Bignonia ‌squalus was described by Vellozo (1829) as ‘Habitat maritimis’, which is recognized as a place in Rio de Janeiro state. Amphilophium ‌crucigerum is a native plant, but not endemic to Brazil. According to Gentry (1977), Lohmann & Taylor (2014), and Thode (2023), this species is found in semideciduous to evergreen forests, including disturbed vegetation. In Brazil, it is found in Amazônia, Caatinga, Cerrado, Brazilian Atlantic Forest, Pampa, and Pantanal.

A copy of the illustration of Bignonia ‌squalus stored in the Manuscript Section of Torre do Tombo can be accessed under the code PT-TT-MSLIV-2776_m0033 [photo!].

Gentry (1975a) considered Bignonia ‌squalus as a synonym of Pithecoctenium ‌crucigerum (L.) A.H.Gentry [Amphilophium ‌crucigerum]. This position was also accepted in later treatments (Pool 2007; Zuntini 2014), and it is reinforced here.

The Bignonia ‌squalus plate well represents Amphilophium ‌crucigerum and includes relevant diagnostic features, such as the equinate capsule of the fruit (Reiche et al. 2020). In addition to the fruit morphology, 2 or 3 leaves with multifid terminal tendril, cordate base, acuminate apex, 5-dentate calyx, and corolla shape form also agree with A. ‌crucigerum ‌(Reiche et al. 2020). However, there is a likely mistake in the line drawing of one leaf, where two small leaves were illustrated in the central foliole. Besides, the leaves’ venation was not represented in the illustration. The original descriptions include features, such as ‘caulis volubilis’ [twining vine] and ‘Corolla flavescens’ [yellow corolla].

Bignonia ‌triphylla ‌Vell. was described by Vellozo (1829) as ‘Habitat maritimis Pharmacopolitanis’, which is recognized as someplace in the Brazilian Atlantic Forest in Rio de Janeiro state, likely in Paraty municipality, but there are no recent samples of this species from this area. Pleonotoma ‌tetraquetra is a native plant endemic to Brazil. Following Lohmann & Taylor (2014) and Gomes (2023), this species is found in the Cerrado and the Brazilian Atlantic Forest.

Bignonia ‌triphylla ‌Vell. is a later homonym of Bignonia ‌triphylla L. (1763), therefore it is illegitimate (Art. 53.1 of the ICN, Turland et al. 2025). Although it seems like a citation of a Linnean name, B. ‌triphylla ‌Vell. and B. ‌triphylla L. are taxonomically different, since the protologue of the species presents contrasting descriptions, as the leaves: ‘foliis ternatis’, ‘foliolis ovatis’ in B. ‌triphylla L., and ‘foliis triternatis’, ‘foliolis lanceolatis’ in B. ‌triphylla ‌Vell. Furthermore, the Vellozo plant is a Pleonotoma ‌tetraquetra, and this genus is not distributed in Mexico.

A copy of the illustration of Bignonia ‌triphylla ‌Vell. stored in the Manuscript Section of Torre do Tombo is accessed under the code PT-TT-MSLIV-2776_m0035 [photo!].

In addition, Bignonia ‌tetraquetra was published without the indication of a holotype, only mentioning the collector and collection locality: ‘Brasilia ‌tropica. Sellow.’ Thus, following McNeill (2014), it is not possible to identify a holotype. It is therefore necessary to designate a lectotype from available materials or, if no original material exists, a neotype.

The principal herbarium of Friedrich Sellow’s collections was deposited in B, but this material was destroyed along with the great part of his collection (Merrill 1943). Gomes (2006) referred to a material as the ‘holotype’, but the specimen wasn’t directly examined by the author. Subsequently, Lohmann & Taylor (2014) supported the selection of the specimen from LE as the ‘holotype,’ but it is likely that these authors also did not examine the voucher mentioned by Gomes (2006). It remains unclear why they chose a specimen from LE without an available image, especially since LE is not the principal herbarium for Sellow’s collections (Stafleu & Cowan 1976-1988).

Based on all collections we reviewed, we have selected the specimen at BM (BM000882582 [photo!]) as the lectotype. This specimen bears a Berlin label and was examined and identified by Beatriz M. Gomes in 2008 – a specialist in Pleonotoma.

Gentry (1975a) relates Vellozo’s line drawing belonging to Pleonotoma ‌tetraquetra, and in the revision of Pleonotoma, ‌Miers maintained this decision (Gomes 2006), and it is reinforced here.

Bignonia ‌triphylla ‌Vell. well represents Pleonotoma ‌tetraquetra and includes relevant diagnostic features, such as the trifoliate leaves and the corolla that is narrowly bell-shaped with rounded apices (Gomes 2006). Although a branch with a possible four-divided tendril is detailed, it is interpreted as a mistake once this species is known by its trifid tendril (Reiche et al. 2020). The original description includes features, such as ‘caule quadrangulari, cirrhoso’ [quadrangular stem, with tendrils].

Bignonia ‌ignea was described by Vellozo (1829) as ‘Habitat fruticetis mediterraneis’, which is recognized as a Cerrado vegetation in São Paulo state, likely Cunha municipality, but there are no recent samples of this taxon. Pyrostegia ‌venusta is a native plant, but not endemic to Brazil. Following Lohmann & Taylor (2014) and Lohmann (2023e), this species is found in dry to humid lowland forests. In Brazil, it is found in the Amazon Forest, Brazilian Atlantic Forest, Caatinga, Cerrado, Pampa, and Pantanal.

Presl (1845) did the first treatment of Bignonia ‌ignea and proposed Pyrostegia ‌ignea. Later, Gentry (1975a) included B. ignea and P. ‌ignea as synonyms of Pyrostegia ‌venusta.

A copy of the illustration of Bignonia ‌ignea stored in the Manuscript Section of Torre do Tombo can be accessed under the code PT-TT-MSLIV-2776_m0037 [photo!].

The lectotype of Bignonia ‌ignea well represents Pyrostegia ‌venusta and includes relevant diagnostic features, such as the trifoliate leaves with terminal trifid tendril, inflorescence morphology, and flower color (Reiche et al. 2020). Although the plate also details what is interpreted as a mistake: two simple tendrils, and the not-often-cordate base of the leaflets. The original descriptions include features, such as ‘Caulis ‌teres, volubilis’ [Cylindrical stem, twining], and ‘corolla ignea’ [red corolla].

Bignonia ‌trifoliata was described by Vellozo (1829) as ‘Habitat maritimis Regii Prædii S. Crucis’, which is recognized as a place in Rio de Janeiro state, in the Southwest at Royal Santa Cruz farm region. In the past, the farm occupied an area of 50 square kilometers, located in the Province of Rio de Janeiro, extending from the coast (east) to reach the river Paraíba do Sul (west), including the regions called Piraí and Itagui (Pedroza 2018). Adenocalymma ‌trifoliatum is a native plant endemic to Brazil. Following Fonseca & Lohmann (2019) and Fonseca (2023a), this species is found in the Brazilian Atlantic Forest Restinga and Cerrado. It is distributed only in the Rio de Janeiro state.

