ON THE EARTHQUAKE PHENOMENA OF JAPAN. 243 



5. There has been no marked connection between the occurrence of 

 earthquakes and the position of the moon. 



6. Earthquakes have been 11-2 per cent, more numerous at low 

 water than at high water. It is frequently assumed that earthquakes are 

 more frequent at one time rather than at another. I have spent much 

 time m the tabulation of the earthquakes of Japan and other countries 

 comparing together the frequency of earthquakes at certain phases of the 

 moon, at particular seasons, during the day as compared with the night 

 relatively to the state of the barometer and other meteorological changes' 

 &c., with tne general result that there are no strongly marked periods 

 when earthquakes may be expected, the exceptions to rules which may 

 be formulated being almost as numerous as the cases which were the 

 foundation for the rules. The most marked rule about earthquakes is 

 that they chiefly occur during the cold months. 



7. With regard to the nature of earthquake motion as deduced from 

 the numerous diagrams which have been obtained, I cannot say that 

 they do more than confirm the results which I have already communi- 

 cated to the British Association. The greater number of shocks had a 

 duration of from twenty to sixty seconds, but some lasted more than four 

 minutes. 1 he duration recorded depended on the situation of the 

 observer, and on the nature of the instruments. Two observers, with 

 similar instruments, two or three hundred yards apart, might con- 

 siderably differ as to the length of time assigned for the duration of a 

 disturbance. If one observer was situated on a marsh whilst the other 

 was on hard ground, the former would record the longer time for the 

 duration ot sensible motion. 



An instrument with a large multiplying index, and sensitive to small 

 mt quick movements, will often commence to write a record before an 

 ^rument which has only a small multiplying power. Again, an instru- 

 ment with very little friction, and susceptible to very slow movements 

 •ill continue to write a record, after an instrument with considerable 

 friction has ceased to move Strictly speaking it would appear that 

 tne whole of an earthquake has never vet been recorded; many of the 

 preliminary tremors at the commencement of a disturbance and the slow 

 pulsations which bring a disturbance to a close being lost. 



I he preliminary tremors have an amplitude which is a small fraction 



of a millimeter, and a period of twenty-five seconds to sixteen seconds. 



I he tremors may be followed by a shock which consists of three or four 



v rapidly performed back and forth motions, having an amplitude of 



£m one to ten millimeters. The maximum acceleration during such a 



ZZtZ e l' C i Cn] ? e * on .* he assumption of harmonic motion, sometimes 



~ % I'- / red milhmeters P^ ^cond. Such a shock is on the 



Z g t i i ng - 1 f aT) S erous - In ordinary disturbances it is from ten to 



rreiukv ;^V m f rS Per ^^ After the shock we * et a serie * of 



moS »™ I 'J 6 ?/ P ! accom P amed h 7 other shocks. These irregular 

 Swi , 1 CafcUreS m ° rdinar ^ disturbances, and the tremors 



wKf may -,r e abSGnt - The maxim ™ amplitude recorded is from a 



ive seco° I?/ 111 1 1 " 16 ' 61 ' t0 ?** £ tw0 millimeters. The period is from 

 vrtrET, I ? Se "° UdS - Tbe direCti0n of motion of the * e ^regular 



conntZ Z SI aDtIy - Chan x? e f- The ^' d0 not a PP ear to ha ™ a "y direct 

 drrStio^ Z I h t ha t 7 WhlCh tbe ea rtMuake is being propagated. The 

 propagation ' howe7er » S ° CmS t0 C ° incide ™ fch the direction of 



e2 



