Portok anp Leonarp—On the Spectra of Cerium, &c. 261 
ammonia, filtered, washed, and redissolved in the least possible 
quantity of hydrochloric acid, a little water added, then a quantity 
of strong solution of potassium sulphate, a bag of crystals being 
suspended in the beaker as formerly; but the total volume of the 
solution was about one quarter of that present in the previous 
precipitation. After two days the precipitate was filtered off, and 
the solution found to be colourless, and almost free from absorption- 
bands. On precipitating with ammonia, washing, redissolving, 
precipitating with oxalic acid, filtering, washing, drying, and 
igniting, it gave an oxide of a pale buff colour; that, on sparking, 
gave a splendid spectrum of yttrium free from all lines of 
lanthanum or didymium, and containing only a trace of iron, the 
residuary lines of which can be seen in the ultra-violet end of the 
spectrum of the strong solutions of cerium and yttrium ; but they 
disappeared in the dilution-spectra, and did not interfere with 
the determination of the quantitative spectra; but it is remarkable 
that even any iron should remain after so many precipitations 
with oxalic acid. This preparation was used for the quantitative 
spectrum of yttrium. 
PREPARATION OF Pure LANTHANA. 
The first fraction from the fusion of the nitrates—that is, that 
portion which had been twice fused without decomposition—was 
now further treated for the preparation of pure lanthana. The 
solution was nearly colourless, and gave no spark-spectrum of 
yttrium ; and the lines of didymium shown by the spark-spectrum 
were very feeble; but when viewed through any depth, the 
solution gave quite strong absorption-bands. The oxides from 
this fraction were precipitated by ammonia, and weighed about 
50 grams; they were dissolved in the least possible quantity of 
hydrochloric acid; then treated with one-half the quantity of 
sulphuric acid necessary to precipitate the whole. ‘The solution 
gave no immediate precipitate in the cold ; so an equal volume of 
alcohol was added, when a good precipitate came down; this was 
filtered off, washed with alcohol, dissolved in strong hydrochloric 
acid, precipitated by ammonia to get rid of the sulphuric acid 
present, then again dissolved in the least possible quantity of 
hydrochloric acid, diluted to about 500 c.c., then again treated 
