146 CARNEGIE INSTITUTION 



layer exerting the same pressure would transmit the same percent- 

 age A of the beam, so that after passing the two layers the percent- 

 age A^ would remain. This argument, which was that of Pouillet, 

 leads to serious error, for it is now a well-determined fact that the 

 layers of the air do not possess uniform properties of absorption 

 and do not possess a uniform transmission coefficient for all w^ave- 

 lengths. 



It has thus far proved, and, so far as can be seen, always will 

 prove, impossible to determine from near sea-level with any precis- 

 ion by anj^ observations, however careful or long continued, the 

 " constant " of solar radiation. There is no good way to eliminate 

 the complex effect of atmospheric absorption except to observe at 

 the highest practicable altitude, preferably near the tropics, but 

 most certainly in a dry and clear atmosphere, and preferably where 

 there are two stations in view of each other, the first of which is at 

 a notably greater altitude than the second, though the latter is itself 

 at least some thousands of feet above sea-level. 



Temporar}' expeditions with meagre outfits have gone from time 

 to time to high mountain stations for solar observations, and small 

 meteorological stations have even been longer continued. What is 

 needed is rather a permanent astrophysical observatory equipped 

 with the most powerful and refined modern apparatus for solar re- 

 search and located at the highest and clearest station it is practicable 

 to occupy. 



The remarkable complexity of atmospheric absorption has been 

 hinted at in what has just been said, but much remains unsaid and 

 even unknown in this connection. Hardly less importance attaches 

 to the absorption of the air for the radiations emitted by the Earth 

 than to its absorption for the rays received from the Sun, and this 

 part of the subject has been relatively unknown and till recently 

 beyond the possibility of thorough study. It is apparent, however, 

 that the nice balance of the receipt and outgo of radiation by the 

 Karth is that which maintains the range of climate under which we 

 live, and that a thorough understanding of the matter requires the 

 study of the outgo as well as that of the incoming of radiation. 



It is for this purpose and for extending our knowledge of the 

 direct absorption of the solar rays in our atmosphere, and even 

 largely for the determination of the solar constant itself, that the 

 station at a relatively low altitude, but within sight of the elevated 

 station, and which should be provided with a duplicate set of ap- 

 paratus for energy of radiation work, is needed. 



