438 Mr. J. Hopps on the Alterations in the ' 



one in the other, the signs became reversed, and this continued 

 throughout the remaining series of the two wires. 



On page 74, vol. vi. of the i Proceedings of the Physical 

 Society,' I find a paper by Mr. Herbert Tomlinson, in which 

 he says that " the behaviour of nickel is especially worthy of 

 notice, because in it longitudinal traction, when not carried 

 beyond a certain limit, diminishes the electrical resistance in 

 spite of the increase of length and diminution of section which 

 ensues." 



I have not been able to obtain wires in nickel, but, accord- 

 ing to the results obtained, some samples of magnesium appear 

 to have the same property. 



At the last moment Prof. Stocker has shown me that in 

 Die Electricitat, by Wiedemann, it is noted that " the coiling 

 of a copper or iron wire increases, whilst the uncoiling dimi- 

 nishes the specific resistance about 0*003 of the total value of 

 the same." 



And, again, in Die Physik, vol. iii., by Mousson, that " the 

 coiling of a wire upon a cylinder slightly increases the resist- 

 ance, in consequence, as is supposed, of tension; uncoiling 

 diminishes it again. Under similar circumstances, the changes 

 amounted to 0*0032 in the case of copper and to 0056 in 

 iron, of their total resistances." 



It will be observed that both these descriptions are contrary 

 to what I must conclude from my own experiments. 



Having made some inquiries from a friend of mine who 

 had charge of the testing department in the cable factory of 

 J. B. Pirelli, F. Casassa et C ie , of Milan, he has supplied 

 me with data which show that the specific conductivity of the 

 copper conductors have fallen in many cases 1J per cent, 

 during manufacture, and when encased in lead tubes they 

 have often fallen over 2 per cent. In all cases the necessary 

 corrections have been made for increase of length and dimi- 

 nution of section due to strain in passing through the covering 

 machines. Possibly the greater loss in the lead-encased core 

 may be due to a damping-effect from the lead, but the bulk 

 of the increase of resistance must be due to some other cause. 

 These cables were manufactured under very strictly enforced 

 conditions for the Italian State Telegraph-stores ; and the tests 

 taken by government officials, ten weeks after they had been 

 sent from the factory, were in all cases within 0*2 per cent, 

 of the final tests in the factory, showing conclusively that the 

 changes were in a great measure permanent. 



I have therefore come to a very decided conclusion that the 

 changes in the resistance are not altogether the resultof changes 

 in form, and that their cause must be sought for elsewhere. 



