552 Mr. G. A. Shakespear on an Interference-Method for 



Another effect to be noticed is that, in the case of steel and 

 brass at least, after the weight had been left on for some hours, 

 in the succeeding observations the " up " numbers (i. e. con- 

 traction on removing load) exceeded the "down" numbers 

 (i. e. extension on putting on load), and after the weight had 

 been left off for a length of time the u down " numbers were 

 greater than the " up,' v as though in the former case the wires 

 were gradually recovering and in the latter case gradually 

 yielding slightly. It may be remarked here that no gradual 

 yielding was observable when the load was kept on. (The 

 load was about 2 kilos to each No. 20 b.w.g. wire.) In the 

 case of copper the " up " readings showed a tendency to be 

 greater than the " down " throughout. No such tendency 

 was observed either way in the soft iron. 



If the weight was left on for some time the modulus was, in 

 general, temporarily affected — decreased in the case of copper, 

 increased in iron ; but in the latter metal if the load was left 

 on at 100° 0. for several hours and then observations made, 

 and a rest of about half an hour allowed with the weight off, 

 on resuming observations there seemed to be a tendency, after 

 the first two or three extensions, to a sudden decrease of the 

 modulus of about 1 per cent, (this occurred on several occa- 

 sions). Generally, at each change of temperature eight or 

 ten observations were made, of which the above quoted figures 

 are examples. 



During the delay caused by waiting for the construction of 

 the heating-apparatus, some experiments were made on various 

 magnetic effects on iron wires. The form of apparatus above 

 described was modified for the purpose. Three similar solenoids 

 were made, one for each of the three wires, and an electric 

 current was sent through them in series. 



Observations on the effect of a magnetic field on the elasti- 

 city were not quite satisfactory (these are being repeated under 

 more favourable conditions at the present time); there seemed, 

 however, to be an undoubted decrease in the modulus on 

 putting on the magnetic field. 



Other experiments were made, at some length, on the 

 changes in length produced by magnetization in iron and 

 steel wires under different stresses. 



For these experiments the wires were 75 cm. long, and 

 the solenoids extended all their length. The number of turns 

 in each solenoid was 13U0 and the resistance 1*21 ohm ; 

 a current of one ampere would thus produce a field of 

 about 23 c.G.s. units. The interference-bands in this case 

 were got by transmitted light, the upper plate being supported 

 by a tube from the upper disk. 