Adenocalymma ‌longeracemosum ‌DC. is considered a superfluous name (Art. 52.1 of the ICN, Turland et al. 2025), because Candolle (1845), when publishing this name, included Bignonia ‌trifoliata as a synonym, instead of providing the combination Adenocalymma ‌trifoliatum. Laroche (1973) provided the combination of Adenocalymma ‌trifoliatum, and this position was followed by later authors (Gentry 1975a; Udulutsch et al. 2013; Lohmann & Taylor 2014; Zuntini 2014; Fonseca & Lohmann 2019), and it is reinforced here.

A copy of the illustration of Bignonia ‌trifoliata stored in the Manuscript Section of Torre do Tombo can be accessed under the code PT-TT-MSLIV-2776_m0039 [photo!].

The illustration includes relevant diagnostic features, such as the 2 or 3-leaves, lanceolate, entire margin, without tendril, short-pedunculate flowers (Laroche 1973). However, the venation of the leaves, which is a diagnostic feature for the species, was not represented in the illustration. To address this, Fonseca & Lohmann (2019) designated an epitype to stabilize the use name, but in our typification, the correct lectotype was chosen, and the epitype was also designated. Following their approach, we have re-designated the same voucher as the epitype in this study. The original descriptions include other features, such as ‘Caulis ‌scandens, teres’ [Scandent stem, cylindrical].

Bignonia ‌falcata was described by Vellozo (1829) as ‘Habitat silvis maritimis Pharmacopolitanis’, which is interpreted as someplace in Rio de Janeiro state (see comments under Bignonia ‌triphylla). Amphilophium ‌falcatum is a native plant, but not endemic to Brazil. Following Lohmann & Taylor (2014) and Thode (2023), this species is found in humid vegetation. In Brazil, it is found in the Brazilian Atlantic Forest and Cerrado.

Pool (2007) decided to treat Bignonia ‌falcata as a legitimate name and lectotypified part of the illustration. In view of Article 8.2 of the ICN for the typification of a species, it is possible to use a complete sample or part of it (Turland et al. 2025). Thus, Pool (2007) made the typification of part of the illustration [branchlet, leaves, fruit, and seeds] and excluded inflorescence and flowers that, obviously, are very different. Amphilophium ‌falcatum can have a long inflorescence with 25-28 flowers, and the calyx of the flowers is noticeably larger than that depicted by Vellozo. Thus, to address the lack of information about the flowers, Pool (2007) designates an epitype. Following her approach, we have designated the same voucher as the epitype in this study because we also designated the lectotype.

Lohmann & Taylor (2014) made generic delimitation changes and proposed a new combination, Amphilophium ‌falcatum, maintaining the wrongly assigned lectotype plus the epitype selected by Pool (2007). However, even considering the author’s expertise in delimiting the typification to legitimize the use of Bignonia ‌falcata, she ended up typifying the species based on the published set of illustrations (Vellozo 1831; see comments under Bignonia ‌squalus). Due to that, following the same separation made in Pool (2007) and the recommendations of Pastore et al. (2022), a part of the original illustration present at Biblioteca Nacional is chosen as the lectotype (Fig. 2).

A copy of the illustration of Bignonia ‌falcata stored in the Manuscript Section of Torre do Tombo can be accessed under the code PT-TT-MSLIV-2776_m0041 (Noting that the species is recognized only by the branchlet, leaves, fruit, and seeds and excluding inflorescence and flowers) [photo!].

Bureau & Schumann (1896-1897) recognized the identity of Bignonia ‌falcata as a synonym of Neves-armondia cordifolia (Mart. ex DC.) K.Schum. However, Gentry (1975a) interpreted Bignonia ‌falcata as a chimera of two species, once the vegetative part well represents Neves-armondia cordifolia, whereas the flowers illustrated were very dissimilar and justified the rejection under Article 40 of ICN Code (Turland et al. 2025). Later, Gentry (1979) revised the case of N. ‌cordifolia and Pithecoctenium ‌cordifolium ‌DC., proposing a new name: Pithecoctenium ‌hatschbachii ‌A.H.Gentry.

Pool (2007) affirmed the arguments previously used that there is no support for Bignonia ‌falcata rejection, thus she proposed the combination ‘Pithecoctenium ‌falcatum’. To validate the application of the name, she anchored to the interpretation of the illustration, choosing part of the published plate (Vellozo 1831) as lectotype. Since then, many authors have followed this organization (Lohmann & Taylor 2014; Zuntini 2014), and it is reinforced here.

Bignonia ‌falcata provides information about the leaves, fruits, and seeds, but lacks information on the flowers. In fact, upon examining all Bignonia illustrations, the appearance of the flowers closely resembles the species now known as Lundia species. Therefore, it is possible that the illustrator became confused and made this mistake.

Bignonia ‌conjugata was described by Vellozo (1829) as ‘Habitat maritimis Regii Prædii S. Crucis’, which is recognized as someplace in Rio de Janeiro state (see comments under Bignonia ‌trifoliata). Fridericia ‌conjugata is a native plant, but not endemic to Brazil. Following Lohmann & Taylor (2014), and Kaehler (2023a), this species is found in humid forests. In Brazil, it is found in the Amazon Forest, Brazilian Atlantic Forest, Cerrado, and Pantanal.

Gentry (1975a) recognized Bignonia ‌conjugata citing the combination: Arrabidaea ‌conjugata. Later, Lohmann & Taylor (2014) made generic delimitation changes and proposed Fridericia ‌conjugata. This position is reinforced here.

A copy of the illustration of Bignonia ‌conjugata stored in the Manuscript Section of Torre do Tombo can be accessed under the code PT-TT-MSLIV-2776_m0043 [photo!].

Bignonia ‌conjugata well represents Fridericia ‌conjugata and includes relevant diagnostic features, such as the axillary inflorescence and campanulate flowers (Reiche et al. 2020). Besides, the illustrated leaves also seem to agree with many collections, for example: Amorim et al. 1252 (NY), Paniagua et al. 1212 (MBM), Tamashiro et al. 208 (UEC). Although there are leaves without venation, absence of simple tendril, and tubular calyx instead of campanulate (Reiche et al. 2020). Even so, the lack of this characteristic does not harm the recognition of the species. In addition, the original description includes features, such as ‘Corolla violacea’ [violet Corolla].

Bignonia ‌coito was described by Vellozo (1829) as ‘Habitat silvis maritimis ad declivia Alpium Jerissinò’, which is recognized as a reference to the [base of the] Gericinó Massif, located in the now called ‘Serra do Mendanha’ in Rio de Janeiro state. Bignonia ‌corymbosa is a native plant, but not endemic to Brazil. Following Lohmann & Taylor (2014) and Zuntini (2023), this plant is found in disturbed vegetation or semideciduous to evergreen forests. In Brazil, it is found in the Amazon Forest, Brazilian Atlantic Forest, and Cerrado.

The combination Bignonia ‌corymbosa ‌(Vent.) L.G.Lohmann (2008) would be proposed based on Spathodea ‌corymbosa ‌(Vent.) Vent. (Ventenat 1807). However, the name B. corymbosa ‌Vent. (1804) already existed and therefore Bignonia ‌corymbosa ‌(Vent.) L.G.Lohmann is a later homonym, hence illegitimate (Art. 53.1 of the ICN, Turland et al. 2025). Nonetheless, several studies mentioned the inappropriate name instead of the correct one (Lohmann & Taylor 2014; Lohmann et al. 2018; Costa et al. 2019; Costa et al. 2021; Santoset al. 2021). This situation needs to be resolved, and Bignonia ‌corymbosa ‌Vent. must be appropriately used.

