August 2, 1901.] 



SCIENCE. 



179 



duluni observations of earthquakes during 1898 

 and 1899, together with a detailed description 

 of the horizontal pendulum seismographs used 

 and some discussion of the results. 



In the 'compounrl,' or park, at Hongo, 

 Tokyo, in which the buildings of the Japanese 

 Imperial University are located, there has 

 been erected an 'earthquake-proof house in 

 which two of the large horizontal pendulums 

 are placed. They are of rather unusual dimen- 

 sions, the vertical distance from the point of 

 suspension to the point of support being nearly 

 three meters. The pendulum is excellently 

 planned for stability and convenience of adjust- 

 ment, and is capable of a vibration period of as 

 long as three minutes, being, therefore, ex- 

 tremely sensitive. When adjusted to a period of 

 two minutes it will show an index displacement 

 of about 1 mm. for a change of level of about 

 one two-hundredth of a second. Two of these 

 seismographs, exactly alike, are mounted, one 

 to give the N. and S. component of the disturb- 

 ance and the other E. and W. component. 

 They are designed, of course, for the registra- 

 tion of very feeble but often long-continued 

 disturbances whose seismic origin is remote 

 from the point of observation. Concerning the 

 transmission of such disturbances many ex- 

 tremely interesting observations have been 

 made, only a few of the most important of 

 which can be referred to here. 



Broadly speaking, the motion of an earth- 

 quake may be divided into three successive 

 stages, which are always more or less well de- 

 fined ; the preliminary tremor, consisting usually 

 of vibrations of small amplitude and short 

 period ; the principal portion, the active and 

 often destructive part of the earthquake, during 

 which the amplitude is much greater and the 

 period usually longer ; the end portion, which 

 is the feeble, small amplitude finishing of the dis- 

 turbance. The first of these is again usually quite 

 sharply divided into two stages, which may be 

 designated as the first preliminary tremor and 

 the second preliminary tremor, and the principal 

 portion is also made up of three tolerably dis- 

 tinct phases, of which the last, designated as 

 the quick-period phase, is of the greatest impor- 

 tance in the present discussion. 



Several seismologists have already pointed 



out the fact that the duration of the preliminary 

 tremor of an earthquake increases with the dis- 

 tance of the seismic center from the place of 

 observation, and this proposition is not only 

 confirmed by Dr. Omori's observations, but 

 from them and some others he has been able 

 to deduce an equation which represents the 

 relation between duration and distance with 

 astonishing closeness. Sometimes the total 

 duration of these disturbances, due to distant 

 earthquakes, is as long as four hours, but tlie 

 equation here given is based upon first pre- 

 liminary tremors varying in duration from five 

 minutes to eleven minutes. It is as follows : 

 x = 17.1y — 1360 



in which x is distance in kilometres (on the 

 surface of the earth along the arc of a great 

 circle passing through the two points) and y is 

 the duration of the first preliminary tremor in 

 seconds. This equation is based on long-dis- 

 tance observations of several earthquakes, three 

 of which, occurring respectively in Alaska, 

 Smyrna and Japan were observed in Japan, 

 while others originating in Japan were observed 

 in Potsdam and Italy. The agreement of the 

 calculated values of x with the actual values is 

 shown in the following table : 



Actual Calculiited 



Edrthqunke. Distaace. Distance. 



Alasl^a, 6,100 6,500 



Smvrua, 4.800 4,560 



Java 9,200 9,240 



Japan, 8,999 8,840 



Japan, 9,580 9,540 



Perhaps the most interesting subject dis- 

 cussed in these volumes is the transmission ve- 

 locities of the various phases of an earthquake 

 disturbance. Dr. Omori has given much atten- 

 tion to this important question, considering it 

 in considerable detail in the light of his own 

 and other recent observations. It has long 

 been known that all seismic waves did not 

 travel with the same velocity, and it was long 

 ago suspected that different phases of the same 

 disturbance might be transmitted at very dif- 

 ferent speeds. During the past few years the 

 development of 'long-distance seismology,' 

 as it may be called, has added greatly to our 

 knowledge of velocity of transmission, and at 

 the same time it has materially increased the 

 difficulties by which the subject is surrounded. 



