[ '47 I 



as poffible to the original ones, which alone can give us reafon- 

 able hope of a right conclufion, where the difference of temperature, 

 even in a great number of years, will probably be exceedingly 

 fmall. 



Experience fhews us, that the furface of the earth enereafes 

 in its heat during the fummer, and decreafes during winter : Of 

 this our fenfes give us fufficient notice j and were the greatcft 

 heat in fummer, and the leaft in winter obferved, the mean of 

 thefe would give a refult fomewhat approaching to the mean 

 temperature of the furface of the earth during that year. 



But the quantity thus refulting, would feldom correfpond 

 accurately with the truth ; becaufe, the furface of the earth being 

 expofed to the immediate efFcds of the atmofphere, to rains 

 and wind, to clouds and funfhine, it is evidently fubjed to 

 rapid diurnal variations of heat, which render it an unfit ftandard 

 from whence to derive the mean temperature of the year, by 

 any indudion from a fmall number of experiments. 



At the depth of thirty, forty, or fifty feet beneath the furface, 

 thefe diurnal variations feem to lofe their influence ; while a flow 

 and gradual change of temperature takes place, which is not 

 fenfible except after intervals of many days. At fuch depths 

 the earth, in our climate, is ufually found at its loweft degree of 

 cold about one month after the vernal equinox ; from that time, 

 it flowly and progreflively acquires heat, until about a month 

 after the autumnal equinox, at which period it commonly pofl^eflcs 

 its maximum of temperature. After this, gradually parting with 



U 2 heat 



