NO. 15 IMPROVED PYRHELIOMETER ABBOT AND ALDRICH / 



In the third place, we pointed out, as Shulgin has done also, that 

 irregularities of flow of the water current in the use of standard 

 pyrheliometer No. 3 caused galvanometer drift and wiggle, and 

 difficulty in deciding as to the true deflections due to solar heating. 

 Personal equation might well enter in such a case, and there may have 

 been a tendency to read the deflections too large. 



As compared with all these complex and numerous sources of error 

 in the older standard pyrheliometer, the extreme simplicity and perfect 

 operation of standard pyrheliometer No. 5 must strongly incline us 

 to attribute very much greater weight to its results. 



In the fourth place, we have altered the silver-disk pyrheliometer 

 vestibules in recent years to expose only to an angular diameter of 

 5° 48'. In making this change of vestibules of A.P.O.Sbis on Decem- 

 ber I, 1927, it IS possible that some slight structural modification of 

 the pyrheliometer occurred. We did, indeed, make careful compari- 

 sons of A.P.O.Sbis with silver-disk instrument S.I. 5 before and after 

 the change. These comparisons resulted as follows : 



Date 



Number 



of 



comparisons 



Ratio 



S.I.5 



A.P.O.8. 



Ratio 



S.I.s 



A.P.O.8. 



Apr. 15, 191 1 

 Dec. 20, 1912 . 

 Oct. 14, 1921 . 

 Mar. 6, 1922. 

 Oct. 10, 1927. 

 Oct. 22, 1927. 

 Nov. I, 1927. 

 Dec. I, 1927 . 

 Dec. 27, 1927 . 

 Feb. 27, 1928 . 



18 

 12 

 20 

 8 

 20 

 20 



0389 

 0281 



0339 

 0291 



0457 

 0305 

 0395 



1 .0328 

 1.0311 

 I .0422 



Allowing for the almost unavoidable irregularities of such compari- 

 sons over so long a time, due to sky conditions, personal equation, 

 and watch eccentricity, there is nothing here to indicate any change 

 of constant of A.P.O.Sbis with respect to S.I.5. The latter instrument 

 had not been changed at all during all this time. The weighted mean 

 values are as follows : 