A copy of the illustration of Bignonia ‌coito stored in the Manuscript Section of Torre do Tombo can be accessed under the code PT-TT-MSLIV-2776_m0045 [photo!].

Gentry (1975a) treated Bignonia ‌coito as a synonym of Phryganocydia ‌corymbosa. The lectotype of B. ‌coito well represents Bignonia ‌corymbosa ‌Vent. and includes relevant diagnostic features, such as the cylindrical branchlets, 2-leaves, spathaceous calyx, and infundibuliform corolla (Costa et al. 2019), although one of the terminal branches seems to have two petioles growing in the same place. Maybe the illustrator wanted to represent a 2-leaf but misinterpreted it. The leaves are not fully represented; they are always folded (or part of them). Nevertheless, this situation does not harm the recognition of the species. The original description includes an important feature: ‘Corolla ampla, flavescens’ [Wide corolla, yellow].

Bignonia ‌unguis ‌Vell. was described by Vellozo (1829) as ‘Habitat silvis maritimis Paratyensibus’, which is recognized as the Brazilian Atlantic Forest in the Rio de Janeiro state, and here interpreted as the Paraty municipality region. In 1667, the ‘Villa de Nossa Senhora dos ‌Remédios de ‘Parati’ (or ‘Paraty’) was established, situated near the Paratiguaçu River (now Perequê-Açu) and the Patitiba River (Nascimento 2005; IBGE 2023). Thus, Vellozo (1829) may be referring to the region near the city because he refers to Paraty, as noted by Pastore et al. (2021), using ‘Pharmacopolitanis’. Dolichandra ‌unguis-cati is a plant native but not endemic to Brazil. Following Gentry (1973a), Fonseca et al. (2017), and Fonseca (2023b), this species is found in dry forests, moist forests, or rarely in wet forests and tropical wet forests. In Brazil, it is found in the Amazon Forest, Brazilian Atlantic Forest, Caatinga, Cerrado, Pampa, and Pantanal.

Bignonia ‌unguis ‌Vell. is here interpreted as a citation to Bignonia ‌unguis L. (1759). Linnaeus initially published Bignonia ‌unguis-cati, and, in 1759, changed its spelling, removing ‘-cati’ (Linnaeus 1753; 1759). Therefore, Vellozo (1829) gave preference to (or eventually, only known) the later spelling ‘Bignonia ‌unguis’ from Linnaeus (1759). Thus, in this case, there would be no need to designate a ‘type material’, but rather to mention the ‘reference material’ deposited in the BN (mss1198655_024 [photo!]). A copy of the illustration of Bignonia ‌unguis ‌Vell. stored in the Manuscript Section of Torre do Tombo can be accessed under the code PT-TT-MSLIV-2776_m0047 [photo!].

This decision follows the ICN code, as we assume that Bignonia ‌unguis L. and B. ‌unguis ‌Vell. share the same type, allowing both names to be inserted as orthographic variants (Art. 61.2 of the ICN, Turland et al. 2025).

Gentry (1975a) considered Bignonia ‌unguis ‌Vell. as a synonym of Macfadyena ‌unguis-cati, but he asserted, according to Gentry (1972), that Dolichandra ‌unguis-cati was the most used name. This position was also accepted in later treatments (Lohmann & Taylor 2014; Fonseca et al. 2017), and it is reinforced here.

The plate of Bignonia ‌unguis well represents Dolichandra ‌unguis-cati and includes relevant diagnostic features, such as the 2-leaves with a terminal modification into a trifid tendril and campanulate calyx, inflated and truncate with a straight to sinuous margin (Fonseca et al. 2017). Although the illustration lacks representations of striate prophylls, these omissions do not hinder species recognition. The original description includes features, such as ‘Caulis ‌scandens, teres’ [Scandent stem, cylindrical].

Bignonia ‌ternata was described by Vellozo (1829) from ‘Habitat silvis maritimis’, which is recognized as someplace in Rio de Janeiro state (see comments under Bignonia ‌squalus).

Gentry (1975a) mentioned Adenocalymma ‌ternatum as an established name. However, the author was unsure about the name without observing the collection mentioned by Bureau & Schumann (1896-1897). Later, Lohmann & Taylor (2014) and Zuntini (2014) also accepted Adenocalymma ‌ternatum.

Nevertheless, Gentry (1975a) himself already said that the species is rarely collected, and he has never seen it. Recently, Fonseca & Lohmann (2019) published an update about the novelties of Adenocalymma, but Adenocalymma ‌ternatum is not addressed to any species. Fonseca (2023a) mentioned Bignonia ‌ternata as Adenocalymma ‌acutissimum ‌(Cham.) Miers synonym, even treating Adenocalymma ‌ternatum as ‘in doubt’. Indeed, Bignonia ‌ternata is a name with difficult resolution, which explains the different interpretations. The reason is that the illustration does not have conclusive details (L.H.M. Fonseca, pers. comm.), meaning it is devoid of specific characters.

A copy of the illustration of Bignonia ‌ternata stored in the Manuscript Section of Torre do Tombo can be accessed under the code PT-TT-MSLIV-2776_m0049 [photo!].

Previous analyses (Bureau & Schumann (1896-1897); Lohmann & Taylor 2014; Zuntini 2014) accepted Bignonia ‌ternata [Adenocalymma ‌ternatum]. Meanwhile, other authorities believe that the illustration is insufficient for full recognition (L.H.M. Fonseca, pers. comm.).

The lectotype of Bignonia ‌ternata does not present diagnostic features to recognize an Adenocalymma species. The 3-leaves modified (or not) terminal foliole on the simple tendril, axillary inflorescence, and long flowers are common in several species. Therefore, we agree with L.H.M. Fonseca (pers. comm.) and believe it is better to leave Bignonia ‌ternata without identification (incertae sedis) until a conclusive interpretation emerges.

Bignonia ‌scandens was described by Vellozo (1829) as ‘Habitat silvis maritimis’, which is recognized as someplace in Rio de Janeiro state (see comments under Bignonia ‌squalus). Anemopaegma ‌chamberlaynii is a native plant, but not endemic to Brazil. Following Lohmann & Taylor (2014) and Firetti (2023), this species is found in dry to wet forests. In Brazil, it is found in the Brazilian Atlantic Forest, Caatinga, and Cerrado.

A copy of the illustration of Bignonia ‌scandens stored in the Manuscript Section of Torre do Tombo can be accessed under the code PT-TT-MSLIV-2776_m0051 [photo!].

Gentry (1975a) considered Bignonia ‌scandens as a synonym of Anemopaegma ‌chamberlaynii. This position was also accepted in later treatments (Lohmann & Taylor 2014; Zuntini 2014), and it is reinforced here.

