624 1'HE!<IDE^t's ADDREt>S 8ECT10-\ H. 



will have lost its tendency to easy slip. Slip can then only take 

 place on that surface by the application of a greater force, or it 

 must occur on other surfaces on which slip did not at first take 

 place so easily. The second reason advanced is that where severe 

 strain occurs slip takes place along more than one set of parallel 

 surfaces in each grain. The intersection of these causes the surfaces 

 to become stepped, and thus slipping cannot take place so readily as 

 before. 



To me none of the reasons that have been put foi'ward appear 

 satisfactoiy. Bound up with this question we have the remarkable 

 discovery of Bauschinger that whenever by tensile overstrain we 

 raise the elastic limit in tension we always simultaneously lower 

 the elastic limit in compression. If iron that has been stretched 

 beyond the yield point in tension be immediately subjected to com- 

 p]-e.ssion, before the sliding- surfaces have time to heal up, the 

 elastic limit in compression may be zero. But, even after the lapse 

 of time, or after the iron has been moderately warmed, so that the 

 slip surfaces are restored, the elastic limit in compression is always 

 considerably lower than it was before the tensile overstrain. This 

 means that once the grains have been distorted in one direction, and 

 have been allowed to settle down into their new shapes, it is harder 

 to distort them still more in that direction than it is to push them 

 back again towards their original form. One w^ould expect that 

 any explanation of the raising of the yield point by overstrain must 

 at the same time explain this phenomenon, but this is still wanting. 



In this connection the effect of temperature is interesting. Thus, 

 Dewar and Hadfield* have shown that ordinary soft ductile iron be^ 

 comes at the temperature of liquid air quite another substance, 

 brittle, and possessing little or no ductility, biit mth its tenacity 

 more than doubled. It seems reasonable to suppose that at this 

 temperature the molecules cannot get that freedom of motion 

 necessary to form the mobile layer at slip surfaces, and so ordinaiy 

 plastic flow cannot so readily take place. Thus, one specimen of ii'on 

 tested by Hadfield carried at ordinaiy temperatures a maximum 

 stress of 21 tons per square inch, with 25 per cent, elongation; at 

 — -182 degrees C. its tensile strength because 54 tons per square inch, 

 with elongation nil. On the other hand, steel at hig^h temperatures, 

 such as 750 degrees C, lias its yield point lowered, and it behaves 

 like indiarubber or glass, t The time effects we have previously 

 i^poken of become very marked. If it is stressed fo)- a time, and the 

 stress then removed, it does not at once recover, but. after the 

 immediate elastic recovery, there is a slow contraction perceptible 

 for many minutes. Such creeping can be barely detected at ordinary 

 temperatures, but at a red heat it attains a different order of mag- 

 nitude, becoming in its total amount a substantial fi'action of the 

 whole deformation. Such time effects are ascribed l)y Ewing to the 

 existence of gi'oups of molecules in a position of feeble stability. In 

 the region adjacent to the plane of slip it is probable that many 

 of the molecules will be swung round by the disturl)ance due to 

 strain, and may not swing right back again into their original place, 



* Jmirnal Iron and Steel Inst., I., 1905. 



t Hopkinson and Rogers, Proc. R.S., Vol. 7(5a, 190"). 



