Ductile Materials under Combined Stress. 101 



(as in the Adjuster) ; to this was hinged by an axis resting in 

 two V's E, F, and a point at the end of the axis resting on 

 a surface Gr, a lever 2. The lever 2 has a projection H 

 which was kept in contact with the tube by a spring K. The 

 piece 1 also carried the V-bearings L, M of a mirror- frame 

 3, the axis of the bearings being perpendicular to the axis of 

 the specimen. The mirror-frame was connected to a screw 

 working in the lever 2 by means of a link 4 ; the link having 

 sharp knife-edges of V-form which worked on knife-edges 

 on the screw and mirror-frame. The instrument was sup- 

 ported partly by springs, and partly by the friction at the 

 surface of the tube. The distance of the knife-edge N 

 fastened to the mirror-frame, and upon which the link 4 

 worked, being small, its distance from the axis of rotation of 

 the mirror-frame was determined by clamping the piece 1 

 and a micrometer-screw in such a way that turning the 

 screw pushed, by means of a small distance-piece, the knife- 

 edge through a definite amount. The twist of the mirror 

 was then determined optically. The amount of the strain 

 corresponding to any twist of the mirror can then be easily 

 calculated ; it being remembered that, owing to the use of 

 V's for the tube to rest in, the movement of H relative to the 

 piece 1 is not simply the increase of the diameter, but is to 

 be considered as the effect of an enlargement of the tube. If 

 a is the semi-angle of the V's, the distance from the point of 

 the V to the point of contact of the lever is ?'(l + coseca), 

 where /• is the radius of the tube. Hence an increase of 

 radius 8r corresponds to a movement of the lever at the point 

 of contact with the tube of Sr(l + cosec a). 



40. Method of making the Tests. — In making the tests, a 

 tube, after its dimensions had been determined, was soldered 

 on to the holders and then placed in the testing-machine, and a 

 load of 250 lbs. applied, and the shackle-bearings tapped in 

 order to make the specimen line itself up as accurately as 

 possible. This load of 250 lbs. was employed as it pulled 

 into contact the surface of the lower head of the testing- 

 machine, which would not be in contact permanently until 

 the load on the specimen exceeded the weight of all the parts 

 between it and the straining-gear of the machine. [There 

 was also a slight amount of backlash in the nut o£ the screw.] 

 The torsion-bars and steel points were placed in their positions, 

 and the screws in the bracket adjusted to just touch the lower 

 cross-bar. The grips of the extensometer were then fixed to 

 the specimen by the use of the adjuster, and the lower torsion- 

 mirror applied. The telescope for the torsion-tests was then 

 adjusted opposite the lower mirror with its optical axis hori- 

 zontal, and the lower mirror was adjusted by the screws in 



