NO. l8 SMITHSONIAN PYRHELIOMETRY- — ABBOT AND ALDRICH 3 



Intercomparisons between the secondary pyrheliometers named in 

 the preceding table are as follows : 



Dates. 



1910 — June 21 



1911— Apr. 25 



1912 — Dec. 20 



1909 — Oct. 23, 31 



1911 — June 22, 28; Sept. 14. 



1910 — July 28 



191 1 — June 28 



1911 — June 28 



1912 — Feb. 6, 10 



1912 — Nov. 8 



1912 — Nov. 11 



Ratio 



i . 0242 

 [.0389 

 [.0281 

 [.3611 

 I-35I8 

 [.3820 

 [.4102 

 [.0469 

 [.0328 

 [ .0470 

 [-0445 



Probable 

 error. 



0.0019 

 0.0019 

 0.0013 

 0.0021 

 0.0017 

 o . 0008 

 0.0046 



O.OOIO 

 O.OOII 



0.0013 

 0.0014 



Combining the results of the two preceding tables we obtain the 

 following constants which we now adopt for the secondary pyrhelio- 

 meters named above. These constants and others which are derived 

 from them in the remainder of this paper we designate as Smithsonian 

 Revised Pyrheliometry of 1913 ' 



From 



S.I. 5 



A.P.O. 8 



S.I. 5 



A.P.O. 8bis 



we find 



A.P.O. 8 _ i.Q3ii _ 

 A.P.O. 8bis 1.0242 



1 . 0068 



„ A.P.O. 8 , A.P.O. 



prom — and - 



A.P.O.. 8 _i.36n_ 



A.P.O. IV 



we find 

 A.P.O. IV A.P.O. 8bis 1. 3518 



= 1 . 0068 



From A - R0 - 9 (49 values) we find A - P -°- 9_ — l 042(5 



A.P.O. 8bis A.P.O. 8bis 



u A.P.O.9 , A.P.O. 8bis r , A.P.O. 9 _i.398l_ T M „ 



Prom — and - we find ■— — = — — — =1.0343 



A.P.O. IV A.P.O. IV A.P.O. 8bis 1. 35i8 



From A - R0 - ** 

 A.P.O. IV 



(19 values) we find 



A.P.CUV . 

 A.P.O. 8bis 



0.7398 



These results leave no choice as to the values to be adopted for 



A.P.O. 8 . A.P.O. IV , . , .. ( , A 



— — — and — — ' but are less satisfactory as regards 



A.P.O. 8 bis A.P.O. 8 bis 



A P O o 



- ' ' ' ^ . However it will be seen that the discordance in this 

 A.P.O. 8 bis 



ratio almost wholly depends on the six comparisons ' ' ' -^— of 



