(3029) Ludovia subacaulis Poit. in Mém. Mus. Hist. Nat. 9: 31. 1822 [Angiosp.: Cyclanth.], nom. utique rej. prop. Typus: non designatus. Ludovia subacaulis Poit. (in Mém. Mus. Hist. Nat. 9: 31. 1822) was validly published through the provision of a short Latin diagnosis "Subacaulis erecta: foliis rigidis bisectis" and some brief descriptive commentary in French: "[…] elle a le port d'un jeune palmier dont la tige n'est pas encore développée; je me suis assuré que sa fleur est semblable à celle de l'espèce précédente [L. funifera Poit.]; elle en diffère d'ailleurs par sa tige droite, haute d'un pied au plus, et qui ne tend jamais à grimper sur les arbres" (it has the appearance of a young palm tree whose stem is not yet developed; I assured myself that the flower is similar to that of the preceding species [L. funifera Poit.]; it differs also by its straight stem, at most one foot high, and which never tends to climb trees). Poiteau (l.c.) also attests that the species is common in humid forests in the district of Roura, French Guiana. No specimens or illustrations were cited. Sprengel (Syst. Veg. 3: 772. 1826) reduced L. subacaulis to a synonym of Salmia trigona (Ruiz & Pav.) Willd. (in Mag. Neuesten Entdeck. Gesammten Naturk. Ges. Naturf. Freunde Berlin 5: 401. 1811). Subsequent authors proposed new combinations for the name: S. subacaulis (Poit.) Schult. & Schult. f. (Mant. 3: 341. 1827) and Carludovica subacaulis (Poit.) Kunth (Enum. Pl. 3: 107. 1841), without presenting any amendments to the original description. To frame his new species, Poiteau (l.c.) used the generic concept of Ludovia Pers. (Syn. Pl. 2: 576. 1807; non Ludovia Brongn. in Ann. Sci. Nat., Bot., sér. 4, 15: 361. 1861, nom. cons.), which although not technically a superfluous name under ICN Art. 52 (Turland & al. in Regnum Veg. 159. 2018) for the heterogeneous Carludovica Ruiz & Pav. (Fl. Peruv. Prodr.: 146. 1794), at the time of publication included all five of the subsequently published species (Ruiz & Pavón, Syst. Veg. Fl. Peruv. Chil.: 291–293. 1798) of Carludovica. Throughout the 19th century, the name Ludovia subacaulis, and combinations based on it, are cited at least 10 times in the literature, mainly in lists of species cultivated in botanical gardens (e.g., Planchon, Hort. Donat.: 42. 1858; Groenland in Rev. Hort. (Paris) 1860: 543. 1860; Verloet in Rev. Hort. (Paris) 1861: 39. 1861; Schaedtler in Hamburger Garten- Blumenzeitung 31: 301. 1875; Desfontaines, Cat. Pl. Hort. Paris.: 8. 1929). In the 20th century, in his monumental "Monograph of the Cyclanthaceae", Harling (in Acta Horti Berg. 18: 1–458. 1958), even under doubt, treated L. subacaulis as a terrestrial, short-stemmed form of Evodianthus funifer (Poit.) Lindm. sensu lato (in Bih. Kongl. Svenska Vetensk.-Akad. Handl. 26: 8. 1900), whose basionym is L. funifera Poit. (l.c.: 28), proposing its synonymization (Harling, l.c 1958: 265). Evodianthus funifer is a nomadic vine, rarely terrestrial, common and well distributed from Nicaragua throughout the Amazon basin, to the Brazilian Atlantic forest (Bahia and Espírito Santo states) and on the islands of Trinidad & Tobago (Harling & al. in Kubitzki, Fam. & Gen. Vasc. Pl. 3: 202–215. 1998). This species presents wide morphological variation, mainly in relation to the size, shape and depth of the partition of the leaves (Harling, l.c. 1958: 264). The main reason for Harling's synonymization was the statement by Poiteau (l.c.: 31): "je me suis assuré que sa fleur est semblable à celle de l'espèce précédente". The preceding species in question was Ludovia funifera Poit. (= E. funifer), which was published in the same work, with flowers described in detail. Harling's (l.c. 1958: 270) conclusion was that "it seems very improbable that he [Poiteau] would have overlooked the important dissimilarity which must have been present in respect of these organs, if the material of L. subacaulis had belonged to any other of the later-established genera than Evodianthus". Since then, L. subacaulis has appeared scarcely in the literature, always as a synonym of E. funifer (Harling in Opera Bot., B, 1: 1–48. 1973; Gomes & Mello-Silva in Rodriguésia 57: 159–170. 2006; Leal in Fl. Funga Brasil. 2024 [https://floradobrasil.jbrj.gov.br/FB48850, accessed on: 15 Mar 2024]), followed by other databases of scientific names (POWO. 2024 [https://powo.science.kew.org/taxon/urn:lsid:ipni.org:names:297690-1]; Tropicos. 2024 [https://tropicos.org/name/9700367]; WFO. 2024 [http://www.worldfloraonline.org/taxon/wfo-0000683644], all accessed on 15 Mar 2024). Three facts stand out when analyzing Harling's point of view. (1) Right after the statement that supported Harling's decision, after a semicolon mark, Poiteau continues (l.c.: 31): "elle en diffère d'ailleurs par sa tige droite, haute d'un pied au plus, et qui ne tend jamais à grimper sur les arbres". By joining the two clauses, a new interpretation becomes possible, suggesting that Ludovia subacaulis differs from L. funifera, not in its reproductive structures but due to its terrestrial habit. (2) Poiteau adopted Persoon's Ludovia, which coincided with Carludovica, which at the time was heterogeneous and had members with very variable floral structures that currently represent four different genera. Noting that the reproductive structures of the two species were similar, it seems unlikely that he would describe his L. subacaulis (= C. subacaulis) in any other genus than Ludovia Pers. (3) When describing his new genus Ludovia, Brongniart (l.c.) compared