328 Professor John Milne [Feb. 12, 



lish in a suitable district a triangular arrangement of three sets of 

 levels, the distance between each set being several miles. All the 

 instruments should be on the rock, and displacements parallel and at 

 right angles to the dip should be recorded. 



A summary of all the advances which have of late years been 

 made in the study of earthquakes would, in great measure, be found in 

 an epitome of the twenty volumes which since 1880 have been published 

 by the Seismological Society of Japan, a work which is being 

 actively continued by a committee supported by the Japanese 

 Government. 



Previous to 1878 our knowledge of the charActer of earthquake 

 motion was largely dependent upon the effects snch motion produced 

 upon various bodies and upon our senses. To correct and extend 

 this knowledge, students of earthquakes in Japan at about this time 

 devoted nearly their whole attention to seismometry, first testing pre- 

 existing forms of apparatus, and then experimenting with forms 

 which were new. Those instruments which were intended to record 

 the rapid and violent movements of the ground, whether these were 

 in a vertical or horizontal direction, did this relatively to a mass so 

 suspended that, although its suj)ports v/ere moved, a point in this 

 mass remained practically at rest. An account of these seismograj^hs 

 was in 1888 given to this Institution by Prof. J. A. Ewing, F.R.8. 



For earthquakes in which there was a vertical component of motion, 

 however, it was soon noticed tliat these " steady points " were swung 

 from side to side by tilting, and instruments had then to be devised 

 to measure angular displacements. Following these came a class of 

 instruments intended to record the slow undulatory and often unfelt 

 earthquake motions. These, together with a group of tromometers or 

 tremor measurers — apparatus to record the time at which shocks had 

 occurred — resulted in the development of a group of instruments 

 which would require for their description a volume on Seismometry, 

 and it is fair to say that the seismometry of Japan revolutionised the 

 seismometry of the world. 



After the new inventions, the story of which forms one of the 

 most important in Japanese seismology, records were obtained which 

 showed that the impressions we had with regard to earthquake move- 

 ments had been widely incorrect, whilst they also indicated that our 

 estimates in mechanical units of seismic destructivity had been 

 founded on a wrong hypothesis. Having given the dimensions of a 

 body that has been overturned, or the dimensions and tensile strength 

 of a wall or column-like structure that has been shattered, we are 

 now in a position to calculate the acceleration to which the same has 

 been subjected, and the result arrived at is not far removed from 

 calculations of the same quantity derived from the diagrams obtained 

 at the same time and at the same place from a seismograph. Inves- 

 tigations of this description have been applied with marked success to 

 construction, and as new engineering works and new buildings spring 

 up in Japan, we see that rules and formulae are followed which are 



