There are at least two oxides of cobalt, the one blue, the other black. Authors differ as to the proportions of the elements. Proust states the blue oxide to consist of 100 metal, and 19 or 20 oxygen, and the black of 25 or 26 oxygen. Klaproth finds in the blue, 100 metal and 18 oxygen. But Rolhoff according to Berzelius, finds 100 metal and 27.3 oxygen in the blue oxide, and 40.9 in the black. I have taken some pains to investigate these oxides, and have been able to satisfy myself in a good degree, respecting their constitution. The blue or protoxide consists of 100 metal and 19 oxygen, and the black oxide of 100 metal, and 25 or 26, very nearly as Proust determined. _Protoxide._ By repeated trials I have found, that if 37 parts of metallic cobalt be treated with the due quantity of nitro-muriatic acid, and a heat of 150°, a rapid solution takes place, and a disengagement of pure nitrous gas; this being carefully collected, it will be found to weigh 8 grains, and of course corresponds to 7 grains of oxygen; hence 37 cobalt, unite to 7 oxygen, to form 44 of the blue oxide; and as this is the only oxide that combines with acids, it must be considered as the most simple or protoxide, being 1 atom of metal (37), and 1 of oxygen (7). The estimation of the atom of cobalt at 50 or 60, (page 265), must therefore be corrected. _Compound oxides._ When the blue oxide of cobalt is precipitated from a solution, by an alkali or lime water, and oxymuriate of lime is gradually dropped in, the precipitate changes colour rapidly; it passes from blue to green and olive, thence to a dark bottle green, and finally becomes black; oxygen gas is given out copiously when an excess of oxymuriate of lime is used. I find the additional oxygen requisite to convert the blue to the black oxide is what Proust states it, namely, ⅓ of that necessary to form the blue; hence it must be considered as a compound of 1 atom of oxygen and 3 of the protoxide. Probably the other coloured oxides are 1 to 4, 1 to 5, &c. The protoxide is blue when precipitated, but it is supposed to contain water, or to be a hydrate; as it is dark grey when heated. The blue oxide in a short time after precipitation being still under water, changes to a yellowish or dead-leaf colour; which also appears to be a hydrate of the protoxide, as it dissolves in acids without giving out gas, and yields the blue oxide by an alkali. According to Proust, this hydrate contains 20 or 21 per cent. water. If we suppose the blue to be 1 atom oxide, and 1 water, the yellow hydrate may be 1 water and 2 of the proto-hydrate; or 88 oxide, and 24 water, which will be nearly 21 per cent. water. The black oxide gives out oxygen gas by a red heat, and is reduced to the grey oxide: it forms oxymuriatic acid, with muriatic acid, and the protoxide remains in solution. (See Tassaert.--An. de Chimie 28; Thenard, 42; and Proust, 60.) 22. _Oxides of manganese._