And they are too frequent and general to be afcribed to ac- cidental fires. I am perfuaded their caufe muft be fought for in the atmofphere it- felf, to aid us in which I know but of thefe conftant circumftances; a dry air; a tem- J| perature as warm at leafl as that of the fpring or autumn; and a moderate currunt of wind. They are moft frequent about fun-fet; rare in the middle parts of the day; and I do not recolleft having ever met with them in the morning.
The variation in the weight of our atmofphere, as indicated by the barometer, is not equal to two inches of mercury.
During twelve months obfervation at Wil- liamfburgh, the extremes were 29, and 30.86 inches, the difference being 1.86 of an inch: and in nine months, during which the heighth of the mercury was noted at Monticello, the extremes were 28.48 and 29.69 inches, the variation being 1. 21 of an inch. A gentleman, w^ho has obferved his barometer many years, aflures me it has never varied two inches.
Contemporary obfervations, made at Mon- ticello and WilliamIburgh, proved the variations in the weight of air to be fimul- taneous and correfponding in thefe two places.
Our changes from heat to cold, and cold to heat, are very fudden and great. The mercury m Farenheit's thermometer has been known to defcend from 92" to 47° hi thirteen hours.
» It was taken for granted, that the pre- ceding table of average heat will not give a falfe idea on this fubjeft, as it pro- pofes to ftate only the ordinary heat and cold of each month, and not thofe which are extraordinary. At Williamfburgh in Auguft ly 66, the mercury in Farenheit's thermometer was at ^8"" correfponding with 29! of Reaumur. At the fame place in January 1780, it was 6** corre- fponding with I li below o, of Reaumur.