128 BEPORT— 1892. 



which stand np in serrated ridges and overlook the plain, are composed 

 of slates, schists, and other metamorphic rocks. Here and there beds of 

 limestone are found, and rising from the midst of these hills are several 

 larger granite bosses. Volcanic rocks do not exist in this jaart of Japan. 

 From ancient maps and historical accounts we know that the southern 

 portion of this plain has rapidly been encroaching on the sea. This is, 

 no doubt, lai'gely due to sedimentation ; but because evidences of elevation 

 exist at so many places along the eastern coast of Japan, it is reasonable 

 to infer that the growth of land may in part be attributable to this 

 cause. A certain number of earthquakes are every year recorded in the 

 Nagoya-Gifu Plain, but it is by no means so often shaken as many other 

 parts of the Empire. A somewhat remarkable observation connected 

 with the seismological history of this portion of Japan is the fact that, 

 although written records of natural phenomena are usually fewer the 

 further we go back in time, yet, from what has been chronicled, great 

 earthquakes were more frequent in the district between Nagoya and 

 Osaka in bygone times than they have been during more recent times. 

 The last severe shakings at and near Gifu took place in 1826, 1827, and 

 in 1859. Many ordinary buildings and even mountains suS'ered, people 

 and animals were killed, rivers were stopped up, and floods occasioned. 

 The shocks lasted for several days. A rather severe shock was felt on 

 May 12, 1889. In 1880 there were shocks and sounds coming froni the 

 north-west. From 1885 to 1890 the number of shocks annualTy recorded 

 in that district was respectively 9, 4, 10, 12, 15, and 36. In 1888, in 

 one locality near to the centre of the late disturbance, 19 shocks were 

 recorded ; in 1889 the number was 15 ; in ] 890 there were 20 shocks ; 

 and between January and October 1891, that is, up to the time of the 

 great distui'bance, 26 shocks were noted. These figures suggest the 

 idea that for four years before the Gi-eat Earthquake there was a marked 

 increase in seismic activity, and that an unusual number of small dis- 

 turbances had heralded the great collapse. 



Even if it is only sometimes true that small shakings warn us of 

 larger ones to follow, because the latter are so terrible in their effects, it 

 would seem well to carefully study districts in which from time to time 

 there are definite indications of an increase in underground activity. 



Earthquakes generally occur in mountainous countries where the 

 mountains are geologically young, or in countries where there is evidence 

 of slow secular movements like elevation. These latter movements are 

 usually well marked in volcanic countries, and it is not unlikely that 

 the majority of earthquakes, even in volcanic countries, are the result of 

 the sudden yielding of rocky masses which have been bent until they 

 have reached a limit of elasticity. The after-shocks are suggestive of 

 the settling of disjointed strata. 



In Japan, the majority of the earthquakes which we experience do 

 not come from the volcanoes, nor do they seem to have any direct con- 

 nection with them. Assuming that the greater number of earthquakes 

 represent interruptions in the general process of rock crumpling, it would 

 appear that light might le thrown upon the time of their occurrence by 

 careful observations on the change of level in a district where seismic 

 disturbances were frequent. To accomplish this it is suggested that 

 several miles of water-pipes be laid at right angles to a known axis of 

 elevation, and that continuous photographic records be kept showing the 

 height of the water in standards at each end of the line. A more com- 



