﻿xxvi Proceedings. [February ytk, 1905. 



were nominated auditors of the Society's Accounts for the 

 session 1904-1905. 



Mr. Thomas Thorp, F.R.A.S., exhibited his new direct-vision 

 spectroscope and explained its construction. 



The dispersion of direct-vision spectroscopes of the ordinary 

 kind is effected by means of a fixed compound prism, the 

 surface nearest the eye being at an angle of 45 ° with the axis 

 of the instrument. This surface is utilised to reflect a scale or 

 pointer into the field of view, which can be made to travel along 

 the spectrum, and so enable readings of the positions of the 

 well-known bright and dark lines to be made. The accuracy 

 of the readings is, however, not uniform, as the spectrum is not 

 a normal one, being extended at the blue end and contracted at 

 the red end. 



In Mr. Thorp's instrument the dispersion is effected by 

 means of a transparent grating of about 14,500 lines to the inch, 

 mounted on the long face of a light crown prism having a 

 refracting angle of about 37 to secure direct vision. This 

 prism-grating is mounted in a hinged frame and adjusted so 

 that the grating face is at an angle of 45° with the axis of the 

 instrument when the frame is at the centre of its range of 

 motion. A spring holds the frame tightly against the end of a 

 micrometer screw having a graduated head., this head being in 

 the focus of a lens placed near the ocular of the spectroscope 

 so that it can be read off without taking the instrument away 

 from the eye. The pointer, which is a small triangular aper- 

 ture in a metal plate, illuminated by the source of light under 

 observation, is placed, as usual, in the focus of a collimating 

 lens and deflected on to the posterior surface of the grating by 

 means of a total reflecting prism. On now turning the micro- 

 meter screw by means of a milled head, the pointer is seen to 

 move along the spectrum by reason of the alteration in the 

 position of the prism-grating which, however, does not affect 

 the spectrum in any appreciable degree, and the positions of 

 the bright and dark lines can be read off with the same degree 

 of accuracy from end to end of the spectrum, it being almost 



