566 HENRY A. EOWLAISTD 



Unite and weak in samarskite. The solution has the absorption bands 

 of " erbium " and most of these seem to belong to & rather than i. How- 

 ever, we can readily prove that the absorption bands of erbium belong 

 to two substances, as we can produce a decided variation in it. 



I cannot reconcile this with my spectrum work without assuming a 

 fourth ingredient in " erbium." 



Substance & is in the sun, but not i. With 6 and i the substance d 

 always occurs. 



Substance d 



This is the principal impurity of a sample of yttrium, kindly furnished 

 me by Dr. Kriiss, which my process of making yttrium separates out. It 

 has not been obtained pure, but occurs strongly in the yellow part of 

 the oxides. It is in the sun. 



By aid of ferrocyanide of potassium the substance a can be obtained 

 pure from d. With this exception d occurs in all the preparations of 

 the yttrium group and cannot be separated from &, i, c, n, h, or any 

 of the other substances. Indeed, I have found it in some specimens of 

 cerium and lanthanium, although in traces only. 



On account of the trouble caused by it and its universal presence, I 

 propose the name demonium for it. 



Its principal spectrum line is at w. 1. 4000-6 nearly. 



Substance h 



This occurs mainly in samarskite. Hints toward its separation will 

 be given below, but I have otherwise obtained none of its properties. 



Substances n, Ic and c 



These always occur with d and form a group intermediate between the 

 yttrium and cerium groups. They can be separated from these by sul- 

 phate of potassium or sodium by always taking in intermediate portions 

 of the precipitate. They seem to have a weak absorption spectrum in 

 the visible spectrum and strong in the ultra violet, especially If. 



Chemical Separation 



The first process that suggests itself is that by the sulphates of soda 

 or potash. This is the usual method for separating the cerium from 

 the yttrium groups. When the solution of earth and the sulphate 

 solution are both hot and concentrated, everything except some scan- 

 dium comes down. When done in the cold with weaker solutions, there 

 is more or less complete separation of the cerium group. Let the mixed 



