PERMANENT DISPLACEMENTS OF THE GROUNDS. 23 
moved. ‘There is no direct evidence that forces brought into play by the general com- 
pression of the earth thru cooling or otherwise were involved, for there is no evidence 
that the surface of the earth was diminisht by the fault.. It is true that the surveys 
did not extend over the whole length of the fault, and therefore are not decisive on this 
point, but so far as they went they show an extension of the region between San Fran- 
cisco and Monterey Bay, between the surveys of 1851-1865 and 1906-1907. 
A strong, shearing force would be produced along the fault-plane by forces making an 
angle in the neighborhood of 45° with it; that is, by either tensions or compressions in 
directions roughly north and south or east and west, or by a combination of the two. A 
tension alone could not have caused the rupture, for then the sides would have been pulled 
apart; an east-west compression would have brought Mount Diablo and the Farallon 
Islands nearer together and would have reversed the observed relative movements on 
opposite sides of the fault. The surveys, altho not entirely decisive, are against a north- 
south compression; and, moreover, the elastic distortion accompanying a compression 
which could produce a fracture 435 km. long would not have been restricted to a zone 
extending only 6 or 8 km. from the fault-plane. A shear exerted by forces parallel with 
the fault-plane on the eastern and western boundaries (which is equivalent to a north- 
south compression and an east-west tension at the boundaries) with no resistance at the 
under surface would have produced an even shearing strain thruout the region between 
them; and straight lines would have been changed into other straight lines, exactly as 
occurred in the experiments described above and illustrated in figs. 8 and 9. An additional 
compression or tension in any direction would not have 

altered this characteristic. Similar forcesontheeastern “ 
and western boundaries with forces at the under surface m C 
resisting the movements would have produced some 
such distortion of the straight line AC into A’C’ as nal at c 
shown in fig. 10. The tendency to rupture would be 
greatest at A’ and C” and least in the neighborhood of O; it is evident that such forces 
could not have produced a rupture at O, and the displacements are not like the dis- 
placements observed. 
The only other boundary is the under surface of the moved region, and it is here that 
we must suppose the disturbing forces applied; and they must be distributed over this 
surface so as to produce the distortions observed. 
Nors. — Mr. Gilbert has suggested a modification of the experiments described above; 
instead of making the cut, which represents the fault, all the way thru the jelly, he suggested 
that it extend only a part way thru, and that it would thus more nearly represent the 
true conditions of the earthquake fault. This was tried, but the jelly was not strong 
enough to resist the forces developt during the displacement and the break was quickly 
extended all the way thru the jelly. It is not difficult, however, to see what forces would 
be developt under these circumstances. There are two cases: first, suppose there exists 
below the crust a region practically devoid of elasticity, in which only viscous forces can 
act, and suppose the fault extends to this region; we then come back to the last case con- 
sidered. Second, suppose the elastic character of the rock extends well below the lower 
limit of the fault; such a case could easily exist if the strength of the rock increased with 
depth, even tho the strains continued far below the fault as great as they were within its 
limits. Let us consider the nature of the distortion produced in this case. We shall sup- 
pose the rock under elastic shearing strain, and when the rupture occurs, the shearing forces 
across the fault-plane, which upheld the strain, are annulled and the rock takes a new 
position of equilibrium under the new forces brought into action, in such a way that the 
surface line A’OC’ (fig. 11), straight just before the rupture, afterwards takes the position 
A’0’, D’C’. Below the limit of the fault no change takes place, but the original vertical 
plane thru A’O’C’ has been broken and warped, suffering no displacement below the fault, 
but gradually increasing its distortion until it corresponds to AO’ and D’C’ at the surface. 
