ON DEW, &c. 269 



temperature of 27, though the air in the im- 

 mediate neighbourhood was purposely much 

 agitated by him. 



The conclusions here given by me, respecting 

 the cold produced by the evaporation of water, 

 were drawn from experiments made in the day, 

 while the sky was clear, the air very calm, and 

 the temperature of the atmosphere stationary. 

 At night, and during a cloudy day, the differ- 

 ences were less. On the other hand, if there 

 was any perceptible motion in the air, they 

 were greater. They were also greater if the 

 heat of the atmosphere was increasing; but 

 less, if this was decreasing. 



Having thus, I think, placed beyond doubt, 

 that the formation of ice in Bengal is not occa- 

 sioned by evaporation, I shall now state several 

 reasons, which have induced me to believe, that 

 it depends upon the radiation of heat to the 

 heavens. 



1. This cause not only exists, but exists in a 

 degree, sufficient for the production of the effect, 

 which I attribute to it. For Mr. Wilson found 

 the surface of snow, during a clear and calm 

 night, to be 16 colder than air feet above it, 

 the temperature of the latter being taken by a 

 naked thermometer ; whereas the greatest heat 

 of the atmosphere ever observed by Mr, Wil- 

 liams, at the distance of 5j feet from the ground, 



