responds, without doubt, with the deep-seated center of impulsion, and has been called the epicenter. Here is where the vertical shocks or succussions are most usually felt. From these radiate the undulatory movements, the speed of propagation of which has been estimated at from 1,100 to 1,500 feet per second, or about that of sound in the air. Sometimes the area of disturbance is very limited, even when the convulsion is most violent; at other times it is very extensive, as was the case in the Lisbon earthquake of 1755, which affected a surface equivalent to one-thirteenth of that of the globe. Most frequently the linear dimensions of the agitated surfaces are included within from 65 to 600 miles, or from one four-hundredth to one fortieth of the circumference of the earth.
The area of disturbance is usually irregular in form, and not circular, as was once supposed. Sometimes it is much lengthened in one direction, following the lines of neighboring mountains or other important accidents of structure. During the earthquakes of 1783 and October, 1876, the chain of the Apennines served as a protecting wall to the eastern provinces of the Italian Peninsula. While one side of the chain was assailed by thousands of shocks, which caused great disasters, nothing whatever was felt on the opposite slope. In the Alps, most of the earthquakes take place in the lateral, north and south spurs, which are formed of sedimentary beds, while the central chain, composed of crystalline rocks, is not disturbed. The Andes of South America form a natural bulwark which the strongest convulsions of the Pacific littoral, while they extend a long distance parallel to the chain, hardly ever cross; and, if occasionally a few shocks are propagated beyond it, they become extremely weak.
The movements are very unequally perceptible within the area of disturbances; and between two points shaken by the same impulsion there may be intermediate points that continue quiet. These are sometimes called bridges or arches. The shocks are frequently accompanied by noises resembling heavily loaded wagons rolling over the pavement, or subterranean thunders or roarings; but their intensity bears no kind of proportion to that of the agitation. The great earthquake of Riobamba, in 1797, was silent. But the sounds have relation to the rocks that transmit them. Then there are subterranean rumblings that are associated with shocks like the bramidos of Guanajuato, in Mexico, in 1784, continuing for a month, under terror of which the inhabitants left the city. There were flashes as of lightning, alternating with long rollings, like that of distant thunder. This phenomenon gradually passed away. The noises associated with earthquakes seem to be of the same nature as those that accompany eruptions. The latter are propagated through the ground, not the air, for hundreds of miles. But nothing else that is known of this kind reaches the proportions of what took place on the 26th of August, 1883, at the eruption of Krakatoa, the sounds of which were heard within the whole
