1^ On the Influence of the Sun 



The obfervatlons of thefe three men afford fome indications 

 of the influence which the heavenly bodies have on the baro- 

 meter, and to which M. Planer afcribes part of the efi'efts : 

 but he is far from proving this; for the number of the places 

 the obfervations were made at, which M. Steiglehner com- 

 pared, were not above three^ and confequently loo few to de- 

 termine the quedion. The cafe is the fame with the fix ob- 

 fervations of M. Planer, made daily, as has been obferved 

 by M. Chiminello. And might not what is afcribed to the 

 period of four hours have happened in the lad hour, or the 

 lafi. half-hour ? Chiminello's experiments are of more im- 

 portance, but for want of night obfervations they are Hill 

 defective and imperfeft. 



In this refpedt, how-ever, I was more fortunate, as T had 

 an eafy and fure method of followine; all the variations of the 

 barometer, and even the fmalleft, by night as well as by day. 

 I had in mv poffcffion, at that lime, a barometer, belonging 

 to the Electoral Phyfical Mufeum, invented lately by M. 

 Changeux, and called a barometrograph, which being con- 

 nefted with a clock fhows the flate of the mercury every 

 four minutes, by means of a point impreffed on a moveable 

 plate. I began on the aift of May 1785 to make ufe of 

 this excellent inftrument, and it appeared to me, by repeated 

 obfervation, that a continual variation took place towards 

 noon as well as towards midnight. By more accurate ob- 

 fervation, I found that this variation in the mercury confift*, 

 at leaft, in an inclination to fall. But that I might be more 

 certain in treating a quettion of fo much importance, I re- 

 fohed to compare all the obfervations from the firft dav to the 

 end of the year, which are printed in five volumes of our 

 Ephemerides. In that time I found that the pafl'age of the 

 fun over the meridian by day and by night had taken place 

 446 times, and confequently as many variations of the baro- 

 meter were to be coniidered. From the moft accurate coni- 

 parifon, the following three rules prefented themfelves : 



il\, When the fun pafles the meridian, the barometer, if 

 in the aft of falling, continues to fall, and the falling is 

 accelerated, 



2d, When the fun paffes the meridian, the barometer, if 

 in the aft oi riling, falls, or becomes (lationary, or riles more 

 fiowly. 



3d, When the fun paffes the meridian, the barometer, 

 which is (iaiionary, falls, if it has not rifen before or after 

 being fl:ationary; in which cafe it ufually becomes ftationary 

 duiing the fun's paOage. 



The firft rule had only one exception, the fccoud none, 

 4 and 