Bignonia ‌scandens well represents Anemopaegma ‌chamberlaynii and includes relevant diagnostic features, such as the 3 or 2 leaves with modified terminal foliole on the simple tendril, ovate leaves, acuminate apex, inflorescence with many flowers, infundibuliform corolla, fruit morphology, and winged seeds (Reiche et al. 2020). Although the complete leaf venation was not represented in the illustration, the lack of this characteristic does not harm the recognition of the species. Even though it is common to find descriptions that indicate the presence of ‘trifid tendrils’ instead of ‘simple tendrils’ (Reiche et al. 2020), some materials of this species presented leaves with simple tendrils (e.g., Hatschbach 39842 (MBM); Furlan et al. CFSC6728 (SPF); Zuntini et al. 213 (SP).

Bignonia ‌angrensis was described by Vellozo (1829) as ‘Habitat silvis maritimis Angrensibus’, which is recognized as a place in Rio de Janeiro state, here interpreted as Angra ‌dos Reis municipality region.

Although Angra ‌dos Reis was only officially elevated to the status of a city with this name in 1835 (Capaz 2006; Soares & Rodrigues 2020), the name ‘Angra ‌dos Reis’ was already in use long before that. Around 1593, the ‘Freguesia de Nossa Senhora da Conceição de Angra ‌dos Reis’ was established. After that, in 1608, the locality was elevated to ‘Vila de Angra ‌dos Santos Reis Magos da Ilha Grande’ (APERJ 2012; Rocha 2013). On the official documents, the locality is commonly found under the names ‘Nossa Senhora da Conceição’ and ‘Vila da Ilha Grande’ (Capaz 2006; Soares & Rodrigues 2020). Besides, the use of the name ‘Angra ‌dos Reis’ in the region before the realization of Floræ Fluminensis can be proven through parts of letters documented by Rocha (2013) in documentary analysis, such as ‘Angra ‌dos Reis da Ilha Grande servimos este presente anno de 1749…’ and ‘A Igreja Parochial da Villa de Angra ‌dos Reys da Ilha grande se acha…’. Thus, based on these explanations, it was concluded that Bignonia ‌angrensis was collected in the current territory of Angra ‌dos Reis municipality.

A copy of the illustration of Bignonia ‌angrensis stored in the Manuscript Section of Torre do Tombo can be accessed under the code PT-TT-MSLIV-2776_m0053 [photo!].

Gentry (1975a) highlights certain characteristics that may cause confusion due to anomalous details but suggests Bignonia ‌angrensis as a synonym of Arrabidaea ‌corallina ‌(Jacq.) Sandwith (Tanaecium ‌dichotomum ‌(Jacq.) Kaehler & L.G.Lohmann). However, Bignonia ‌angrensis has not appeared recently as a synonym of any current names (Lohmann & Taylor 2014), being placed as ‘nomina dubia’ in Zuntini (2014), and it is not mentioned in the Tanaecium synopsis (Frazão & Lohmann 2019).

Nevertheless, it was recently ascertained that Tanaecium ‌dichotomum has no occurrence in RJ, RS, SP, and SC states (Frazão 2023). Thus, Gentry’s (1975a) hypothesis is put in doubt. Frazão & Lohmann (2019) discussed the large morphological variation of T. ‌dichotomum, and the species may be confused with others, such as Tanaecium ‌selloi. Therefore, the illustration and plant collections from Rio de Janeiro were carefully analyzed, but we did not find morphological characteristics to suggest a synonymization assertion.

The lectotype of Bignonia ‌angrensis resembles Tanaecium ‌selloi and exhibits relevant features, such as 3-leaves and a terminal thyrsus. However, the calyx is not truncate, and the corolla also does not match that of Tanaecium ‌selloi ‌(Frazão & Lohmann 2019). It seems like the plate has not been well drawn, and the original descriptions also lack diagnostic features. Thus, in favor of nomenclatural stability, the absence of debates about the true identity of the illustration, and in agreement with Gentry (1975a), this name should remain unidentified (incertae sedis).

Bignonia ‌fluminensis was described by Vellozo (1829) as ‘Habitat silvis maritimis’, which is recognized as someplace in Rio de Janeiro state (see comments under Bignonia ‌squalus). Pleonotoma ‌fluminensis is a native plant endemic to Brazil. Following Lohmann & Taylor (2014) and Gomes (2023), this species is found in the Brazilian Atlantic Forest.

A copy of the illustration of Bignonia ‌fluminensis stored in the Manuscript Section of Torre do Tombo can be accessed under the code PT-TT-MSLIV-2776_m0055 [photo!].

Miers (1863) noted Bignonia ‌fluminensis and proposed the combination Memora ‌fluminensis. Later, Sandwith (1962) commented that Memora ‌fluminensis is identifiable with Vellozo’s description and illustration. Gentry (1975a) transferred Bignonia ‌fluminensis to Pleonotoma but made it clear that this could be a very rare species and perhaps already extinct. This position was also accepted in later treatments (Gomes 2006; Lohmann & Taylor 2014; Zuntini 2014), and it is reinforced here. However, due to the low number of collections (only two), it has been classified as ‘questionable’ in the genus review (Gomes 2006).

Bignonia ‌fluminensis well represents Pleonotoma ‌fluminensis and includes relevant diagnostic features, such as the trifoliate leaves with a retuse or rounded apex (Gomes 2006). Although the venation was not represented in the illustration, the lack of this characteristic does not harm the recognition of the species.

Pleonotoma ‌fluminensis exhibits morphological similarities with Pleonotoma ‌tetraquetra ‌(Cham.) Bureau (1868), a resemblance previously observed and documented by Gentry in a Herbarium voucher (Fig. 3) and also seen by Gomes (2006). Additionally, this study identified a resemblance between P. ‌fluminensis and P. ‌tetraquetra var. tetraquetra proposed in Gomes (2006). It would not be surprising if, in the future, they were to be recognized as the same species.

Bignonia ‌fasciculata was described by Vellozo (1829) as ‘Habitat silvis maritimis Regii Prædii S. Crucis’, which is recognized as someplace in Rio de Janeiro state (see comments under Bignonia ‌trifoliata). Tynanthus ‌fasciculatus is a native plant endemic to Brazil. Following Lohmann & Taylor (2014) and Medeiros (2023), this species is found in humid forests. In Brazil, it is found in the Brazilian Atlantic Forest and Cerrado.

Nomenclatural notes

A copy of the illustration of Bignonia ‌fasciculata stored in the Manuscript Section of Torre do Tombo can be accessed under the code PT-TT-MSLIV-2776_m0057 [photo!].

Some authors arrange Bignonia ‌fasciculata in different genera of Bignoniaceae (Candolle 1845; Sonder 1849; Bureau 1864-1865). Later, Miers (1863) made the combination, which is accepted for old and recent authors, and it is reinforced here (Sandwith 1962; Gentry 1975a; Lohmann & Taylor 2014; Zuntini 2014; Medeiros & Lohmann 2015).

Bignonia ‌fasciculata well represents Tynanthus ‌fasciculatus and includes relevant diagnostic features, such as the 3-leaves, acuminate apex, cuneate base, terminal thyrse that resembles corymbose or subcorymbose aspect, and denticulate calyx apices (Medeiros & Lohmann 2015). Although the venation leaves were not represented in the illustration, the lack of information does not harm the recognition of the species. The original description includes features, such as ‘Corolla ex-luteo-albida, tubo brevissimo, vix calycis longitudine’ (Yellowish-white corolla, short tube, almost as long as the calyx).

