Ekaboron

In 1879 Nilson was preparing ytterbia from euxenite and gadolinite by the procedure that had been given by Marignac, the discoverer of the earth. In so doing, he made the discovery of a new earth, present in very small quantity, and characterised by its very feeble basicity (less than that of ytterbia), its very low chemical equivalent, and its spark spectrum. Only 0.3 gram of the earth was obtained, and that in an impure state. To the new element present in the earth, Nilson gave the name scandium. A few weeks after the announcement of the discovery, Cleve reported that he had isolated 0.8 gram of scandia from 4 kilos, of gadolinite, and 1.2 grams of scandia from 3 kilos, of keilhauite. Cleve also described a few compounds of scandium. A little later, Nilson obtained a few grams of scandia, described several scandium salts, and determined the atomic weight of the element.

Scandium corresponds very closely with ekaboron, one of the elements predicted by Mendeleeff when he put forward the periodic classification of the elements; and its discovery, coming only four years after the discovery of eka-aluminium or gallium, was a matter of great interest and assisted largely in the recognition of the merits of the periodic classification by chemists in general. The identity of ekaboron and scandium was pointed out by Cleve; the following table will serve to show how closely the properties of scandium were predicted by Mendeleeff: -

Ekaboron	Scandium
Atomic weight, 44.	Atomic weight, 44.
Should give one oxide, Eb2O3, density, 3.5; more basic than Al2O3, less basic than MgO; insoluble in alkalies.	Scandium oxide, Sc2O3, density, 3.86, is more basic than Al2O3 and probably less than MgO; insoluble in alkalies.
Carbonate should be insoluble in water, and probably be precipitated as a basic salt.	Scandium carbonate, insoluble in water, readily loses carbon dioxide.
The double sulphates with alkali sulphates will probably not be alums.	Double sulphates are known, but they are not alums.
Anhydrous chloride, EbCl3, should be more difficultly volatile than AlCl3. In aqueous solution it should hydrolyse easier than magnesium chloride.	Scandium chloride, ScCl3, begins to sublime appreciably at 850°. In aqueous solution it is decidedly hydrolysed.
Ekaboron will probably not be discovered spectroscopically.	Scandium was not detected by means of its spectrum.