420 On the Retardation of Electric Signals on Land Lines. 



the insulators (which will probably vary considerably according 

 to the weather), and possibly by the effects of polarization at the 

 points of support. (The value does not vary rapidly with the 

 variation of height above the ground.) The apparent capacity, 

 measured by the method M. Guillemin adopts, will also be 

 affected by the dynamic induction from wire to wire, the effect 

 of which will be to make the capacity appear larger. Consider- 

 able variations may therefore be expected to be observed between 

 different lines, even where a wire of the same diameter is used ; 

 and the state of the weather may sensibly affect this capacity. 



The number of signals which may be sent on a land line, of 

 the construction used in M. Guillemin's experiments, may be 

 deduced from the Table calculated by Professor Thomson, pub- 

 lished by the writer, and now given in an abridged form. 



Table V. 

 Period of dots in 

 function of tx,, Amplitude in percentage 



_ Jccl 2 j jqto of maximum current. 



3-0 a 2-97 



40 „ 6-31 



5-0,, 10-42 



6-0 „ 14-85 



7-0 „ ....... 19-67 



80 „ 24-42 



9-0 „ 29-11 



10-0,, . 33-68 



The first column represents the period of time occupied by two 

 equal contacts at the sending-end of a telegraph wire, the first 

 contact being made with the battery end, and the second with 

 earth. The time is expressed in seconds, and must be calculated 

 from u for any given conductor. The second column repre- 

 sents the total variation in the received current which will be 

 produced when a regular succession of contacts is sent at the 

 speed given in the first column. The maximum variation, i. e. 

 that produced bv a permanent contact with the battery, is called 

 100. 



The value u may be calculated for a given line, using for k 

 the resistance per metre of the conductor in B. A. measure, di- 

 vided by 9'656 x 10 9 , and using the values of c given in this 

 paper, i. e. from 0*15 to 0-25. As an example, if we take c=0'20, 

 then a on a line 500 kilometres long would be 0*00112 second. 

 By inspection of the Table, we find that a period of 10a is re- 

 quired to produce a variation in the received current equal to 

 one-third of the maximum current that would be received when 



