Po.tiok anp Leonarp—On the Spectra of Cerium, &c. 259 
ceria being converted into the insoluble yellow dioxide and 
remaining suspended in the fluid, the oxides of lanthanum, 
didymium, and yttrium dissolving completely when the solution 
is absolutely saturated with chlorine. In operating on large 
quantities it is necessary to repeat the process three times. ‘The 
whole quantity of the mixed oxides was dissolved in hydrochloric 
acid, treated with excess of caustic soda, and a steady stream of 
chlorine passed through the liquid for two days. ‘The cerium 
was converted into dioxide and remained undissolved; all the 
other earths dissolved. After filtration and washing, the cerium 
dioxide was suspended in fresh caustic soda and again treated 
with chlorine, and the operation thereafter repeated a third time, 
when no more dissolved. 
The dioxide after the third treatment was dissolved in hydro- 
chloric acid, precipitated as oxalate, and ignited to oxide, and 
preserved for future experiments. The oxide was of a light reddish- 
brown colour; its solution in hydrochloric acid was quite free 
from absorption-bands, and, on sparking a concentrated solution, 
gave no trace of lanthanum, yttrium, or didymium Jines. This 
preparation was used for the quantitative spectrum of cerium. 
The solutions were boiled, then precipitated by oxalic acid, 
and the oxalates ignited, giving mixed oxides free from cerium ; 
and these were used for the subsequent preparations. It was 
remarkable that the particular specimen of cerite operated on was 
not of the usual composition ; for it contained over 50 per cent. of 
silica, and gave an almost equal quantity of oxalate of cerium, 
and of the mixed oxalates of the other rare earths. These 
latter consisted for the most part of lanthanum, with a large 
proportion of didymium, and a small proportion of yttrium. 
After a great number of preliminary experiments, in which the 
spectra of all the products were examined, the following method of 
preparing pure yttria and lanthana was adopted, but no efficient 
method for preparing neodymium and praseodymium free from 
lanthanum was discovered; and we did not get them in a 
sufficient state of purity to determine their quantitative spectra. 
Cerite of normal composition will not fuse with as little as its 
own weight of caustic soda, but is nevertheless completely decom- 
posed, and easily soluble in hydrochloric acid, after treatment 
with this quantity of alkali. 
2M 2 
