HARDENED BY OVERSTRAIN. 
21 
Specimen A of Diagram 5 gave an ultimate strength of 31 tons, with an elongation 
of 14i per cent, on 4 inches. The virgin material tested in the usual fashion, that is, 
without effecting recoveries from overstrain, broke at 23 tons, with an elongation of 
27 or 28 per cent, on 4 inches. 
Distinctions between various Hardnesses. 
The examinations of both iron and steel which have been described above show 
that thin bars of material in the condition as supplied by the makers are more or less 
hard ; abnormally high yield-points are exhibited, ultimate strengths are somewhat 
higher, and ultimate elongations somewhat lower than those obtained with material 
which has been annealed. Tliis hardness is probably due to conditions of manufacture. 
The bars may leave the rolling mills at a confparatively low temperature (say not 
higher than 600° C.), and so he subjected, in passing through the rolls, to a species of 
overstraining at a moderately high temperature. The material may also be suddenly 
cooled through some range of temperature more or less high. It has been shown 
above (Diagram 3) that this hardness of thin rods of iron and steel, when in the 
condition as supplied by the makers, may be gradually reduced by the process of 
tempering or gradual annealing. 
The hardness just referred to is not such as coidd be produced by simple tensile 
overstrain, that is, by giving the material a permanent stretch; nor is it such as is 
produced in steel by quenching in cold water. For, as compared with the projDerties 
of annealed material, the hardness of thin rods in the condition as supplied is 
characterised by a high primary yield-jDoint, by the large amount of stretching which 
occurs at a yield-point, and by the large stej) through which the yield-point is raised 
after recovery from tensile overstrain. This is shown by the conqiarison of Diagrams 1 
and 2, or of Diagrams 4 and 5. Material which has been hardened by simple 
stretching is characterised by a high primary yield-point, but by no change (as 
compared with the properties of annealed material) in the amount of yielding at the 
yield-point or in the step by which tlie yield-point is raised after recovery from tensile 
overstrain. Steel which has been hardened by quenching exhibits no distinct yield- 
point, and the structure of quenched steel, as revealed by the microscope, is entirely 
different in character from tliat of annealed steel. 
The Tempering of Steel Hardened by Stretching. 
Perhaps the simplest method of showing the tempering of steel hardened by 
stretching would have been to have taken a series of annealed specimens from the 
same rod of steel, to have hardened them all to the same extent by tensile overstrain, 
and then to have tried the effect of heating to various temperatures, a different 
temperature being used for each specimen, just as was done to obtain Diagram No. 3- 
