On Structures in an Earthquake Country. 209 



measurements of the heat of the electric-light spectrum, that 

 in the normal spectrum the maximum was near A. 



Lundquist arrived at these results by a mathematical pro- 

 cess, based on the same general principles as the graphical 

 one I have employed. 



XXIV. On Structures in an Earthquake Country. By John 

 Perry and W. E. Ayrton, Professors in the Imperial Col- 

 lege of Engineering, Japan*. 



IN a country like Japan, where several sharp earthquakes 

 occur yearly, where there are between three and four 

 hundred destructive earthquakes on record, and where over a 

 hundred thousand people are said to have been killed in one 

 almost continued earthquake lasting for a month, and which 

 occurred so recently as 1855, the question of the stability of 

 structure is all-important. 



When working at our paper " On a neglected Principle that 

 may be employed in Earthquake Measurements," read before 

 the Asiatic Society of Japan on the 23rd of May, 1877, and 

 which appeared in the Number of the Philosophical Magazine 

 for July 1879, we were led to consider how the effect produced 

 by an earthquake on a structure is influenced by the time of 

 vibration of the structure. 



It follows from that principle that if a number of quickly 

 vibrating bodies form part of the same structure, they all 

 vibrate in much the same way; that is, the periods of their 

 swings are all approximately equal to one another and equal 

 to the periods of the earthquake ; and although they differ in 

 the amount of their motions, these amounts and their differ- 

 ences are all exceedingly small ; whereas if one or more of the 

 parts of the structure are only capable of vibrating slowly, the 

 periods of vibration of the different parts vary very much, the 

 amounts of the motions are all comparatively great, and their dif- 

 ferences are all relatively considerable. If, however, there is a 

 sufficiently great viscous resistance to motion of such slowly vi- 

 brating parts, these parts will be found during an earthquake to 

 behave much as if their natural periods of vibration were quick. 

 Supposing the foundation of a structure to vibrate with the 

 earth which encloses it, we see that a slowly vibrating struc- 

 ture which is fastened to these foundations is during an earth- 

 quake subjected to stresses which may be excessively great 

 and of a very complicated kind, whereas a quickly vibrating 

 structure is subjected to stresses which may be said to be de- 



* Communicated by the Authors. 



