ON MASS-MOVEMENTS IN TECTONIC EARTHQUAKES, 31 
extend downwards, and the ends of the broken rock would fly upwards, just as do the 
ends of a stick broken by bending, and an open fissure would be formed at the principal 
fracture; but along the side cracks the relative elastic rebounds might be in opposite 
directions and the parts might remain in contact. The principal fracture would be that 
in Disenchantment Bay, but no soundings have been made there to discover the existence 
of a fissure. Fissures and displacements of this character, due probably merely to com- 
pression, but on a very small scale, have been described.’ 
We know very little about the interior of the earth or of the origin of the forces which 
produce such great changes at the surface. Great thrust faults exist which indicate 
tangential compressions; and normal faults, which indicate expansion. Great uplifts 
have occurred unaccompanied by compressions, due, apparently, to vertical forces; and 
the California earthquake has emphasized the existence of horizontal drags below the 
crust. Future study may reveal forces applied in other ways; but it is not going too far 
to say that whenever ruptures occur, they result from elastic strain, and the sudden 
movements produced are merely elastic rebounds; and moreover, except in the case of 
earthquakes connected directly with volcanic action, the strains have not been set up 
suddenly, but are gradually developed by the slow displacements of adjacent areas. 
And severe earthquakes caused by shearing strains, vertical, horizontal, or oblique, 
where the elastic rebounds are in opposite directions on opposite sides of the fault, which 
remain in contact, will be more common than those due to the tensional strains of bend- 
ing, where the elastic rebounds are in the same direction and a gaping fissure is opened. 
THE PREDICTION OF EARTHQUAKES. 
As strains always precede the rupture and as the strains are sufficiently great to be 
easily detected before the rupture occurs, in order to foresee tectonic earthquakes it is 
merely necessary to devise a method of determining the existence of the strains; and the 
rupture will in general occur in the neighborhood of the line where the strains are greatest, 
or along an older fault-line where the rock is weakest. To measure the growth of strains, 
we should build a line of piers, say a kilometer apart, at right angles to the direction which 
a geological examination of the region, or past experience, indicates the fault will take 
when the rupture occurs; a careful determination from time to time, of the directions of 
the lines joining successive piers, their differences of level, and the exact distance between 
them, would reveal any strains which might be developing along the region the line of 
piers crosses. In the case of vertical, horizontal, or oblique shears, if the surface becomes 
strained thru an angle of about 1/2000, we should expect a strong shock. It would be 
necessary to start with the rock in an unstrained condition; this could readily be done 
now in the neighborhood of the San Andreas fault. The monuments set up close to the 
fault-line (vol. 1, pp. 152-159) were not placed with this object in view, but with the 
object of measuring actual slips on the old fault-line. Measures of the class described 
would be extremely useful, not only for the purpose of prediction, but also to reveal the 
nature of the earth-movements taking place, and thus lead to a better understanding of 
the causes of earthquakes. Less definite, but still valuable, information could be obtained 
by the simpler process of determining, from time to time, the absolute directions of 
Farallon Light-house and Mount Diablo from Mount Tamalpais; by this means northerly 
or southerly movements of 1 foot of either of the first two stations relative to the third 
could be detected; and we should know if strains were being set up in the intermediate 
region; but we could not tell where the strain was a maximum nor to what extent it may 
have been relieved by small displacements on intervening fault-planes. 

1¥F, Cramer, Am. Jr. Sci., 3d Series, 1890, vol. xxxrx, pp. 220-225; and 1891, vol. xt, pp. 482-434. 
Mr. H. P. Cushing has shown me pictures of similar cracks with elevated lips in central New York, 
