24 



DE. A. E. H. TUTTON: 



red 



that surface which bears at its centre the silvered reference ring nearest the micro- 

 scope, and with the direction of the edge of the wedge horizontal. Each of these 

 wedo-e-discs is engraved with two dots at the thick end of the diameter perpendicular 

 to the edge of the wedge, and with one dot at the thin end, so that it is only necessary 

 to arrange the disc with the diametral line joining these two marks vertical, say the 

 two dots at the top. 



This should be done while the mount-fitting is some little distance away from the 

 microscope, and as soon as g 2 is in position the fitting is pushed along its dovetailed 

 bed until the black-glass disc </., and the polourless disc g 2 are just within a millimetre 

 of each other. They should be approximately parallel, and, if this is not the case, 

 f/2 should be adjusted by means of the three brightly lacquered screws h until 



parallelism is approximately attained. There will be two 

 images of the signal-stop reflected from this plate, one from 

 the important outer surface near to the black-glass surface 

 and another from the back surface within the mount ; the 

 one will be immediately under the other, as at B and C in 

 fi<r. 8, B referring to the surface nearer to the black-glass 



o o o 



and C to the back surface, while A refers to the image from 

 the black-glass surface. This latter image will be quite 

 close to B if parallelism of the two surfaces has been nearly 

 attained, and the closer the more perfect the adjustment. 

 At the worst it should not be further away than is indicated 

 by the dotted image A' in fig. 8. 



The reason for arranging the images thus vertically sepa- 

 rated is, that when they are broadened out into spectra by 

 the dispersion apparatus, the two spectra, being horizontal, 

 will not overlap, as will be clear from the horizontal lines 

 indicating the spectra in fig. 8 ; whereas, if the images were 

 horizontally separated by placing the line joining the two 

 dots and one dot of the disc horizontal, their spectra would overlap and interfere with 

 each other. 



If the images A and B do not already overlap, the images B and C (which move 

 together) should be moved by manipulation of the bright screws until B overlaps A, 

 the relative positions of A, B, and C being then as shown in fig. 8 ; the image A can 

 readily be distinguished as being that one of the three which does not move when one 

 of the bright screws is gently touched. The countervailing wedge-disc g t is then to 

 be similarly attached in the right mount, but with the two dots now below and one 

 dot above, on the vertical diametrical line. On observing the images again, we now 

 see five the three already referred to, B and C belonging to the colourless disc g 2 

 and A to the black glass g a> and A more or less overlapping B and two new ones, 

 D and E, belonging to g lf These two latter should be arranged as in fig. 8, by 



yellow 



red 



Fig. 8. Positions of signal- 

 images from the five sur- 

 faces. 



