Prof. Dove on the Electrical Light. 385 
scarcely differs from that of a small spark. The comparison is 
best made when the second conductor in contact with the point 
is alternately brought in contact with the conductor and removed 
from it. The spectrum recognized at the basal point of the 
brush is then transferred to the point of interruption, where the 
sparks pass. The phenomenon in the electrical ege is exactly 
analogous. The perpendicular stream of light gives a many- 
coloured spectrum, consisting of a very broad blue border, a 
broad green and a narrow red streak; it is faintly visible through 
a red glass, whilst the band-like streaks are seen very clearly 
through the cobalt glass, but are completely absorbed in the red 
glass. 
The light of an exhausted glass tube containing a little mer- 
cury, which appears brilliantly white in the dark, and in day- 
light a bluish-green, is not visible through the red glass, but 
very distinct through a green one, and rather less bright through 
the deep blue. If the tube be held to the conductor, it shines 
for a long time uninterruptedly; and besides blue and green, 
the spectrum contains a slight trace of red. I have sketched 
coloured spirals upon a white ground, which, when looked at 
through certain glasses of the same colour, disappear in such 
a way, that, on turning them round, the perfectly white hinder 
surface cannot be distinguished from that on which the coloured 
spiral has been drawn. This experiment succeeds without any 
coloured glass in the case of a spiral sketched with Schweinfurt 
green, when illuminated in the dark by the mercurial tube ; the 
light in the mercurial tube consequently hasthe colour of this spiral. 
The electric spark is distinctly visible through any coloured 
glasses, with the colour of the latter. Catoptrie colours mo- 
mentarily illuminated by it appear distinctly, as also do colours 
of interference when I concentrate the sparks of a self-dischar- 
ging Leyden flask, by the object-lens of my polarizing apparatus, 
upon the aperture of the polarizing Nicol’s prism; and the cale- 
spar plate, which appears colourless when the Nicol’s prism is 
rapidly rotated in continual illumination, then exhibits the an- 
nular system distinctly, and therefore behaves exactly like the 
coloured sectors of a rotating colour-circle. Whilst the nature 
of the metals exerted an influence upon the spectrum of the 
spark, the absorption-phenomena of the rays of the brush 
remained unaltered, when I developed it from’ gold, platinum, 
iridium, nickel, iron, bismuth, tin, zinc and copper, or from a 
drop of water sprinkled on the conductor; this is in accordance 
with Faraday’s observations. Whilst the introduction of a moist 
thread essentially modifies the light of the spark, the brush pro- 
duced by a conductor united with the primary conductor by a 
wet thread remains unchanged. On the other hand, the lumi- 
Phil, Mag. 8. 4, Vol, 14, No. 94, Nov, 1857, 2C 
