186 REPORTS ON THE STATE OF SCIENCE.—1913. 
certainly possess ‘natural’ elastic limits when it has been subjected to a 
number of repetitions of stress of approximately the amount of the limiting 
range ; and it has been shown by Bairstow that a limit exists above which 
the tension elastic limit cannot be raised, so long as the stress is entirely 
removed in each cycle. The piece may possess natural elastic limits when 
the process of overstrain and recovery attempted by Bauschinger, and 
carried out with more (though not complete) success by Bairstow, has been 
applied to it. It has not yet been proved that the natural elastic limits 
for equal + stress cycles are the same as would be found in static tension 
and compression by the use of an exceedingly delicate extensometer on the 
piece in its primitive state. The question has not been definitely settled 
whether, when the primitive elastic limits have been altered, and the granu- 
lar structure distorted, by cold working, the natural elastic limits will or 
will not remain the same; though it is certain that the part of the f, n 
curve for small values of n will be made to fall above the corresponding 
part of the curve for the unworked stuff. It appears from the Table, 
Appendix II., that the annealing to which the specimens of published 
repeated stress tests have been subjected produces, in general, some little 
lowering of the ‘natural’ elastic limits; though the primitive elastic 
limits may be very much altered by this heat treatment. 
There appears to be no definite relation between the ‘ natural’ elastic 
limits and either the primitive elastic limit, the yield point or the ultimate 
tensile stress. 
Recovery of Elasticity. 
It is a well-known fact, discovered by Weber in 1835, that ‘ when a body 
is strained beyond the elastic limit and is set free, part of the strain 
disappears at once, and the strain that does not disappear gradually 
dimmishes. The body never returns to its primitive condition, and the 
ultimate deformation is the permanent set; the part of the strain that 
disappears is called elastic after-strain.’* In the case of the metals of 
engineering construction, the immediate re-application of the stress after 
such overstrain shows the metal to be in an imperfectly elastic state ; but 
if the stress be re-applied after a considerable period of rest, during which 
the elastic after-strain disappears, the elasticity is found to be restored. 
The period of rest may be shortened to one of a few minutes only, if the 
temperature be raised to 100° C.; presumably the elastic after-strain dis- 
appears in this short interval, though, so far as the writer is aware, this has 
not been verified experimentally. After this recovery the elastic limit is 
somewhat higher than at the first overstrainig. Provided the recovery 
is complete, further exposure to this temperature, or to considerably higher 
temperature in the case of many metals, produces no further effect ; 
and the additional exposure has no more effect than on a piece of the 
unstrained material. 
It may be noticed here that the limit of proportionality of wrought iron 
is practically the same at 0° C. and 250° C., there being some little varia- 
tion between these temperatures with a maximum about 200° C.+ 
It has been proved that, in general, non-elastic strain is effected by 
cleavage plane slipping in the crystalline grains. The parts of a crystal 
not immediately contiguous to the slipped surfaces are, so far as can be 
* Quoted from Love’s Theory oj Elasticity. 
+ A. Martens, Proc. Inst. C.E., vol. civ. 
