20 i< M. R. Thalen on the Determination of- 



manent elongations are in general small, passes then through 

 the point at which the curvature attains its maximum, and here- 

 upon is almost horizontal. At this latter part of the curve the 

 bar is most susceptible of loads. The previous tolerably stable 

 position of equilibrium of the particles appears now to be quite 

 displaced. If the bar has been so much lengthened that the 

 curve is convex as regards the abscissa axis, the new positions 

 of equilibrium are certainly more stable than they were before ; 

 but the smallest increase of load, even the same load, can dis- 

 turb anew the relative position of the particles, and therefore 

 produce permanent elongations. The curve which was first 

 convex is again concave ; it more and more endeavours to become 

 horizontal, and thus indicates that the permanent elongations 

 increase more and more until fracture ensues.., 



To this description, in its general features, of what takes place 

 in stretching might be added, that if, after the bar has been torn, 

 the stretching-experiments be continued with the pieces, the 

 curve continues in the old direction, provided blisters in the bars 

 did not cause the break*. If the curve on first breaking was as 

 good as horizontal, the new tearing cannot be specially greater 

 than that just observed, but the permanent elongation may of 

 course increase so much the more ; for it has been seen that if a 

 bar has been torn twice successively, and both times w r ith the same 

 load, the addition in the permanent elongation can amount in 

 the hardest steel to 1*5 decimal line for a Swedish foot, and in 

 soft steel to 6 decimal lines. It is therefore clear how inap- 

 propriate it is to attach so much importance, as is often done in 

 practice, to the elongation on tearing. 



If it is a wire with which the stretching-experiment has been 

 made, and if this wire has been greatly stretched on drawing, 

 the limit of elasticity must have a high position, and likewise the 

 point of curvature; it will then often be found that the curve 

 bends quickly in a horizontal direction, upon which tearing 

 asunder directly follows f. 



Many ores exhibit a difference as regards the form of the 

 curve, which moreover also depends on the degree of hardness 

 (proportion of carbon ?). While iron and the softer steel from all 

 ores have a strong curvature at the point of the greatest curve, 

 and immediately after a distinctly convex part, in the hardest 

 kinds of steel from certain ores the curvature is considerably 

 smaller, and the convex part is quite wanting. Hence in them 



* From the irregularity in the form of the curve on tearing, it can be 

 immediately seen whether it took place too soon owing to defective homo- 

 geneity in the bar. 



t See Wertheim's experiments with wires of cast steel (acierfondu etire) 

 loc. cit. p. 41. 



