ON DE\V, &c. 209 



air, after having been extremely foggy for 

 nearly a week, had become clear enough to 

 allow the sun's being distinctly seen during the 

 whole of the afternoon, though there was still 

 a sufficient obscurity in the lowermost parts of 

 the atmosphere, to obstruct considerably the 

 view of objects on the ground and very near to 

 it. 3. On the day following the observation, 

 the fog was again much less ; on the next it 

 disappeared, and was succeeded by snow. It 

 is to be mentioned likewise, that on the even- 

 ing in question the state of the grass, which 

 was the subject of experiment, was unusually 

 favourable to the production of cold ; since, 

 contrary to general experience, it was as cold 

 as swandown. If, then, the latter substance, 

 from the much greater regularity of the ap- 

 pearances exhibited by it, be taken as the 

 standard, by which the occurrences of different 

 nights are to be compared together, it will 

 follow, that the fog of which I am speaking, 

 though it did not prevent, must have lessened, 

 the production of cold from radiation. For, 

 on the preceding evening, when there was little 

 fog, the atmosphere being equally still on both, 

 the difference between swandown and the air 

 was 12; and on another, a fortnight after, the 

 difference at the same place of observation, be- 

 tween thermometers in the same situations, was 



