n-Butyraldehyde and D-glucose condensed to give 4,6-O-butylidene-~glucose , which on reduction yielded 4, B-O-butylidene-~-glucitol ( 1,3-0butylidene-L-gulitol) . Periodate oxidation of either the glucose or glucitol acetal gave 2,4-O-butylidene-~-erythrose, and thence on reduction 1,3-0butylidene-L-erythritol.THIS Paper describes the synthesis and proof of structure of 4,6-0-butylidene-~-glucitol (I) (1,3-0-butylidene-~-gulitol), an acetal which, as expected, was not obtained as an endproduct by direct butylidenation of D-glucitol.n-Butyraldehyde and D-glucose condense in the presence of a trace of concentrated sulphuric acid to yield the expected 4,6-0-butylidene-~-glucose (24-38%).This acetal appears to be dimorphic, one form melting at 150" and the other at 168"; but in chemical reactions or physical properties (infrared spectra in a Nujol mull, paper chromatography and specific rotation) other than melting point the two forms are identical. Both the dimorphs would be expected to have the same conformation as 4,6-0-benzylidene-~-glucose ,2 and from the high initial positive rotation which exhibits a downward directed mutarotation, we would infer that the solid dimorphs are both a-glucoses. The a-configuration was also tentatively assigned to them for their infrared spectra 3 in Nujol mulls. The acetal was characterised through its crystalline 9-nitrophenylhydrazone, the formation of which indicates that C-1 is not involved in the acetal linkage.The acetal consumed 2 mol. of periodate in a slightly acid solution, and liberated 2 mol. of formic acid but no formaldehyde. The major periodate oxidation product was solid 2,4-0-butylidene-~-erythrose. Evidence that the erythrose acetal was not monomeric was (a) that a Nujol mull of the compound showed no carbonyl absorption in the infrared, (b) that the acetal mutarotated in solution, and (c) that a cryoscopic determination of the molecular weight using dry benzene as solvent showed that the acetal was a dimer at infinite dilution. The material became increasingly associated with increase in concentration.Using dry acetic acid as solvent again showed that the solute was associated. Schaffer found that 2,4-0-ethylidene-~erythrose in formaniide has a molecular weight corresponding to 2-16 monomer units (no concentration of solute was stated), and he proposed that two monomer units link together through a 1,3-dioxan ring. A similar structure could exist for the butylidene acetal, for which structure (11) could presumably be a favoured conformation, since the large substituents on all three rings are equatorial.The association was rapidly reversed on dilution.