52 



THE SEPAKATION OF THE EAEE EAETHS 



[Johns Hopkins University Circulars, No. 112, pp. 73, 74, 1894] 



In the course of several years' investigations of the so-called " rare 

 earths," such as yttrium, erbium, holmium, cerium, etc., I have devised 

 several methods for their separation. I wish to give an account of these 

 now, and hope soon to be able to publish a complete description of my 

 work and its results. 



It was evident very early in the work that cerium, lanthanium, praseo- 

 dymium, neodymium and thorium differed from the yttrium group, and 

 I have seen no reason to suppose that they can be divided any further. 

 All of these "earths" appear, in varying proportions, in such minerals as 

 gadolinite, samarskite, yttrialite, cerite, etc. Besides the elements of 

 the cerium group here present there are at least seven other substances. 

 For the present I shall speak of them as 



a, 6, i, d, h, n, c, fc. 

 Their properties are as follows: 



PKOPEBTIES OF ELEMENTS 



Substance a 



This is the principal element of yttrium and may possibly be divided 

 into two in the future, as I have observed a variation in the arc spec- 

 trum on adding potash or soda. However, this is no more evidence than 

 occurs in the case of iron or zirconium. I give a process below for pro- 

 ducing this pure. 



Properties. No absorption bands. Oxalate and oxide pure white. 

 It occurs in the sun. Its properties are those of yttrium as hitherto ob- 

 tained, but I am the first to obtain it with any approach to purity. 



Mixture of I, i and d 



These seem to be the principal ingredients in so-called "erbium." 

 Oxalate is red. Oxide is pure white. Absorption band is that of 

 "erbium." It colors the electric arc green, and shows the "erbium" 

 emission bands on heating white hot. The substance b is strong in gado- 



