IN QUAETZ OF LIGHT IN THE VISIBLE REGION OF THE SPECTRUM. 
275 
spectroscope, the slit of which was adjacent to a small draught-hood in which the arc 
was placed. 
The rotating arc shown in fig. 1 was a modification of that described by Baly 
(‘ Spectroscopy,’ 1912, p. 39l). Silver rods l-j-inch long and ^-inch in diameter 
were screwed into copper cylinders 1-inch long and |-inch diameter, which served to 
cool as well as hold the rods. The copper cylinders in their turn were screwed on 
to f-inch iron spindles which were rotated in opposite directions by means of pulley- 
wheels driven by belts from a small power shaft at the back of the machine, this shaft 
being run from a fan-motor in series with an adjustable resistance. One of the 
spindles was mounted on a pedestal provided with a slide, so that it could be moved 
inwards as the electrodes were burned away. In order to maintain a steady arc, free 
from serious flickering, it was necessary to exercise some care in centering the silver 
electrodes and also to run the machine at the highest convenient speed. A current of 
5 amperes was usually employed but there is little doubt that in the case of the silver 
arc a heavier current up to 10 amperes might have been used. 
(5) (a) Silver Dark Green. Ag 5209’25. 
This line was read with the extra dense prisms on the constant deviation spectro¬ 
scope but no grating and with a Rutherford prism in front of the eye-piece. Tlie 
half-shadow angle was very small, about 3°, and the readings were amongst the easiest 
of those taken with the rotating arc. 
(5) (b) Silver Light Green. Ag 547172 and Ag 5465'66. 
Tliis doublet, separated only by 6'06 units as compared with 5‘97 units for the 
sodium doublet, was read quite easily with the system of highest dispersion, including 
the extra dense 30° prisms, grating, and Rutherford prism. The intensity of the 
light was so great that even in the case of the less refrangible weaker component, a 
half-shadow angle of only 5° could he used ; the readings were among the most 
concordant of the whole series. 
(6) Cadmium. 6438-4722, 5085-8240, 4799-9107, 4678-37. 
Although an enclosed arc would undoubtedly be the ideal source of cadmium light, 
a very efficient suljstitute is afibrded l)y an open arc with electrodes of a cadmium 
silver alloy (Lowry, ‘ Phil. Mag.,’ 1909, vol. 18, 320). This can be run very satisfac¬ 
torily during about an hour with a current of 4 amperes, the current being adjusted 
so that the cadmium distils out from the end of one of the electrodes, occasionally in 
such a way as to leave a considerable cavity in the interior of the rod. Before 
starting up again it is necessary to file away the tip of the electrode in order to bring 
to the surface a fresh supply of cadmium. 
(6) (a) Cadmium Red. Cd 6438-4722. 
On account of the presence of continuous red light, this line is not easy to read 
2 N 2 