Bignonia ‌longisiliqua ‌Vell. has no indication of locality, vegetation, or month of the year in the original manuscript of Floræ FluminensisStizophyllum ‌perforatum is a native plant, but it is not endemic to Brazil. Following Lohmann & Taylor (2014) and Beyer (2023), this species is found in humid forests or disturbed vegetation. In Brazil, it is found in the Brazilian Atlantic Forest and Cerrado.

A copy of the illustration of Bignonia ‌longisiliqua ‌Vell. stored in the Manuscript Section of Torre do Tombo can be accessed under the code PT-TT-MSLIV-2776_m0059 [photo!].

In addition, Bignonia ‌perforata was described with only a mention of the collector and collection locality: ‘E Brasilia ‌calidiori ‌misit ‌Sellowius’. Thus, the name lacks a holotype and requires typification (see comments on Bignonia ‌triphylla ‌Vell.).

Lohmann & Taylor (2014) addressed this name and referred to a specimen in LE as the ‘holotype’, but the authors did not examine this material. Therefore, based on the materials we studied and the information about Sellow’s herbaria (see comments on Bignonia ‌triphylla ‌Vell.), we selected the specimen in the US herbarium as the lectotype. This specimen bears the Berlin label and was examined and identified by Alwyn H. Gentry, a specialist in Bignoniaceae.

Gentry (1975a) observed that Bignonia ‌longisiliqua ‌Vell. was a later homonym of Bignonia ‌longisiliqua ‌Jacq. (1780), and co-specific with Stizophyllum ‌perforatum. Later, Zuntini (2014) also considered B. ‌longisiliqua ‌Vell. as a synonym of S. ‌perforatum. Finally, Kataoka & Lohmann (2021) interpreted B. ‌longisiliqua ‌Vell. as an attempt to cite Bignonia ‌longisiliqua ‌Jacq. (Catalpa ‌longissima ‌(Jacq.) Dum.Cours.). Our analysis supports Gentry’s (1975a) and Zuntini (2014) position and Bignonia ‌longisiliqua ‌Vell. is here considered a synonym of Stizophyllum ‌perforatum.

Bignonia ‌longisiliqua ‌Vell. well represents Stizophyllum ‌perforatum and includes relevant diagnostic features, such as 3-leaves with presence (or not) of simple tendril, cordate base, entire margin and long fruit, although the illustration does not seem to have been well executed, since in the protologue it is stated ‘pedicellis unifloris’ [uniparous pedicel] and the characteristic was not depicted (Vellozo 1829). Besides, there are confusing details, as the curved structure resembles a simple tendril, but without leaves, and lacks the venation representation on the majority of leaves. Nevertheless, this detail does not harm the recognition of the species.

Bignonia ‌elegans was described by Vellozo (1829) as ‘Oʃʃendi ad viam publicam prope Molendimum Sacchariferum vulgo dictum Lamarão’, which is recognized as someplace in the Rio de Janeiro state. Fridericia ‌elegans is a native plant endemic to Brazil. Following Lohmann & Taylor (2014) and Kaehler (2023a), this species is found in seasonal Deciduous Forests or Ombrophilous Forests. In Brazil, it is found in the Brazilian Atlantic Forest.

When we carried out research about ‘sugar mills’ or ‘sugar mills Lamarão’ in Rio de Janeiro, it was shown that the information related to ‘Freguesia de Campo Grande’. This place is currently a neighborhood in Rio de Janeiro (capital), but the region already has many sugar mills, with ‘Lamarão’ being one of these mills (Pedroza 2010). Hence, this collection locality is interpreted here in reference to the city of Rio de Janeiro.

A copy of the illustration of Bignonia ‌elegans stored in the Manuscript Section of Torre do Tombo can be accessed under the code PT-TT-MSLIV-2776_m0061 [photo!].

Schumann (1894) made the first treatment with Bignonia ‌elegans and proposed the combination: Adenocalymma ‌elegans. Kuhlmann (1941) suggested a new generic delimitation and transferred the species to Pseudocalymma ‌A.Samp. & Kuhlm. [Mansoa ‌DC.]. Gentry (1975a) made a comprehensive treatment, explained the right position of Bignonia ‌elegans, and made a combination: Arrabidaea ‌elegans. This position was also accepted in later treatments (Gentry 1980, 1992a). Recently, Lohmann & Taylor (2014) proposed a new generic delimitation, transferring the species to Fridericia ‌Mart. emend. L.G.Lohmann. This decision was accepted in subsequent studies (e.g., Kaehler et al. 2019), and it is reinforced here.

Bignonia ‌elegans well represents Fridericia ‌elegans and includes relevant diagnostic features, such as the 2-leaves, rounded base, acute apex, and yellow corolla. Although the fruits are not illustrated, this detail does not prejudice the species recognition.

Bignonia ‌grandifolia has no indication of locality, vegetation, or month of the year in the original manuscript of Floræ FluminensisAdenocalymma ‌grandifolium is a native plant endemic to Brazil. Following Lohmann & Taylor (2014) and Fonseca (2023a), this species is found in the Ombrophilous Forests. In Brazil, it is found in the Brazilian Atlantic Forest.

Bignonia ‌grandifolia ‌Vell. is a later homonym of Bignonia ‌grandifolia ‌Jacq. (1798). Candolle (1845) published Adenocalymma ‌grandifolium as a replacement name for Bignonia ‌grandifolia ‌Vell. (Art. 6.11 of the ICN, Turland et al. 2025). Later, Miers (1863) proposed Adenocalymma ‌prasinum to Bignonia ‌grandifolia ‌Vell. However, Miers (1863) did not see that Candolle (1845) had already proposed Adenocalymma ‌grandifolium as a substitute name, and therefore, Adenocalymma ‌prasinum is superfluous. The appropriate treatment is presented in Udulutsch et al. (2013).

A copy of the illustration of Bignonia ‌grandifolia ‌Vell. stored in the Manuscript Section of Torre do Tombo can be accessed under the code PT-TT-MSLIV-2776_m0063 [photo!].

Since Laroche (1973), the name of Vellozo has been frequently associated with Adenocalymma ‌grandifolium (Gentry 1975a; Udulutsch et al. 2013; Lohmann & Taylor 2014; Fonseca & Lohmann 2019), Zuntini (2014) being the only exception who proposes a connection with Adenocalymma ‌ternatum. In addition, Bignonia ‌grandifolia ‌Vell. received a confusing treatment in some databases. The POWO (2025), anchored in Lohmann & Taylor (2014) and Govaerts (1996), put Bignonia ‌grandifolia ‌Vell. as a synonym of Adenocalymma ‌ternatum. However, on the Flora and Funga of Brazil, anchored to Fonseca & Lohmann (2019), Adenocalymma ‌grandifolium accept the name, and this position is reinforced here.

Bignonia ‌grandifolia ‌Vell. well represents Adenocalymma ‌grandifolium and includes relevant diagnostic features, such as the big 3-leaves, and terminal inflorescence with long flowers.

