THE SEPARATION OF THE RARE EARTHS 567 



earths be dissolved in a very slight excess of nitric acid and diluted some- 

 what (possibly 1 k. to 2 or 3 litres). Place in a warm place, add lumps 

 of sulphate of soda, and stir until no more will dissolve. Continue to 

 add and stir for a day or two until the absorption lines of neodymium 

 disappear from the solution. Filter off and call the solution No. 1. 

 Add caustic potash to the precipitated sulphates and wash so as to leave 

 the oxides once more. Dissolve in nitric acid and precipitate again with 

 sulphate of soda, calling the nitrate No. 2. Proceed in this way pos- 

 sibly 10 or more times. The nitrates contain less and less earths; and 

 the precipitate is more and more the pure cerium group; but a dozen 

 precipitations still leave some impurity. 



The portions 1, 2, 3, etc., show decreasing "erbium" absorption bands, 

 and the spectrum shows that the substances a, 6, d, i are gradually sepa- 

 rated out with parts 1, 2, etc., while the numerous fine lines belonging to 

 d, n, c, etc., with the cerium group, fill the spectrum of the portions 

 8, 9, 10, etc. This intermediate group has only very weak absorption 

 bands and evidently has three or four elements in it, as I have produced 

 at least that number of variations in its spectrum. The group can be 

 obtained fairly free from , 5, and i, but the substance d persists in all 

 the filtrates and in the precipitated cerium group also. This interme- 

 diate group d, n, etc., seems to be in greater proportion in samarskite 

 than in gadolinite, and there seem to be more elements in samarskite 

 than in gadolinite. One of these I have called li. 



The oxides, especially for samarskite, are very yellow and dark. 



Sulphate of potash has a decided action in separating a and i from 6, 

 a and i coming down first. After two months, the solution gradually 

 drying, the proportion of & to a in the filtrate increased many times. 

 Sulphate of soda has an action of the same kind, but much weaker. 

 After leaving two months over sulphate of potash and soda, the follow- 

 ing was the result of analysis of the soluble part as compared with the 

 original mixture: 



Sulphate of Potash. Sulphate of Soda. 



Ge., La., etc. o o 



a Weak Medium weak 



b Much stronger Stronger 



c oo 



d Unchanged Unchanged 



t Weaker Medium strong 



o Stronger Weaker 



The oxide of the members of this group which are only slightly pre- 

 cipitated by the sulphates of soda and potash is pure snow-white, and 

 hence those of & and i must be so. 



