ON DEW, &c. 263 



was commonly 4 higher than that among the 

 pans. It is possible, therefore, that Mr. Wil- 

 liams may have seen ice, a little before 'sunrise, 

 when the temperature of the air was 46. But 

 granting this were the fact, it would not hence 

 follow, that the ice was formed, while the air 

 possessed that heat. For, although the air is 

 generally held to be in all countries colder 

 about sunrise, than at any other time, I know 

 from my own observations, that this is not 

 always the case in England ; and similar ex- 

 ceptions may occur in Bengal. Sir H. Davy 

 has said, in his Elements of Chemistry, that ice 

 will form in Bengal, when the temperature of 

 the air is not below 50 ; but he has given no 

 authority for this assertion. 



The formation of ice, in the circumstances 

 which have been just mentioned, was attributed 

 by Sir R. Barker altogether, and by Mr. Wil- 

 liams in great measure, to cold produced by 

 evaporation. Sir R. Barker's opinion has since 

 been adopted by some of our most distinguished 

 writers on Natural Philosophy, as Watson, 

 Thompson, Young, Davy and Leslie, apparently, 

 however, without their having fully considered 

 it, as I shall now attempt to show. 



1. It is necessary for the complete success of 

 the process, that the air should be very still ; 

 wind, which so greatly promotes evaporation, 



