Examination of the holotype of Globorotalia barisanensis Le Roy, 1939, has revealed that this species from the Middle Miocene of central Sumatra is non-keeled, contradicting the synonymizing of G. barisanensis with Globorotalia fohsi. Because the type sample is untraceable, we examined the available type specimens (holotype and eight paratypes), notes and data presented by previous researchers, and material from the near-topotype locality of Paragloborotalia siakensis, to clarify Le Roy's concept of G. barisanensis. Our investigation has revealed that the Middle Miocene of central Sumatra contains two different groups of globorotaliids: the Globorotalia menardii group and the G. fohsi group, and that G. barisanensis belongs to the latter.
The paratypes examined belong to Globorotalia peripheroacuta and to forms transitional between G. peripheroacuta and Globorotalia praefohsi. The holotype, as described by Le Roy, is included in G. peripheroacuta as well, despite the observation that it differs from the paratypes in the abnormally inflated final chamber. In retrospect, the specimen selected as the holotype was an unfortunate choice as it has long been misinterpreted and used to denominate the earliest representatives of the G. fohsi group (viz. G. fohsi barisanensis) until G. barisanensis was reinterpreted as a junior synonym of G. fohsi based on a flawed redrawing of a fully keeled holotype of G. barisanensis. We conclude that Le Roy's concept of G. barisanensis was broad and may have included G. fohsi and Globorotalia peripheroronda as well. Remarkably, after 67 years of misinterpretation since Bolli (1950), G. barisanensis appears to be a valid name and a senior synonym of G. peripheroacuta. However, we recommend that it will be more practical to conserve the latter name, which is well established in the literature.