From the experiments of D’Elhuiarts, Bucholz[13] and Berzelius[14] it seems very probable that the tungstic acid is composed of about 100 metal + 25 oxygen. It is a _yellow_ powder of the sp. gr. 6.12, and is best obtained from the native tungstate of lime (a scarce mineral). One part tungstate of lime and four of carbonate of potash are fused together, dissolved in water, and then the tungstic acid may be precipitated by nitric acid. There is an inferior oxide that is black or dark brown; Berzelius reduced the yellow oxide to a flea-brown colour, by sending a current of hydrogen gas through it in a glass tube heated red hot. 100 parts of this oxide burnt be 107 yellow oxide. Hence 100 metal must combine with about 16½ or 17 oxygen to form this oxide, which is ⅔ of that in the yellow or tungstic acid.--Upon the whole it does not seem improbable, considering the great sp. gravity of this metal, that it forms three oxides and that the acid or yellow oxide is the 3d. Hence the atom of tungsten must be 84, that of the protoxide 91, the deutoxide 98, and the tritoxide or tungstic acid 105. The native tungstate of lime, if pure, according to this would be 81.4 acid + 18.6 lime, which is not far from Klaproth’s analysis; he having found 18.7 lime in one specimen; nor from that of Berzelius, he having found 80.4 tungstic acid and 19.4 lime in 99.8 tungstate of lime.[15] [Footnote 13: An. of Philos. 6--198] [Footnote 14: An. of Philos. 3-244] [Footnote 15: An. of Philos. 8--237] There is another view however, which would accord with the experiments and perhaps will be found preferable in other respects; that is, to suppose the tungstic acid to be composed of 1 atom deutoxide and 1 atom protoxide united; in this case the atom of tungsten = 42, that of the protoxide = 49, that of the deutoxide = 56, and the tungstic acid = 105 as before. 27. _Oxides of titanium._ Nothing certain is known respecting the oxides of titanium. An observation of Richter, quoted by Berzelius (An. of Philos. 3--251), if it could be relied upon, furnishes an important fact, namely, that a solution of muriate of titanium containing 84.4 oxide, gave 150 muriate of silver. Now 150 muriate of silver contain 28 acid; hence 28 acid must have combined with 84.4 oxide; but if 28 ∶ 84.4 ∷ 22 ∶ 66 nearly for the weight of an atom of the oxide. This would indicate 59 for an atom of the metal. 28. _Oxides of columbium._