en the State of the Barometer. 1^5 



and the third fix. The laft were all in the night-time ex- 

 cept one. On this occafion, the barometer during the fun's 

 jjaliage over the meridian ftood ftill, though it fell before 

 and after. When thefe feven exceptions are dedufted from 

 the 446 variations there remain 439, which accord with the 

 rules. When I lay, however, that thefe variations took 

 place during the fun's paflage over the meridian, I do not 

 jiiean by that expreffion the very mt)ment when the fun is 

 exaftly in the meridian, but the whole period when he is 

 near it. 



All thefe variations took place between eleven o'clock and 

 half after one, confcquently within the fpace of an hour and 

 a half; except fix, five of which happened half an hour or 

 a -quarter before eleven, and one about an hour after noon ; 

 which, indeed, muft be afcribed to a particular caufe, which 

 either haftened or retarded the c(le6l of the general caufe. Of 

 the above-mentioned 439 variations, 54 took place after twelve 

 o'clock, and all the relt before one. 



As it is therefore certain that the barometer during the 

 paflage of the fun over the meridian is always inclined to fall, 

 it cannot be doubted that the caufe of fo invariable a phae- 

 nomenon, which follows fo remarkable laws, and recurs at 

 the fame time under the like circumftances, muft be conilaut 

 and invariable. It is at prefent well known that the prin- 

 cipal atmofpheric caufe of the falling of the barometer is 

 heat and evaporation. But it appears, on clofer examination, 

 that neither of thefe can produce the phaenomenon in quef- 

 tion. What M. De Luc, however, fays *, that the leaft 

 height of the barometer, when no particular caufe intervenes, 

 always takes place about the hotted part of the day, which 

 in every feafon is about |- of the artificial day, might lead to 

 afufpicion that the falling of the barometer about noon pro- 

 ceeds from the heat. P^or, as there is a confiderable increafe 

 of heat about noon, a perceptible falling of the mercury about 

 the fame time murt be conncfted with it, until, in the above- 

 mentioned part of the day, the mercury reaches its loweft 

 Hate, and the heat attains to its maximum. But if we admit 

 that the greaieft heat commonly takes place about the ^ of 

 the day, experience forbids us to afcribe to the fame period 

 the grcatcft falling of the mercury. I have therefore con- 

 ftru6led a table in which the loweft ftatc of the mercury for 

 each day of the month, between noon and the evening, or 

 »)ight, is given from the above-mentioned barometrographic 



'' Recherche s fur Us MoJifvations dc I'Alniofpbc'Cf vol-ii. p- 94. 



obfervaiions. 



