303 



this system five or six distinct letters per second, or sixty words per 

 minute, may be readily delivered through air lines and submarine 

 lines up to 100 miles, or perhaps even considerably more, of length, 

 and recorded by a self-acting apparatus, which I shall describe in a 

 communication I hope to make to the Royal Society before its next 

 meeting. 



V. " On Practical Methods for Rapid Signalling by the Electric 

 Telegraph." (Second communication.) By Prof. W. 

 THOMSON, F.R.S. Received December 11, 1856. 



I. Further remarks on proposed method for great distances. 



Since my former communication on this subject I have worked out 

 the determination of operations performed at one extremity of a sub- 

 marine wire, so adjusted, that when the other extremity is kept con- 

 stantly uninsulated, the subsidence of the electricity in the wire 

 shall follow the triple harmonic law (that is to say, the electrical 

 potential shall ultimately vary along the wire in proportion to the 

 sine of the distance from either end, one-third of the length of the wire 

 being taken as 180). The condensation of the electrical pulse at the 

 receiving extremity, due to such operations, is of course considerably 

 greater than that which is obtained from operations leading only to 

 the double harmonic as described in my last communication ; but 

 experience will be necessary to test whether or not the precision of 

 adjustment in the operations required to obtain the advantages which 

 the theory indicates, can be attained in practice when so high a 

 degree of condensation is aimed at. The theory shows exactly what 

 amount and duration of residual charge in the wire would result 

 from stated deviations from perfect accuracy in the adjustments of 

 the operations ; but it cannot be known for certain, without actual 

 trial, within what limit such deviations can be kept in practice. 

 From Weber's experiments on the electric conductivity of copper, 

 and from measurements which I have made on specimens of the 

 cable now in process of manufacture for the Atlantic telegraph, I 

 think it highly probable that, with an alphabet of twenty letters, one 

 letter could be delivered every two seconds between Newfoundland 

 and Ireland (which would give, without any condensed code, six 



