HARDENED BY OVERSTRAIN. 
23 
reached, and on the removal of a load less than that of the yield-point, a slight 
permanent set was often recorded. A second loading usually brought the material 
into a cyclic state, a hysteresis cycle such as that shown by this Curve No. 4 being 
obtained. 
The specimen under consideration was now heated to C., allowed to cool 
slowly and then tested. It was not expected that this temperature would have 
had any tempering or annealing effect, but Curve No. 5, Diagram 6, shows that 
the hardness of the material had been somewhat reduced, a yield-j^oint being reached 
at 47 tons per scp inch, that is, the position of the yield-point had been lowered 
by about 3 tons per sc[. inch as a consequence of the heating at 360° C. 
The test illustrated by Curve No. 5 involved of course a further hardening of the 
material, so to effect recovery from overstrain the specimen was heated to 310° C. 
The furnace had been kept hot, so the specimen was only in for alxjut 15 minutes 
before the gas was turned off, and the furnace and specimen were allowed to cool 
down together. On testing the specimen it was found that 52 tons per sq. inch 
could now be safely applied, and as a yield-point would be expected rather under 
54 tons (he., 47 + 7) tons per sq. inch if only recovery of elasticity had been 
effected, it follows that very little if any annealing action had been produced by 
raising the material to 310° C. The specimen was again raised to 310° C. and kept 
at that temperature for 2 hours. Curve No. 6 shows that still there had been no 
appreciable tempering or annealing action. A temperature very little higher tlian 
this would probably have produced a measurable lowering of the yield-point, hut in 
order to avoid danger of fracture the specimen was raised to 500° C., and Curve 
No. 7 shows the considerable lowering of the yield-point which resulted. At a load 
of 40 tons jjer sq. inch large yielding set in, and with 41 tons a yield-point was 
certainly passed. 
The specimen was now first heated to 350° C., and no appreciable annealing effect 
was found to have been produced. A temperature of 380° C. was then tried, and 
still it was found that the specimen could withstand a load of 46 tons per sq. inch 
without yielding. In consequence of overstrain and recovery from overstrain a 
yield-point would be expected at 48 {i.e., 41 + 7) tons per sq. inch. The specimen 
w'as then kept for one hour at a temperature from 380 to 390° C., cooled and 
tested with the same result as before. A temperature of 410° C. was then tried, 
and finally 430° C. was found to produce only very slight annealing, a yield-point 
occurring just under 46 tons per sq. inch. Curve No. 9, Diagram 6, illustrates 
this test. 
The subsequent history of this specimen will be understood by reference to 
Diagram No. 6. The temperatures marked at the top of the various curves in the 
diagram are those to which the specimen had been raised immediately preceding 
the test from which the curve in question was obtained. The specimen was of 
course always tested at the temperature of the laboratory. The permanent exten- 