Bignonia ‌cordata was described by Vellozo (1829) as ‘Habitat silvis maritimis Pharmacopolitanis’, which is recognized as someplace in Rio de Janeiro state (see comments under Bignonia ‌triphylla). Lundia ‌corymbifera is a native plant, but not endemic to Brazil. Following Lohmann & Taylor (2014), Kaehler & Lohmann (2021), and Kaehler (2023b), this species is found in open and disturbed areas of wet forest environments or seasonal forests. In Brazil, it is found in the Amazon Forest and the Brazilian Atlantic Forest.

A copy of the illustration of Bignonia ‌cordata stored in the Manuscript Section of Torre do Tombo can be accessed under the code PT-TT-MSLIV-2776_m0065 [photo!].

Gentry (1975a) considered Bignonia ‌cordata as a synonym of Bignonia ‌corymbifera (Lundia ‌corymbifera). Other authors accepted the combination Lundia ‌cordata ‌(Zuntini 2014). Later, Lohmann & Taylor (2014) treated Bignonia ‌cordata as a heterotypic synonym of Lundia ‌corymbifera and this position is reinforced here.

Bignonia ‌cordata well represents Lundia ‌corymbifera and includes relevant diagnostic features, such as the 3-leaves with terminal foliole modified in a simple tendril, acuminate apex, cordate base, axillary inflorescence, and infundibuliform corolla with acuminate apices. These characteristics agree with Lundia ‌corymbifera ‌(Kaehler & Lohmann 2021). Although existing information lacks fruits and seeds, it does not harm the recognition of the species.

Bignonia ‌exoleta has no indication of locality, vegetation, or month of the year in the original manuscript of Floræ FluminensisDolichandra ‌unguis-cati is a native plant, but not endemic to Brazil. Following Gentry (1973a), Fonseca et al. (2017), and Fonseca (2023b), this species is found in dry forests, moist forests, or rarely in wet forests and tropical wet forests. In Brazil, it is found in the Amazon Forest, Brazilian Atlantic Forest, Caatinga, Cerrado, Pampa, and Pantanal.

A copy of the illustration of Bignonia ‌exoleta stored in the Manuscript Section of Torre do Tombo can be accessed under the code PT-TT-MSLIV-2776_m0067 [photo!].

Gentry (1975a) listed Bignonia ‌exoleta as a synonym of Macfadyena ‌unguis-cati. Later, Zuntini (2014) accepted the connection between Bignonia ‌exoleta and Bignonia ‌unguis-cati. More recently, Fonseca et al. (2017) treated and also accepted these concepts. Thus, our study, following Fonseca et al. (2017), reinforces this position.

Bignonia ‌exoleta well represents Dolichandra ‌unguis-cati and includes relevant diagnostic features, such as the folioles with acute apex, cuneate base, serrated margin, campanulate calyces, inflated, with truncate to sinuous margin. These characteristics agree with Dolichandra ‌unguis-cati ‌(Fonseca et al. 2017). Besides, the illustration shows bifoliolate leaves without tendrils, a fact that disagrees with the D. ‌unguis-cati description (Fonseca et al. 2017). However, consulting the D. ‌unguis-cati collections, it is possible to note the morphological variation already cited by Gentry (1975a). In this example, the Bueno & Siqueira 86 (SPF) and Hoehne s.n. (NY) present gatherings lacking the terminal foliole modified on tendril.

Bignonia ‌convoluta has no indication of locality, vegetation, or month of the year in the original manuscript of Floræ Fluminensis Cuspidaria ‌convoluta is a native plant, but not endemic to Brazil. Following Lohmann & Taylor (2014) and Francisco (2023), this species is found in dry forests. In Brazil, it is found in the Brazilian Atlantic Forest, Cerrado, Pampa, and Pantanal.

A copy of the illustration of Bignonia ‌convoluta stored in the Manuscript Section of Torre do Tombo can be accessed under the code PT-TT-MSLIV-2776_m0069 [photo!].

Gentry (1975a) interpreted Bignonia ‌convoluta as Cuspidaria species for reasons of ‘reflexed anther thecae’. He commented that Bureau & Schumann (1896-1897) treated B. ‌convoluta as Arrabidaea ‌puberula ‌(Mart. ex DC.) Bureau (Cuspidaria ‌simplicifolia ‌DC.). Nevertheless, Gentry concluded that, for reasons of open panicule and small deeply dissected calyx illustrated, this species should be best associated with Cuspidaria ‌pterocarpa ‌(Cham.) DC. Thus, he finished the analysis by proposing a new (and currently accepted) combination: Cuspidaria ‌convoluta. Later treatments also accepted this position (Zuntini 2014; Lohmann & Taylor 2014) and it is reinforced here.

Bignonia ‌dichotoma ‌Vell. (Fig. 4) has no indication of locality, vegetation, or month of the year in the original manuscript of Floræ Fluminensis.Tanaecium ‌pyramidatum is a native plant, but not endemic to Brazil. Following Frazão & Lohmann (2019) and Frazão (2023), this species is found in dry and wet vegetation. In Brazil, it is found in the Amazon Forest, Brazilian Atlantic Forest, Caatinga, Cerrado, Pampa, and Pantanal.

Brazil•Rio de Janeiro, Nova Iguaçu, Reserva Biológica do Tinguá; 15.I.2002; Silva-Neto 1613; RBR[RBR00028554, photo!]•Petrópolis, Cortiço, à margem de um caminho, na capoeira; 13.I.1942; Monteiro 2433; RBR[RBR00002745, photo!]•Rio Claro; s.loc.; 6.XII.2003; Assis 1695; SPF[SPF00229616, photo!]•Santa Maria Madalena, Estrada para Sossego, via Cachoeirão; 20.XII.2017; Ferreira et al. 514; HUEFS[HUEFS246501, photo!]; VIES[VIES028311, photo!]•São Paulo, Pariquera-Açú, Estação Experimental do Instituto Agronômico; 16.XI.1996; Ivanauskas & Gomes 883; UEC[UEC082594, photo!]•Riolândia, a 5 km de Riolândia em direção a Cardoso 1º parada; 11.X.1994; Souza et al. 29; UEC[UEC127386, photo!]•São José do Rio Preto, Mirassol; 5.X.1995; Rezende 210; SJRP[SJRP00002067, photo!], UEC[UEC082593, photo!]•Ubatuba, Picinguaba; 15.XII.1995; Pedroni & Sanchez 2340; UEC[UEC026034, photo!].

Bignonia ‌dichotoma ‌Vell. is a later homonym of Bignonia ‌dichotoma ‌Jacq. (1760), therefore it is illegitimate (Art. 53.1 of the ICN, Turland et al. 2025).

A copy of the illustration of Bignonia ‌dichotoma ‌Vell. stored in the Manuscript Section of Torre do Tombo can be accessed under the code PT-TT-MSLIV-2776_m0071 [photo!].

Sandwith & Hunt (1974) treated Bignonia ‌dichotoma ‌Vell. as a synonym of Arrabidaea ‌selloi ‌(Spreng.) Sandwith (Tanaecium ‌selloi). Gentry (1975a) also treated this name as a basionym of some species of Arrabidaea ‌DC., especially A. selloi (Tanaecium ‌selloi). However, still considering the big variation in morphology in the species, the analysis of the illustration and plants is a lot like Tanaecium ‌pyramidatum. Although the protologue does not indicate locality, the search for information was concentrated in Rio de Janeiro and São Paulo states. Thus, based on the evaluation of characteristics of the illustration, description of species, and plant collections in RJ and SP, a new synonym is detected here. The illustrated characteristics are 2-foliolate leaves with an indication of a terminal foliole often modified on the simple tendril, axillary thyrse, and (apparently) tubular calyx; these are some characteristics that agree with Tanaecium ‌Sw. (Frazão & Lohmann 2019).

