360 PROFESSOK J. A. EWING AND MR. W. ROSENHAIN 



the crystalline grains of metals are built up of crystalline elements which are similarly 

 oriented throughout the mass of each grain. 



The experiments now to he described were intended to throw light on the nature 

 of plastic strain in metals, and the results obtained are interpreted by the aid of the 

 theory of the crystalline structure of metals, of which an outline has been stated at 

 the beginning of this paper. Their complete agreement with that theory affords 

 further confirmation of its truth. 



It has long been known that when a specimen of iron or steel with a bright smitli 

 surface is strained sufficiently to give it permanent set the smoothness of the surface 

 is destroyed. A microscopic examination of the surface shows, as CHARPY has pointed 

 out, that the crystalline grains become visibly differentiated from one another by 

 straining, and the effect is in this respect not unlike that which etching would have 

 produced. There is, however, a further effect of straining, which will be descriljed 

 immediately. 



It is also well known that when a piece of severely -strained metal is polished and 

 etched, as for instance a bar which has had its section reduced by hammering or 

 rolling in the cold state, the crystalline grains are found to have changed their form. 

 They are lengthened in the direction in which the piece was extended and shortened 

 in the direction in which the section has contracted. Accordingly, a severely-strained 

 piece is readily recognised on polishing and etching it, on account of the shape of its 

 crystalline grains. In the strained specimen these are found to have a direction of 

 predominating length according to the direction of strain, while in the unstrained 

 specimen there is no direction of predominating length. Further, it is well known 

 that in a strained specimen elongated grains are not found after the strained state 

 has been relieved by annealing. The effect of rolling or hammering is often spoken 

 of as a conversion of the metal from a crystalline to a " fibrous " state. In the 

 present writers' view this language is misleading, and the physical conception under- 

 lying it is a mistaken one. 



In the first experiments on the effects of strain we aimed at keeping a particular 

 place on the polished and etched surface of the specimen under continuous observation 

 while the specimen was being strained. This was accomplished by constructing a 

 small straining machine which could be attached to the stage of the microscope, and 

 by which strips of sheet metal could be gradually extended until they broke. Fig. I 1 

 shows the arrangement which was used. The stress was applied by a screw which 

 could be turned by hand while the specimen was under observation, and any displace- 

 ment of the particular grains on which the microscope was focussed could be made 

 good by means of the traversing screws in the mechanical stage. With this apparatus 

 it was easy to keep the same group of crystalline grains under observation from the 

 first application of stress until the specimen was broken, and to obtain photographs of 

 the same group at any number of stages during the strain. The first specimens 

 observed in this way were strips of sheet iron, of the nearly pure kind supplied for 



