JULT 19, 1907] 



SCIENCE 



91 



land, 1886, the main cleft was about 6 miles 

 long, nearly straight; in the Sonora district of 

 Arizona and Mexico, 1887, an irregular fissure, 

 35 miles long, with a displacement from 8 to 

 over 20 feet; in the Neo valley, Japan, the 

 Mino-Owari earthquake of 1891 produced a 

 scarp for 40 miles, with a displacement of 

 more than 10 meters; northwestern India, 

 1892, a fissure 120 miles in length; and finally 

 the long fissure of the San Francisco earth- 

 quake of 1906. In several cases, the new 

 fissures followed lines of depression or subdued 

 scarps, presumably formed by earlier earth- 

 quakes. 



It would be of interest, in comparing these 

 seismic features with the straight lines repre- 

 senting " lineaments " which Hobbs in this 

 and other papers draws through points where 

 earthquakes have been recorded and along 

 lines of coasts or valleys, to inquire carefully 

 into the course followed by observed fissures 

 and scarps, in order to determine how far they 

 would give warrant for the rectilinear course 

 of hypothetical lineaments (rectilinear, at 

 least, on the maps employed). As far as 

 data are at hand, it does not appear that ob- 

 served fissures and scarps are straight enough 

 to give support to Hobbs's lines, which in any 

 case seem, as far as earthquakes are concerned, 

 to be largely influenced in location and direc- 

 tion by the evidently subjective element of 

 the location of cities and villages where ob- 

 servers are numerous. Por example, in the 

 absence of evidence as to recent or ancient 

 fault lines, the fact that earthquakes have 

 been recorded at Springfield, Hartford and 

 New Haven is no sufiicient reason for think- 

 ing that seismotectonic information can be 

 gained by drawing a straight line across Con- 

 necticut into Massachusetts through these 

 three cities (see Fig. 7, p. 268) ; indeed, there 

 is even less reason for thinking that seismo- 

 tectonic lines should be closely related to 

 centers of urban population than that rivers 

 should run by large cities. 



FAULT SCARPS AND FAULT-LINE SCARPS 



The relation of earthquake scarps to mod- 

 ern and to ancient faults is to-day well proved. 



The San Francisco earthquake fracture fol- 

 lowed, for at least part of its length, a pre- 

 viously known fault line of thousands of feet 

 displacement, along which the signs of geo- 

 logically recent movement were so manifest 

 that those familiar with the ground had for 

 some years expected the occurrence of further 

 disturbance. The long fault line at the base 

 of the Wasatch range in Utah, with its recent 

 scarps across alluvial fans and flood plains, is 

 well known through Gilbert's reports. These 

 and other geologically modern fault lines are 

 all more or less curved or irregular. Many 

 other faults are so ancient that the faulted 

 mass may have been baseleveled and after- 

 wards broadly uplifted (without renewed dis- 

 placement) so as to sufl^er revived erosion, 

 whereby the weaker rocks, whether in the 

 heaved or in the thrown block, have been 

 worn away; thus fault-line scarps, as they 

 may be called, are produced. Such a scarp 

 differs from a fault scarp in various signifi- 

 cant respects. A foMlt scarp is a direct meas- 

 ure of differential displacement, except in so 

 far as it is defaced and dissected by erosion; 

 its altitude equals the vertical displacement 

 of the fault ; its length equals the leng-th of the 

 fault; it always faces from the heaved to the 

 thrown block. A fault-line scarp in its most 

 characteristic development — namely, where the 

 original displacement has been baseleveled in 

 a completed cycle of erosion, and where the 

 succeeding cycle has reached early maturity — 

 faces the side of the weaker rocks ; its altitude 

 has practically no relation to the original dis- 

 placement, but depends on the amount of ele- 

 vation by which the new cycle is introduced, 

 or on the thickness of the body of weak rocks. 

 The length of such a scarp is not a measure 

 of the original fault length, but of the dis- 

 tance over ^which rocks of unlike resistance 

 happened to be brought next to each other 

 by the faulting. When a fault-line scarp 

 faces the heaved block it may be described as 

 a topographically reversed fault-line scarp; 

 but care must be taken not to confound it 

 with a "reversed fault" of geological no- 

 menclature. When the rocks on the two sides 

 of a baseleveled fault are of the same hard- 



