010 PHILOSOPHICAL TRANSACTIONS. [aNNO J 784. 



hended the local variation of the thermometers might, in a certain degree, cor- 

 respond with some particular change in the state of the atmosphere. The event 

 answered my expectation in a singular manner in respect to the nocturnal varia- 

 tion; for it generally happened, that when the sky was dark and cloudy, what- 

 ever was the condition of the atmosphere with regard to the other particulars 

 above enumerated, the thermometers agreed pretty nearly with each other; but, 

 on the contrary, whenever the sky became clear, the cold of the night at the 

 lowest station in the garden constantly exceeded the cold at the top of the cathe- 

 dral tower, where the instrument was placed 220 feet from the ground, entirely 

 exposed to the open air, wind, dews, and rain, in a shady northern aspect. 



The local variations in the day-time seemed to be regulated by the general 

 degree of heat only, without being affected by any other particular disposition of 

 the atmosphere, or the clearness or cloudiness of the sky, as the nocturnal varia- 

 tions were. In the month of September, when the glasses rose from 6o° to 70°, 

 the heat at the lower station constantly exceeded the heat at the upper station ; 

 and in some measure proportionably, as the weather was hotter.* In December 

 and January, when from below 30° they seldom rose to 40°, the local variation 

 in the day-time nearly ceased, or was found in very small degrees inclining some- 

 times one way, sometimes the other. 



That the clearness of the sky should contribute to the coolness of the air in 

 the night, is not at all surprizing; but that, whenever the sky becomes clear, 

 the cold should seem to arise from the earth, and be found in the greatest degree, 

 as long as it continues clear, in the lowest situation, seems a little extraordinary: 

 this however appeared to be the case, both in the warmer and in the colder wea- 

 ther, during the whole time these observations were taken. About noon, on 

 the 3d of January, the sky becoming clear, the air got cooler; and going into 

 my garden, about 8 in the evening, I perceived the surface of the ground, 

 which had been wet by the rain in the forenoon, began to be frozen. Looking 

 immediately at the thermometer, I saw the mercury at 33^°; and observing a 

 piece of wet linen hanging near the glass, not 5 feet from the ground, I took it 

 into my hand, and found it not in the least frozen; by which it appeared, that 

 the degree of cold which had frozen the surface of the ground, had not then as- 

 cended to the glass, nor to the linen, and consequently had not been communi- 

 cated to the air 5 or 6 feet above the earth. The next day I found, as expected, 

 a considerable local variation; the index for the cold of the night in the garden 

 being at 32°, that on the hill being at 354°, and that on the top of the tower at 



* As the heat at the lower station exceeded the heat at the upper ones, when the weather was hot j 

 and equally so, whenever the sky was cloudy, as well as when it was clear ; it appears, that the o-lass 

 at the lower station was not materially affected by the reflection of the sun's rays from the earth, as 

 at first I apprehended it would be. — Orig. 



