in an Earthquake Country. 211 



above referred to, of placing a heavy top on a chimney, as- 

 sumes that the shock is an impact, and consequently that a 

 definite quantity of momentum is given to the structure ; but 

 it must be quite evident that it is the relative velocity of the 

 base of the structure with regard to the other parts which is 

 the fixed quantity, and therefore that the more massive the 

 structure, the more momentum enters it through the base. 



There is no easy way of judging what are the forces which 

 cause an ordinary Japanese house to return to the perpendi- 

 cular position after it has received a push or blow ; and so we 

 cannot calculate its natural time of vibration ; but it is well 

 known that it vibrates very slowly, an ordinary Japanese two- 

 storied house with the usual heavy roof taking perhaps four 

 seconds to make a complete vibration. The restoring forces 

 are due merely to stiffness of the joints, there being no 

 rigid connexion with the ground, since the vertical posts of the 

 house are all supported on detached stones, and there are also 

 no diagonal stays in the building. Such a structure is there- 

 fore capable of being displaced very far from its position of 

 equilibrium without fracture occurring ; and as its time of vi- 

 bration is very long, it has a very great amplitude of swing 

 during most ordinary earthquakes. That this amplitude is not 

 even greater is most probably due to the fact that there is a 

 sort of viscous resistance to motion at all its joints. Such a 

 viscous resistance must greatly diminish the motion, and will 

 be especially useful in an earthquake consisting of regular vi- 

 brations; but the most severe test of such a structure consists 

 in an earthquake-shock which begins with a sharpe impulse, 

 or which has a very irregular motion. The slowly vibrating 

 structure would register the shock in a longer period of time 

 than that in which the blow was delivered ; but it would pro- 

 bably have an exceeding great first swing from its position of 

 rest. 



We think that the important elements of safety in ordinary 

 Japanese structures is this viscous resistance which they oppose 

 to motion, and which is mainly due to the great multiplicity o 

 joints (all of which are compelled to move) and to the absence o 

 diagonal pieces ; for we deduced from the principle in our ori- 

 ginal paper, that if the restoring forces are weak there ought 

 to be a great viscous resistance to motion if we wish the strains 

 of the structure to be small. But it must be remembered that 

 this safety is only gained by a very great expenditure of timber; 

 so that, although such slowly vibrating structures as many of 

 the temples in this country may be regarded as exceedingly 

 safe during earthquakes, it must not be concluded that all 

 heavily roofed houses are secure. 



