the Reflective Goniometer, 283 



screws at the points Q, R, S, 38, 13, 25 millims. distant from 

 the centre, for securing the crystal-holder to the instrument 

 in a convenient position. This disk is capable of revolving 

 round a pin the ends of which enter holes in the centres of 

 the disk and graduated circle. A rectangular piece of metal 

 T U, having each arm 95 millims. long, 30 millims. wide and 5 

 millims. thick, with slits the width of which is very little 

 larger than the diameter of the screw, can be securely clamped 

 to the disk by a screw passing through the slit in the hori- 

 zontal arm. A bar in which are inserted two screws, by which 

 it can be secured to the vertical branch of T U, carries at its 

 upper end the ordinary Wollaston's branch, to which the crystal 

 is secured by a cement employed by Wollaston, consisting of 

 beeswax melted with a little olive-oil, honey, and lampblack, 

 and stirred while cooling to prevent the separation of the 

 components. The crystal being now adjusted above every 

 part of the instrument, about 195 millims. above the plane on 

 which it rests, is brought into the axis by making T U slide 

 and revolve. The screw being now tightened, the crystal 

 revolves with the circle without any danger of the small angu- 

 lar displacement which is liable to occur, according to Dauber 

 (Poggenclorff's Annalen, vol. ciii. 1858, p. 107), in the best 

 constructed goniometers when the circle is vertical and the 

 adjustment is by sliders in grooves making right angles with 

 one another. This method of bringing the crystal into the 

 axis of the circle can also be applied with advantage to the 

 small Wollaston's goniometer. 



The image of the bright signal is much more distinct when 

 the incidence is as direct as possible, more especially when the 

 face under observation is striated or partially coated with any 

 foreign matter. A small angle of incidence is hardly obtain- 

 able with a vertical circle or with a horizontal circle in most 

 localities, as long as the faint signal is seen directly. 



Instead of employing an object seen directly for the faint 

 signal, it was suggested independently by A. F. E. Degen 

 (PoggendorfFs Annalen, vol. xxvii. 1833, p. 557) and E. Sang 

 (Edinb. New Phil. Journ. vol. xxii. 1837, p. 213) that the 

 image of a line seen by reflexion in a mirror could be employed 

 for the faint signal. They failed, however, to obtain the great- 

 est advantage from the employment of the mirror ; for though 

 theoretically correct, this is practically inconvenient, because 

 the brightness of the reflection in the mirror very commonly 

 extinguishes the reflected image in the face of the crystal. A 

 better mode of procedure is to place a screen with a narrow 

 vertical slit covered on the side furthest from the observer 

 with thin paper, making at the crystal an angle of from 3° to 



