and the Conditions which they Indicate. 17 



thrust, but the process is a very slow one, characterised "by 

 an infinite number of minute movements, small fractures, 

 and slight shocks, and a limited number of large displace- 

 ments and violent shocks. 



It is then possible that the series of tremors to which the 

 last earthquake belongs indicates that such an operation is 

 in progress under foot, and, if so, it may have been going on, 

 intermittently, for ages past, and it may last for ages to 

 come. But if it has been so we ought to see its superficial 

 effects in the shape of chains of mountains occupying certain 

 determinate lines, and the plains upon which they stand 

 increasing their height relatively to the sea-level. 



Let us test this hypothesis by applying it to the present 

 case. 



A tract, the centre of which is near to 150° E, and 40° S, 

 has been undermined by the shrinking of the subjacent 

 strata. As fast as the shallow cavity is formed, the 

 suspended stratum crushes down into it, simultaneously 

 sending outwards in all directions, and with every move, 

 tremors of varying intensity ; for it cannot be supposed 

 that the hardest rock, however thick, is rigid enough to 

 withstand long the forces of gravity acting on it, directly, 

 through its own mass, and, indirectly, through that of the 

 overlying ocean and atmosphere. Of the latter two elements, 

 the atmosphere is probably the more potent crushing agent ; 

 for the ocean is a fixed dead-weight, and cannot try the 

 tenacity of a stretched stratum in the same degree as the 

 pressure of the unstable lively atmosphere does. The 

 difference between the weights indicated by a high and a 

 low barometer is very great. Two inches of mercury 

 represent a pressure of 2,000,000 tons to the square mile ; 

 a fluctuating barometer, therefore, means the fluctuation of 

 enormous pressures. The atmospheric waves as they dance 

 on and off the rocky floor must cause it to spring again, like 

 the scaffold plank undulating under the hodman's foot, the 

 cohesion of its particles lessening with every vibration, 

 until at last a higher air-wave than usual crushes down 

 the weakened mass. Looked at from this point of view, the 

 presence of an area of high pressure over the seismic centre, 

 at the time of the shock, becomes significant ; and as volcanic 

 activity is repressed by a high and promoted by a low 

 barometer, this evidence tells as much against the theory 

 of the volcanic origin of the shocks as it does in favour of 

 their being due to secular contraction. 



