245 
FRACTURE OF METALS UNDER REPEATED ALTERNATIONS OF STRESS. 
lilies appear, and that after many reversals they become accentuated and broaden in 
the manner that has been described. Tliere is, therefore, a strong presumption that 
reversals of a stress of 8, or even 7, tons would ultimately develop cracks in the same 
manner as tliey are developed by stresses of 9 tons per sq. inch and over. 
The destructive process, of which the above is a brief account, is illustrated in the 
accompanying micro-photographs (Plates 7-9). 
The first series show part of a specimen vdiich had been subjected to a stress the 
maximum value of vdiich, close to tlie grips, vms 14'3 tons per sq. inch. In this 
series the magnification is 150 diameters. Fig. 3 was taken after 5000 reversals had 
been given. A few of the crystals exhibit signs of slip, the slip-lines being still fairly 
fine and sharp. Two crystals in the photograph are seen to have yielded more tlian 
the others. Figs. 4 and 5 show the same set of crystals after 40,000 and GO,000 
reversals of stress res})ectively. In fig. 4 a good many more crystals show signs of 
slip, and the slip-lines vdiich appeared in fig. 3 are far more strongly defined. In 
some cases it vdll he seen that the lines have so broadened out as to run together and 
form dark patches on the surface of the crystals. In fig. 5 a still further breaking up 
of the surface by slip has occurred. At tliis stage, and probably earlier, some of the 
slip-hands have developed into small cracks. Such a crack is seen near the top ligfit- 
hand coiner of the figure, first of all close to a crystalline boundary, and then in steps 
across the next crystal. On the latter crystal two systems of slip-liands had formed 
at right angles to one another and at about 45° to the direction of stress. Only one 
system is clearly visible in the photograph, but it will be seen that the crack runs in 
steps along both. Fig. 5 shows not only the development of this crack, but also 
a general increase in width, length, and number of the slip-bands. This was 
practically the final stage, so far as this portion of the surface was concerned, for the 
specimen broke after another 10,000 reversals along a crack outside the field of these 
photographs. 
When viewed in this state it is not practicable to tell how many of the slip-bands 
have actually developed into cracks, but this is readily seen when the specimen is 
re-polished and re-etched. This treatment olditerates any ordinary slip-bands which 
are steps marking differences of surface level, Imt any cracks remain visiljle. The 
specimen which gave this series of photographs was accordingly re-polished sufliciently 
to clear away the slip-bands, and vns re-etched. This left the cracks alone visible, 
rather accentuated indeed, for the sides of the cracks are to some extent eaten away 
by the acid, and hence the width of the crack is increased. Fig. G shoves the same 
part of the specimen as figs. 3, 4, and 5, after re-polishing and re-etching. The slip- 
bands have disappeared, except wliere they have formed cracks. A careful comparison 
of this vfith fig. 3 will show vdiere cracks have formed. The most conspicuous crack 
is at AA, and its zig-zag character as it follows the two directions of slip-bands across 
one of the crystals is specially noticealfie. 
The specimen illustiated in the next photograpli, fig. 7, had been subjectetl to 
