370 PROFESSOR J. A. EWING AND MR. W. ROSENHAIN 



casting, fig. 43. The second system of boundaries consists, like the first, in polygonal 

 markings, and has such an obvious relation to the first that it almost appears as thoiigh 

 the upper layer of crystalline grains were transparent, and that we were seeing their 

 lower edges. To decide as to which of the markings on the surface constituted the 

 true surface Ijoundaries, two methods were available : etching with an acid and slightly 

 straining the specimen so as to develop the slip-bands. The latter method is much 

 the more instructive, and its results are confirmed by the etching process. When a 

 specimen having this characteristic was strained, it was seen at once that some of the 

 apparent boundaries were consistently ignored by the slip-bands, the others being the 

 real junction lines of the grains. But the true nature of the pseudo-boundaries comes 

 out on examining them under a high power. Although under low powers there is no 

 obvious difference in definition between the genuine and pseudo-lxmndaries, under 

 greater magnifications it becomes impossible to focus the pseudo-boundaries at all 

 they are seen to be more or less ill-defined slopes or changes of level, whereas the real 

 Ixmndaries are sharply defined. In general the real boundaries show some 

 accumulation of gas bubbles along them, and they are never crossed by slip-bands. 

 The pseudo-boundaries are found to consist in small variations of level in the surfaces 

 of the grains in which they occur. Fig. 44 is a high-power photograph of a set of 

 real and pseudo-boundaries showing slip-bands. 



It will be noticed that on the two sides of a real boundary the slip-bands are 

 independent of one another, whereas the slip-bands cross a pseudo-boundary with 

 only a slight change of inclination, which is to be ascribed to the fact that the 

 surface under examination is not a true plane. There is a slight slope on each side 

 of the pseudo-boundary, and the lines are consequently more or less inclined to one 

 another. Again, as we have noticed in other examples, the slope is not as a rule 

 constant and hence the lines are slightly curved. 



An explanation of this appearance of pseudo-boundaries is, we think, to be found 

 in the strains set up by contraction on cooling. If we suppose the outer layer of 

 crystals to cool more rapidly than the inner ones, the resulting contraction will drive 

 the projecting edges of the lower layer into the outside grains and thus cause slight 

 local deformation, which will pi*oject itself on the surface, probably by means of twin 

 bands running through the grains and appearing on the surface. The effect 

 resembles that of a Japanese "magic" mirror, in which slight inequalities of the 

 surface, corresponding to a pattern behind, cause light reflected from the mirror to 

 produce an image in which a ghost of the pattern may be traced. 



The foregoing conclusions refer to experiments on pure or nearly pure metals. We 

 have also examined the effects of strain on various alloys. The micro-structure of 

 alloys has received attention at the hands of most of the workers already named, 

 especially BEHRENS, CHARPY, GUILLEMIN, OSMOND, ROBERTS-AUSTEN, and STEAD. 

 Our observations have been directed towards supplementing theirs, in respect particu- 

 larly of the effects of strain. 