Bignonia ‌dichotoma ‌Vell. well represents Tanaecium ‌pyramidatum and includes relevant diagnostic features, such as the 2-leaves with a terminal leaflet modified into a simple tendril, rounded base, acute apex, entire margin, and flower morphology (Frazão & Lohmann 2019; Fernando et al. 2021). Although Tanaecium ‌pyramidatum is described with a terminal inflorescence, in the lectotype, the inflorescence is axillary, which can also be seen in other gatherings, such as Hatschbach & Guimarães 56149 (HUEFS, INPA, MBM), Salis & Joly 78 (UEC), Lombardi 6426 (SPF).

Bignonia ‌unguiculata has no indication of locality, vegetation, or month of the year in the original manuscript of Floræ Fluminensis.Dolichandra ‌unguiculata is native and endemic to Brazil. Following Fonseca et al. (2017) and Fonseca (2023b), this species is found in tropical wet and coastal forests, mainly near riverbanks or at ‘restinga’ vegetation. In Brazil, it is found in the Brazilian Atlantic Forest.

A copy of the illustration of Bignonia ‌unguiculata stored in the Manuscript Section of Torre do Tombo can be accessed under the code PT-TT-MSLIV-2776_m0073 [photo!]. The search for the name on Biblioteca Nacional do Rio de Janeiro, uses ‘Bignonia ‌unquiculata’ with a ‘Q’ instead of a ‘G’.

Gentry (1975a) proposed the combination Parabignonia ‌unguiculata. Since then, the basionym Bignonia ‌unguiculata has always been considered an accepted name, but in different genera (Lohmann & Taylor 2014; Zuntini 2014; Fonseca et al. 2017), and its actual organization is reinforced here.

Bignonia ‌unguiculata well represents Dolichandra ‌unguiculata and includes relevant diagnostic features, such as 2-leaves (terminal trifid tendril), apiculate apex, rounded base, inflated calyx, and infundibuliform corolla (Fonseca et al. 2017). Besides, it is interesting to note that a resemblance was perceived between B. unguiculata and B. unguis, when he wrote ‘B. unguis assimilatur’ [resembling Bignonia ‌unguis].

Bignonia ‌perianthomega has no indication of locality, vegetation, or month of the year in the original manuscript of Floræ Fluminensis Perianthomega vellozoi is a native plant, but not endemic to Brazil. Following Lohmann & Taylor (2014) and Lohmann (2023d), this species is found in dry vegetation. In Brazil, it is found in the Brazilian Atlantic Forest, Cerrado, and Pantanal.

A copy of the illustration of Bignonia ‌perianthomega stored in the Manuscript Section of Torre do Tombo can be accessed under the code PT-TT-MSLIV-2776_m0075 [photo!].

Miers (1863) treated Bignonia ‌perianthomega and made the combination: Memora ‌perianthomega. Later, Bureau (1893) recognized the basionym B. perianthomega and combined the species in the monotypic genus Perianthomega ‌Bureau ex Baill. This position was also accepted in later treatments (Gomes 1957; Gentry 1975a; Lohmann & Taylor 2014; Zuntini 2014), and it is reinforced here, and the illustration includes relevant diagnostic features, such as the 9-leaves, the folioles have an acuminate apex, rounded base, entire margin, and there is a simple tendril. Besides, the very large and campanulate calyx is diagnosed (Gentry 1992b).

Bignonia ‌hirta (Fig. 5) has no indication of locality, vegetation, or month of the year in the original manuscript of Floræ Fluminensis. Cuspidaria ‌simplicifolia is a native plant, but not endemic to Brazil. Following Lohmann & Taylor (2014) and Francisco (2023), this species is found in dry forests. In Brazil, it is found in the Brazilian Atlantic Forest, Caatinga, and Cerrado.

Initially, Bureau & Schumann (1896-1897) tentatively associated the name with Stizophyllum ‌perforatum ‌(Cham.) Miers. Disagreeing with this position, Gentry (1975a) emphasized the resemblance with Clytostoma ‌campanulatum ‌(Cham.) Bureau & K.Schum. due to the illustrated pseudostipules. Another resemblance highlighted by Gentry (1975a) is with Cuspidaria ‌puberula (C. ‌simplicifolia), on account of the paniculate inflorescence and simple lower leaves of each branchlet. Nevertheless, he does not consider the anther thecae illustrated as reflexed and concludes by saying that he could not resolve it at that moment. Although Gentry (1975a) praised Vellozo’s original illustration and noted some similarities, he did not address the name Bignonia ‌hirta.

Nascimento et al. (2024b) explained that the name Cuspidaria ‌simplicifolia was established almost two centuries ago, while the species has been transferred among several different genera (e.g., Arrabidaea ‌DC., Setilobus ‌Baill., and Saldanhaea ‌Bureau). In addition, the authors mentioned that in other cases it has remained under Cuspidaria ‌DC. as C. ‌puberula, but in recent decades a consensus has developed in favor of adopting Cuspidaria ‌simplicifolia as the correct name (e.g., Kaehler et al. 2019; Francisco et al. 2023). Therefore, Nascimento et al. (2024b) concluded that a new combination would cause nomenclatural confusion and possible disputes that would compromise nomenclatural stability. Thus, the authors proposed a rejection to maintain the use of Cuspidaria ‌simplicifolia. In fact, the resolution case was better treated in Nascimento et al. (2024b), where things such as priority and nomenclatural stability are explained.

A copy of the illustration of Bignonia ‌hirta stored in the Manuscript Section of Torre do Tombo can be accessed under the code PT-TT-MSLIV-2776_m0077 [photo!].

Bignonia ‌hirta was frequently regarded as a heterotypic synonym of Fridericia ‌mollis ‌(Vahl) L.G.Lohmann, as documented by various sources (Zuntini 2014; BFG et al. 2022; Kaehler 2023a). However, in the Brazilian territory, F. ‌mollis is only collected in the North. Thus, it does not make sense to associate it with Bignonia ‌hirta (collected in the Southeast). Clytostoma ‌campanulatum (Bignonia ‌campanulata ‌Cham.) has more elongated/lanceolate leaves with terminal foliole often modified on simple tendril – e.g., Benson & Parentoni 5 (MO), Gentry 49155 (NY), Hoehne s.n. (MO). Besides, B. ‌campanulata can have persistent foliaceous bracts (Zuntini 2014), and Bignonia ‌hirta’s illustration also does not show that.

The only detail that made Gentry (1975a) not relate Bignonia ‌hirta with Cuspidaria ‌simplicifolia was the less curved anthers present in the illustration. In fact, members of Cuspidaria often have anthers with forward-curved thecae (Lohmann & Taylor 2014), but this characteristic was not drawn in Bignonia ‌hirta. The imprecise or insufficient illustrations are common in FF, but this lack of information does not prejudice the recognition of this species.

