Natural History of Volcanos and Earthquakes. 51 
confined to any particular season of the year, although it is cer- 
tainly remarkable, that of fifty-seven earthquakes, which were 
felt at Palermo during a period of forty years, almost a fourth 
part happened in the month of March.* Perhaps the best means 
of ascertaining whether any connection exists between earth- 
quakes and meteorological phenomena, is the observation of the 
barometer. But Hoffman was unable to discover anything pe- 
culiar or extraordinary, either in the relative height of the barom- 
eter, in the direction of its motion, or in the extent of the oscilla- 
tion, during the fifty-seven earthquakes above alluded to. The 
oscillations never went beyond their ordinary limits; indeed, in 
most cases they were very inconsiderable.t Von Humboldt also 
says that between the Tropics, on days when the earth is agitated 
by violent earthquakes, the regularity of the hourly variations of 
the barometer is not disturbed.t 
If aqueous vapors and compressed gases are the cause of earth- 
quakes, there can be no doubt that hot springs and exhalations of 
Sloane’s Letter with several accounts of the earthquake in Peru, October 20th, 
1687, at Jamaica, 10th February, 1688, 7th June, 1692; ibid, y. 1694, p.78; Hist. 
des Trembl. de Terre, t. ii, p. 442; Collect. of the Massachusetts Hist. Soc., t. v, 
. 223. 
* Hoffman, loco cit. p. 52. It is also well known that in other countries, es- 
pecially in Chili and the Moluccas, the periods of the equinox, for reasons of which 
We are ignorant, are considered as those most favorable to earthquakes. During 
the above named period of forty years, this law does not seem to have been appli- 
cable to the autumnal equinox in that part of Europe. 
t During the earthquakes the barometer stood decidedly oftener above the mean 
than under it. However, Hoffman remarks, p. 56, that during the only shock of 
importance which occurred in this period at Palermo, viz., in March, 1823, the bar- 
ometer remained the whole month constantly below the monthly mean. : 
+ Reise, t. i, p 487; also Relat. hist. t. iv, 19. Likewise Boussingault in Ann. 
J 
academy of Turin, during the earthquakes in the year 1808. state of the 
barometer was also inv ariable, whilst the shocks at Lisbon, the 9th December, 
1755, were very strongly felt at Turin Philos. Trans. ix. The observations 
made on the island of Meleda. near the coast of Dalmatia, from the 15th November, 
1824, to the 28th February, 1826, which likewise prove, that no connection exists 
between earthquakes and the pressure of the atmosphere, are very important, the 
shocks felt on this island having been the only ones of their kind as regards length 
of duration.—Die Detonations-Phinomene auf der Insel Meleda von P. Partsch, 
Wien, 1826, P- 204, 
