or the earth's surface. 155 



sink below the medium level, as at k and i. An annular de- 

 pression would thus be produced, surrounding the central ele- 

 vation, vertical sections of which are exhibited by d kf and 

 hie j that depression below the medium level being exactly 

 equal in bulk to the additional elevation at the centre, occa- 

 sioned by the momentum, the vertical section of which is 

 fghc. The middle elevation, and the annular depression, 

 would continue to increase during a period, and to an 

 amount regulated by that class of laws, which determines 

 the magnitude and duration of waves, as relative to the wind 

 which produces them ; and the water at the end of that 

 period would be thrown into the position denoted by the 

 dotted line in fig. 2., and by the single continued line in 

 fig. 3., indicated in both figures by the letters dlcfghie. 



During a third period, the w T ater flowing from without the 

 annular depression, to fill it up, a wave of depression would be 

 propagated outwards in all directions, and would visibly reach 

 to places where the elevation produced by the direct impulse 

 in the first moment, had not been at all perceptible, in con- 

 sequence of their distance from the centre. It is owing to 

 this cause, I am convinced, that the catastrophes at Lisbon, 

 at Cadiz, at Madeira, and at Lima, in so far as the position 

 of the surface of the sea was concerned, were all first announ- 

 ced by a retreat of the sea, as shewn at m, in fig. 4. 



During another successive period, the undulatory motion 

 continuing, the wave of elevation would advance towards the 

 shore at ra, and being met by the retiring water, would produce 

 the curling form called the breaking of a wave, as shewn 

 at m in figure 5. At that instant the calamitous effects 

 described in all the great earthquakes, as resulting from the sea, 

 have been produced. 



U2 I 