the female flowers of L. lancifolia Brongn. (l.c.: 362) with those of species of Carludovica cultivated in the "Jardin des Plantes" of the National Museum of Natural History in Paris. He pointed out differences between the tepals of the pistillate flowers of C. subacaulis, which are short, fleshy and truncated, and those of C. funifera, which are oval and more prominent. This evidence convinces us that L. subacaulis and L. funifera (= E. funifer) are not conspecific and that the difference between the two species was clear to both Poiteau and subsequent authors. The scarce morphological information provided by Poiteau (l.c.) for his Ludovia subacaulis is far from sufficient to link the entity to a cyclanth genus. However, we found an illustration in the Collection des vélins du Muséum national d'Histoire naturelle, Portefeuille 75, folio 18, 1834, which is labelled in pencil "Carludovica subacaulis Poit." (available at https://bibliotheques.mnhn.fr/EXPLOITATION/infodoc/digitalCollections/viewerpopup.aspx?seid=MNHN_VEL_PORTEFEUILLE075&i=MNHN_VEL_PORTEFEUILLE075_FOL018.jpg). This illustration was probably painted from material grown in Paris, sent by Poiteau during his stay in French Guiana (1819–1822) (Bureau in Nouv. Arch. Mus. Hist. Nat. 3(9): 1–94. 1897). A careful analysis of this illustration clearly shows that the species linked to the name L. subacaulis belongs to the genus Asplundia Harling (in Acta Horti Berg. 17: 41. 1954). Bifid leaves with broad, strongly tricostate segments, the lateral costae running far inside the leaf margin, are not found in the genus Evodianthus Oerst. (in Vidensk. Meddel. Naturhist. Foren. Kjøbenhavn 1857: 194. 1857). In addition, the general appearance of the inflorescences and infructescence is characteristic of a typical Asplundia species. The identity of L. subacaulis must be attributed to a terrestrial, broad-leaved species possibly occurring in French Guiana. Four (maybe two) species are possible: A. flavovaginata Harling (l.c. 1958: 194) and A. gardneri (Hook.) Harling (l.c. 1954: 42, based on Carludovica gardneri Hook. in J. Bot. (Hooker) 2: 29. 1840), with the epidermis at the base of the sheath and peduncle, yellow, shiny and brittle; or A. brachyphylla Harling (l.c. 1958: 231) and A. maguirei Harling (l.c. 1958: 228), with the epidermis at the base of the sheath dull in color and consistency, not differentiated from other parts of the petiole, sheath or peduncle. The species Asplundia flavovaginata was originally described from the montane forests of central Peru (Harling, l.c. 1958: 194). Harling (in Steyermark & al., Fl. Venez. Guayana 4: 471–486. 1998) recorded the species for the Venezuelan states of Amazonas and Bolívar. Some materials from French Guiana have been identified with this name in the CAY herbarium (first author, pers. obs.). Asplundia gardneri is a species that occurs in Brazil and has been recorded in the states of Amapá and Pará (Leal in Fl. Funga Brasil. 2024 [https://floradobrasil.jbrj.gov.br/FB7133, accessed on 15 Mar 2024]), which are adjacent to French Guiana. The species A. brachyphylla is very common and well distributed in the Guianas (Eriksson & Harling in Contr. U.S. Natl. Herb. 55: 87. 2007) and is also recorded in the Brazilian states of Amazonas, Amapá, Pará and Maranhão (Leal in Fl. Funga Brasil. 2024 [https://floradobrasil.jbrj.gov.br/FB135423, accessed on 15 Mar 2024]). Finally, A. maguirei is recorded for Venezuela, Guyana and Suriname (Harling, l.c. 1998; Eriksson & Harling, l.c.) and some specimens from French Guiana have been identified with this name (first author, pers. obs.). Our taxonomic analysis shows, therefore, that the name Ludovia subacaulis if combined in Asplundia, could possibly be applied to any of the species listed above. Since the illustration is insufficient for unequivocal determination at the species rank, the name L. subacaulis could possibly be neotypified with a specimen belonging to any of those species (see Art. 9.8 of the ICN). Such a typification would not be in conflict with its protologue, and thus would need to be accepted. In this way, the name L. subacaulis, if transferred to Asplundia, could become the correct name for one of the species currently called A. flavovaginata, A. gardneri, A. brachyphylla, or A. maguirei because the name L. subacaulis predates all of these names. However, substituting any Asplundia name in current usage for an obscure and little-used name, synonymized for almost 70 years under a species of another genus, would be undesirable and evidently disadvantageous to nomenclatural stability. Thus, with the objective of protecting the current use of these scientific names in Asplundia and to prevent an unwanted nomenclatural change, and in accordance with Art. 56 of the ICN, the name L. subacaulis is here proposed for rejection. It will most likely remain impossible to ascertain the identity of L. subacaulis and we consider it more appropriate to have this name rejected than to propose an arbitrary typification. Acceptance of this proposal would allow Asplundia flavovaginata, A. gardneri, A. brachyphylla and A. maguirei to continue to be used as the correct names for these species, without threat from an arbitrary application or typification of Ludovia subacaulis. Rejection of this proposal would mean that L. subacaulis could eventually replace any Asplundia species cited above as its correct name. We are grateful to John Wiersema and John McNeill for their advice and valuable comments that improved this proposal.