Type. In Serra de Spilon, Minas Gerais, Brazil, May 1818, Pohl 2827.
Schumann [Mart. Fl. Bras. 6(6): 206. 1889] based his section Pattalosia on this species, supposedly characterized by its entire stigmas, included stamens, and erect corolla lobes. Schumann (p 207) stated that Pohl’s description and plate were incorrect in the features emphasized in Schumann’s delimitation of the section Pattalosia. However, an examination of the isotype specimen of Pohl’s collection at NY, verified by additional collections made by Dr. Howard Irwin et al, prove the correctness of Pohl’s original observations. The stigma is found to consist of two lobes and is not entire, as presumed by Schumann. Pohl correctly depicted this bilobed stigma in his original plate 108. Although the stigmas do stick together, being applanate, they could be falsely thought to represent a single unit, but they can be separated by dissection.
On the type specimen the stamens are exserted beyond the orifice, but reach about 3/4 the length of the corolla lobes, these latter when spreading easily displaying the exserted stamens. Schumann likewise maintained that the corolla lobes were erect and not spreading, as shown by Pohl’s plate, and concluded that the stamens were included. In all these respects Schumann erred and his key for Ferdinandusa in Martius’s Flora Brasiliensis was based on his erroneous assumptions that led him to separate F. speciosa on characters supposedly different from the other species assigned to his section Gomphosia. As Dr. Irwin’s ample collections further corroborate Pohl’s description and plate, the stamens are exserted, the corolla lobes are spreading, and the stigmas are two-lobed.
Typical F. speciosa has the lower surface, midrib, and lateral nerves either completely glabrous, or the midrib and lateral nerves may show some pilosity. Pohl’s description was based on a specimen having completely glabrous lower leaf surfaces, nerves, and midrib, as well as glabrous branches of the inflorescence.