that may be employed in Earthquake Measurements. 49 



of H J be, say, one fifth of that of the earthquake, then any 

 rotatory vibration of the earth will be well recorded*. 



The observations of Mr. Mallet, made at the scene of the 

 Neapolitan earthquake of 1857, are of great value in connexion 

 with the science of seismometry, which owes its growth in a 

 great measure to the labours of that gentleman. But we have 

 no hesitation in saying that three recording seismographs, such 

 as we have described, suitably placed in the plain of Yedo 

 and with clocks in telegraphic communication with one an- 

 other, would give more information regarding earthquakes in 

 a few months than could be obtained by the most experienced 

 observers from the remains of many destroyed cities. We are 

 aware of the great interest now being taken by the German 

 Asiatic Society in the subject of seismometry; and it is to a 

 certain extent in consequence of this that we have been led to 

 publish this paper. A not very extended series of experi- 

 ments would probably be all that would be required before we 

 could furnish working drawings of an almost perfect record- 

 ing instrument ; and after such instruments had been con- 

 structed the Japanese Government might possibly be induced 

 to allow them to be used at their telegraph offices. With very 

 little extra expense these seismometric records might be sup- 

 plemented by regular observations of the natural currents in 

 the telegraph-lines, to the importance of observing which we 

 have recently directed attention in a former paper f. 



Addition, May 1879. 



It is interesting to notice that the principles developed ma- 

 thematically in this paper have, since it was written, been 

 arrived at by natural selection in the relatively rigid naturally 



* Since writing this paper a rather sharp earthquake has been experi- 

 enced in Tokio, which caused the scale-pans of a balance in the Physical 

 Laboratory of the Imperial College of Engineers to describe perfect circles, 

 the chains (about 35 centimetres long) which supported the pans and the 

 pans themselves moving like a conical pendulum. The radius of the circle 

 described by each pan at the beginning was about 5 centimetres ; and the 

 motion continued for a long time after the earthquake had ceased. The 

 circular motion was probably produced by two successive shocks being 

 nearly at right angles to one another. 



Some time after this earthquake we constructed an exceedingly sensi- 

 tive rotatory seismograph on the general principle shown in tig. 7, the 

 wire K L being about ten feet long ; but up to the date of one of the authors 

 leaving Japan last year the instrument gave no evidence of any earthquake 

 producing a turning action by a single shock. 



f " The Importance of a General System of Simultaneous Observations 

 of Atmospheric Electricity," Transactions of the Asiatic Society of Japan, 

 vol. v. part 1, p. 131, and Journal of the Society of Telegraph Engineers, 

 vol. vi. p. 242. 



Phil. Mag. S. 5. Vol. 8. No. 46. July 1879. E 



