22 
MR. J. MUIR ON THE TEMPERING OF IRON 
Instead of doing this, the extended series of experiments described below, and 
illustrated by Diagram 6, were performed on a single sjDecimen, and the results 
obtained were corroborated by a few simple experiments, using other specimens. It 
was shown in this manner that a temperature as low as 300° C. could effect tempering 
or partial annealing, provided the material was sufficiently hardened by tensile 
overstrain. The harder the material was made, that is, the higher the yield-]3oint 
was raised l)y the preliminary stretching, the lower was the temperature which could 
be shown to have a tempering or annealing action ; and the higher the temperature 
employed, the greater was the tempering or annealing effect produced. 
The material employed to obtain Diagram No. 6 (winch shows the tempering of 
steel liardened l)y stretching) was the semi-mild steel whose analysis is given on a 
previous page, and whose elastic properties are illustrated by Diagrams 1 and 2. A 
specimen of this material was turned down (except at the ends) to a diameter of 
about 0’4 of an inch, and was then annealed by being heated to 820° C., and allowed 
to cool slowly. It was necessary to anneal the specimen, for otherwise the hardness 
produced by tensile overstrain would have been superposed on the hardness referred 
to above as having been produced in the process of manufacture. A 4-inch length 
was then marked off on the specimen by the aid of the marking instrument, the 
extensometer was attached, load was applied, and Curve No. 1 of Diagram 6 was 
plotted from the readings taken. The loading was continued until a yield-point was 
passed at 28 tons per sq. inch, 
Ptecovery from this first overstrain was effected by heating the sj^ecimen to over 
200° C., and allowing it to cool slowly. The specimen was then hardened still 
further by the two successive overstrains illustrated by Curves 2 and 3, Diagram 6. 
After recovery from the third overstrain it was known that the material should 
bear a load about 7 tons higher than the last overstraining load (7 tons l^eing the 
step between the yield-points shown by Curves 1, 2, and 3) ; that is, the material 
should bear a load of about 50 tons per sq. inch before a yield-point was passed. 
As, however, experience had shown that fracture was liable to occur when the 
material was stressed to this extent, no further hardening of the material was 
attempted.^' 
Curve No. 4, Diagram 6, is given in order to show that after recovery from the third 
overstrain (by heating to 270° C.) the material could bear a load of at least 48 tons 
per sq. inch without reaching a yield-point. As indicated above, j^rohably a load 
just under 50 tons could have been safely applied. Slight imperfection of elasticity 
is shown by Curve No. 4, hut tliis coidd not be got rid of by the ordinary means 
adopted to procure restoration of elasticity after overstrain. With this material 
there was always a slight departure from Hooke’s law before a yield-point was 
* Another specimen of the same rod, after annealing at 775° C., was fonnd to give yield-points at 
about 29, 37, 44, and 52 tons per square inch; but after the yielding at the last stress had just spread 
throughout the 4-inch length under test, a neck formed and fracture occurred. 
