F. W. Very — Transmission of Terrestrial Radiation. 51! 



Art. XXX1Y. — Fundamental Distinctions Special to the 

 Process of Transmission of Terrestrial Radiation by the 

 Atmosphere, and the Value which is Obtained for the 

 Coefficient of Transmission when these are Considered / by 

 Frank W. Very. 



In finding- the transmission of terrestrial radiation by the 

 atmosphere, I have hitherto assumed that it is unnecessary to 

 measure more than the zenithal radiation, for the following 

 reason : If we consider the radiation to a limited area of the 

 sky from an instrument having a very small surface, normal to 

 the direction of the rays, the radiation falls off as the pointing 

 departs from the zenith and is practically zero close to the 

 horizon. Hence some have supposed that a complex summa- 

 tion of the radiation by zones is needed in order to get the 

 actual " sky radiation." But in nature, the absorption is 

 exerted in a comparatively shallow layer of air close to the 

 surface and the radiator is not a minute area, or almost a 

 point, but is an indefinitely extended surface faced by a 

 parallel and closely juxtaposed absorbent layer. It is, there- 

 fore, only necessary to consider the transference between two 

 layers in a normal direction, that is, vertically. 



In proof that this is so, we have only to try a still more 

 direct experiment by making the moon our distant radiant 

 source, after eliminating that part of its rays (both visible and 

 invisible) which is simply reflected solar radiation, and getting 

 the transmission of the true lunar emission through an entire 

 atmosphere. On trial, the result demonstrates that the zenithal 

 observation with a radiator of limited surface, and radiating to 

 a limited angular area, is the one to be used ; and that the 

 zonal integration which gives a smaller apparent transmission 

 and one which does not represent a natural fact, is at fault, 

 because the details with which the summation is concerned are 

 purely those of a laboratory experiment, while the conditions 

 in nature are entirely different. 



The fallacy just noted may have arisen from a common prac- 

 tice of measuring what is sometimes called "nocturnal radia- 

 tion," or the radiation of a thermometer-bulb to either a part 

 or the whole of the night sky. As an instrumental method, 

 this is perfectly legitimate, but the result should not be con- 

 founded with the radiation of a widely extended surface of the 

 earth to space, which is the true terrestrial noctural radiation. 



In this Journal for April, 1913, I gave a curve of apparent 

 sky temperature obtained by a thermopile of limited aperture 

 (fig. 2, p. 383). Here the readings were entirely similar to 

 those which might be obtained by a nocturnal radiation ther- 



