ON SEISMOLOGICAL INVESTIGATION. 191 



established, it seems likely that the cause of cable interruptions may be 

 better understood. Because the fifteen repairs indicated in the previous 

 table possibly cost half a million sterling, the advisability of localising 

 areas that should be avoided, and that we should be able to attribute 

 effects to their real cause, are evidently desiderata of great importance. 



St. Louis — Fernando Noronlia. — From a paper read at the Institution of Electrical 

 Engineers by Mr. H. Benest, A.M.Inst.C.E., ' On some repairs to the South American 

 Company's cables off Cape Verde in 1893 and 1895,' it seems that the St. Louis — 

 Fernando Noronha cable has been twice broken. The first break occurred on 

 December 26, 1892, about 130 miles from St. Louis du Senegal, in a depth of 1,220 

 fathoms, at the time of a heavy gale. The tape covering for 140 fathoms was rubbed 

 bare to the sheathing wires, but on one side only. The sheathing wires at the break 

 were drawn out as if they had been broken in a testing-machine. The Fernando side 

 of the break also showed the e.fect^ of rubbing, and the character of the fracture 

 was similar to the other end. In picking up these two ends there was at first a strain 

 in one case not exceeding 26 tons, and the other of 4 tons ; but as the ends were 

 approached this rose to about 6 tons, when the cable evidently cleared itself from 

 some obstruction, and came easily on board. 



Although we have here evidence of what may possibly have been a 

 submarine landslip, I am not aware that at that time any disturbance was 

 noted in Europe. 



The second date is March 10, 1895. Here, again, great difficulty was 

 experienced in breaking out the cable from beneath the mud, detritus, or 

 whatever the materials were that had covered it. The position of this 

 break was about 20 miles south-west from that of 1893. 



On March 5, at 22 hours G.M.T., a very large unfelt disturbance was 

 recorded in Europe, and one of moderate intensity at several places in 

 Italy on May 10, at 10.4 p.m. 



Mr. Benest holds the opinion that these fractures are connected with 

 submarine river outlets and gully formations in the ocean beds. The 

 gradients in the vicinity of the fractures vary from 1 in 34 (1° 30') to 

 1 in 7 (8°). 



Pemanibnco—Cape Verde.— To the north-west of St. Paul's (lat. 2° 41' 45" N., 

 and long. 30° 29' 15" W.), which is a volcanic centre, two cables broke simul- 

 taneously in a depth of 1,675 fathoms, indicating that the rupture was due to a 

 widespread cause. This was on September 21, 1893. Here, in the deep ocean, this 

 was the only failure in nineteen years. 



Madras — Peiiang and Aden — Bomlay. — These interruptions are referred to on 

 pp. 198, 199. 



hiterruptions to Cables on or near to Sub-oceanic Continental slopes. 



West Coast of Central and South Ainerica. — As illustrative of conditions 

 which may exist round many parts of tlie west coast of South America, 

 where there have been sudden and gradual upliftings of the land within 

 historical time, a portion of a chart showing contours near to the mouth 

 of the river Esmeralda is reproduced. The soundings are in fathoms. 

 Those in ordinary figures are from information received prior to June 

 1895, whilst those in larger type are from soundings taken in March 1896. 

 Changes from 13 or 20 fathoms to upwards of 200 fathoms in this short 

 interval of time are certainly remarkable ; and as the position of the cable- 

 repairing vessel ' Relay,' belonging to the Central and South American 

 Telegraph Company, which made the observations, was ensured by cross- 

 bearings on the land, their general accuracy cannot be doubted. 



The figures surrounded by a circle were taken many years ago, and 



