ON TIIK Cl; V>T\I.I.INK STRUCTURE OF MI-TVLS. 



uniform slip occurs is not necessarily plane ; it may be the trace of a straight line 

 inn\iiiu r pandlfl t<> its.-lf: a line which is in the direction of slip and always lies 

 parallel to one of the cleavage planes. For convenience in argument we may for the 

 moment assume that the surl':ires <>!' easiest sliding are determined by some accident, 

 such as the presence there of minute layers of some impurity, such as occluded gig, 

 If, as the straightness of the slip-bands seems to indicate, these surfaces are true 

 planes in plastic metals like lead or gold (so far as each individual grain is concerned), 

 sliding might take place in two directions on each of these planes. But when sliding 

 has once taken place, the intersecting layers of impurity would no longer be distri- 

 buted over planes, and further sliding on transverse planes would necessitate the 

 starting of fresh slips in surfaces that had no special tendency to facilitate sliding. 

 This suggests how such a metal may be hardened by previous straining ; and how, 

 also, the slip-bands formed on re-straining a piece of metal hardened in this way 

 would be less straight and more liable to sudden steps and branches than those that 

 are found in the virgin material ; the slips would, as far as possible, follow the old 

 surfaces of easy sliding, but these would now be stepped instead of plane as Ixjfore. 



A striking proof of the persistence of crystalline structure m metals which have 

 been submitted to severe distortion is found in the existence of geometrical etched 

 pita. These are readily developed in sections cut from cold-rolled iron, and they differ 

 in no way from the etched pits developed in the virgin material ; like these, they 

 appear as similar and similarly oriented geometrical figures over the face of each 

 grain. Fig. 32 is the photograph of a group of crystalline grains in the specimen of cold- 

 rolled Swedish iron referred to above (cold-rolled from three-quarters of an inch to half 

 an inch diameter). Among them is a large grain (showing light in the figure) with an 

 outline so unlike those found in unstrained metal that its form is evidently due to 

 violent distortion in the process of rolling. The face of this grain is covered with 

 minute etched pits, and an examination of these under high powers shows that they 

 liave preserved their similarity of shape and orientation in spite of the violent dis- 

 tnrtiiin which the grain, as a whole, has undergone. Fig. 33 is a photograph under 

 800 diameters of a portion of the large grain in question, which appears near the 

 middle of fig. 32. 



From the various lines of evidence here indicated we conclude that the charac- 

 teristic crystalline structure of metals is not destroyed by strain, no matter how 

 severe, and tliat jilastie drt'<innat i< >n OMOM l>y DMMM t' -lips al..n^ th0 clwflgft Q> 

 gliding planes of the crystalline grains, the crystalline elements which build up each 

 grain remaining unaltered both as to shape and orientation. This statement, how- 

 ever, is subject to the following <|ualitioation. We have found in certain metals, 

 nutal>ly copper, gold, silver, lead, cadmium, tin, zinc, and nickel, that twin crystals are 

 liable to be developed by straining. Hence in such cases it is not exact to say that 

 straining produces no change in the orientation of the crystalline elements, for twin- 

 ning implies a rotation of one group of elements with respect to the rest through a 



