[ 373 ] 



towards fun-fet both thermometers approach to an equality.' This 

 {hews that the higher thermometer at 75 feet communicates its 

 heat upwards ftill more flowly than the low"er communicates its 

 heat to it ; which proves that the higher heat afcends, the more 

 flowly it is propagated to ftill fuperior heights. This explains the 

 conftancy of the heat at the fummit of the mole while the heat 

 below increafed as already mentioned. But after fun- fet the lower 

 thermometer cools quicker than the upper, and towards the end 

 of twilight is often about four degrees colder, notwithftanding 

 that the earth a little below its furface is much warmer. The 

 cold which thus aiFeds the lower thermometer is evidently caufed 

 by evaporation ; this excefs of heat of the upper over the lower 

 thermometer lafts the whole night, and until two hours after fun- 

 rife, when, notwithftanding the evaporation, the earth is more 

 heate'd by the fun's rays than it is cooled by evaporation, and fo 

 alfo is the thermometer fufpended five feet over it, and thus ac- 

 cumulates heat fafter than it can reach the upper thermometer, 

 until after the hottefl hour of the day, as already mentioned. 



These obfervations happily point out the hours of the day 

 moft proper for difcovering the mean temperature of the atraof- 

 phere, a matter of great importance in barometrical menfuration. 

 It is evident that the hour of fun-rife (or before fun-rife) is an 

 improper time, finc^ the heat at the height of 75 feet is fome 



Vol VIII. 3 A degrees 



I 



