NO. 4 THE BRIGHTNESS OF THE SKY MOORE AND ABBOT 35 



normally. As stated above, a considerable portion of reflected ligbt 

 from the clouds observed was lost, but there was compensation from 

 other parts of the sky. On the following day, however, we used sun- 

 light and a cone of sky light 30° around it, so that compensation 

 could come only from the rest of the sky, this outside radiation 

 amounting perhaps to one-thirtieth, or possibly one-twentieth of 

 that admitted by the thimble. Sets b, c, d and e show that the radia- 

 tion from the snow was fairly uniform over a considerable area. 

 Indeed, the highest readings are found in c where the instrument 

 was pointed 30° east from the position of b in which the most direct 

 reflections from the sun were measured ; but this is probably because 

 the snow of c was slightly cleaner than that of b. We can obtain 

 an approximation, however, of the total radiation reflected to a 

 point, by considering the efl^ect if a whole hemisphere of snow were 

 observed, or in other words, if no thimble were used, it being assumed 

 that this were possible without the admission of the stray light. For 

 if a ray coming from the sun is deflected a certain number of degrees 

 west from the plane normal to the main snow surface and containing 

 the sun, then at some other spot to the west of the first another ray 

 is deflected by a snow crystal a corresponding number of degrees 

 to the east of a normal plane containing the sun — or if in such a 

 normal plane a certain ray makes a greater angle with the face of a 

 crystal than with the main surface, another ray makes a correspond- 

 ingly smaller angle with the face of another crystal ; so that there 

 is a general tendency toward compensation. The rays most directly 

 reflected will be found the most effective, as they approach nearest 

 to normal incidence on the pyranometer strips. We may now esti- 

 mate the entire reflection by multiplying the results obtained with 

 the thimble by 4 (the derivation of this ratio being given in the 

 section on thimbles). Proceeding in this manner with sets h, c, d, e, 

 and h, and adding 4 per cent to the sun and sky values, to allow for 

 the portions of the sky not observed, we conclude that on the day 

 in question about 50 per cent of the radiation coming from the sun 

 and sky was reflected diffusely by the snow. 



It seems very probable that considerably higher radiation would be 

 obtained with snow that was freshly fallen. Unfortunately there 

 was no appreciable fall of snow after this date, so that this opinion 

 could not be confirmed. 



SUMMARY 



Quite an extensive set of observations was made at Hump Moun- 

 tain, near Elk Park, North Carolina, from July, 1917 to April, 1918. 

 with the pyranometer, an instrument constructed by the Smithsonian 



