear and appeared perfectly uniform. It will 

 be shown later on, that there is also strong experimental evidence 

 for the perfect uniformity of the sky. 



In order to obtain from the observations a more detailed idea 

 of the effective radiation to different parts of the sky, I proceeded 

 in the following way: In a system of coordinates, where the 

 zenith angle is plotted along the x-axis, the magnitude of the 

 radiation along the y-axis, every measurement with the instrument 

 corresponds to an integral extending over 32 ° and limited by the 

 .r-axis and a certain curve — the distribution curve of radiation. If 

 the measurements are plotted as rectangular surfaces, whose widths 

 are 32 ° and whose heights are proportional to the magnitude of the 

 radiation, we obtain from the observations a system of rectangles like 

 those in figure 10. A curve drawn so that the integrals between the 

 limits corresponding to the sides of the rectangles are equal to the 

 areas of these rectangles will evidently be a curve representing the 

 radiation as a function of the zenith angle. 



(Note. — Against this procedure it can be objected that the observations do 

 not really correspond to rectangular surfaces, the opening being circular and 

 not square. The consequence will be that the real distribution curve will cut 

 the rectangles in points lying nearer their central line than the section points 

 defined by the procedure described. In fact this will alter the form of the 

 curves very slightly; in drawing them the conditions just mentioned have 

 been taken into consideration.) 



In figures iia and iib the curves are shown. They indicate the 

 fact — which has already been pointed out by Homen — that the effec- 

 tive radiation to a constant area of the sky decreases with an increase 

 in the zenith distance. My observations indicate very strongly that 

 the radiation approaches the zero value, when the zenith angle ap- 

 proaches 90 , which shows that the lower atmosphere, taken in very 

 thick layers, radiates like a black body. If there were no radiating 

 atmosphere at all, the distribution curve would be a straight line 

 parallel to the .ar-axis. 



A comparison between the different curves shows, further, that 

 they differ in a very marked way from one another in regard to their 

 form. It is also evident that trus difference in form is very closely 

 connected with the density conditions of the atmosphere and espe- 

 cially with its content of water vapor. 



