;o6 



NA TURE 



[July 26, i960 



is found to cover not more than 1,000,000 square miles ; while 

 if the shock of June 12, 1897, is treated in a similar manner, 

 we obtain a total seismic area of over 1,750,000 square miles. 



Owing to the paucity of good records, the. course of the 

 isoseists could not be traced in detail. The outermost isoseist 

 was, however, determined with approximate accuracy for about 

 half of its circumference. The seismic area presents a pecu- 

 liarity in that there is a detached area in the alluvium about 

 Ahmedabad over which the shock was felt, though it was unfelt 

 over a tract of about one hundred miles separating this alluvial 

 area from the furthest limit at which the shock was felt on rock. 

 It is also reported to have been felt at Burhanpur, on the border 

 of the Tapti valley alluvium, though it was felt nowhere else in 

 the neighbourhood. Outside the area over which it was felt 

 there are records, in India, of the passage of the earthquake 

 wave as indicated by the swinging of lamps, &c. 



Apart from the records in India, there is good evidence that 

 it was felt in Italy ; the observers at Catania, Leghorn and 

 Spinea di Mestre all record having felt a slight shock at the 

 exact time when the instruments throughout Italy recorded the 

 advent of the first phase of the disturbance due to this earth- 

 quake. Had there been only a single record, it might have 

 been attributed to a distinct local shock ; but these three sepa- 

 rate records, all agreeing with each other in time, and also 

 with the advent of the first tremors, which, having a ])eriod of 

 about '5 second, might have been sensible, leaves little possi- 

 bility ot doubt that the Indian earthquake was actually felt. The 

 observers are, however, to be complimented on their acuteness 

 of observation. 



The epifocal area is of a peculiar shape. Situated in Western 

 Assam and North-eastern Bengal, it is bounded on the south by 

 a straight line running about E.S.E, for some 200 miles ; on the 

 north it is bounded by a nearly symmetrical double sigmoid curve, 

 the maximum breadth being not less than 50 miles, and possibly 

 as much as twice this amount. Over the whole of this area 

 of not less than 6000 square miles, the intensity of the shock 

 was in excess of 10 degrees of the Rossi-Forel scale, and alter- 

 ations of level have taken place ; while for a year and more 

 afterwards earthquake shocks — some severe, but most feeble and 

 local — were very frequent. The changes of level were not only 

 shown by faults, one of which was traced for a distance of over 

 12 miles, and had in places a measured throw of over 30 feet, 

 and by differential changes of level, whereby streams were 

 dammed up into lakelets, but also by a remeasurement of some of 

 the triangles of the great Trigonometrical Survey. As the whole 

 of the triangles reobserved lay within the epifocal area, it is not 

 possible to say what amount of actual change has taken place ; 

 but changes of position of one hill relative to another were de- 

 termined, which reach as much as 24 feet in a vertical, and 

 12 feet in a horizontal, direction. 



. The results of the triangulation, as published by the Trigono- 

 metrical Survey of India, indicate an increase in the horizontal 

 distances between the stations ; but in the geological report it is 

 shown that this is probably due to a shortening, by compression, 

 of that side which was assumed as an unaltered base-line. The 

 true nature of the focus is regarded as a thrust plane, from which 

 minor faults branched off, and in places appeared as such, while 

 elsewhere they died out before reaching the surface and merely 

 caused those changes of level which, where other circumstances 

 were favourable, led to the formation of lakes. No less than 

 thirty of these were observed, the largest having a length of 

 \\ miles and a breadth of f mile, and the smallest a few yards 

 across ; the depth varied from i to over 20 feet. 



Within this epifocal area the violence was everywhere great, 

 though subject to great local increase in the neighbourhood of the 

 fault planes which extended upwards to the surface. Not only 

 were upright stones broken, but sound hardwood trees of a 

 diameter of 6 to 7 inches were snapped across by the violence of 

 the motion they were subjected tc ; no masonry building was 

 left standing, and the hill sides were scarred by landslips. In 

 many places it was noticed that stones lying on the ground had 

 been projected into and through the air. 



