Notes on the Earthquake of October, 1S60. 367 



of gravity. Geology teaches us to refer sucli effects to the slow 

 expansion or contraction of great masses of rock under the influ- 

 ence of heat, to the disengagement of elastic gases under pres- 

 sure, to the removal of matter from the interior to the surface by 

 volcanoes, to the transference of sediment from the land to the 

 sea basins. Such causes are constant and secular, and of course 

 the precise time at which the tension or unsupported weight shall 

 give way can scarcely be calculated, and may occur with sudden- 

 ness and at irregular intervals; and so nice may be the balancing 

 of opposing forces, that observation shows us that the attraction 

 of the moon or an unusually low state of atmospheric pressure 

 may overset the equilibrium and induce an extensive vibration of 

 the solid crust of the earth, yet the actual causes of the pheno- 

 menon may have been for ages slowly preparing for it. 



The fractured condition of the rocks of the earth shows that 

 earthquakes have been occurring throughout all geological time, 

 and they are by no means rare phenomena at present. For the 

 whole earth their rate of occurrence is stated to be nearly 3 per 

 month or 36 per annum ; and no doubt very many are unrecorded 

 and would considerably increase the average. But their distribu- 

 tion locally is very unequal. While in some spots slight earth- 

 quakes are of almost constant recurrence and in others great 

 agitations of the earth are not infrequent, in other extensive 

 regions no earthquakes are known to have occurred. Earthquakes 

 are manifestly connected with the'causes of volcanic action, and 

 follow the same law of distribution on the surface of the globe ; 

 though in volcanic regions earthquakes and volcanic eruptions 

 sometimes alternate, as if the suppression of the latter gave 

 increased energy to the former. Hence volcanic vents have been 

 regarded as safety valves to those pent-up Seismic agencies, as 

 they have been called, which shake the pillars of the solid land. 



In Mallet's map of the distribution of eartliquakes, in the 

 Report of the British Association for 1858, a belt of intense 

 seismic activity runs from the Falkland Islands and Cape Horn 

 along the Andes and Rocky Mountains, giving otf a branch 

 through Columbia to the West India Islands. It crosses over to 

 Asia by the Peninsula of Alaska and the Aleutian Islands, and 

 runs down through Kamtschatka, the Kurile and Japan Islands, 

 from which it gives off a branch along the Ladrone Islands, but 

 the main body crosses over to the Philippines, an<i from these a 

 great crescent shaped patch stretches around Celebes, Java, and 

 Sumatra. This crescent of the East India Islands seems to be 



