NO. 4 SOLAR RADIATION ABBOT, FOWLE, AND ALDRICH 



23 



observations, using for this purpose a reduced scale of abscissae, in 

 which 1 cm. = 0.5 air-mass. 



We have read off from the plots so obtained the inclination of the 

 best straight lines, giving logarithms of transmission coefficients ; 

 and also the intercepts on the axis of ordinates, giving logarithms of 

 intensities outside the atmosphere. The plots were read up inde- 

 pendently for three different ranges of air-masses. The first range 

 is that which we customarily employ, from about 1.3 to about 4.5 air- 

 masses. The second reading includes all points from 1.3 to 20 

 air-masses or thereabouts. The third reading was made with the 

 portion of the curve which Mr. Very states to be the best, namely, 

 from air-mass 4 to air-mass 10 or thereabouts. The results of all 

 three readings are given in table 11. For September 20 this table 

 gives also the percentage deviations, in ordinates, of the observed 

 points from the natural numbers corresponding to the straight lines 

 of the logarithmic plots which were chosen in the second reading to 

 represent them. In order to show that the somewhat Jarge percent- 

 age errors at some places are not inconsistent with experimental 

 error of very moderate amount, we give for two holographs the 

 deviations expressed in millimeters on the original holographs. The 

 reader should bear in mind that the holographic trace itself is nearly 

 1 millimeter wide, and subject to tremor. Also the line of zero 

 radiation is interpolated between zero marks 1 minute of time, or 

 8 centimeters of plate, apart. 



We then determined the area which the holographic curve would 

 include if it were taken outside the atmosphere, and we multiplied 

 this area by the appropriate constant (see table 9) to give the result 

 in calories per sq. cm. per minute. To this we added the small cor- 

 rections to reduce the result to mean solar distance, and to zero 

 atmospheric humidity, as explained in Annals, Vol. Ill, p. 43. All the 

 details of the foregoing processes have been described and investi- 

 gated in Vols. II and III of the Annals of the Astrophysical 

 Observatory, and to these the reader is referred. 



The following are the solar constant values obtained : 



Table 10 — Solar Constant Values 

 In standard calories (15°) per sq. cm. per minute at mean solar distance 



Air-masses 

 Sept. 20. . . 

 Sept. 21 . . . 



4 to 12 



1.909 

 1.929 



