761 



mercury appears to remain stationary during a 

 very considerable time. That time varies from 

 15' to 2 h ; and in determining with precision 

 the half -duration of the stationary state, we 

 should distinguish between the real instant of 

 the maximum, and the epocha when the baro- 

 meter ceases, as to our senses, to rise or fall. 



5°. In the torrid zone in general, between the 

 equator and the parallel of 15° north and south, 

 the strongest winds, storms, earthquakes, the 

 most sudden variations of temperature and hu- 

 midity, neither interrupt nor modify the period- 

 icity of the variations. This is the more wor- 

 thy of attention, as in some parts of equatorial 

 Asia, where the monsoons blow with violence, 

 (for instance in India,) the rainy season entirely 

 masks the type of the horary variations, and 

 that at the same period when these variations 

 are insensible in the interior of the continent, 

 on the coast, and in the straits, they are mani- 

 fested without any alteration within the same 

 parallel, in the open sea. 



6°. Between the tropics, one day and one 

 night suffice to know the limit-hours, and the 

 duration of the small atmospheric tides ; in the 

 temperate zone, in 44° and 48° of latitude, the 

 phenomena of periodicity are manifested at all 

 seasons with great clearness, in the mean of 

 from fifteen to twenty days. 



7°. The unequal extent of the diurnal varia- 



