EFFECT OF PRESSURE UPON ARC SPECTRA. 
127 
4. The System of Mirrors. 
It was found by some preliminary work that special means were essential for 
continually adjusting the image of the arc upon the slit; as soon as the air in the 
cylinder was compressed the arc became unsteady, and moved about the ends of the 
electrodes in an uncertain manner. Without a means of training the image on to the 
slit only a very small proportion of the length of exposure was effective, but with the 
apparatus to be described it was easy to follow the movements of the arc, and to 
ensure its almost continual focussing on the slit. 
Two mirrors were silvered and polished on the surface, and fitted into frames which 
held them at an angle of 45° with the horizontal; the upper was fixed with its centre 
in the line joining the centre of the grating with the centre of the slit. The second 
was placed 4 inches below the first, opposite the window of the pressure cylinder, and 
could be moved parallel to itself by means of the rack and pinion shown in fig. 10. 
Vertical displacements of the arc from its central position, caused by unequal 
consumption of the electrodes or unequal feeding, were counterbalanced by this 
motion. If we confine our attention to the bundle of rays symmetrical about a 
horizontal ray and for the moment neglect the lens K, we see that a lowering of the 
arc (which occasions a lowering of the position of the spot of light received upon the 
upper mirror) may be counterbalanced by moving the mirror M parallel to itself away 
from the arc. This restores the reflected beam to its original position on mirror N ; 
similarly, when the arc is higher than usual, the re-adjustment is effected by moving 
the mirror M forward. 
