Ixvi PROCEEDINGS. 
another remarkable Swedish chemist, isolated Scandium, and since that 
time Cleve claims to have found in Erbia a threefold group for one mem- 
ber of which he retains the name Erbia, calling the others Thulia and 
Holmia. 
In the meantime, from a sample of Didymium obtained from the 
mineral Samarskite, first found in North Carolina, a new oxide, Samaria, 
was separated in 1878 by Boisbaudran and Delafontaine, acting 
separately, the latter calling it Decipium. Finally, in 1883, Didymium, 
which since 1842 had held its place as an element, and from which a 
metal had been isolated and which had played a prominent part in 
several quite bitter chemical discussions, all parties basing their argu- 
ments on its being an element, was, by Welsbach, a chemist of Vienna, 
best known as the inventor of the Auer or Welsbach Light, disintregated 
into what, for want of better name, are called, (or perhaps I should say 
into tentative substances,) Praseodidymium and Neodidymium. Still 
more recently, Bettendorf has obtained evidence of the presence of still 
another oxide in Yttria, which he proposes to name in honor of the 
original discover of Yttria, Gadolinum. 
It has been twice thought that Zirconia was not elemental, once in 
1845, when Svanberg thought he had isolated Noria, and again in 1866-7, 
when Sorby thought he had found Jargonia. Both were subsequently 
proven to be Zirconia, or it was shown, at least, that there was not enough 
evidence to consider them elemental earths. 
You will note then, that from the two original earths, the Yttria of 
1789 and the Ceria of 1804, not less than eleven earths have been 
isolated and probably two or three more, though the evidence is less 
conclusive. When I mention that the knowledge of these, though very 
accurate, is less than that known of our ordinary elements at, say 100 
years ago, it will be seen that a wide field exists here for investigators. 
Why is so little known concerning them? The answer might be 
hazarded that it is because of their rarity. This is not so, however, as 
several of them have been proved to exist in considerable quantities. 
The trouble lies in their close resemblance to one another, chemical 
reagents acting similarly toward them all, and thus the only means of 
separating them is by taking advantage of the difference iu basicity of 
their compounds. This is a very slow process and uncertain, for, being 
fractional, it is only made exact by numberless repetitions, and so it is 
extremely difficult to get pure material on which to experiment. This 
