Glass, and Metal Wires subjected to Constant Pull. 529 



the stretch no farther crystals arc broken, and that after this 

 the resistance to the load is provided by ordinary viscous 

 force between the faces of the cloven crystals according to 

 the law which sometimes obtains in very viscous liquids, that 

 the viscous force is proportional to the square of the velocity, 

 This would suggest that even in pure metals there may be a 

 non-crystalline matrix in which the crystals are embedded, 

 and that this matrix acting as a very viscous liquid is re- 

 sponsible for this law of stretching. In the Phil. Mag. Oct. 

 1904, Prof. Trouton and Mr. Rankine have published a 

 research in which they find this logarithmic law of decay of 

 stress and strain in the case of lead wires. They did not 

 require to anneal in any way, probably because any internal 

 strain would decay very rapidly in such a soft metal. In the 

 case of the metals which I used the logarithmic law did not 

 obtain unless they had been annealed. 



Prof. Trouton and Mr. Rankine also found that after about 

 the first hour the stretching became a linear function of the 

 time, and this also is probably due to their using such a soft 

 metal. 



In the case of gold, I found that after about a day the 

 stretch approximated to a linear function of the time, but of 

 course the linear stretch was a very much smaller proportion 

 of the logarithmic stretch than was found with lead. The 

 copper, platinum, and silver wires made no attempt towards 

 approximation to a linear function, even after two days or 

 more, but when the stretch became so slow that it was difficult 

 to measure, there were considerable deviations from the 

 logarithmic law. 



There do not appear to be any absolute values of the stretch 

 given in their paper, so that I cannot compare the size of the 

 stretch in lead wires with that in copper, platinum, silver, 

 and gold. 



The}' also indicate that the assumptions which one usually 

 makes with regard to the stretching of solid bodies do not 

 give the stretch as a logarithmic but as an exponential function 

 of the time, and it seems difficult to make any reasonable 

 assumption which shall give the experimental law. 



The Slow Stretch '/// I run and Steel Wires. 



In the case of iron and steel wires a bundle of each kind 

 of wire was annealed for nie in one of the annealing furnaces 

 at Messrs. Perry cV Co.'s pen work.-, but in all other respects 

 the wires were experimented upon in exactly the same way 

 a- the other wires. The results, however, are entirely 

 different. As before, there seems to be no slow stretching 



