202 REPORT— 1897. 



blished, the number of interruptions Avhich have occurred have been at 

 least 245. For certain lines it would appear that fractures were more 

 frequent at one season than at others, and that therefore a proper analysis 

 of the table or its parts — such, for example, as those to which earthquake 

 statistics have been subjected — might lead to the discovery of periodicities 

 in cable-interruptions. Unfortunately, because the material in our 

 possession is yet so meagre, such discussions must for the present be 

 reserved. 



Out of the 245 breaks, 87 of them, each marked with an asterisk, 

 occurred at the times when instruments were in operation which would 

 record unfelt earthquake effects. Fifty-eight of the 87 cable-interrup- 

 tions occurred at or about the times when Europe was agitated by these 

 unfelt movements. The fractures accompanying earthquake, or, as it is 

 sometimes called, volcanic movement — which could be felt, and which 

 in two instances caused destruction on neighbouring shores — were at 

 least 10 in number, which may be raised to 24 by including the Java 

 records. In three of these instances, two or three cables were broken 

 simultaneously. With the latter the submarine dislocations extended 

 over a wide area ; in the Gulf of Corinth great changes in ocean 

 depth were brought about, and from this latter place we know the motion 

 to have radiated so that a few minutes after the interruption well-defined 

 diagrams of earth- waves were obtained at localities 1,000 miles distant, 

 at places where no movement could be felt. 



Instances like the latter clearly establish a connection between cable- 

 interruptions, earthquake-motion which has been felt, submarine disloca- 

 tion, and the records of horizontal pendulums in distant localities. This 

 being the case, and because earthquake-motion cannot be felt at great 

 distances from its origin, it is reasonable to conclude that the records of 

 unfelt earthquakes which approximately coincide in time to those at which 

 cables have been interrupted may sometimes indicate that submarine 

 geological changes have accompanied seismic efforts. 



Although certain conclusions arrived at in this paper are definite, 

 until the materials necessary for analysis can be obtained, others remain 

 matters of inference. The records of interruptions for the lines men- 

 tioned are, we have reason to believe, incomplete. The horizontal 

 pendulum records with which to make comparisons have not only been 

 few in number, but, because they are confined to Europe, could only be 

 expected to throw light upon disturbances originating at a great distance, 

 which were exceptionally large. The records of earthquakes which have 

 been felt ai-e confined to an imperfect list for Java, a few from the Medi- 

 terranean, and a few reported from the west coast of South America. 

 Lastly, the hours, and in some cases even the days, on which cable- 

 interruptions have taken place, together with the probable cause of these 

 interruptions, are unknown. These latter facts are no doubt to be found 

 in the archives of many cable companies, and it would be to the interest 

 of all who desire to increase our knowledge of sub-oceanic change if com- 

 parisons could be made between the records of unfelt earthquakes now 

 published, and the times and circumstances at and under which corre- 

 sponding cable-ruptures have taken place.' 



' The writer, whose address is Shide Hill House, Newport, I.W., England, would 

 be glad to receive any information respecting the day, hour, and probable causes of 

 failure, connected with cable-interruption. 



