ON DEW, &c. 219 



Should what has been said be thought suf- 

 ficient to establish, that the air arrests part of 

 the sun's heat, which is radiated into it bound 

 up with light, two consequences must also be 

 allowed. The first is, that air will exert a 

 greater power of the same kind upon heat ra- 

 diated into it without light, since the sun's heat 

 passes instantaneously through many bodies, 

 which refuse a similar way to heat radiated by 

 terrestrial substances ; the other, that air must 

 be as capable of becoming cold by radiating its 

 own heat*, as of becoming warm from heat 

 radiated into it, as these two properties are uni- 

 formly observed to exist together, and to be 

 proportional to each other. The truth of the 

 latter conclusion may also be inferred from this 

 fact, that in still and calm weather the heat of 

 the air, a few feet above the earth, will some- 

 times decrease, even in this country, 18 or 20 

 degrees between sunset and sunrise, though no 

 change of wind has in the meantime occurred > 

 for the inconsiderable conducting power, which 

 air is now known to possess, will permit only 

 a small part of this diminution to arise from 

 heat passing, by means of that power, from the 



* Mr. Prevost says : " On peut supposer queles molecules 

 3e 1'air rayonnent." Du Calorique Rayonnant, p. 24. 



