January 20, 1898] 



NA TURE 



273 



the same rate of 2 or 3 km. per second over paths which 

 are long or short. The prehminary vibrations, in all 

 likelihood representing waves of compression, appear to 

 have velocities of propagation closely related to the length 

 of the path as measured on an arc over or across which 

 they have been propagated. Such figures as we have, 

 which are open to correction after observations have 

 been extended, are easily remembered. Starting with a 

 velocity of 2 km. per second for a path of 2000 km., 

 then for a path of 4000 km. the velocity would be 

 4 km. per second, from Japan to England, or 9000 

 km., the velocity would be 9 km. per second ; whilst, if 

 we may carry conclusions beyond the limits of observa- 

 tion, we may imagine that the velocity of propagation 

 between an origin and its antipodes may be 18 km. per 

 second. As Mr. J.-Larmor pointed out, if our records 

 are correct, then outside our first 2000 km. coseismal, 

 great disturbances will be recorded at all points on the 

 surface of the world at approximately the same instant. 

 Undoubtedly velocities of 9 and 10 km. per second have 

 been noted, which indicate that the preliminary vibra- 

 tions, at least, can not have been transmitted round the 

 heterogeneous and broken materials constituting the 

 earth's crust, but rather, that the movements have passed 

 along direct, or by refraction along curvilinear paths 

 through the interior of our earth, This fact, taken in 

 conjunction with the fact that velocities of transmission 

 increase with an increase in the length of the wave 

 path, and that at any given station two sets of vibration 

 separated by a time interval proportional to the difiference 

 in the great circle distances between an epicentre and 

 the point of observation have not been observed, tend 

 to strengthen this same view. 



Rules connected with the transmission of preliminary 

 tremors which apparently promise to throw new light on 

 the physical condition of the interior of the earth are— 



(i) The velocity in kilometres per second with which these 

 movements are propagated is equal to one-quarter of the square 

 root of the mean depth of the chord or path (in kilometres) along 

 which we may suppose they travelled. 



(2) The duration of preliminary tremors, or the interval of 

 time expressed in minutes by which they outrace the longer 

 period waves (as shown on a seismogram), equals the square 

 root of the mean depth of the supposed wave path expressed in 

 kilometres. 



These rules, taken in conjunction with a map showing 

 the surface configuration of the globe, enable an individual 

 observer not only to locate an origin but also to determine 

 the time at which an earthquake originated. 



Inasmuch as a Committee of the British Association 

 have lent their weight to the establishment of a number 

 of horizontal pendulums at various observatories around 

 the world, within the next few years it is likely that our 

 present knowledge of the seismic breathings of the 

 earth's crust will be placed on a more extended and 

 accurate basis than that on which it at present rests ; 

 and, amongst other things, we shall have accumulated 

 new facts bearing upon the efifective rigidity of our 

 planet, which is evidently greater than that usually 

 assumed. 



The practical outcome of this work is already many- 

 sided. Observers at magnetic observatories, like those 

 at Kew, Potsdam and Batavia, are now aware that 

 greater or less disturbances in the uniformity of their 

 records sometimes accompany unfelt earth waves. Be- 

 cause the earth waves do not always leave a record of 

 their occurrence, whilst there are magnetographs of the 

 Kew type, as, for example, at Greenwich, Falmouth and 

 Stonyhurst, which, so far as 1 can learn, are but very 

 rarely disturbed, we can not say with certainty that the 

 movements of magnetometers are altogether the eflfect 

 of mechanical disturbance. At about 5 a.m. on Sep- 

 tember 21, very marked movements were recorded by 

 the magnetographs in Batavia, evidently the result of 



NO. 1473. VOL. 57] 



a movement recorded in the Isle of Wight (see Fig. 8). 

 Because the preliminary tremors have a duration of 

 forty-three minutes, the distance of their origin from the 

 south of England is probably about 112°. This, together 

 with the fact that on this date a " slight tremor " was 

 felt along the coast of North Borneo, makes it toler- 

 ably certain that the origin of this convulsion was one 

 or two hundred miles from that island. Strange to say, 

 at 7 p m. on the previous evening a disturbance, in all 

 its main features identical with the one here illustrated, 

 is also to be seen on the Isle of Wight seismograms, 

 from which it may be inferred that we have the records 

 of two earthquakes from the same origin. 



Assuming this to be the case then, so far as we car> 

 judge from information received from Batavia, it was 

 only the second of these unfelt motions that caused 

 disturbances in the magnetographs at that place. Not 

 only are magnetographs affected at the time when huge 

 earth waves slowly move the areas on which they are 

 situated, but in Japan similar instruments situated near 

 to an earthquake centrum have been perturbed some 

 days before the occurrence of a world-shaking shock. 

 Then, again, there are the remarkable secular changes in 

 declination and "3ip observed between 1880 and 1885, 

 at several eastern stations, respecting which Captain E. 

 W. Creak, F.R S., gives the following notes :— 



Bombay. — Until 1883-85 the needle was moving east- 

 wards. It then stopped, since which it has been moving 

 westwards at an increasing rate. In 188 1 there was. 

 a sudden change in the dip, and the needle is now going 

 down. 



Hong Kong. — Until 1875 the needle was moving east- 

 wards. Then there was a rest until 1880, when it 

 turned westwards. The dip was upwards until about 

 1880, since which it has turned downwards. 



Batavia. — Until 1884 the needle was moving east- 

 wards, when it became stationary. It is now moving" 

 westwards. The dip was moderately upwards until 1881, 

 but it has now greatly increased. 



When it is remembered that these remarkable changes 

 took place about the time of large earthquakes in Japan, 

 as, for example, that which led to the inauguration of 

 the Seismological Society in 1880, the Krakatoa eruption 

 of 1883, and that the illustrations of more immediate 

 possible connections between seismic and magnetic 

 phenomena may be multiplied, we recognise that the 

 seismic survey of the world may not only throw new 

 light upon its physical condition, but also, perhaps, it 

 may lead to inferences respecting gravitational re- 

 arrangements of external materials and internal magmas. 

 The sites of these, at present, hypothetical hypogenic 

 changes we should expect to find in districts where 

 secular movements and superficial loading, due to sedi- 

 mentation, are most pronounced. Because earthquakes 

 are apparently the results of critical conditions in these 

 processes, the records from horizontal pendulums tell 

 us that the sites we search for are to be found submerged 

 beneath deep water at or near the base of steeply- 

 sloping continental areas. The enormous size of the 

 superficial displacements which accompany certain of 

 these suboceanic changes is indicated by the creation 

 of sea-waves, which have often agitated the whole of the 

 Pacific for a period of one or two days ; and if these 

 rearrangements of material on the bed of an ocean are 

 related to changes in the state of stress or accelerations 

 in the movement of an internal quasi rigid matter in- 

 fluenced by continental load, we see in the operations 

 which culminate as earthquakes, causes which should, at 

 least, have a local magnetic influence. 



Captain Creak, writing on " the general bearing of 

 magnetic observations," in Science Progress., April 1896, 

 says : "It may be remarked in passing that a remarkable 

 alteration in the amount of secular change has been 

 noticed in the declination and inclination at the following 



