1 6 THE ADDRESSES, LECTURES, ETC., OF 



all the results obtained. Bat the conductivity of gutta-percha 

 varies in an extraordinary ratio with change of temperature, as will 

 be seen by the diagram, Fig. 1, Plate 1, in which the abscissas repre- 

 sent temperatures, and the ordinates the corresponding electrical 

 resistance ; it also varies by electrification of the covering, accord- 

 ing to the length of time the current has been active, in the ratio 

 represented in diagram, Fig. 3, Plate 1. In order to obtain 

 standard tests, it was necessary to have them all taken at a uni- 

 form temperature, which was fixed at 75 Fahr., being the highest 

 temperature to which the cable was likely to be exposed when laid 

 in a tropical sea. 



For this purpose the core to be tested was immersed for twenty- 

 four hours in water-cisterns, which were kept at the standard 

 temperature, when each coil of a mile in length was required to 

 show a gutta-percha resistance of not less than 90 millions of 

 units. The coils of core were then transferred to Reid's pressure 

 tanks, and again tested at the standard temperature, and under a 

 pressure of 600 pounds per square inch. I stated before that in- 

 crease of pressure ought to increase the electrical resistance of gutta- 

 percha, and accordingly it was found that there was an increase of 

 something like 20 per cent, each time when the pressure was 

 applied, and unless that increase took place, it was inferred that 

 the gutta-percha coating was not as perfect as it ought to be. 

 Each coil was tested separately, and those that did not fulfil the 

 conditions insisted upon, were put aside. The actual resistance 

 obtained in each coil carrying a distinctive number was marked 

 against it, and it was then sent to the cable works. There the cable 

 was tested as each coil was added to its length, and it was required 

 that the total resistance of the whole should be equal to the calcu- 

 lated total resistance of all the parts of which it was composed. 

 There correction had to be made for temperature, because it was 

 not found possible to heat the tanks into which the finished cable 

 was received, to the standard temperature. A very perfect core 

 was thus obtained, and the result has proved that though the 

 external covering may be faulty, being an ordinary iron sheathing, 

 the insulation has never given any trouble, and I believe it is now 

 as good as it was at the beginning. The principal instruments 

 used were Professor Wheatstone's bridge arrangement, suitably 

 modified, and Dubois' galvanometer, which is so delicate as to 



