236 



ABBOT— SOLAR CONSTANT Ol' RADIATION. [Ap'i' ->. 



tainty as to the value of the sohir constant of radiation between the 

 limits of Pouillet's value, 1.76 calories, and Anj^strom's value, 4.0 

 calorics per s(|uarc centimeter per minute, i)revailed at the beginning 

 of the twenlieth century. 



Professor Pringsheim has collected the following table- of solar- 

 constant values, as dcterniined by differer.t observers: 



Tie omits Angstrom's 4.0. published in 1890 and withdrawn 

 in lyoo, but which is even yet sometimes ((uoted. lie omits also 

 Very's 3.1, pu])lishe(] in i<)Oi and independently obtained in 1910. 

 Kecentlv pu])lished values oi Kini])all, (Imwynski and others, ap- 

 proximately 2.0, are based in part on work of .\bbot and Fowle. 



The determination of the solar constant involves: (i) correct 

 measurements of the heat e(|ui\alent of the solar radiati<^n at the 

 earth's surface; (J) a correct estimate of the losses which the rays 

 have suffered in the atmosphere before they reached the measuring 

 apparatus. We shall now discuss these two branches of the work. 



Pouillel invented, about 1S35, his well-known instrument, the 

 pyrheliometer, for measuring the solar rays at the earth's surface. 

 Many criticisms have been justly made in regard to the accuracy 

 of this pioneer instrument, and atleiupts have l)een made by many to 

 improve on it, or to substitute a better. In our practice at the 

 Smithsonian Astrophysical Observatory, we have substituted a silver 

 disk for Pouillet's water chamber ; inserted a cylindrical bulb thcr- 

 mometer, radially instead of axiallw in the disk; provided a metal- 

 lined wooden chamber to screen the instrument from the wind; and 

 added convenient adjuncts for shading and exptx^ing the instru- 



■ " Phvsik dor Sonne," p. 417. 



" Til is value was expressed in terms oi a provisional scale of pyrlicli- 

 ometry which h;is since been proved too hii^h. 



