SUBMARINE TELEGRAPHY 237 



city, gave certificates that seven or eight words per minute might 

 be sent along the new cables. New and complex instruments 

 were devised to insure even this result, and now eighteen or 

 twenty words per minute have been obtained ; with the omis- 

 sion of many of these inventions, twelve words per minute is 

 the fair average speed. Nevertheless, the engineers and elec- 

 tricians were not to blame. On the contrary, they deserve 

 praise for their moderation. To explain how their estimate was 

 formed, a sketch of the theory of the transmission of submarine 

 signals is required ; and here again Sir William Thomson must 

 be named as the first to state that theory, and draw the main 

 conclusions from it. In a letter to the ' Athenasum,' dated 

 November 1, 1857, Sir William Thomson pointed out, in oppo- 

 sition to Mr. AVhitehouse, then electrician to the Atlantic Tele- 

 graph Company, that the number of words which in a given 

 time could be sent through a long submarine cable varied in- 

 versely as the square of the length of that cable ; that when 

 the length of a cable was doubled, only one quarter the number 

 of messages per diem could be sent through it. In a paper 

 published in the ' Proceedings of the Royal Society,' Sir William 

 Thomson gave the complete theory, showing that on all lines a 

 limit existed to the speed of transmission, and giving an esti- 

 mate of the probable speed through the 1858 Atlantic Cable as 

 three words per minute. 



The speed of electricity used to be given as 288,000 miles 

 per second, but in reality Professor Wheatstone's beautiful ex- 

 periments only proved that this speed might in given circum- 

 stances be attained. Electricity seems to have no proper speed, 

 in the usual sense of the word. The speed depends in each case 

 on the condition of the conductor, 1 and may on certain conceiv- 



1 This fact, and the increased retardation observed in underground wires 

 and therefore in submarine cables, is guessed, or rather foreseen, in a very 

 curious proposal for an electric telegraph, by Francis Ronalds, published in 

 1823, containing an account of experiments made in 1816, long before the 

 days of Gauss or Cooke and Wheatstone, btfore even the discoveries of 

 Oersted and Ampere, which have rendered our present system of telegraphy 

 possible. The writer is indebted to Mr. Latimer Clark for the knowledge of 

 this fact. Mr. Ronalds' proposal, based on actual experiment throuo-h eight 

 miles of wire, deserves to be better known than it is. His book is called 

 Descriptions of an Electrical Telegraph, and was published for R. Hunter 72 

 St. Pauls Churchyard, in 123. 



