F. W. Very— Sky Badiation and Isothermal Layer. 387 



Altitude (km.) 



45 



50 55 



60 



65 



70 



75 



80 





T (isothermal) 



T(by regular fall). 



186 

 40 



143 120 

 30 22 



92 

 16 



52 



12 



35 



8 



20 

 5 



10 

 3 





Difference 



146 113 98 



76 



40 



27 



15 



7 





Comparing the two curves, we see that at an altitude of 40 

 km. the temperature is probably about 150° C. higher than it 

 would be were it not for the heat derived from the absorption 

 of solar radiation. 



Although the air transmits a variable but considerable frac- 

 tion of telluric radiation to outer space and thus possesses 

 selective absorption, it is a curious fact that for an ultimate 

 remnant of this radiation, the atmosphere behaves as if it were 

 an impervious or perfectly reflecting solid for all wave-lengths 

 emitted by a body at about —60° or —70° C. The greatest 

 cold observed in the Arctic regions in winter seldom falls 

 below the latter temperature. This, therefore, is the effective 

 temperature of that atmospheric layer which limits the out- 

 put of terrestrial radiation. The action would be the same if 

 the air layer were a perfectly absorbent one and at the given 

 temperature, or if the phenomenon were one of reflection and 

 the reflecting layer at any temperature whatever within a wide 

 range. For my present purpose it is not necessary to discrim- 

 inate between these two cases. 



The outcome of the process is that in temperate latitudes 

 with the clearest skies there is a seasonal change in sky radi- 

 ation determined by the presence or absence of aqueous vapor. 

 Just as the abundance of aqueous vapor in the lower air in 

 summer enables this substance to play the part of a conserva- 

 tor of surface heat, so there must be a similar conservator for 

 the heat of the upper air, the heat in either case being derived 

 from the sun's rays. The water vapor limits surface radiation 

 mainly by the absorption in the great H band situated on the 

 short-wave side of the terrestrial radiant maximum which in 

 sunshine lies between 9/ul and 10/*. The great A band of carbon 

 dioxide at 14*7/* occupies a somewhat similar position in respect 

 to the maximum radiation of the upper air, but since this 

 gas is too dense to reach the greatest heights, its action must 

 be confined to the lower part of the layer, and if the upper 

 air behaves like an opaque solid, its absorption spectrum 

 beyond A must be a continuous succession of broad and intense 

 bands. Water vapor has a number of bands in this part of 



