296 Messrs. J. H. Gray and J. B. Henderson. 



In this table the percentage stretchings are given instead of the 

 weights producing the stretching, as, copper being so soft, the effect 

 of the weights depends altogether on the times during which they 

 were applied, and as these were very variable, a record of the 

 weights would be of no value. For example, if a weight were put 

 on, and allowed to hang for three minutes, then taken off, and a 

 balance immediately found, and if the same weight be again put on, 

 left for another three minutes, and a balance again found, it would 

 be considerably different from before, showing that the wire had 

 received an additional stretching. For this reason it is advisable to 

 give the percentage stretching produced. In the case of steel wire it 

 was quite different, as, throughout all the series, till the wire was 

 just about to break, the elastic limit was not exceeded, so that the 

 wire attained its ultimate stretching as soon as the weight was ap- 

 plied. With regard to this table, the results given are one set out 

 of a great many which agreed very closely, T ^ per cent, being 

 the greatest variations for corresponding stretchings ; this set was 

 chosen as having one peculiarity very well marked. This is with 

 reference to the high value 1 '01083, that is, an increase of 1 per cent, 

 for a stretching of 22 '3 per cent. The wire was almost at breaking 

 point, and, on being examined, the surface was found to contain 

 numerous cracks^ many of which were so large that they could be 

 easily felt by the fingers. These cracks were no doubt caused by the 

 rapidity of stretching, as in other series in which the stretching was 

 conducted more slowly they were not so apparent, and the change of 

 specific resistance was not so great. 



It was found quite impossible to smoothe out all the little deviations 

 from perfect straightness in the copper wire. Although the first 

 14 Ibs. were sufficient to keep it fairly straight, there were still little 

 irregularities which could be seen with the eye. Not till about 56 Ibs. 

 were applied were these quite removed, and by that time there was 

 also considerable stretching. The result was to make the apparent 

 initial length shorter than it really was, and therefore the specific 

 resistance seemed to diminish at first with the addition of weight. 

 After the irregularities were taken out by stretching the length ap- 

 peared to become greater, while the resistance did not increase, the 

 result being that the specific resistance would seem to have dimin- 

 ished. This showed itself to a greater or less degree in all the series, 

 according to the condition of the wire as to straightness. The 

 values 0-99969, 0*99985 are clearly due to this. It might, therefore, 

 have been more accurate to give the results, taking as the initial 

 reading the lengths after the irregularities had been removed ; but 

 this would only very slightly alter the values, in fact by -^ per cent. 



Leaving out the value of the alteration just before breaking, wo 

 find in all our trials that the greatest increase of weight specific 



