218 Scientific Proceedings, Royal Dublin Society. 
above are not amongst them, and as the authors are engaged in an 
investigation of the methods of separating the elements of the 
cerium and yttrium groups, they find that a knowledge of the 
residuary lines of all the common elements of the ammonia and 
ammonium sulphide groups is necessary. Gold electrodes have 
been substituted for graphite to bring the spectra into uniformity 
with other work by the authors, and in particular to make them 
readily comparable with the general index of spectra published 
by one of the authorsin the Proceedings of this Society. On com- 
parison of the dilution spectra of zine and silicon with those of 
Hartley, it would seem that gold electrodes are not so sensitive as 
graphite. The results are otherwise substantially the same, a few 
more lines having been observed with a 1 per cent. solution of zine, 
and all the lines disappear more rapidly ; the last lines to disappear 
on dilution are identical with those observed by Hartley. That 
graphite should be more delicate than gold appears very natural, as 
the graphite tends to absorb the solution-sparks over a larger surface, 
and hence yields more vapour of the element under examination ; 
but for all ordinary analytical purposes, gold is more convenient. 
The work was done with a one-prism quartz spectrograph, by 
Hilger, using the spark produced by a Ruhmkorfi’s coil and 
condenser, with a Hemsalech self-induction coil placed in the 
circuit for the removal of air-lines. ‘The plates were Rainbow 
Fast, made by the Warwick Photo Co., and in every case the 
exposure was 1’—this photographed clearly from X 4792-8 to 
d 2544 on one plate. To go farther down in the ultra-violet, it 
would be necessary to readjust the instrument and take another 
set of photographs. 
The general method of procedure was to make a strong or 
saturated solution, of the chloride of the element under considera- 
tion; also solutions containing one gramme of the element in 
every 100, 1000, 10,000, 100,000 parts of solution. A photo- 
graph was taken of the gold electrodes with a long slit; the slit was 
then shortened and the metal sparked, thus giving the spectrum 
of gold with long lines, and the spectrum of the metal with short 
lines. ‘The process was then reversed, the metal taken long, and 
the gold short, so that any lines coincident with gold lines might 
be seen. A photograph was then taken with both the strong 
solution and gold electrodes long, and the metals short, to show 
