W. A. Norton — Uocperimenfs on Wood, Iron, and Steel bars. 289 



or repeatedly. But with a load of 500 lbs. a permanent set was 

 obtained, as the result of a single application of the stress; and 

 repetitions of the stress were attended with a continual increase 

 in the depression of the middle of the bar. It may accordingly 

 be affirmed that a practical limit of elasticity exists, but not a 

 theoretical one. 



12. If a bar, on the withdrawal of a transverse stress, fails to 

 recover its original line of position, or, technically speaking, has 

 a set, it is plain that its integrant molecules have not returned 

 precisely to their original positions, and that the distances be- 

 tween contiguous molecules have either increased or diminished 

 —increased in the line of the longitudinal fibers that have 

 experienced a tensile strain, and decreased in the line of those 

 which have experienced a compressive strain. Now we have 

 seen that, as the result of a series of increasing transverse stresses, 

 the set increases continuously with the stress, from the lowest 

 amount capable of detection with the measuring apparatus em- 

 ployed. We must therefore conclude that, after the application 

 of a series of increasing strains, in which the molecules are 

 relatively displaced by minute fractions of their intervening 

 distances, they take up, when the strain is removed, a series of 

 new positions of equilibrium, differing by excessively minute 

 degrees from those previously occupied. We may draw the same 

 conclusion from the experiments on the set produced by a series 

 of direct tensile and compressive stresses, made by Hodgkinson, 

 Chevandier and Wertheim, and other experimenters. This 

 general conclusion, to which experiments on set, under every 

 variety of strain, conduct, leads to the inevitable inference that 

 the effective forces exerted by the molecules on one another have 

 suffered some change of intensity, in consequence of the stress a'ppbed 

 to the bar under experiment. Viewing the residual displacenunt 

 of the molecules, in their relative positions, as a mechanual 

 problem, we are constrained to regard the effective molecular 

 forces, that take effect at a given distance, as having acquired a 

 different intensity. We have confirmatory evidence of this 

 induced molecular condition of the bar in the fact that all the 

 diverse effects, which may ensue on subsequent applications of 

 a transverse stress, are found to be either less or greater than 

 those previously observed under similar conditions. 



13. The fluctuations that have been noticed as occurring in 

 the I ■ ■ 

 in the i 



from the temporary application of the stress, is not permanent 

 but fluctuating; and may, according to the amount of the stress 

 applied, rapidly pass off, or, after a partial collapse, be^slow^ly 

 It should be observed,' 



