In Northern Chile, from the extreme infrequency of rain, or even of weather foreboding rain, the probability of accidental coincidences becomes very small; yet the inhabitants are here most firmly convinced of some connexion between the state of the atmosphere and of the trembling 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 out, "How fortunate! there will be plenty of pasture 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' time producing a thin sprinkling of grass.

At other times rain has followed earthquakes at a period of the year when it is a far greater prodigy than the earthquake itself: this happened after the shock of November 1822, and again in 1829 at Valparaiso; also after that of September 1833, at Tacna.

A person must be somewhat habituated to the climate of these countries to perceive the extreme improbability of rain falling at such seasons, except as a consequence of some law quite unconnected with the ordinary course of the weather.

In the cases of great volcanic eruptions, as that of Coseguina, where torrents of rain fell at a time of the year most unusual for it, and "almost unprecedented in Central America," it is not difficult to understand that the volumes of vapour and clouds of ashes might have disturbed the atmospheric equilibrium.
Humboldt extends this view to the case of earthquakes unaccompanied by eruptions; but I can hardly conceive it possible that the small quantity of aeriform fluids which then escape from the fissured ground can produce such remarkable effects.