ON SEISMOLOGICAL INVESTIGATION. 251 



time of the same, magnetic perturbations should be observable. In Japan 

 such appears to have been the case. 



The large sudden subterranean adjustments may not occur on the 

 average more than twenty times per year ; but if we attribute the smaller 

 earthquakes to similar activities, one of these may, on the average, take 

 place every half-hour ; and although none of these latter is likely to 

 produce an appreciable magnetic effect on the surface of our earth, their 

 cumulative effect after a sufficient interval of time, as representing a re- 

 arrangement and new condition of magnetic materials, might possibly 

 result in measurable changes in magnetic elements. 



VII. Sub-oceanic Changes. 



In Section 9 of the Report for 1897 (p. 181) it was stated that off' coast 

 lines there was a tendency for sediments and detritus derived from the 

 land accumulating under the influence of gravity to assume unstable 

 contours. That such contours had an existence was shown by reference to 

 soundings. By excessive deposition of sediments, the sub-oceanic escape 

 of waters from subterranean sources, the sudden release of waters backed 

 up in bays by gales, changes in the magnitude and direction of ocean 

 currents, and by sub-oceanic seismic and volcanic action, sudden and 

 extensive yieldings might take place along the faces of slopes in a critical 

 condition. That such sub-oceanic landslides had often taken place was 

 proved by an appeal to the experience of cable engineers, who often found 

 that cable interruptions were the result of their burial along lengths of 

 several miles, the materials covering the lost section;! having fallen from 

 the faces of slopes along the base of which the cables had been laid. 

 In a few instances it was noted that there had been a considerable 

 increase in the depth of the ocean along a line of slip. Many examples 

 were given where cable interruption accompanied an earthquake which 

 had a submarine origin, and therefore it may be presumed that it was the 

 earthquake which caused the landslide beneath the ocean, in the same 

 manner that severe earthquakes result in similar displacements of what are 

 probably much more stable surfaces on the land. 



It is believed that most of the deep-water cable interruptions on the 

 west side of South America are attributable to sub-oceanic activities of 

 this description, and it was shown that in the Mediterranean, off' the coast 

 of Java, and in other parts of the globe, we had from time to time evidences 

 of a very close relationship between seismic activity and the failure of 

 cables. 



The fact that earthquakes originating in deep water, as, for example, 

 at a depth of 4,000 fathoms off' the N.E. coast of Japan, have been 

 accompanied by a series of sea waves which may agitate an ocean for 

 24 hours tells us that there must have been a sudden sub-oceanic displace- 

 ment of a very large body of material, accompanying some form of brady- 

 seismical adjustment. 



Although the earthquakes which result from these sudden movements 

 may not be felt or be recordable on a coast at a distance of 200 or 300 

 miles from their origin, they may often be noted in the records obtained 

 from instruments which are capable of registering the slower movements 

 of the earth's surface at distances of many thousands of miles from their 

 origin. The object of the following table, the materials for which were 

 almost entirely gathered together by my friend Mr. INI. H. Gray, of 

 Silvertown, is to indicate the frequency of sub-oceanic disturbance, but by 

 no means to attribute more than a fractional portion of the same to 



