EARTHQUAKE 367 may be somewhat to the right or left of the line of propagation from the origin of the im- pulse, owing to unequal strains or other dis- turbing causes; but in the main it is simply forward during the first half of the orbit (the first semiphase), and backward during the latter half, and therefore nearly in a straight line. The forward and the return movements when at their maximum velocities are generally almost precisely in a line with the direction of the propagation, and it is at these moments of maximum velocity that the destruction of walls, &c., by the earthquake is accomplished. During the forward movement occurs also the greatest velocity in the upward direction. The down- ward movement of the earth accompanies the backward or the latter half of the pulse ; but the backward movements are generally slo'wer than the forward, and therefore have a less destructive effect. The twisting of isolated blocks, which had by many been ascribed to a circular or vorticose movement of the earth, is in reality, as Mallet shows, the effect of a sim- ple direct stroke acting in connection with the inertia of the body and the friction at its base as a mechanical couple. When the origin of the impulse is vertically beneath the observer, the movement of each particle of the earth about him will be mainly in the vertical direction ; when the origin is at a distance, the movement will be more or less inclined to the horizon ; and if by a proper seismometer we determine for several places the azimuthal direction and the angle of emergence of the pulse, it becomes possible to fix exactly the position of the origin. The destruction of walls, moving of solid bodies, &c., are therefore due not to the velocity of transit or that with which the shock passes along over the earth's surface, but to the velocity of the movement Of each particle of the surface of the earth in its small elliptical orbit. The velocity of translation depends on the elasticity of the rock or earth, but rapidly diminishes in proportion as the rock is discontinuous or non -homogeneous ; it is therefore generally different in each direc- tion. The wave of shock or the pulse of com- pression is reflected and refracted on encoun- tering a new stratum of different elasticity, precisely as in the case of sound. So also phenomena of interference can arise, and it is in such cases that the pulses cease to be recti- linear or elliptical, and become more compli- cated curves. When the earthquake origi- nates under the bed of the sea, the vertical shock communicated to the waters causes a wave or a swell, often of large extent, that spreads outward in all directions with a velo- city varying with the depth of the ocean at the point over which it may be passing, and which on reaching shallow water rushes along as a breaker preceded by a fall or a recession of the water. In earthquakes of this kind, therefore, there may proceed from a single origin (the seismic focus) within the earth a great wave of compression (the wave of shock) 280 VOL. vi. 24 and, accompanying it, a wave of sound ; both waves will before they reach the observer be transformed by the influence of discontinuity and non-homogeneity into a complicated mix- ture of strong and weak shocks and loud and feeble noises. The sea water first affected by the shock will receive and propagate waves of sound, small forced waves of water, and the large swell which ultimately becomes the great earthquake wave, all of which will reach the observer on the land at different moments, depending on their respective velocities. The disturbance having passed through the ocean upward to the atmosphere must communicate to it the three sorts of waves already existing in the water, but of these the sound wave is the only one that is perceived. .This completes the entire phenomena of an earthquake so far as relates to the shocks and sounds. When the impulse originates under the land, there is wanting from the preceding enumeration of six classes of waves or pulses only the great sea wave ; all the others may be present, but if the shock extends with severity to the shore of a sea, the small forced waves that will arise may be of considerable magnitude, while the vertical stroke that was in the former case expended in producing the great sea wave now expends itself on terrestrial objects. The preceding views, as elaborated by Mallet, and in part adopted by Hopkins in his report of 1847 to the British association, were very generally accepted, but partook somewhat of a hypothet- ical nature, because of the scarcity of accurate observations of actual earthquakes, till 1862, when Mallet published the details of the meth- ods and results of his investigation of the Nea- politan earthquake of 1857, to the study of which he had devoted much time at the expense of the royal society. In this investigation, the great importance of which as well as its com- plete success justifies our detailed attention, Mallet has established the science of seismolo- gy upon an immutable basis of accurate ob- servation. By using the fissures in buildings, the disturbances of heavy objects, &c., as natu- ral seismometers, and by proper selections of representative cases, he was able, with the mathematical assistance of the Eev. Samuel Haughton, to fix from 177 determinations the position of the seismic focus or origin of the dis- turbance as being a cavity beneath the village of Caggiora near Potenza. From 26 determi- nations of the angle of emergence he found the mean depth of the cavity to be 5| m. beneath the sea. Lines of equal seismic disturbance being drawn, he showed the earthquake to have desolated a region larger than any of which we have exact knowledge. The refractions and reflections of the wave of shock in encounter- ing various strata were clearly shown. From the number and directions of the sounds and shocks he deduced the general form and position of the focal cavity, and showed that it must have been of the nature of a curved fis- sure whose height was 3 m., length 9 m., and
