190 REPORT— 1897. 



3. Observations by Prof. G. Vicentini, at Padua, February 1 to August 29, 1895. 



4. Catalogues of Prof. P. Tacchini, January 1895 to October 16, 1896. 



5. Observations at Shide, Isle of Wight, by John Milne, August 19, 1895, to May 

 1897. 



Fracture of Cables in Deep Ocemis. 

 The times of earthquakes are given in G.M.T. astronomical time. Noon = 24 hours. 



North Atlantic. — Through the kindness of an engineer, whose experi- 

 ence in the laying and repairing of cables has extended over many years, 

 I am enabled to give the dates at which various cables have become 

 ruptured, or been restored to working order. The only case of alteration 

 in depth which he noticed was during the repairs of November 1884, but 

 this was not great. It seemed as if the picked-up cable had to be pulled 

 from under a bank of earth which had slipped down from the eastern 

 slope of the Newfoundland Bank. 



The following is a table of North Atlantic cable-interruptions : — 



North-eaitern Slope of Flemish Cap.— (37° W. to 44° W. long.) July 1894 (about); 

 June 1888 (about) ; September 1889 ; September 1881 ; June 10, 1894* ; July 28, 

 4.40 A.M., 1885 ; April 18, 8 p.m., 1885 ; July 25, 8 A.M., 1887 ; June 1895. 



Near South-eastern Slope of the Newfoundland Bank. — (46° W. and 50° W. long.) 

 September 1887 (about) ; October 3, 9.15 p.m., 1884 ; October 4, 4.8 A.m., 1884 ; 

 October 4, 4 and 8 a.m., 1884 ; September 1889. 



An unfelt eairthquake was recorded, June 11, 7h. 22m., 1894, very strong at 

 Charkow. 



A striking feature connected with these Atlantic troubles is that 

 nearly all have occurred in deep water near to the base of the eastern 

 slope of the Flemish Cap, 330 miles from St. John's, Newfoundland, or 

 the south-eastern slope of the Newfoundland Bank. Off the Flemish Cap 

 in lat. 49° N. and long. 43° E. there is a .slope, in a distance of 60 miles, 

 from a depth of 708 fathoms to 2,400 fathoms, or 1 in 35. Another 

 slope, over a distance of 30 miles, is fi'om 275 to 1,946 fathoms, or 1 in 17. 

 Off the eastern side of the Newfoundland Bank, in a distance of 25 miles, 

 the depth changes from 27 to 1,300 fathoms, indicating a slope of 1 in 19. 



These slopes are all well within the limits at which from time to time 

 yielding, due to bradyseismical thrust or secular crush, should be expected ; 

 and the further a cable can be kept away from the scene of such action, if 

 we may judge from experience, the longer will be its life. 



In one case only has the cause of failure been attributed to a land- 

 slide, which it is just possible was caused by, or accompanied with, seismic 

 phenomena. A very significant fact is the case when three cables running 

 in parallel lines about 10 miles apart broke, at points nearly opposite to 

 «ach other, on the same straight lines. This was on October 4, 1884. 

 At first the accidents were attributed to the grapnel of a cable vessel, but 

 as no grappling was done then this hypothesis had to be abandoned. 

 Because three cables broke apparently at the same time in the same 

 locality, one inference is, that the cause resulting in rupture was common 

 to all, and this may have been a sudden cliange in the configuration of the 

 ocean bed. Such a change does not necessitate any alteration in depth, 

 such as could be detected by sounding, but either a landslip along a line 

 of considerable length or simply a line of fracture like that which was 

 suddenly formed along the Neo valley in Japan in 1891. 



When, on the American and English coasts, types of seismometers 

 which will record the unfelt movements of the earth's crust have been 



