EARTHQUAKE OF CENTRAL JAPAN, 1891. ;53C) 



as^serted that the earth^^uake could not have originated in this vallev. 

 Had they come to Kimbara, where I left off the description of the 

 earth-rent, this notion would soon liave vanished. 



Kimbara itself hes on the left of the fissure, but so near to it 

 that all the buildings in the village entirely collapsed. Ar the north 

 end of it the road descends the steep slope of Dando-zaka. and brinu's 

 us to the beginning of the phice of greatest destruction in the vallev. 

 Beyond the steep descent there was no trace of road left bv (he 

 earthquake ; rubbish and blocks of hornstone had oljstructed and 

 buried it under them. For a few days after the grejit shock there 

 was still everv now and then, the crash of some mass of rocks falling 

 from precipices on either side of the valley that had l)een left by a 

 landslip. 



The rent now runs east of Hinata, where the river bed makes 

 crooked turns, and at Hirano, the next village, passes on the right, 

 and at the south of ]\Iidori, it appears on the left side of the Xeo-gawa. 

 and here one of the most wonderful effects of the earth* piake 

 was to be seen. A fine new road (see PI. XXXI \'.) leading to (üfu 

 had been obliipiely cut into two, and the lower end with the 

 surrounding fields had sunk about 6 metres below the upper end. 

 The road is here on the west side of the river, and the eastern 

 half of the valley, which includes the Xeo-gawa and opposite hills, 

 shows the tilting edge, or cutting of the fault. That tlie east JudfJuul 

 been puslied é metres northwards, in C(jnformity with the general rule, is 

 well seen in the photograph by an abrupt change in the direction of 

 the displaced road. So far as I am aware, the formation of such 

 a colossal fault on the surface, as this in Midori, is exception- 

 ally rare, and finds its equal only in the ' Ullah-bund,' or God's 

 dam at the Runn of Cutch, in the lower course of the Indus. 



Those who have followed me with patience to this point in my 



