Lead tetraacetate is highly selective for oxidation of a-hydroxy-hemiacetal groups and hence most readily attacks cyclic forms of the sugars. The reaction proceeds stepwise: the hemiacetal a-glycol being cleaved and the monoester of a correspondingly shorter-chained sugar formed. After cyclization the new sugar in turn is oxidized a t the hemiacetal a-glycol group t o yield a diester of a still-lowerorder member of the series. I n this manner D -~~L I C O S~ first yields mono-0-formyl-D-arabinose and then di-0-formyl-D-erythrose. Similarly, D -~~U C~O S~ is degraded t o a glycolate ester of D-erythrose and finally t o a formate-glycolate diester of D-glyceraldehyde. Some relatively rare sugars thus may conveniently be prepared directly from abundant higher-order members of the series. The reactions appear to involve preferential attack of the furanose form of a sugar rather than of the normally-predominant pyranose form, or possibly migration of ester groups towards the reducing end of the sugars.Criegee noted that D-glucose rapidly consumes about three moles of lead tetraacetate i11 warm acetic acid without concurrent production of formaldehyde (5). He suggested that the apparent failure of the oxidant to attack the 5,6-glycol group was due to the presence of the 1,5-hemiacetal oxygen bridge, and that the sugar must therefore be oxidized a s a ring, rather than in the open-chain form. According to Hockett and Zief (14) D-glucose a t 20°C. quickly consumes only two moles of lead tetraacetate and further oxidation is very slow. If the sugar exists in solution predomiilailtly as D -g l~~~p y r a I I~s e (I) it would be expected to consume a t least three moles of oxidant, as had been found by Criegee, and possibly yield 2-0-formyl-D-glyceraldehyde (11). For the related oxidant, periodic acid, this possibility has been verified recently by Schopf and Wild (28) who isolated D-glyceraldehyde moiloformate by treating D-glucose with a limited quantity of periodate. In contrast to these results, Abraham (1) reported that D-glucose in warm aqueous acetic acid is completely but slowly degraded by lead tetraacetate to five moles of formic acid and one mole of formaldehyde, the reaction being recommended for radioactive-carbon assay. Hence the literature records a t least three different sets of d a t a for the reaction of D-glucose with lead tetraacetate. These variations are attributable most probably to the differences in the oxidation conditions employed but they invited further examination of the reaction.The results of Hocltett and Zief were of particular interest since the consumption of only two moles of oxidant would be expected to afford a derivative of a tetrose, rather than of a triose. Accordingly, D-glucose, in glacial acetic acid or in acetic acid containing 1-2% of water (2), was treated a t room temperature with excess lead tetraacetate. Two moles of oxidant were consumed within three minutes and oxidation proceeded subsequently a t a 1Manuscript