96 abbot: solar constant of radiation 



He made some attempts to extrapolate his observations to the 

 limit of the atmosphere, but these, like other solar constant val- 

 ues obtained by pyrheliometry alone, are not definitive. Great 

 value, however, attaches to the long series of direct observations 

 continued from 1883 to 1900 at Mont-Pellier. These show plainly 

 the influence of the volcano Krakatau and others. 



K. Angstrom observed with the electrical compensation pyr- 

 heliometer at several stations at different altitudes on the island 

 of Teneriffe in the years 1895 and 1896. Some of his measure- 

 ments were made at the altitude of 3700 meters, and give direct 

 readings' of solar radiation as high as 1.63 calories per square centi- 

 meter per minute. Angstrom declined to determine from these 

 a value of the solar constant of radiation, recognizing that this 

 demanded observations of the solar spectrum as well as pyrhe- 

 liometric work. In later years he prepared spectro-bolometric 

 apparatus for this purpose, and made many solar constant meas- 

 urements therewith at Upsala. These measurements are still 

 being continued there by his successors. It is hoped that this 

 long and interesting series will soon be published. 



Passing from this work of Angstrom, which belongs in a later 

 period, and omitting mention of valuable pyrheliometric obser- 

 vations by numerous observers in Italy, Switzerland, and Russia, 

 which I regret that space forbids me here to discuss, attention 

 must now be directed to the work of Langley, which marked an 

 epoch in this kind of investigation. 



Langley 's observations. Prior to Langley's observations, there 

 had been numerous attempts to determine the solar constant, 

 which are well summed up in the excellent little book of Radau, 

 entitled Actinometrie. It is shown that nearly all observers were 

 in comparative agreement, so far as their actual observations go, 

 and if the transmission of radiation by the atmosphere be esti- 



_B 



mated by the simple formula A = Aop ''''^ which was em- 

 ployed by Pouillet and many others, the value of the solar con- 

 stant would be found in the neighborhood of 1.75 calories. 



But Forbes, Desains, Violle, Crova and others showed con- 

 vincingly that this equation does not accurately express the 



