24 
PROF. 0. W. RICHARDSON ON THE EMISSION OF ELECTRONS 
Some useful data pertaining to the lines used are collected in the following 
Table :— 
Table IV.— Mercury Spectrum {visible). 
Colour. 
Wave-length between 
(cm. X 10 5 ). 
Frequency (prer second). 
Relative intensity.* 
Yellow. 
5-769 and 
5-19 X 10 14 
23-6 
5 • 790 
Green.. 
5-460 
5-49 X 10 14 
18-7 
Blue. 
4-916 
6-10 X 10 u 
1-0 
Violet 1. 
4-347 and 
6-89 X 10 14 
14-7 
4-358 
Violet 2. 
4-046 and 
7-40 X 10 14 
8-2 
4-077 
In addition to the illuminator, I had the use of a large selection of colour filters kindly 
lent by Messrs. Hilger. Unfortunately, those which might have been useful cut down 
the working radiation too much. For monochromatism the dispersion of the illumi¬ 
nator has therefore alone been relied on. As a matter of fact, the lines were all sharp 
and widely separated, and no trouble has been experienced, owing either to overlapping 
or to stray light of other wave-lengths. The lamp was run at the maximum brightness 
at which it could be depended on to run smoothly and remain constant. To prevent 
trouble due to stray light it was necessary to enclose the lamp in a large tin box, and 
to prevent over-heating a water-cooling system had to be provided. Once the apparatus 
was adjusted the successive lines could be rapidly thrown on the drops simply by 
adjusting the screw setting of the illuminator. In every case the satisfactoriness of the 
mechanical setting was confirmed by visual observation. 
Very steady conditions have been obtained with these arrangements. The data for 
two determinations of the characteristic curves with this apparatus are shown in fig. 12. 
One of these, x, was taken the same day as and immediately preceding the determina¬ 
tion of the true zero to be described in a moment, and the other, O > as early as possible 
the following day. The smooth curve is drawn to the first set of points marked x , 
and mil be seen to cover them very closely. The first set marked x agree with the 
second, except that (1) they lie about 0 T2 volt to the right, indicating a change by 
this amount in the contact potential difference between the two experiments, and (2) 
they give higher values for the small currents at the foot of the curve on the left-hand 
* These energy measurements are taken from a paper by E. S. .Johansen in ‘ Strahlentherapie,’ vol. 6, 
p. 55 (1915), and are for a different design of lamp which ran on 220 volts, taking 2 • 6 amperes with 128 volts 
between the proles of the arc. The quartz mercury vapour lamp used in these experiments ran on 200 
volts and took 3-0 amperes with 150 volts between the poles. No doubt the relative intensities of the 
lines will vary to some extent with different installations, but for the present purpose it is only necessary 
to know the relative energies approximately. 
