40 PRESIDENTIAL ADDRESS SECTION A. 



be able to give a fairly precise description of the various layers 

 that compose it, at least, as regards their elastic strength. 



To record earthquake waves completely, and to find their 

 true meaning, we need three recording instruments at each sta- 

 tion — one to record the N. and S. horizontal motions of the 

 ground, another to record. the E. and W. horizontal motions, and 

 the third to record the vertical motions. The earth motions, much 

 .iiagnified, are recorded on a uniformly moving paper, on which 

 the intervals of time are also marked. 



Many earthquakes have their origin near the surface, and 

 these are seldom felt at great distances. But the so-called tec- 

 tonic earthquakes which seem to be caused by those underground 

 shiftings of rock masses, of which we have so many geological 

 evidences in the foldings and faultings of strata, often originate 

 at depths such as six to twelve miles, and their shock is recorded 

 all over the world. Many such occur every year. 



When we examine the record of any such earth-shaking 

 earthquake, we find that the large waves that chiefly constitute it 

 are invariably preceded by a gentler rippling motion known as 

 the preliminary tremors. When the recording station is near 

 the seat of the earthquake, these tremors precede the main waves 

 by a very short interval of time, but the interval grows longer 

 and longer the further away the seismometer, till at last the 

 tremors outstrip the main waves by more than an hour. These 

 tremors predicted by Knott and first observed clearly by Rebeur- 

 Paschwitz, were shown by Oldham to fall with more or less 

 clearness into two stages. These become most distinct when we 

 compare the records given by the three instruments at any one 

 station. The instinnments show clearly that in the first stage of 

 the preliminary tremors the ground is pushed backwards and 

 forwards along the line in which the quake is travelling. For 

 example, an earthquake originating in Italy would at first affect 

 the N.S. seismometer at Johannesburg much more than the 

 E.W. one. This part of the tremor is clearly due to longitudinal 

 vibrations. In the second phase, however, the earth shakes across 

 the line of advance of the waves, i.e., the vibration is transverse, 

 and would in the case supposed mainly affect the E.W. in- 

 strument. In the chief waves, the motion is partly vertical an3 

 partly horizontal, and somewhat resembles the motion of sea- 

 waves. 



The times taken by earthquake waves to reach stations at 

 dift'erent distances from the origin have been examined with 

 great care. It is found that the main waves take a time very 

 nearly proportional to the distance travelled measured along the 

 earth's surface, i.e., to the arc. We therefore conclude that they 

 travel along the surface, and this agrees with the mathematical 

 theory of their progress as worked out by Rayleigh and Lamb. 

 On the other hand, the times taken by the two parts of the 

 preliminary tremors are much more nearly proportional to the 

 direct distance right through the earth. If we assume that they 

 go along the surface we must imagine that they go faster and 



