102 Mr. J. J. Guest on the Strength of 



its frame until the reflexion of the optical axis was parallel to 

 the specimen. The top mirror was then fixed to the top grip 

 and the torsion-scale placed at the same level : the screws of 

 the top mirror were then adjusted until the image of the scale 

 appeared horizontal and at the centre of the field. The scale 

 was then moved horizontally until the reading was nearly 

 zero, the final adjustment to zero being made just before a 

 test by the azimuth-screw of the telescope. The adjustment 

 of the mirrors is such as to satisfy the condition found above 

 (§ 36) for the vanishing of the terms of the second order in the 

 expansion of the twist. The remaining parts of the extenso- 

 meter were then placed in position, and the scale for axial 

 extensions being moved so that its centre was opposite the 

 mirror, the second screw Q was used to adjust the reading to 

 nearly zero, the final adjustment being made by the altitude- 

 screw of the corresponding telescope. The diametral extonso- 

 meter was then adjusted on the specimen, above the other and 

 strain apparatus. 



The various forces were applied by suitable increments, 

 and readings of the extensometer taken at each application : 

 the time-effect at the elastic limit was looked for, and at its 

 appearance the forces were applied by smaller amounts until 

 it was judged that the yield-point had been reached. In tests 

 involving tension and torsion the torque was applied first ; 

 during the application of the tension-load the first reading 

 seldom changed appreciably, though it usually changed a few 

 divisions. On only one occasion was the change (indicating 

 a bending— moment) considerable. In all tests involving in- 

 ternal pressure, the pressure was applied last, as it was some- 

 what difficult to keep it constant at a desired amount. In the 

 Tables the tension-loads are denominated by the letter P, and 

 expressed in pounds-weight ; the loads producing torque by 

 W, so that the torques are 15 W inch-lbs., the cross-bar 

 being 30 inches long between the points ; the fluid pressure 

 as recorded by the gauge by p , the corresponding pressure 

 being obtained from the table of gauge corrections. The 

 reading of the axial extensometer is called the axial reading, 

 the factor to reduce this is changed in the course of the tests; 

 otherwise all scales were placed at standard distances. This 

 factor was 0-000001286 for tubes I., II., III., X., XI., XII., 

 XIII., and 0-000001265 for the remainder. The reading to 

 determine the twist is called the twist-reading; the factor to 

 reduce it to shearing-strain or slide varies with the diameter 

 of the tube, for a 1^-inch diameter tube it is 0-00000377. The 

 diametral extensometer readings have not been tabulated, it 

 having been deemed more reliable to deduce the strain from 

 the elastic constants ; the factor to reduce scale-readings to 



