The 2-desoxy-ribose nucleosides, first isolated by Levene and London’ by the hydrolysis of thymus nucleic acid, are characterized by their extreme ease of hydrolysis by very dilute mineral acids. Thus, guanine-2-desoxy-riboside is completely hydrolyzed by heating with 0.01 N hydrochloric acid during 5 minutes. Since the rate of hydrolysis is of the same order as that of the furanosides, it was considered possible that the 2-desoxy-ribonucleosides, similarly to the ribonucleosides, have the furanoside ring structure. When, however, it was discovered2 that methyl-2-desoxyglucopyranoside is almost as readily hydrolyzed, it appeared just as feasible that the natural derivatives might have the pyranoside ring structure. Hence the question of their ring structure was in need of further special investigation. We have now studied the ring structure of the sugar portion of thymidine (thymine-2-desoxy-riboside). This substance is found to react with triphenylmethyl chloride in pyridine to give a monotrityl derivative. Since it was shown by us3 that the analogous monotrityl uridine is the 5-substituted derivative, the formation of a monotrityl thymidine might, of itself, be considered a good indication that thymidine is, likewise, a furanoside. Definite proof of this conclusion was provided by substituting the remaining free hydroxyl group by a tosyl group and examining the behavior of the resulting monotosyl trityl thymidine on treat-