HARDENED BY OVERSTRAIN. 
25 
sions which occurred at tlie various yield-points are also marked (within l)rackets) 
below the yield-points. These extensions implied of course a diminution in the 
diameter of the specimen, this was allowed for in the loading, the loading being 
always in tons per scp inch of actual section. 
The tests illustrated by Curves 12, 13, 14, 15, and 16 were made in order to try 
whether the method or the amount of the overstrain given to tlie specimen altered 
the temperature required to produce a certain annealing effect. The experiments 
seemed to slmw that a definite annealing temperature corresponded to a definite 
position of tlie yield-point, no matter how or to what the extent the material had 
been overstrained. 
After Curve No. 16 was obtained, the specimen was suljjected to a treatment 
which practically resulted in a repetition of Curves Nos. 13, 14, and 15, the tem¬ 
perature of 600^^ C. being still employed, and finally the specimen was heated to 
700° C., and Curve No. 17 obtained. The yield-point was now lowered to 30 tons 
per sq. inch, which is only 2 tons higher than the yield-point obtained when first 
testing this specimen after it had been annealed by heating to 820° C. 
Experiments were made with another specimen of the same steel, and these 
served to corroborate the results which have just been described and which are 
illustrated by Diagram No. 6. It was thus shown tliat steel hardened by successive 
tensile overstrains may be softened to a greater or less degree by heating to 
temperatures ranging between 350° C. and 750 or 800° C. ; and that the higher the 
temperature, the greater is the softening produced, or the lower is the stress to 
which the yield-point is brought. 
In order to show tempering produced by comparatively low temperatures, it is 
necessary to severely overstrain tlie steel, so as to bring it into a condition liaving 
a large elastic range. In Diagram No. 6 the lowest temperature shown to have 
had a tempering or annealing effect is 360° C., hut there is no doidit that a lowei' 
temperature than this could have been shown to have liad a tempering action, had 
the specimen in the first place lieen more severely overstrained. That it was 
possible to further harden this material is shown by Diagram No. 2, where a 
fourth yield-point was safely passed, and the material brought into the condition 
having an elastic range of from zero to almost 59 tons per sq. inch, instead of only 
to 50 tons as in Diagram 6. If then, as shown by Diagram 6, a temperature 
of 360° C. sufficed to lower the yield-point to 47 tons per sq. inch when the 
elastic range was from zero to 50 tons, it seems probable that temperatures even 
under 300° C. would suffice to produce a slight tempering or annealing efiect when 
the range of elasticity extended to 59 tons. It may he remarked, however, that 
the temperature of rather over 300° C. which was employed to effect tlie last 
restoration of elasticity illustrated in Diagram 2, is shown by Curve No. 8 of that 
diagram to have had very little if any annealing action. In this curve large 
yielding is shown to have started at 59 tons per sq. inch, and a yield-point would 
VOL. CXCVIII.— A. E 