The acceleration necessary to cause the fracture of standing 

 monoliths, or sound hardwood trees, must have been great — 

 much greater than the measured accelerations, as determined by 

 West's formula from overturned tombstones, which range up to 

 32 feet per sec, per sec. It is doubtful, however, whether 

 West's formula is applicable to cases where the height of the over- 

 turned column is less than three or four times its diameter ; in 

 the earthquake of 1897 all the high accelerations were obtained 



NO. 1604, VOL. 62 ] 



in places where there must have been a large vertical component 

 in the wave motion, and the overthrow of squat pillars is 

 regarded as a modified form of projection. It is improbable 

 that accelerations of over 6 feet per sec. per sec. can occur, 

 except in the vicinity of the epicentre, where there is a consider- 

 able vertical component in the wave motion, and the excessive 

 accelerations which have been supposed to have been measured 

 in the case of other earthquakes must be regarded with suspicion. 

 Opportunity was taken to review the various formula for 

 deducting the accleration and velocity of movement of the wave- 

 particle ; these have been all collected in an appendix and dis- 

 cussed. ■ One result of the discussion is in a manner reactionary, 

 for the one quantity which it was believed could be determined 

 with real accuracy, the velocity as deduced from projection, is 

 shown not to be due to wave motion at all. The velocities 

 deduced from observed projections are shown to lead to im- 

 possible results if combined either with the deduced acclerations 

 or with any conceivable amplitude or period, and the conclusion 

 is come to that the projection of solid objects was due, not to 

 molecular wave motion, but to a molar displacement of the 

 ground, resulting in permanent changes of level. 



Instances of the rotation of objects, both within and without 

 the epicentre, were numerous. As many as possible of these were 

 carefully measured to determine, not only the angular rotation, 

 but also the direction and amount of displacement of the centre 

 of gravity. From a careful examination of the data, it is shown 

 that none of the attempts to explain rotation by simple rectilinear 

 niotion are in accordance with the observed facts, and that it is 

 necessary to accept the explanation of vorticose motion. This 

 vorticose motion does not, however, take the form of angular 

 rotation as has been assumed by some investigators, but the 

 whole ground either moves in a more or less circular track, or is 

 subject to a more or less rectilinear to-and-fro motion, whose 

 direction changes continually in azimuth. 



Over a large alluvial area the river channels were narrowed, 

 railway lines bent into sharp curves, and bridges compressed 

 and destroyed, much as in the Japanese earthquake of 1891. 

 This compression is shown to have been due, in all cases, to dis 

 placement of the superficial alluvium, and not to any general 

 compression. Over this same alluvial area fissures and sand 

 vents were opened in myriads. With regard to the fissures, it is 

 shown that Mallet's explanation of their formation by unsupported 

 masses of clay being thrown off" from free surfaces by their own 

 inertia is incomplete, and that they were formed in places where 

 no such action could have taken place. It is suggested that in 

 such cases the fissures were due to the visible surface undulations 

 which were noticed by many observers. The sand vents were 

 formed in such numbers that large areas were temporarily 

 flooded by the volumes of water which issued from them with 

 such force that it rose in solid columns to a height of 3 feet 

 and more from the ground, while splashes and spouts are said 

 to have reached 18 or 20 feet in height. It is noteworthy 

 that in several cases these sand vents are said to have been 

 formed after the passage of the shock, and flowed for a period 

 of half an hour or, according to some, several hours. This is 

 attributed to the settling of clay beds on to underlying quick- 

 sands, which supported the overlying beds as long as these were 

 continuous, but would not do so after they had been broken up 

 by the earthquake. 



Earthquake sounds were very loud and. conspicuous, but the 

 data available do not allow of much advance in this difficult 

 branch of seismology. In some cases explosive sounds of short 

 duration were heard after the earthquake had passed, and the 

 connection of these with the " Barisal guns," " mistpoeff'ers," 

 "marina" and other similar phenomena is discussed, all being 

 regarded as probably in the main seismic. 



The most important results obtained are probably those con- 

 nected with the rate of propagation. Numerous time observa- 

 tions in India yield a time curve with double curvature 

 like that of Schmidt's "hodograph," but the curvature is 

 too slight to accord with it, and the true time curve is 

 shown to be most probably a straight line indicating a uniform rate 

 of propagation of 3'o km. per sec. Turning from the observed 

 rate of propagation of the sensible shock to the distant records, 

 it is shown that the records of the Italian seismographs exhibit 

 three principal phases of motion, after each of which there 

 is a marked diminution of movement. The first of these gives 

 an average rate of propagation of 9 6 km. per sec, the second of 

 5-6 km. per sec, and the third, the phase of long period undu- 

 lation, accompanied by marked tilting of the ground, a rate of 



