NO. IO REFLECTING POWER OF CLOUDS ALDRICH 5 



table) three current calibrations 1 were made — just before the first 

 ascension, between the first and second, and after the second ascen- 

 sion. The galvanometer circuit was unchanged throughout the 

 observations, so that the calibrations were made under the same con- 

 ditions as the cloud observations, save that the balloon was near the 

 ground for the former and above the fog for the latter. 



Column 5 in the table (total solar radiation per sq. cm. normal to 

 the beam) was obtained as follows : On the morning of September 

 1 6, the usual solar constant observations, which include pyrheliometer 

 measurements of the total solar radiation on normal surface, were 

 made on Mt. Wilson. Then on September 17, simultaneously with 

 the cloud reflection observations, Mr. H. Benioff of the Mt. Wilson 

 Solar Observatory staff very kindly made pyrheliometer readings on 

 Mt. Wilson with Pyrheliometers IV and' VII. He made eight 

 determinations, the mean of which gave for an air mass 1.5 the value 

 1.46 calories, total solar radiation received per square centimeter 

 of normal surface. The plotted values of September 16 give for the 

 same air mass the practically identical value, 1.452 calories. Further- 

 more the solar constants determined at the recently established Smith- 

 sonian station in Chile are : 



September 16, 1918 1.960 



September 17, 1918 I-9SI 



As far as visual observations of the sky could indicate the two days 

 were identical. Thus, since the two days show nearly identical solar 

 constant values and nearly identical pyrheliometer values at a given 

 air mass, it is to be assumed that the pyrheliometer values for the 

 whole range of air masses would have been nearly identical. Values 

 of column 5 are therefore taken from the pyrheliometer curve of 

 September 16. 



Column 4, the sky brightness, was not so easily obtained. Un- 

 fortunately, owing both to delay in the return of instruments and to 

 an unprecedented amount of cloudy weather, sky brightness values 

 on a day with sky conditions similar to September 17 were not avail- 

 able. 2 The pyranometer data of previous years was examined and 



1 The first-swing method was used. See Smith. Misc. Coll., Vol. 66, No. it, 

 p. 8. 



* It will be possible to obtain such values at some future time, however. 



