Poitok AND Leonarp—On the Spectra of Cerium, &c. 263 
the oxide was dissolved in hydrochloric acid, and then diluted, so 
as to contain 10 per cent., 5 per cent., 1 per cent., ‘lL per cent., 
‘O01 per cent., and 001 per cent. of the element. These solutions 
were then successively sparked, using a Ruhmkorff coil, condenser, 
and self-induction coil for removal of air-lines, and photographs 
were taken of the spectra giving an exposure of 1 minute each. 
The photographs of all the dilutions of lanthanum are reproduced 
in Plate XXI.; but as the lines of the more dilute solutions do 
not show in the reproductions, only the spectra of the 10 per cent. 
and 1 per cent. solutions of cerium and yttrium are shown in 
Plates XXII. and XXIII. The most persistent lines are indicated 
by one, two, or three dots placed under them according to their 
persistency. 
In the tables of wave-lengths, double lines, the wave-length otf 
only one of which is given, are indicated by two dashes after the 
intensity ; so 10”, the persistency of the lines, is indicated by Greek 
letters, thus :— 
o seen with a strong solution only. 
d Stel Rally i not seen with 0-1 / 
at ee Cont 7 i i 0-01 
lh La eal OILE, F i 0-001 ¥ 
Ww Py) 0-001 Is 9 
The photographs of spectra extend from A 5890 to A 25443 
on Plates XXI. and XXII., and from A 2825°6 to 2100 on 
Plate XXIII.; and they were taken by a one-prism quartz spectro- 
graph by Hilger. A slow-motion tangent-screw was attached to 
the slit to permit of its rotation, so as to bring the slit absolutely 
perpendicular to the edge of the prism, and then clamp it there. 
This attachment was found of great advantage in practical work, 
and should be added to all spectrographs. 
The measurements were made with a micrometer, also by 
Hilger ; but a movable sub-stage, controlled by a slow-motion screw, 
was fitted to it, to enable the micrometer to be set to any particular 
reading. The plate was clamped on to the stage, and then any line 
could be brought exactly to position between the cross-wires of the 
microscope. ‘This is the greatest possible convenience in conjunc- 
tion with the use of gold electrodes in the measurement of spectra, 
as it greatly facilitates and expedites the reduction of the measure- 
