once at the bottom of the silver-mines of Charnacillo, more than six hundred feet underground, during an earthquake that destroyed his own house and several others right above him, while he did not feel the least agitation.
The duration of the shocks is generally very brief, sometimes not more than a second or two. The undulatory movements are more prolonged. A few instants are sufficient to produce the most disastrous effects. Three shocks, each of which was estimated to be not more than four seconds long, destroyed more than 20,000 persons in and around Carácas in March, 1812; and the convulsion at Pdobamba, in 1797, killed as suddenly 30,000 victims.
But, however overwhelming and disastrous in reference to our persons and buildings earthquakes may be, it must be borne in mind that the amplitude of their movements is wholly insignificant in proportion to the dimensions of the globe whose epidermis they shake. The phenomenon is rarely limited to a single shock. Generally, several shocks follow one upon another at short intervals. In many cases, the movements are repeated for months and years, with pauses of a variable duration, so as to form as a whole, till they are totally extinguished, what might be called a seismic period. After the shock which overthrew Thebes, on the 18th of August, 1851, the commotions continued in Bœotia for eleven months, occurring sometimes as often as three times in twenty-four hours. Long series of shocks disturbed a part of Scotland during the two years from the 20th of October, 1839, to the 7th of December, 1841. Hundreds of similar examples might be cited.
The chain of the Alps has furnished examples of seismic periods in many of its parts. Series of shocks were felt at Pignerol, in Piedmont, from the 2d of April, 1808, till the 17th of May following, during which time not a day passed but some movement was felt. Sometimes the tremors were noiseless, sometimes they were accompanied with commotions preceding the destruction of buildings. The commotions were renewed on the 26th of September, the 28th of October, and the 22d of November; and in the next year on the 13th of March and 26th of June. Similar periods of seismic action were observed in le Valais in 1755, on Lake Gardo in 1866, on Monte Baldo in 1868, and at Belluno in 1873. The present period in Andalusia is of the same kind. The prelude of the 23d of December, which disturbed a part of the Spanish Peninsula, the great earthquake of the 25th of December, and that long succession of shocks which still continued on the 9th of March with sufficient violence to cause new ruins, belong to the same series. As Humboldt has remarked, it is noteworthy that series of this kind are produced more especially in countries distant from volcanoes.
In the disturbed regions we may generally remark a tract of limited extent in which the movement is particularly energetic. It cor-
