328 CARNEGIE INSTITUTION OF WASHINGTON. 



will have the best chance of escape into outer space without encountering 

 other molecules capable of absorbing it. Even if all the CO2 in the atmos- 

 phere were engaged in the first absorption, a part of this sixth of the radiation 

 would be absorbed, though the chances would be poor. But if the larger 

 portion of the CO2 were not involved in the first absorption but lay above, 

 even this outward-going sixth would have little chance to escape a second 

 absorption and the contingencies that lie beyond that. 



One of the other sixths of the molecules' re-radiated energy would be directed 

 earthward and would almost certainly either be absorbed by the molecules 

 of CO2 intervening between it and the earth — or by any other molecules 

 capable of such absorption — or would reach the earth, be absorbed again, and 

 start on a new career as new terrestrial radiation. The remaining four 

 sixths would be radiated in a more or less horizontal direction and would 

 penetrate much greater depths of air before reaching open outer space and 

 would encounter correspondingly large chances of absorption. If there is 

 more than enough CO2 in a vertical section of the atmosphere to absorb all 

 of this type of rays in the terrestrial radiation, there would certainly be much 

 more than enough CO2 in a tangential section. The total secondary ab- 

 sorption of this initial re-radiation would be relatively large. 



But, as already observed, the secondary series does not end with this re- 

 radiation. Every unit of energy absorbed in it starts a new radiation of 

 like nature, and so absorption and re-radiation continue until all the thermal 

 energy has escaped in some way. Perhaps the most piquant feature of the 

 case lies in the quasi-paradoxical fact that the smaller the proportion of the 

 CO2 molecules required to make complete first absorption of the CO2 class 

 of wave-lengths the higher is the percentage of contingencies of secondary 

 absorption and the greater the relative prolongation of the thermal activities 

 of this class. 



But two other factors are to be considered. The discussion has proceeded 

 thus far on the standard assumption that the molecules of each gas absorb 

 selectively only a certain set of wave-lengths, and that one gas does not do 

 the work of another, at least so far as their distinctive work — selective 

 absorption and emission — is concerned. It has been urged, however, that 

 the vapor of water can absorb the same portion of the terrestrial radiation 

 that CO2 is known to absorb selectively, and may thus replace its work and 

 destroy its importance as a climatic agency. Whatever may be true, or 

 appear to be true, in the case of warm, fully saturated air on the very borders 

 of precipitation or in its early stages, there is ground to question this claim 

 when the vapor-content of the air is low, as when the air is very dry or very 

 cold. But so far as this analysis is concerned, let this doubt be waived. 

 Let it be assumed that the water-vapor can absorb and radiate precisely 

 the same wave-lengths as carbon dioxide. This assumption implies reciproc- 

 ity of action and carries the corollary that CO2 can absorb and radiate the 

 corresponding wave-lengths radiated by the water-vapor, and that the vapor 

 radiates such wave-lengths because it absorbs them. This amounts to 

 increasing by so much the thermal function usually assigned to the CO2 

 alone. If there is only 14 per cent of the CO2 class of vibrations in the 

 terrestrial radiation, of course no more than 14 per cent will be absorbed by 

 any amount of CO2 or of H2O vapor, so far as the first action of absorption 

 and of bringing this energy into heating service in the air is concerned; but 



