270 DR E. G. COKER ON 



III. — Method of Experimenting. 



The diameter of the bar to be tested was first ascertained by a micrometer caliper, 

 the mean of several readings being taken. The measuring instrument was then applied, 

 and the calibration value of the readings ascertained. The bar was then placed in the 

 testing machine, and the balance weights adjusted to give zero torque. If bending 

 moment had to be applied, this was effected before the final adjustment of the reading 

 microscope, care being taken to bring the wire midway between the two sets of clamping 

 screws, and also to set the wire in the most favourable position for taking accurate 

 readings, viz., the plane in which bending takes place. Unless readings are taken at 

 equal intervals of time, the time effect of a stress will show itself, and it is therefore 

 very necessary that the separate loadings be at equal intervals. It was found that the 

 most convenient interval was one and a half minutes, this being necessary to bring the 

 weigh-beam back to its zero position ; and all readings were taken with this interval, 

 except where otherwise stated. 



IV. — The Form of the Stress-strain Curve. 



Before taking up the detailed examination of the relation of stress to strain, it is of 

 interest to consider the stress-strain curve as a whole. 



Fig. 15 shows the general nature of the stress-strain curve for a wrought-iron or 

 steel bar of circular section when subjected to a gradually increasing torque. Starting 

 from no torque, and gradually increasing the load upon the bar, the relation between 

 stress to strain is found to be a linear one until near the point a, when the defect from 

 linearity is first noticed in the gradual creeping up of the readings, the whole twist 

 upon the bar being at the rate of from 1° to 2° per inch of length. 



At a the yield-point occurs, and there is a large increase in the strain, with no in- 

 crease in the loading. The material has also changed from a nearly perfectly elastic to 

 a semi-plastic condition, and the bar when released from load will no longer go back to 

 zero, but shows a very considerable set. The material has also hardened by the process, 

 and the curve rises at first quickly to a point c, and then more slowly until fracture 

 occurs at d ; the strain then being generally considerably more than one hundred-fold 

 the strain within the perfectly elastic condition. 



V. — The Form of the Curve at the Yield-point. 



The first sign of deviation from the linear law indicates the failure of the elastic state 

 at the fibres most severely strained — viz., the outer ones — and a semi-plastic condition 

 is entered upon, which, as the loading proceeds, extends inwardly until a more or less 

 uniform shear stress is established throughout the section. The passage from the one 



