262 Scientific Proceedings, Royal Dublin Society. 
with half the quantity of sulphuric acid necessary for complete 
precipitation, and an equal bulk of alcohol added. On solution 
and precipitation by ammonia, about 10 grams of pure white 
hydrate of lanthanum were obtained, which gave a pure 
spectrum of lanthanum free from all lines of didymium; and the 
solution showed only very faint indications of absorption-bands. 
This preparation was used for the quantitative spectrum of 
lanthanum. 
PREPARATION OF Dipymia. 
As already mentioned, the didymium was concentrated in the 
fourth fraction of the fused nitrates, that is, that portion which was 
first to decompose in two separations. This fraction was dissolved, 
and the didymia precipitated by potassium sulphate for the prepara- 
tion of pure yttrium. The precipitated basic sulphates were now 
fused with alkaline carbonates, boiled and filtered for the removal 
of the sulphuric acid, and then dissolved in hydrochloric acid. 
The sulphates are not freely soluble in acid without this 
preliminary fusion with alkali, which is a great inconvenience in 
the use of this method of separation. The solution was now 
fractionally precipitated by sulphuric acid and alcohol, as in the 
case of lanthanum; but the precipitate was rejected, the alcohol 
distilled off, the didymium precipitated by ammonia, and the 
process repeated. This yielded about 8 grams of hydrate of a 
pink colour, which, on dissolving, gave a dark pink solution with 
very intense absorption-bands, and a strong spectrum showing the 
characteristic lines of neodymium; but the lines of lanthanum also 
showed strongly; and it was not possible to satisfactorily determine 
the quantitative spectrum of didymium, as its spectrum vanishes 
very rapidly on dilution, whereas that of lanthanum is very 
persistent. 
METHOD OF TAKING THE SPECTRA. 
The quantitative spectra were taken after the manner devised 
by Professor W. N. Hartley,’ with the modifications described in 
our previous papers “‘ On Quantitative Spectra,”? published in the 
Proceedings of this Society. In each case a weighed quantity of 
1 Proc. Roy. Soc., 1882, p. 81. 
* Proc. Roy. Dub. Soc., 1907. Nos. 16, 17, and 18. 
