﻿the Micro- Azimometer. 395 



21. It is generally recognized as desirable that the mirror 

 of a reflecting instrument should be as light as possible, and 

 should have a small moment of inertia ; but I venture to 

 think that, in many cases where the most favourable relation 

 between sensitivity and period is desired, it will be found 

 advisable to tolerate the somewhat increased thickness of 

 mirror which may be essential to accurate planeness. The 

 vertical dimension of the reflecting surface referred to in 

 Table I. could probably have been reduced from 4 to 2 

 or 3 mm. without impairing the closeness of the settings ; 

 but even where the very lightest suspensions have to be used, 

 these micrometric methods can be made available by a slight 

 modification. 



22. This consists in the introduction of a short-focus 

 objective cylindrical lens, with generating lines horizontal, 

 between the object-glass (fig. 1) and the mirror M ; the 

 latter being at the principal focus of the cylindrical lens. 

 Thus the light emerging from a slit r mm. high would be 

 condensed into a horizontal band in the plane of the mirror, 

 the (vertical) breadth of this band being rgjf, where g is the 

 focal length of the objective cylindrical lens. Accordingly 

 the vertical dimension of the mirror might easily be reduced 

 to (say) 05 mm., and yet an eye placed not far behind the 

 wire W (fig. 1) would &ee the whole height of the objective 

 cylindrical lens filled with light ; the horizontal extent of 

 the light-patch being, of course, limited to the breadth of the 

 mirror. For the height of the objective cylindrical lens, 

 4 or 5 mm. should in all ordinary cases be ample ; for con- 

 venience its breadth should be rather greater than that of 

 the mirror. The appearance is then the same as if the 

 height of the mirror were some 4 or 5 mm., and this also 

 remains true when the field is viewed through a cylindrical 

 or other eyepiece. 



23. As regards the aberrations thus introduced, those 

 which concern the vertical components of rays are of no 

 consequence, so long as full illumination is secured over a 

 sufficient vertical height of the diffraction pattern. Even the 

 mirror need not be truly plane, provided all horizontal 

 sections of its reflecting surface are parallel straight lices. 

 To avoid aberrations which might be detrimental, the wire 

 W and slit 8 (fig. 1) should be in the same vertical line, and 

 the objective cylindrical lens should beset with its generating 

 lines horizontal and at right angles to the beam transmitted 

 by the object-glass 0. It is quite easy to arrange that the 

 last-named condition shall still be automatically satisfied 

 when a change in the azimuth of the mirror makes it necessary 



