170 Mr. L. Schwendler on Testing Telegraph Cables 



is an unknown function of the time during which the battery 

 circuit is closed and of the length, and also an unknown func- 

 tion of the temperature. We cannot, therefore, yet calculate 

 the insulation of the whole from its different parts until we are 

 in possession of these unknown functions. Thus at the sheath- 

 ing-works, where the length of the cable increases daily and the 

 temperature may also vary considerably, we are not enabled to 

 judge whether the insulation be normal or not, in the same easy 

 manner as we decided the question for the conductor. How- 

 ever, for some cables the dependence of insulation resistance on 

 temperature has been determined empirically*; and with this 

 function, and insulation-tests giving the maximum insulation 

 for each length of cable, it would of course be possible to cal- 

 culate the required insulation of the whole cable from its differ- 

 ent parts reduced to a standard temperature. But such a cal- 

 culation is always tedious ; and maximum tests are required, 

 which, especially for long cables, are not practicable at the 

 sheathing-works. Besides this, I question if the calculation 

 would be as exact as the high sensibility of the present testing- 

 instruments and the great importance of the case itself require. 



Here the electrician at the sheathing-works finds himself in a 

 very unsatisfactory position. He has all materials collected with 

 the greatest care during the manufacture of the core, he has all 

 necessary means to measure even the smallest change in the insu- 

 lation of the cable, but he is not able to use all this in such a 

 way that he can say, without an elaborate calculation, at any 

 moment during the sheathing-process, that the insulation of the 

 cable is normal, i. e. corresponding with the measured conditions 

 of its component parts. 



On the other hand, we have a test for continuity during the 

 sheathing-process by putting the conductor in circuit with an 

 alarm which gives a signal as soon as the conductor breaks ; 

 while for insulation the only method which I have met with is 

 that suggested by Robert Sabine f — to place a fixed point at some 

 place on the galvanometer-scale, and let the pointer close the cir- 

 cuit of a delicate relay and battery whenever the deflection ex- 

 ceeds a certain amount. But the chief objection which I have 

 to this is, that it cannot be applied simultaneously with the conti- 

 nuity-test. 



* Experiments made by Messrs. Bright and Clark upon four coils of the 

 insulated core destined for the Persian Gulf cable gave the empirical 

 formula 



nearly the same empirical curve as C. W. Siemens found for the Malta- 

 Alexandria cable. 



t The 'Electric Telegraph,' by Robert Sabine, page 404. 



