314 COSMOS. 



aqueous vapor is precipitated, and on the connection existing 

 between these deposits and the changes of temperature, and 

 the direction and succession of winds. 



4. Tlie electric condition of the atmosphere. The primary 

 cause of this condition, when the heavens are serene, is still 

 much contested. Under this head we must consider the re- 

 lation of ascending vapors to the electric charge and the form 

 of the clouds, according to the different periods of the day and 

 year ; the difference between the cold and warm zones of the 

 earth, or low and high lands ; the frequency or rarity of thun- 

 der storms, their periodicity and formation m summer and 

 winter ; the causal connection of electricity, with the infre- 

 quent occurrence of hail in the night, and with the phe- 

 nomena of water and sand spouts, so ably investigated by 

 Peltier. 



The horary oscillations of the barometer, which in the trop- 

 ics present two maxima (viz., at 9 or 9i A.M., and lOi or 

 lOf P.M., and two minima, at 4 or ^ P.M., and 4 A.M., 

 occurring, therefore, in almost the hottest and coldest hours), 

 have long been the object of my most careful diurnal and noc- 

 turnal observations.* Their regularity is so great, that, in 

 the daytime especially, the hour may be ascertained from the 

 height of the mercurial column without an error, on the av- 

 erage, of more than fifteen or seventeen minutes. In the tor- 

 rid zones of the New Continent, on the coasts as welli as at 

 elevations of nearly 13,000 feet above the level of the sea, 

 where the mean temperature falls to 44° -6, I have found the 

 regularity of the ebb and flow of the aerial ocean undisturbed 

 by storms, hurricanes, rain, and earthquakes. The amount 

 of the daily oscillations diminishes from 1'32 to 0'18 French 

 lines from the equator to 70° north latitude, where Bravais 

 made very accurate observations at Bosekop.f The supposi- 

 tion that, much nearer the pole, the height of the barometer 

 is really less at 10 A.M. than at 4 P.M., and, consequently, 

 that the maximum and minimum influences of these hours 



* Observations faites pour conslater la MarcTie des Variations Horaires 

 du Barometre sous les Tropiques, in my Relation Historique du Voyage 

 aux Regions Equinoxiales, t. iii., p. 270-313. 



t Bravais, in Kaemtz and Martins, MHiorologie, p. 263. At Halle 

 (51° 29' N. lat.), the oscillation still amounts to 0-28 lines. It would 

 seem that a great many observations w^ill be required in order to obtain 

 results that can be trusted in regard to the hours of the maximum and 

 minimum on mountains in the temperate zone. See the obser\'ation« 

 of horary variations, collected on the Faulhom in 1832, 1841, and 1842 

 (Martins, M^tiorologie, p. 254.) 



