In 1908 Prof. E. Dubois cited and very briefly diagnosed a new species of emydine turtle, Hardella isoclina, from the Trinil Beds in Java. The very fine unique type shell has never been figured or fully described, the generic assignment appears to be incorrect, and the original diagnosis is insufficient. Prof. Boschma and Dr. Brongersma of the Rijksmuseum van Natuurlijke Historie in Leiden have, as custodians of the Dubois Collection, very generously offered me the opportunity of redescribing this form. The fossil is a very distinct species and a very noteworthy component of the Trinil fauna but even after much study and comparison it proves impossible to refer it with confidence to any known genus, and in the absence of any knowledge of the characters of the skull no new genus can be satisfactorily defined. The Trinil form appears to resemble most closely the living species mutica from southern China, Formosa, Hainan, and Japan, and since mutica is currently though questionably assigned to the genus Clemmys it will be convenient for the present to refer the Dubois species to Clemmys with a query: Clemmys? isoclina (Dubois) Diagnosis: an emydine resembling "Clemmys" mutica but differing in greater size, in having the posterior margin of the carapace not at all serrate, in having the gular region not produced and the inguinal scute larger. Type: Dubois Collection No. 2722, an almost perfect shell, lacking only the anal region of the plastron. Type locality: Kedoeng Panas, Java. Horizon: Pleistocene, Trinil Beds.