FRACTURE OF .METALS UNDER REPEATED ALTERNATIONS OF STRESS. 
247 
is mainly confined to the neighboiniiood of the crack, other crystals (as was the case 
with that illustrated in figs. 9-12) changing hut slightly. 
Fig. 14 (Plate 8) shows a different part of the same specimen (after it had broken), also 
under a magnification of 1000 diameters. The broad band in the middle of the crystal 
is a crack which has develoj^ed along what was originadly a line of slip. The heaving-up 
of the surface along the edges of the crack is well marked and may he compared 
with similar appearances at the edges of slip-bands in other parts of the photograph. 
The stresses which are stated here are in all cases calculated from the observed 
load, as measured by the spring balance, acting at tlie end of })rojecting beam or 
“ cantilever,” and they are the values which (on the ordinary theory of bending) are 
reached at a place close to the clamp. It was observed, ho^vever, that the 
destructive effects of reversals were not confined to the metal immediately adjacent 
to the clamps, but extended in most cases for a considerable way towards the loaded 
end of the specimen. The development of slip-bands, and their gradually widening 
and final conveision into cracks, occurred in some cases at least half-an-inch from the 
clamp, at a place where the fixing of the specimen could not disturb the distribution 
of stress in any way. 
In another experiment the load was such as to produce a maximum stress, close to 
the clamp, of 9’2 tons per sq. inch, and 800,000 reversals were given. It was then 
seen that the greater number (though not all) of those crystals which closely adjoined 
the place where the specimen was clamped showed signs of repeated slip. Further 
away from the clam23 the slij^-lines became less numerous ; but they were plainly seen 
on individual crystals as much as half-an-inch from the clamp. At the most distant 
places where slip-lines were plainly apparent the stress was only 7 "3 tons per sq. inch. 
It was clear that a stress no greater than this was sufficient to develop slips, under 
many reversals, and that the lines so produced became accentiiated as the process 
went on. 
This was confirmed by another experiment in which the maximum stress, close to 
the clamp, was only 6’9 tons per sq. inch. After 3,000,000 reversals of this stress 
one slip-hand was observed on a crystal a little way from tlie clamp. This is shown 
in fig. 15, where the slip-band is seen in the hi'oadened condition which resulted from 
3,000,000 reversals of stress. Prior to this, the same specimen had suffered 1,000,000 
reversals of a stress of 5'3 tons per sq. inch, without showing the smallest sign of 
damage. It was only after increasing the stress to 6 '9 tons per sq. incli that any action 
became apparent. It is an open question whether an isolated slip-band such as tliis 
would have led to fracture, if the process of reversal had been continued. At the 
conclusion of the experiment it was still confined to one crystal and it did not even 
extend all the way across that. 
We have noticed that when lines indicating slip appear during reversals of a 
comparatively small stress, they are generally to be found in the central parts of 
individual crystals, not extending to the boundaries of the crystal. 
