The procedures by which glucose had been isolated were now applied to fraction A4, yielding 52 mgm. of a substance which showed the mobility of cellobiose when run on paper chromatograms in different solvent systems. The substance was acetylated with pyridine and acetic anhydride and the purified derivative had m.p. 191, unchanged by admixture with authentic ^-cellobiose octa-acetate, m.p. 191. (Found : C, 49-7 ; H, 5-7 per cent. Calc. for C28H38Oi9 : C, 49-6 ; H, 5-6 per cent.) The infra-red spectra of this octa-acetate and that of authentic /?-cellobiose octa-acetate were without significant differences. When lactate is metabolized by A. acetigenum in. absence of other sources of carbon it is mainly oxidized through acetate to carbon dioxide, though in certain liquid media it can serve as a source for formation of cellulose3. It seems probable that in our experiments the lactate played a dual part, some of it suffering oxidation and thereby sparing glycerol for cellulose synthesis, and some of it undergoing conversion to a metabolite on the pathway to cellulose. So far as we are aware, neither glucose nor cellobiose has hitherto been isolated from among the products of bacterial action on either glycerol or lactate, although the formation of glucose in cultures of Poria vaillantii on a glycerol medium has been reported by Sison and Schubert4, while Horecker5 has suggested a cyclic mechanism by which this conversion may be effected. It is proposed to report the work in ffurther detail elsewhere.One of us (K. R.) acknowledges the receipt of an Imperial Chemical Industries Research Fellowship.