1895.] Strains in the Testing of Materials and Structures. 133 



tion for about three minutes while this was happening, and when it 

 was removed there was a permanent extension of O'lO inch on the 

 marked length of 9 inches. 



The extensometer was immediately reset to the normal length of 

 9 inches, an operation which occupied two or three minutes more, 

 after which the bar was reloaded as follows : 



(3s.) Same piece after stretching by a load of 6 tons. 



Extensometer. Difference per ton. 



200 

 297 



Load in tons. 

 

 1 

 2 



' 3 

 4 



Unload 



Reload 



Unload 



97 



395 98 



497 102 



600 (Creeping) 



601 After 1 minute. 104 

 202 (Creeping back) 



200 After 1 minute. 



297 97 



395 98 



495 100 



598 103 



705 



707 After 20 seconds. 109 



767 



770 After 20 seconds. 



214 



210 After 3 minutes. 



208 After 45 minutes. 



These figures show how widely different are the elastic qualities of 

 an iron bar in its primitive annealed condition, and in the condition 

 in which it is put by overstrain. In the first condition there is 

 almost no evidence of creeping or " elastische nachwirkung " up to say 

 5 tons of load, and Hooke's law is nearly valid. Immediately 

 after an overstrain there are distinct evidences of creeping at lower 

 loads, and when these lower loads are removed, there is the same 

 kind of thing in the unloaded state, namely, an apparent set which 

 disappears in whole or in part when time is allowed for the creeping 

 to take effect. And further, after the overstrain there is no longer 

 nearly so close an agreement with Hooke's law ; a given increment 

 of load produces notably more strain at high loads than at low ones. 

 In loading and unloading there is now much hysteresis in the relation 

 of strain to stress. 



In the overstrained state the lowest loads produce not much more 

 strain than they did in the primitive state ; the value of Young's 



