DR. W. HUGGINS ON THE PHOTOGRAPHIC SPECTRA OP STARS. 
671 
position was previously found by means of solar light, and the frame was then firmly 
fixed in position before the apparatus was mounted in the telescope. 
The photographic plates are l\ inch long by \ inch wide, and the length of the 
photographic spectrum between the lines G and P in the ultra-violet about \ inch. 
The definition is so good that the photographs can be examined with advantage 
under a low-power microscope, and notwithstanding their small size, about fourteen 
lines may be counted between the lines H and K. 
The apparatus combines very successfully a sufficiently defined separation of the 
parts of the spectrum with a moderate diminution only of the intensity of the star’s 
light. 
Fig. 1. 
I 
a. Slit plate. 
b. Tube for collimation. 
c> d. Quartz lenses. 
e. Prism of Iceland spar. 
/. Photographic plate. 
cj. Bevelled edge. 
h. Screw for adjustment in focns of mirror. 
i, h. Shutters of slit. 
1. Silver plate with slit. 
The width of the slit which was finally adopted was based on a compromise. The 
very narrow slit which gave the best photograph of the solar spectrum was found to 
diminish too seriously the light of the stars, and the slit was then opened until the 
interval between the edges was about - 3 ^-oth of an inch. When the slit is of this 
width of opening the solar lines are still well defined, but the number of lines to be 
counted between H and K is reduced to about seven. I found it was not possible 
to work with a narrower slit. 
The base plate of the apparatus is bevelled at the edges and slides in the grooves 
of a second plate, which is firmly screwed to a wooden platform which is attached to 
the end of the telescope tube (fig. 2 ). The small convex speculum was removed from the 
Cassegrain telescope, and the spectrum apparatus fixed at the end of the tube, as 
already described, was so adjusted that the slit was brought exactly to the position 
4 R 
