1905.] The Effects of Momentary Stresses in Metals. 505 



giving a tension momentarily exceeding 900 Iba., produces a 

 permanent extension of 1 /3500th part, or 1/30 of the ultimate extension 

 caused by a steady load of 700 Ibs. 



Copper Wire. This was 0-129 inch diameter, and of the kind used 

 in electric light cables. It was set up without preparation of any kind. 

 A load of 220 Ibs. stretched it by l/1000th part, corresponding to 

 E = 7500 tons per square inch. The wire weighed 0*0503 Ib. per foot. 

 The steady tension was 200 Ibs. Mass of falling weight (M) 0'945 Ib. ; 

 mass of stop, 0'023 Ib. Velocity of propagation of waves, o 

 = 1 1,800 f.s. 



With a fall of 12 '6 inches the additional extension observed on the 

 top 20 inches was 41 micrometer divisions. The calculated extension 

 (formula (2) above) is 43 divisions. The permanent extension pro : 

 duced by 20 blows was about 2 micrometer divisions. 



With a fall of 2 feet 6 inches the observed extension was 67 divisions. 

 The calculated extension is 70 J divisions. Ten such blows extend the 

 20 inches by 13'5 micrometer divisions, or 6*7 thousandths of an inch. 



The elasticity is therefore practically perfect up to the stresses 

 caused by a fall of 2 feet 6 inches. The greatest strain at the top end 

 caused by this blow is 2V/a = 2-10 thousandths. The tension due to 

 this is 440 Ibs., and the resultant tension, including the initial 200 Ibs., 

 is 640 Ibs. The mean tension on the top 20 inches is 570 Ibs. (calculated 

 from the observed extension of 67 divisions). 



Tested statically with the extensometer, this wire showed failure of 

 elasticity at 500 Ibs. With a load of 590 Ibs. it yielded rapidly and 

 finally broke. 



With a fall of 5 feet, and the same initial tension of 200 Ibs., the 

 observed extension on the top 20 inches was between 100 and 105 

 micrometer divisions, as against 103 calculated ; of these about 

 30 divisions were permanent. The calculated maximum tension in 

 this case (including the 200 Ibs.) is 890 Ibs. But the elasticity here is 

 far from perfect, and the actual stress is probably somewhat less than 

 the calculated value. 



The observed extensions are, in the case of the copper wire, about 

 5 per cent, less than the calculated. I think that this is more than can 

 be accounted for by errors of observation, or by such causes as friction 

 between wire and weight, especially having regard to the much closer 

 agreement in the other wires with which I have experimented. A 

 possible explanation is that in the copper wire the value of Young's 

 modulus for these extremely rapid extensions is 10 per cent, greater 

 than for slowly applied forces. The difference between the adiabatic 

 and isothermal elasticities as calculated from the coefficient of expansion 

 and Young's modulus, is not sufficient to account for the effect, which 

 must be a true time effect if it exists. 



The history of the stress in a section of the wire after one of these 



