HARDENED BY OVERSTRAIN. 
5 
action travels along the bar.^ It was usually found with unturned specimens that 
the yielding started in the grips of the testing machine, and spread upwards or 
downwards as the case might be. Sometimes yielding was observed to begin at both 
ends and to travel towards the centre of the bar. 
The amount of stretching which occurred at the yield-point shown in Curve 1, 
Diagram 1 was very considerable, the specimen having been given a permanent 
extension of 0'13 of an inch on a 4-inch length. The horizontal part of Curve No. 1 
would thus require to be continued for over 8 feet in order to represent this yielding 
to the scale employed in the diagram. 
After Curve No. 1 had been determined and the load removed, the reduced 
diameter of the specimen was measured and the new area of section was calculated. 
A 4-inch length was again marked off by means of the marking instrument, the 
extensometer was readjusted, and the load was reapplied in tons per sq. inch of 
actual section.! Extensometer readings were taken after the addition of every 
4 tons per sq. inch, and Curve No. 2 was plotted from these readings. This curve 
shows the semi-plastic nature of the material immediately after overstrain. 
It was noticed in previous experiments on the recovery from overstrain, that 
different steels, after tensile overstrain, recovered their elasticity at very different 
rades ; so it was decided to examine the rate at which the material in question 
recovered from the semi-plastic condition illustrated l)y Curve 2, Diagram 1. The 
specimen was simply allowed to rest and was tested at intervals. Curves Nos. 3 
and 4 illustrate the progress made towards recovery of elasticity, one and three- 
quarter days, and two weeks after overstrain, respectively. 
In order to effect perfect recovery of elasticity, the specimen was put in the gas 
furnace described above and was heated until the pyrometer recorded about 200° C. 
It is probable that a few minutes at the temperture of boiling water woidd have 
been nearly thougli perhajjs not quite as effective in restoring the lost elasticity.;]; 
After cooling, the specimen was tested by reloading and carefully increasing the 
load above its previous maximum amount. A well-defined yield-point was obtained 
(as is shown by Curve No. 5, Diagram 1) at the stress of 49 tons per sq. inch, the 
yield-point having been raised by the large step of 11 tons per sq. inch. The 
amount by which the material yielded at this second yield-point was, to the nearest 
yjoth of an inch, the same as that by which it yielded at the primaiy yield-jjoint, 
namely, by 0'13 of an inch on the 4-inch length. 
The material after this second overstrain was once more in the semi-plastic state, 
* Professor Ewing, in the paper referred to above, “ On Measurements of Small Strains,” &e., has 
already observed that yielding may begin near one end of the specimen and gradually spread throughout 
the length, ‘Roy. Soc. Proc.,’ vol. 58, p. 135, April, 1895. 
t This procedure was adopted throughout the course of the work. Whenever a yield-point had been 
passed, the specimen was re-measured and the loading altered to suit the new area of section. 
I “ On the Recovery of Iron from Overstrain,” ‘Phil. Trans.,’ A, vol. 193, p. 22, 1899. 
