A wave-length torsometer and its use with the universal interferometer

The author has at last been able to fulfil his intention of extending to the determination of the torsion constants of crystals (and any other small bodies, unprocurable in large pieces) the accuracy and refinement of the measurement by wave-lengths of light. It is comparable to the method of determining Young’s modulus of elasticity by the bending of plates and bars, by the use of the interferential “elasmometer,” which the author described to the Royal Society in 1903. The torsometer now described, used with the universal interferometer described in the preceding memoir, does for torsion what the elasmometer does for bending, so that it is now possible to determine the whole of the elastic constants of crystals, which in the cases of crystals with minimum symmetry (triclinic) are no less than 21 in number, by the refined method of the half-wave-length scale of parallel and rectilinear interference bands, and their micrometrically determined fractions. The torsometer itself is a refinement of the instrument devised and used by the late Prof. W. Voigt, in which the method of the optical lever was employed. The new torsometer transmits the torse movement instead to a Grayson-ruling signal, or fiducial mark, consisting of five engraved lines only one forty-thousandth of an inch apart, the movement of the central line of which is followed by the microscope and measured in interference bands, the torsometer being fixed on the work-table of the universal interferometer for the purpose. This is, indeed, the first practical application of the new interferometer.