THE PREVISION OF EARTHQUAKES 641 



ravages of the fires which followed it. And the fires spread 

 almost unchecked owing to the dislocation of the water-mains 

 by an extraordinary movement along a fault which has been 

 traced in a roughly north-west and south-east line for a total 

 distance of about six hundred miles from near Cape Mendocino 

 on the north to the Colorado Desert on the south. It was only 

 along the northern half of the fault that movement occurred in 

 1906. But, for a total length of 290 miles (including some sub- 

 marine portions), the surface-crust on both sides of the fissure 

 slipped in opposite directions, that on the south-west side to the 

 north-west, and that on the north-east side to the south-east, 

 tearing apart with resistless force every work of human hands 

 that crossed its line. Water-mains were cut through, and the 

 severed ends separated. Roads, fences, bridges and piers were 

 split across and their ends shifted by amounts which at the 

 surface ranged from eight to more than twenty feet. 



These displacements of course represent the sum of those on 

 both sides of the fault. And this is all that could be gathered 

 from the evidence visible to the unaided senses. The precise 

 nature and amount of the several movements could be determined 

 only by a comparison of the trigonometrical surveys carried out 

 in the district some years before and shortly after the earthquake. 

 By this it was found that both sides had moved in the directions 

 mentioned above, that the maximum movement occurred in the 

 immediate neighbourhood of the fault, and that it diminished 

 rapidly with increasing distance from the fault, so that, at a few 

 miles from it on either side of the displacement, if it did not 

 actually die out, it was less than could be detected by the accurate 

 instruments at the disposal of the surveyors. Thus, if a straight 

 line twenty miles long had been traced before the earthquake in 

 a direction at right angles to the fault, the line after the earth- 

 quake would have been severed at the fault, the ends separated 

 by about twenty feet, and the portions near the fault curved, 

 so that on the south-west side the concavity would face to the 

 north-west and on the north-east side to the south-east. 



Taking into account the magnitude of this extraordinary 

 movement and the concentration of the greatest damage wrought 

 by the shock along the line of the fault, there can be no doubt 

 that the earthquake was due partly to the sudden shift at the 

 last moment, partly to the intense friction that must have arisen 

 with the scraping of the rocks on the two sides of the fault. 



