EFFECTS PEODUCED UPON BUILDINGS. 139 



formed the line of a shingle beach. That portion of the 

 settlement back from the sea stands upon ground which 

 was originally marshy. In the central portions of the 

 settlement this bed of gravel is very thick, perhaps 100 

 feet or so, but as you near the edge of the bluff it pro- 

 bably becomes thinner, until it finally dies out upon the 

 flanks of the scarps. 



On the top of the bluff, the beds of gravel will, in every 

 probability, be generally thinner than they are upon the 

 lower level. The beds of tuff, which is a soft grey- 

 coloured clay-like rock, produced by the solidification of 

 volcanic mud, appear, when walking on the seaboard, to 

 be horizontally stratified. If there is a dip inland, it is 

 in all probability very slight. Here and there the beds 

 are slightly faulted. Taken as a whole we may consider 

 these beds as being tolerably homogeneous, and an earth- 

 quake in passing through them would meet with but 

 little reflection or refraction. At the junction of these 

 beds with the overlying gravels, both reflection and refrac- 

 tion would comparatively be very great. 



On entering the gravel, as the wave would be passing 

 into a less elastic medium, the direction of the wave 

 would be bent towards the perpendicular to the line of 

 junction, and the angle of emergence at the surface would 

 consequently be augmented. At the surface certain 

 reflection would also take place, but the chief reflections 

 would be those at the junction of the tuff and the allu- 

 vium. 



Under the settlement it is probable that all the reflec- 

 tions which took place would be single. Thus wave 

 fronts like A^ advancing in a direction parallel to the line 

 a-i would be reflected in a direction a^ and give rise to a 

 series of reflected waves A^. These are shown by thicker 

 lines. Similarly all the neighbouring waves to the right 



