Electric Waves passing through Telegraph Lines. 327 



arm i strikes the key / and throws it from the stud m (which 

 is in connexion with the intermediate point c) on to the stud 

 n, which is connected with the galvanometer Gr and accumu- 

 lator A. The end b of the conductor is to earth E'. 



Through the kindness of Mr. Willoughby Smith, I have 

 been enabled to make some measurements with this appa- 

 ratus upon lengths of gutta-percha cable, lately manufac- 

 tured by the Telegraph Construction and Maintenance Com- 

 pany, Limited, for submersion in the Red Sea. Measurements 

 of this kind upon coiled cable containing iron in the covering- 

 are of less value than upon submerged cable ; but as the 

 amount of iron in the covering of this cable is comparatively 

 small, I do not think that the magnetoelectric currents set up 

 were of great strength *. I hope to be allowed the opportunity 

 of making similar measurements upon a submerged and dupli- 

 cated cable. In the mean time I give the results, by way 

 of illustrating the employment of the method rather than pro- 

 fessedly for their intrinsic value. 



Exp. 1. — The length ac (fig. 1) consisted of 308 knots of 

 coiled gutta-percha cable ; b c of 3000 ohms wire-resistance, the 

 end of which was to earth. The battery B was ten Minotto cells. 

 Excursions of the needle of a mirror-galvanometer were read 

 as follows, the index-hand (i) of the rotating disk being ad- 

 vanced ten or more divisions between each observation. 



Interval of time 



by disk. 

 0-001 second . . 



Excursion of galvanometer- 

 needle. 

 division. 



0-01 

 0-02 





• i „ 



2 divisions. 



0-03 

 0-04 

 0-05 





4 



7 



. 13 



7) 



0-06 

 0-07 

 0-08 



)1 



. 18 

 . 24 



. 30 





0-09 



)) ' 



. 34 



)) 



0-10 

 0-15 





. . 39 

 . . 57 





0-20 



5) 



. . 68 



5? 



* Mr. Culley, experimenting 

 and back, found " no trace of 



on a circuit from Lowestoft to Holland 

 induction from the neighbouring wires 

 through which the ordinary traffic was proceeding " (Journ. Soc. Teleg. 

 Eng. 1875, vol. iv. p. 64). This fact indicates that the effect of coiling 

 observed in some cables is not due to voltaic induction, but to magneto- 

 induction, and is to be ascribed to the mass of iron of the covering, which, 

 when the cable is coiled, acts like an electromagnet. Therefore a cable 

 with very little iron in its covering should not behave very differently 

 whether coiled or straight, whereas in a cable without iron the coiling 

 should make no difference. 



