ON SOME MARVELS IN TELEGRAPHY. 245 



ously connected as shown by the cross-lines in Fig. 3. (It 

 will be understood that a d and b c do not touch each other 

 where they cross.) The current will now flow from p along 

 a d to B, circuiting round the needle n in a contrary direction 

 to that in which it flowed in the former case, returning by 

 b c to N. The upper end of the needle is deflected then to 

 the left while the current continues to flow along this course. 

 I need not here describe the mechanical devices by 

 which the connection at a b and c d can be instantly changed 

 so that the current may flow either along a b and d c, as in 

 Fig. 2, circuiting the needle in one direction, or along a d 

 and b c, as in Fig. 3, circuiting the needle in the other direc- 

 tion. As I said at the outset, this paper is not intended to 

 deal with details of construction, only to describe the genera) 

 principles of telegraphic communication, and especially those 

 points which have to be explained in order that recent in- 

 ventions may be understood. The reader will see that nothing 

 can be easier than so to arrange matters that, by turning a 

 handle, either (i), a b and cdmay be connected, or, (2), ad 

 and c b, or, (3), both lines of communication interrupted. 

 The mechanism for effecting this is called a commutator. 



Two points remain, however, to be explained : First, A 

 must be a receiving station as well as a transmitting station ; 

 secondly, the wire connecting B with N, in Figs. 2 and 3, can 

 be dispensed with, for it is found that if at B the wire is 

 carried down to a large metal plate placed some depth 

 underground, while the wire at c is carried down to another 

 plate similarly buried underground, the circuit is completed 

 even better than along such a return wire as is shown in the 

 figures. The earth either acts the part of a return wire, or 

 else, by continually carrying off the electricity, allows the 

 current to flow continuously along the single wire. We 

 may compare the current carried along the complete wire 

 circuit, to water circulating in a pipe extending continu- 

 ously from a reservoir to a distance and back again to 

 the reservoir. Water sucked up continuously at one end 

 could be carried through the pipe so long as it was continu- 



