NO. 4 SOLAR RADIATION ABBOT, FOWLE, AND ALDRICH 47 



Accordingly we give merely the readings made upon the original, 

 and their reduction. Figure 9 is from a tracing made to represent 

 the march of the record. 



The pyrheliometer record consists of a series of zigzag reaches of 

 shading corresponding to the up and down marches of the mercury 

 column. We shall principally confine attention to those marked 

 A, B, C, D, E, F , G, which represent the solar radiation measured 

 just before the instrument reached maximum elevation. We do this 

 because: (1) As stated above, the earlier part of the records are of 

 little value owing to the bore of the thermometer being foul for 

 temperatures above +io°. (2). A defect in the record occurs just 

 after the balloons began to descend, first owing to a jerkiness, and 

 then owing to crossing the seam in the paper, which renders the next 

 two following readings doubtful. (3) There is doubt as to the 

 elevation at the time of the last descending records, because the 

 barometer arm did not work quite free. (4) The record is finally 

 lost in clouds. All readable records are, however, given for what 

 they may be worth. 



CORRECTION TO REDUCE TO VERTICAL SUN 

 The extreme width of the degree marks on the record during 

 heating B, D, F, was measured and found 1 .40 millimeters. Inclining 

 the pyrheliometer, first 15. 5 N., then 15.5° S., when exposed to the 

 sun, was found to shift the degree marks through a total range of 

 0.89 mm. Subtracting width of trace, 0.31 mm., and dividing by 2, 

 we find the record sheet is within the pyrheliometer at a distance X, 

 such that X tangent 15. 5 =0.29 mm. Hence Z=i.04 mm. From 

 this it follows that the tangent of the half angle of the cone swept 



through by the sun rays was — — . Hence the half angle of 



the cone is 27" 40'. At Omaha, on July 11, at noon the sun's zenith 

 distance was 19 5'. Hence the pyrheliometer was swinging in a 

 cone whose half angle was 2J° 40' — 19 5' = 8° 35'. 



From these data it follows that the mean value of the cosine of the 

 inclination of the sun's rays upon the pyrheliometer disk at noon 

 was 0.934. But if the instrument had been stationary this value 

 would have been 0.945. Hence the conical rotation produced a 

 change of 0.011. This value has been applied as a correction to the 

 values of cosine Z, corresponding to the several sun exposures. It 

 is probable that the correction is a little too small, because the record 



