Rankin Rodriguez, R. & Greuter, W.: Humboldt, Willdenow, and Polygala(Polygalaceae). – Taxon 50: 1231–1247.2001. ‐ ISSN 0040–0262. Based on published sources (including contemporary correspondence) and the Humboldt expedition's field notebooks, some information on Humboldt and Bonpland's plant collections in their relation to Willdenow and his Species plantarumis provided. The genus Polygalaserved as a test sample. The Willdenow Herbarium (B‐W) holds 23 Polygalaspecimens from the Humboldt expedition (several others were erroneously attributed to Humboldt), representing 17 of the 18 species described by Kunth in the Nova genera.The following main conclusions are drawn: (1) three types of handwriting can be recognised on the expedition's field labels in B‐W, which are identified as Bonpland's, Humboldt's, and a third one that is tentatively ascribed to an unknown field assistant; (2) a full set of the plant collections from Venezuela was sent to Willdenow from Havana early in 1801 and had been received in Berlin by September 1802; (3) in vol. 3(2) of Willdenow's Species plantarum, published in early November 1802, a single species (Hypericum caracasanum) was described based on the new Humboldt material; (4) the numbers found on many of Humboldt's labels refer to the numbered entries in the field notebooks, but are not collection numbers in the modern sense, because many collections, on which no notes were taken, remained unnumbered; (5) numbering was fairly consistent, all duplicates of collections on which field notes exist being normally numbered, except when the label was accidentally lost; (6) the set of duplicates given by Humboldt to Willdenow was virtually complete, but the equally complete set in Bonpland's herbarium (P‐Bonpl.), having been studied by Kunth in the first place, should be used to typify the names of new taxa first published in the Nova genera.Based on an improved understanding of the background of these collections, a recently proposed type designation for Polygala tenellaWilld. is rejected. This long forgotten name is here neotypified in such a way as to retain its traditional place in the synonymy of P. paniculataL.