June, 1835. EARTHauAKES. 431 



point: it appears to me to be one of very great interest, 

 and not well understood. Humboldt* has remarked, "It 

 would be difficult for a person, who has lived a long time in 

 New Andalusia, or in the low regions of Peru, to deny that 

 the season, the most to be dreaded from the frequency of 

 earthquakes, is that of the beginning of the rains, which is, 

 however, the time of thunder-storms. The atmosphere, and 

 the state of the surface of the globe, seem to have an in- 

 fluence unknown to us, on the changes produced at great 

 depths." In Northern Chile, from the extreme infrequency 

 of rain, or even of weather foreboding rain, the probability 

 of accidental coincidences between the two phenomena neces- 

 sarily becomes very small ; yet the inhabitants in that part 

 are most firmly convinced of some connexion between the 

 state of the atmosphere and the trembhngs of the ground. 

 I was much struck by this, when mentioning to some 

 people at Copiapo that there had been a sharp shock at 

 Coquimbo : they immediately cried, " How fortunate ! there 

 wiU be plenty of pastvire there this year." To their minds 

 an earthquake foretold rain, as surely, as rain foretold 

 abundant pasture. Certainly it did so happen that on the 

 very day of the earthquake, that shower of rain fell, which I 

 have described as in ten days producing a thin sprinkhng 

 of grass. 



Mr. Scrope has put forth an ingenious idea, that the 

 period of subterranean disturbance, where the force is just on 

 a balance with the resistance, may be determined by a sud- 

 den decrease in atmospheric pressure, which over a wide 

 extent of country might produce a considerable efiiect. 

 According to this explanation, the earthquake comes on at 

 the given period from that state of the weather, which is 

 generally accompanied by rain. But there is another class 

 of phenomena, where the state of the weather evidently 



* Personal Narrative, vol. iv., p. 11. In the fourth chapter of the 

 second volume, p. 217, Humboldt, however, appears to think that such 

 connexion is fanciful. 



