an unknown element which M. K. J. Bayer thought he had found in French bauxite, have returned to nothingness. We mention also merely as a matter of curiosity a kosmium and a neokosmium, deriving their names not from Cosmos, but from Kosmann, who took out a patent for the preparation of their oxides.
Gallium was discovered in August, 1875, by Lecoq de Boisbaudran in the blende of Pierrefitte, through two very distinct lines in the violet of the spectrum of that mineral, which, however, as afterward appeared, contained only a slight proportion of the new metal—not exceeding 0.0001 per cent—while in the richer blende of Bernbryer it amounted to 0.001 per cent. The preparation of gallium in any considerable quantities was attended with great difficulties on account of the want of a proper mineral to be practicably submitted to the extraction process, and none has as yet been found. Still, the study of the new metal was very interesting, in view of the theoretical speculations of Mendeleef. Scandium and germanium had not yet been discovered, and there was therefore nothing to justify or confirm the conclusions drawn from the law of periodicity. As early as 1869, Mendeleef had affirmed the existence of simple bodies still unknown, the atomic weights of which should be comprehended between 65 and 75; he had even gone so far as to describe in detail the properties of the three hypothetical elements—ekaboron, eka-aluminum, and ekasilicon. We can imagine the interest attached to the question whether the properties of gallium corresponded with the anticipations of the Russian chemist.
At first, the correspondence did not seem to exist; the determinations made on the small quantities of gallium that could be obtained gave the specific gravity the unexpected value of 4.7. But as many of the properties of the new metal—such as the precipitation of its solutions by carbonate of barium, its tendency to form basic salts, and its capacity of forming alums—denoted a relationship with aluminum, Mendeleef had no hesitation in declaring that the new element appeared to correspond with the one the existence of which he had indicated in 1874 as similar to aluminum, and which he had called eka-aluminum. A new determination, made with considerable quantities of gallium obtained by electrolysis, brought the value of the specific gravity up to 59, which correspond exactly with the value calculated by Mendeleef for the hypothetical eka-aluminum. The specific heat (0.08) was afterward found to correspond with Mendeleef's estimate, and the justness of his previsions was established. It was therefore shown to be reasonable to deduce from the properties of known elements those of others still unknown, but the existence of which is anticipated. Mendeleef had not expected so quick a confirmation of his
