The author, in the course of study on the oxidative metabolic products of Gluconobacter. Pseu-domonas and Serratia, found that three γ-pyrone derivatives were produced in the culture liquid of Gluconoacetobacter liquefaciens with the medium of 0.8% yeast extract water containing 9.3%. glucose and 2.5% CaCO3 by shaking culture method. One of these γ-pyrone derivatives was studied and confirmed as a new substance to which the name of “Rubiginol” was given. It was isolated in the shape of plate or prismatic needles melting at 203.5°. Easily soluble in pyridine, dioxan, acetone, methanol and soluble in water, ethanol, isopropanol, butanol, glacial acetic acid, ethylacetate and ether, but almost insoluble in toluol and benzol, gaves reddish violet colour with FeCl3 solution, slightly acidic, reduced Fehling's solution and ammoniacal silver nitrate solution, decolourlized bromine water and potassium permanganate solution. The molecular formula was found to be C5H4O4 by elementary analysis and molecular weight. The infrared absorption spectrum showed the existence of secondary or phenolic OH and strong possibility of conjugated carbonyl group, while the existence of primary or tartiary OH, lactone and carboxyl group were denied. No existnce of carboxyl group was also recognized from the potentiometric titration, though the compound acted as dibasic acid, the pK value of stronger acid radical was almost same as that of kojic acid. The substance was easily extracted from its NaHCO3 solution by ether and could be set free by carbon dioxide. Existence of γ-pyrone ring was proved by transforming its dimethylether into pyridone deri-vative, which was obtained as colourless prismatic crystals. Existence of two OH groups were proved by the analysis of dimethylrubiginol (m. p. 143°) and diacetylrubiginol (m. p. 118°) which were both neutral and didn't give FeCl3 color reaction. Four different chemical structure could be given to rubiginol from its molecular formula as γ-pyrone derivative (see Fig. 7). Among them, the fourth substance had been already reported by R. Kaushal in 1940 and clearly different from rubiginol by its melting point (94°). Dimethyl-rubiginol was boiled with baryta and decomposition products were isolated and analysed. Dioxyacetone dimethyl ether was isolated and identified as 2, 4-dinitrophenylhydrazone, and for-mic acid was gained as Ca-salt, while the CO2 and CH3OH was not produced. From one mole of dimethyl rubiginol, one mole of dioxyacetone dimethyl ether and two moles of formic. acid were produced. From the experimental results mentioned above, the chemical structure of rubiginol was de-cided as 3, 5-dihydroxy-1, 4-pyrone.