208 Scientific Proceedings, Royal Dublin Society. 
the copious precipitate of the mixed oxalates of the rare earths 
filtered off, well washed, dried, and preserved for future treatment. 
The residual liquors were now rendered alkaline with ammonia, 
when a further precipitate came down, which, on spectroscopic 
examination, was found to consist of the rare earths mixed with 
much lime and magnesia. The alkaline precipitate was ignited, 
dissolved in the least possible quantity of hydrochloric acid and 
then precipitated by oxalic acid, and the precipitate filtered, 
washed, and dried. On examining photographs of the spark- 
spectra of the two oxalic acid precipitates, they were found to 
contain identically the same lines; so both oxalates were mixed. 
The residual liquors from this second oxalic acid precipitate 
contained all the lime and magnesia; but its spark-spectrum 
gave also many lines of the rare earths, showing that they 
were not completely precipitated; but as the quantity was 
comparatively small, and as there were no lines, other than 
those of lime and magnesia, that were not represented in the 
first two precipitates, the liquors were thrown out. ‘he liquors 
from the alkaline oxalic acid precipitate on concentration and 
examination showed only the lines of magnesium, potassium, 
sodium, and traces of manganese and iron. We may therefore 
conclude that the earths are completely precipitated by oxalic 
acid in alkaline solution, but cannot be completely precipitated 
in an acid solution. The whole quantity of the mixed oxalates 
were nowignited in a muffle at a dull red heat, and yielded about 
4 lb. weight of a light brown oxide, and a great number of experi- 
ments were made on small portions of these oxides to find the best 
and easiest method of separation, in every case photographs being 
taken of the spark-spectra of the precipitates and solutions, or of 
the successive fractions, and the lines in each photograph carefully 
compared. In all no less than thirty spectra were photographed ; 
but the results may be briefly stated as follows : 
PREPARATION OF PuRE CERIA. 
Preliminary experiments showed that the method of separating 
ceria by precipitating with excess of caustic soda, and then 
saturating the solution with chlorine, gave a sharp quantitative 
separation of ceria from all the other rare earths present, the 
