2 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I23 



Thus two questions arise: (i) Is the Smithsonian water-flow 

 pyrheliometer a standard pyrheliometer ? (2) Do silver-disk pyrheh- 

 ometers, used with Smithsonian directions, give the same results as 

 the water-flow pyrheliometer, under usual ranges of conditions? 



If the answers to these questions are yes, then critical discussions 

 of errors of the silver-disk pyrheliometer are merely academic. 



The water-flow pyrheliometer ceased to be a primary standard when 

 in 1932 it was changed to be an electrical compensation instrument, 

 following the suggestion of the Russian meteorologist W. M. Shulgin.* 

 It can now be a standard pyrheliometer only if electrical heating is 



(a) completely measurable and (b) not more favorably applied for 

 measurement than solar heating. 



These requirements are discussed in Smithsonian Misc. Coll., vol. 

 no, No. II, 1948, and in volume 3 of Annals of the Smithsonian 

 Astrophysical Observatory, pages 60 to 67. As the mean of nearly 

 50 determinations, electrical heating was loo.o percent recovered. 

 Electrical heating was appHed in some cases to coils wound on the rear 

 surface of the receiving cone for solar rays, and perhaps more favor- 

 able to measurement than sun heat. But it was also applied in some 

 cases to a coil in air, in front of the cone, where electrical heating was 

 at a disadvantage. No difference between the results of the two wind- 

 ings was found. We therefore hold that the two conditions (a) and 



(b) are satisfactorily met. 



In an absolute standard pyrheliometer, solar heat must be sensibly 

 completely absorbed. This, we think, cannot be seriously questioned 

 regarding our water-flow instrument. For considering the construc- 

 tion of the water-flow chamber, and that about 95 percent of the 

 solar radiation is absorbed at first incidence on the blackened cone, 

 repeated reflections within the blackened chamber must reduce the 

 remaining 5 percent almost to zero. 



The aperture for solar rays, used in 1932 and all subsequent ob- 

 servations with the water-flow pyrheliometer, was measured by means 

 of a fitted plug with more than sufficient accuracy. The aperture 

 was of such a diameter and distance from the receiving cone as to 

 have the same solid angle as the aperture of the long-barreled silver- 

 disk pyrheliometer. Its solid angle is 0.0013 hemisphere. Our pyr- 

 heliometers therefore measure a little sky light besides direct sunlight. 

 What percentage this is must vary with the clearness of the sky. 

 This consideration introduces a small uncertainty as to the absolute 

 value of the solar constant of radiation. But when solar-constant 



* Shulgin, W. M. Monthly Weather Rev., vol. 55, No. 8, p. 361, 1927. 



