HARDENED BY OVERSTRAIN. 
15 
were all taken from a rod of steel similar in quality to that used for Diagrams 1 and 
2, but of a smaller diameter."^ This rod, in the rough, was about fths of an inch in 
diameter, but some of the specimens after being slightly turned down, except at the 
ends, measured only about 0‘32 of an inch. 
Specimen A was tested in the condition as supplied, and the two curves marked A 
in Diagram No. 3 show that the primary yield-point has occurred at 36|- tons per 
sq. inch, and that after recovery from overstrain the yield-point has been raised by 
11 tons j)er sq. inch. 
The second specimen, B, was heated to 600° C. for a few minutes, slowly cooled, 
and then tested, with the result that creeping set in at the stress of 36 tons jDer 
sq. inch. The load was, however, immediately reduced, no time being allowed for 
the extension of the yield-point to take place. It had thus been shown, however, 
that 600° C. was sufficient to produce a very slight annealing action. This specimen, 
B, was again used to obtain the curves marked D in Diagram No. 3. 
Specimen C was heated to 650° C. and slowly cooled. On testing, a yield-point 
was obtained at 33|- tons per sq. inch, and the step by which the yield-point was 
raised after recovery from overstrain was found to be about 9|- tons per sq. inch. 
The curves marked D in Diagram 3 show that material which had been heated to 
700° C. gave a yield-point at about 31|- tons per sq. inch, and that the step by 
which the yield-point was raised by recovery from overstrain was about 9 tons per 
sq. inch. 
Specimen E was raised to 730° C., and the yield-point was found to be thereby 
lowered to 28^ tons per sq. inch and the step reduced to 8 tons; while specimen F, 
which was tested after being heated to 780° and slowly cooled, showed that 780° C. 
brought the primary yield-point to 24 tons per sq. inch, and reduced the step to 
7 tons per sq. inch. 
The extensions which occurred at the various yield-points shown in Diagram 3 are 
all marked within brackets in the diagram, and it will be noticed that the extensions 
at the yield-points obtained with any one specimen are ap^^roximately equal, while 
for different specimens these extensions become less and less as the primary yield- 
point is lowered and the step diminished in consequence of tempering or annealing at 
higher and highei* temperatures. In the following table there will be found the 
various data obtained from the experiments which have just been described. 
* Particulars of this material are given in a footnote on p. 11. 
