In his key to the species of the genus Diploglossus, Boulenger (1885, p. 284) distinguishes between two groups of species, viz., one group in which the digits terminate in "a large compressed sheath, into which the claw may be entirely or nearly entirely retracted", while in the other group such a sheath is absent. Barbour (1910, p. 297) considers the presence or absence of an ungual sheath as a character of generic value ; he separates the species lacking such a sheath from the true Diploglossus, and revives the genus Celestus Gray for them 1). Burt & Burt (1931, pp. 241-242) also stress the importance of this character. Indeed the sheath is absent in Celestes occiduus (Shaw), of which Celestus striatus Gray (the type of the genus) is a synonym. Of the other species included in Celestus (Barbour, 1937, pp. 138-139) I have examined only Celestus de la sagra (Cocteau). Of the two specimens in our collection (Herp. reg. nos. 3626, 3634), one (no. 3626) is a cotype of Scincus (Diploglossus) de la sagra Cocteau (in Cocteau & Bibron, 1839, p. 180, pl. 20). In both specimens the terminal scale on the upper surface of the digits forms a