EARTHQUAKES. 65 



ceases, and tlie sky, clear at the zenith, presents the chap. v. 

 appearance of a reddish vapour near the horizon. But 

 these prognostics are very uncertain, and the dreaded 

 evil has arrived in all kinds of weather. 



Under the tropics the regularity of the horary vaT'ia- EfTect 

 tions of the haronieter is not disturhed on the days when n"ete'r ^^°' 

 violent shocks occur. In like manner, in the temperate 

 zone the aurora horealis does not always modify the 

 variations of the needle, or the intensity of the magnetic 

 forces. 



When the earth is opened and agitated, gaseous Gaseous 

 emanations occasionally escape in places considerably emanatioas. 

 remote from unextinguished volcanoes. At Cumana, 

 flames and sulphureous vapours spring from the arid 

 soil, while in other parts of the same province it throws 

 out water and petroleum. At Riobamba, a muddy 

 inflammable mass called moya issues from crevices which Moya 

 close again, and forms elevated heaps. Flames and 

 smoke were also seen to proceed from the rocks of 

 Alvidras, near Lisbon, during the earthquake of 1755, 

 by which that city was ravaged. But in the greater 

 number of earthquakes it is probable that no elastic 

 fluids escape from the ground, and when gases are 

 evolved, they more frequently accompany or follow 

 than precede the shocks. 



The subterranean noise which so frequently attends Sutten-a- 

 earthquakes is generally not proportionate to the strength "'^'*° """"^ 

 of the shocks. At Cumana it always precedes them ; 

 while at Quito, and for some time past at Caraccas and 

 in the West India Islands, a noise like the discharge of 

 a battery was heard long after the agitation had ceased. 

 The rolling of thunder in the bowels of the earth, which 

 continues for months without being accompanied by the 

 least shaking, is a very remarkable phenomenon. 



In all countries subject to earthquakes the point at 

 which the effects are greatest is considered as the source 

 or focus of the shocks. We forget that the rapidity 

 with which the undulations are propagated to great 

 distances, even across the basin of the ocean, proves the 



D 



