Strictly speaking, the geological effects of earthquakes are of less importance than is usually supposed. The violent shaking of the surface often brings about great land-slips in mountain regions, which precipitate enormous masses of earth and rock from the heights down into the valleys. A striking example of this was given by the earthquakes of northwestern Greece in 1870, in which the rockslides were on a gigantic scale. The falling masses may temporarily or permanently block the valleys, converting their streams into lakes.

On the other hand, the diastrophic forces which produce the earthquake often have other effects of the greatest importance. In all of the more violent quakes cracks and fissures of the ground are formed, which may close again or remain open, and may show a lineal, curved, zigzag, or radiating arrangement. Through these fissures great quantities of water and sand are often forced up from below and form little sand craters, or water-filled funnels on the surface. Frequently the fissures assume the character of faults, or dislocations, one side being raised, the other depressed, so that long scarps, or low cliffs, are left standing. A long list of such faults formed in modern earthquakes might be given, though the limitations of space forbid the mention of more than a few.

In 1811-1812, near New Madrid, Mo., hundreds of faults resulted from the violent earthquakes which shook that region, and a depressed area, 70 X 30 miles in extent, known as the "sunk country," was formed. The earthquake of Owen's Valley, Cal, in 1872, was accompanied by the formation of a fault 40 miles in length and with a vertical displacement, or throw, of 5-20 feet, along the eastern base of the Sierra Nevada. In the Sonora earthquake of 1887, in Arizona and Mexico, a zigzag fault, 35 miles long and with a maximum throw of 20 feet, was produced. The Sierra Teras, in Mexico, appears to have been raised in this movement, for a second fault, with opposite inclination, was formed on the eastern side of that range. The Japanese earthquake of 1891 was accompanied by a fault of 40 miles in length, with throws exceeding 33 feet in height.

Mention has already been made of the great Indian, or Assam, earthquake of 1897; it was accompanied by several large faults, from which lesser ones branch out. As a result of the earthquake of 1899, the region of Yakutat Bay, Alaska, was much disturbed, with an elevation of the coast at one point of 47 feet. "The change of level was differential, indicating a complex system of faulting on a large scale; and shattering of the rocks proves much differential movement on a smaller scale." (Tarr and Martin.) In the San Francisco earthquake of 1906 two long lines of parallel faults were formed, with varying throw up to 20 feet.

Fault scarp in the Neo Valley, Japan, earthquake of 1891.

Fig. 7. - Fault-scarp in the Neo Valley, Japan, earthquake of 1891. The road shows a slight horizontal as well as vertical displacement. (Milne).

Earthquake displacements may be horizontal as well as vertical, though the former are less obvious and have not therefore been so frequently observed. Lateral displacements amounting, in some" localities, to as much as 20 feet, in a direction parallel with the fault-scarps, were a very marked feature of the San Francisco earthquake. Horizontal displacements were also observed after the Indian earthquake of 1897, the Sumatran of 1892, the Japanese of 1891, the Owen's Valley of 1872, and others.

The most remarkable of modern faults are those which have been detected by soundings in the floor of the eastern Mediterranean. The earthquake of October, 1873, off the western coast of Greece resulted in a scarp on the sea-floor with a depth of 2000 feet, where formerly the depth had been 1400 feet. In 1878 the cable to the island of Crete was broken in two places by a violent earthquake and the sea-floor had become so irregular that, in relaying the cable, it was found necessary to make a long detour. The cable from Zante to Crete was broken by an earthquake in August, 1886, and at the break soundings revealed an increased depth of 1300 feet. Some of the Mediterranean scarps are 3000-5000 feet high.

Horizontal shifting of the ground, San Francisco earthquake, 1906.

Fig. 9. - Horizontal shifting of the ground, San Francisco earthquake, 1906. Before the earthquake the road was straight. (Photograph by Sinclair).

A very common result of earthquakes is a change in the circulation of underground waters. Wells and springs go dry, while other springs are formed in new places, or old ones may be increased in volume. The changes in the form of the land surface produce corresponding changes in surface drainage; rivers are diverted into new channels or dammed into lakes, while streams intersected by fault-scarps form new cascades. Many new lakes resulted from the Indian earthquake of 1897.

The general results obtained from the study of the diastrophic movements which accompany earthquakes are thus summed up by Professor Hobbs: -

"1. Appreciable surface dislocations appear to be formed only at the time of macroseisms, and the throws upon these planes stand in some relation to the magnitude of the disturbance.

"2. The evident dislocations produced are generally of two orders of magnitude, those of the higher order being generally very limited in number, while those of the lower order are often quite numerous.

"3. Earthquake dislocations are normal faults with hades approaching the vertical." That is to say, the hade, or inclination, of the fault-plane slopes downward toward the depressed side, as though the latter had merely slipped down the inclined plane.

"4. The crustal movements indicated at the surface at the time of earthquakes appear to be due to an adjustment in position of individual blocks".

The formation of normal faults and open fissures during an earthquake shows that the stresses to which the outer portion of the crust yields are tensional, or stretching. Compression may also occur locally, as appears from the upward or lateral bending of railway rails, so frequent a phenomenon of great disturbances, but such compression is frequently due to the slipping of deep masses of soil and to be compensated by stretching at other points. The measurements of the railway lines after the great Indian earthquake of 1897 proved that whenever the rails had been bent at one point, they had been dragged apart by an equivalent amount at another. This refers to the movements of soil; horizontal faulting, on the other hand, implies a true compression.