F. W. Very— Sky Radiation and Isothermal Layer. 377 



greater than that of the entire air column. Moreover, the 

 temperature of the air, with its clouds and aqueous vapor from 

 which its own radiation partly proceeds, is not purely obtained 

 by radiation exchanges, but is due to a complex of radiation, 

 convection, absorption, cloud precipitation, etc. 



The radiation absorbed by the aqueous vapor of the lower 

 air is not re-radiated directly into space, for it can not pass the 

 barrier of aqueous vapor at higher levels, but this thermal 

 energy is passed on to other atmospheric constituents, to the 

 dust, or to the major part of the air, its oxygen and nitrogen, 

 which, though feeble radiators, must be the final members of 

 the series by which the heat of the atmosphere is dispersed 

 into space. 



The day temperature of sunlit rock, gravel, dry sand, or 

 moderately dry soil is much higher than the air temperature 

 at a height of one meter, or thereabouts, which is set down in 

 the usual meteorological data as the observed terrestrial tem- 

 perature. Thus a surface temperature of 120° F. for the first 

 half centimeter of dry soil is common in summer when the air 

 temperature is nearer 75° or 80° F., and the mean temperature 

 of land surfaces for day and night has been quite generally 

 assumed too low from neglecting this sunshine effect. With 

 the low sun of winter in the following measures, the sunshine 

 effect is of little consequence, and the ground measures were 

 always of a shaded surface. The radiator in these observations 

 has been the measuring instrument itself, kept at a nearly con- 

 stant temperature of about + 60° F. throughout the winter 

 series. In order that there may be the means of computing 

 the radiation of the soil to the atmosphere, measures of the 

 radiation of the instrument towards the ground have also been 

 included. 



The temperature of the ground when frozen and partly cov- 

 ered with ice was assumed to be the same as that of the air one 

 meter above the surface ; but when the ground was covered 

 with snow, the surface temperature was taken by a thermome- 

 ter with its bulb barely covered by the snow. Usually there is 

 scarcely any difference between the temperature of the shaded 

 surface and that of the air in the daytime, nor yet at night if 

 the sky is veiled by even the thinnest cloud. But on the 

 morning of February 4th, 1909, an exception was noted which 

 is of sufficient interest to be recorded. 



At 5 h 15 m a. m., the sky being covered by cirro-stratus cloud 

 dense enough to give a strong lunar halo and corona, the snow 

 had a temperature of — 6'0° F., but the air, which was calm, 

 was at 0° F. I was surprised at the magnitude of the temper- 

 ature depression, which seemed too great to be produced by 

 nocturnal radiation to such a cloudy sky. The explanation 



