EFFECT OF PRESSURE UPON ARC SPECTRA. 
119 
by a spring attachment to the circular metal plate, which can be turned in its own 
plane by the screws Z controlling the position of the lug attached to it. 
Details of the mechanism by which the jaws are operated are not shown in the 
figure, but the system possesses the property of moving the knife-edges through 
equal distances as the width of the slit is altered, so that the position of its centre 
always remains the same. 
The Grating Room. 
The whole of the mounting is supported at such a level above the floor that it does 
not interfere with the floor space, which is therefore available for other purposes. At 
one end of the room a large platform has been erected, so that the camera can be 
easily manipulated, and a source of light placed in position and operated during the 
experiment. 
Behind this platform are apertures in the wall through which sunlight can enter 
after reflexion from a heliostat. This can be placed on one of the ledges projecting 
from the outside wall, access to which is afforded by a balcony. 
A ventilating chimney immediately over the source of light, usually an electric arc 
or spark, prevents obnoxious fumes from filling the room and injurious gases from 
reaching the grating surface. In addition, a screen of black cloth completely encloses 
this portion of the room and prevents stray light from affecting the plate. A hole in 
this screen permits the light to pass from the slit to the grating. 
The room can be very perfectly darkened by means of blinds, and a double door 
permits access to or egress from it during the course of an experiment. 
The electrical equipment of the platform consists of leads from the Corporation 
mains, from the storage batteries, and from the dynamo house, thus giving a wide 
range in the choice of voltage of both continuous and alternating current. 
The Adjustment of the Large Howland Concave Grating. 
The two carriages without their mountings were placed upon their respective 
girders and—- 
I. The grating girder* A, fig. 1, was so adjusted that the turned face of its 
carriage D remained horizontal for all positions along it. A spirit-level was laid 
upon the carriage, and the height of the four screws projecting upwards through the 
supporting brackets F altered. 
II. The line of motion of the centre of the grating; carriage was marked out. 
© © © 
A pianoforte wire, held at each end by a clamp, was stretched above the girder in 
* This girder was found to be considerably out of true alignment, so clamps were applied at intervals 
along its length to twist it until its edges were parallel and horizontal. The employment of a length of a 
lathe bed and saddle for the grating holder would obviate any twisting, and, I believe, possess many 
advantages over existing systems. 
