'I'lIE GREAT EARTHQUAKE OF S. E. JAPAN ]51 



8m. X 5m. was displaced as much as 35 cm. towards N13°E ; a similar 

 one of 10m. x 10m., as much as 130 cm. towards N10°E (PI. XXV 

 A); and another of 7m. x 7m., as much as 75 cm. towards N10°E. 

 In this town, houses as a rule collapsed towards the south and mostly 

 took tire ; only a narrow area escaped the flames and some houses 

 therein made displacements such as I have just described. (PI. XXV 

 B-XXVI). These enormous displacements seem to have taken place in 

 some cases at a single stroke, and in other cases after continued 

 action. Such phenomena can be explained only by assuming a strong 

 horizontal vibration accompanied by a big vertical one. For the 

 first step, let us suppose an earth-movement capable of causing the 

 displacement of a structure from its foundation-stone. The accelera- 

 tion of the earth-movement required must be at least r/ x 0.6, because 

 the coefficient of friction between wood and stone is equal to 0.6. 

 Repetitions of such horizontal vibration, however, would cause the 

 structure to move to and fro from its initial position, a suppositon which 

 contradicts the facts. Again, if we suppose the earth-movement to 

 consist of large horizontal as well as vertical movements, then the 

 critical condition for sliding would be as follows: — 



[X {g — a sin 0) — a cos 6, 



where [i, a and tan 6 represent respectively the coefficient of friction, 

 the acceleration of the resultant earth-movement, and the ratio of the 

 vertical acceleration to the horizontal one. If ^^y=0.6, then d is ap- 

 proximately equal to 31° for the minimum of n; in this case, a, its 

 horizontal and vertical components are equal to (/x051, ;/ x 0.43 and 

 (/xO.26 respectively. If vibration of such magnitude take place 

 repeatedly parallel to a straight line, then the structure would slide 

 and stop alternately according as the sense of the vertical component 

 of the earth-movement was downward or upward, so that such struc- 

 tures standing within a narrow area, would undergo enormously large 

 displacements in a common direction. 



The phenoiuenoua of large displacements were first observed in 

 this country by the late Prof. Omori in the case of the great Mino- 

 Owari Earthquake of 1891. Here, in the most disturbed area, a 

 wooden temple gate having 8 i)illars was found displaced as much as 

 90 cm. along the underlying stone pavement. I had also an ojq)or- 

 tunity of observing the same phenomena in the case of the Anegawa 

 Earthquake of 1909 in the meizoseismic area, where I found two bell- 