Bignonia ‌cymosa has no indication of locality, vegetation, or month of the year in the original manuscript of Floræ Fluminensis.

A copy of the illustration of Bignonia ‌cymosa stored in the Manuscript Section of Torre do Tombo can be accessed under the code PT-TT-MSLIV-2776_m0079 [photo!].

Gentry (1975a) commented that Bignonia ‌cymosa does not correspond to any species already seen until that year. He still saw a possible resemblance to Arrabidaea ‌chica ‌(Bonpl.) Verl. (Fridericia ‌chica ‌(Bonpl.) L.G.Lohmann), but it would be necessary to consider some mistakes.

The interpretation of this drawing is difficult and, as Gentry (1975a) said, ‘would never be positively identifiable without type material’. The species was referred to as a possible junior synonym of Fridericia ‌chica, but the majority of the collections from Rio de Janeiro and São Paulo states disagree with Vellozo’s representation. Some agree with the shape of the leaves, as Ferreira & Patusco 567 (MBM), Santoro 901 (US), Santoro 853 (US), and Stranghetti et al. 1458 (SJRP). However, we were not able to find any collection with 3 leaves or with a terminal foliole modified on a simple tendril. When a tendril appears, the shape of the leaves does not agree with Vellozo’s drawing — e.g., Hoehne 2628 (US), Leitão Filho et al. 32746 (SPF). Besides, Vellozo illustrated flowers with lobes (apparently) crenate (details in two open flowers), but all collections of Fridericia ‌chica seen in our study do not show this detail.

The last resource to obtain a convincing result was a search in collections of a lot of Bignoniaceae in Rio de Janeiro state and observing the resemblances with Vellozo’s drawing. Thus, we noticed an affinity with Fridericia, particularly with the ‘Group Six’ of the identification key available on Kaehler (2023a). Therefore, we agree with Gentry (1975a) and believe it is better to leave Bignonia ‌cymosa without identification (incertae sedis) until an assertive interpretation emerges.

Bignonia ‌longa has no indication of locality, vegetation, or month of the year in the original manuscript of Floræ Fluminensis. Lundia ‌longa is a native plant endemic to Brazil. Following Lohmann & Taylor (2014), Kaehler & Lohmann (2021), and Kaehler (2023b), this species is found in open coastal vegetation, along road banks, forest edges, or wet sites. In Brazil, it is found in the Brazilian Atlantic Forest.

Candolle (1845) cites Bignonia ‌longa at the Lundia ‌longa ‌(Vell.) DC. description. He also mentioned a collection of Martius, but this collection should be better interpreted as ‘examined material’ since he clearly indicates Bignonia ‌longa as the basionym (Kaehler & Lohmann 2021).

Gentry (1975a) misunderstood when considering Bignonia ‌longa as a synonym of Lundia ‌cordata ‌(Vell.) DC. (Lundia ‌corymbifera ‌(Vahl) Sandwith). In his opinion, Lundia ‌longa ‌(Vell.) DC. is based on the Martius collection (Martius s.n. of 1818 from Minas Gerais), which is different from Vellozo’s illustration. However, Gentry (1975a), as well as other people since Bureau & Schumann (1896-1897), have mistaken the species (Kaehler & Lohmann 2021) and identified Lundia ‌longa as Lundia ‌corymbifera.

Bignonia ‌longa, as a synonym of Lundia ‌cordata, was inaccurately associated in some treatments (Zuntini 2014). Nevertheless, the recognition of Bignonialonga, as the basionym of Lundia ‌longa ‌(Vell.) DC., is accepted in other treatments (e.g., Lohmann & Taylor 2014; Zuntini & Lohmann 2016; Kaehler & Lohmann 2021; Costa et al. 2022), and it is reinforced here.

A copy of the illustration of Bignonia ‌longa stored in the Manuscript Section of Torre do Tombo can be accessed under the code PT-TT-MSLIV-2776_m0081 [photo!].

Bignonia ‌longa well represents Lundia ‌longa and includes relevant diagnostic features like the 3-leaves with terminal tendril modified in a trifid tendril, acuminate apex, cordate base, axillary and terminal inflorescence (Kaehler & Lohmann 2021). Although there are leaves without venation, the lack of this characteristic does not harm the recognition of the species.

Bignonia ‌triflora (Fig. 6) has no indication of locality, vegetation, or month of the year in the original manuscript of Floræ Fluminensis. Cuspidaria ‌pulchella is native and endemic to Brazil. Following Lohmann & Taylor (2014) and Francisco (2023), this species is found in the Cerrado.

Gentry (1975a) noted a resemblance between Bignonia ‌triflora and Arrabidaea ‌pulchella (Cuspidaria ‌pulchella) by the few-flowered inflorescence, setae of the calyx, and the simple tendrils. However, the pubescent filaments presented in Vellozo’s drawing led the Gentry to prefer maintaining this species without definitive identification.

Gentry (1975a) observed that the absence of illustrated reflexed thecae in the anthers does not correspond to Arrabidaea ‌pulchella. This unresolved status remains today, and B. ‌triflora has not been confidently assigned to any species (e.g., Zuntini 2014; Nascimento et al. 2024c; POWO 2025; Tropicos 2025).

Kaehler et al. (2019) revised the polyphyletic genus Arrabidaea ‌DC., proposing a new circumscription divided among four genera: Cuspidaria ‌DC., Fridericia ‌Mart. emend. L.G.Lohmann, Tanaecium ‌Sw. emend. L.G.Lohmann, and Xylophragma ‌Sprague. Notably, the species formerly known as Arrabidaea ‌pulchella is now recognized within Cuspidaria, which frequently possess anthers with forward-curved thecae (Francisco et al., 2023). However, this feature is not depicted in Vellozo’s illustration of Bignonia ‌triflora. Gentry (1975a) himself acknowledged occasional inaccuracies in Bignoniaceae illustrations, suggesting that the anther drawing may reflect an illustration error rather than a true morphological discrepancy.

Nascimento et al. (2024c) critically examined Bignonia ‌triflora and compared it with all Cuspidaria species occurring in the regions of Rio de Janeiro and São Paulo. Evaluation of the illustrated diagnostic traits such as the apparent liana habit, 2-foliolate leaves with a terminal simple tendril, axillary inflorescences bearing few flowers, and the geographical distribution of Cuspidaria ‌pulchella – along with the anther profile in Vellozo (1831), aligns closely with the depiction of A. pulchella in Lohmann & Pirani (1998). These observations confirm the strong similarity between Bignonia ‌triflora and Cuspidaria ‌pulchella, supporting the proposal of a new combination and the reassignment of C. ‌pulchella.

In the end, Nascimento et al. (2024c) conclude that, although lacking some important features, Bignonia ‌trifloral makes clear the possibility of it being applied as the correct name for Cuspidaria ‌pulchella. Thus, a new combination would be a continuing source of dispute on the application of the name, and the situation is disadvantageous to nomenclatural stability. For this situation and other details presented in Nascimento et al. (2024c), the authors propose to reject the name.

In addition, A copy of the illustration of Bignonia ‌triflora stored in the Manuscript Section of Torre do Tombo can be accessed under the code PT-TT-MSLIV-2776_m0083 [photo!].

In collections of Cuspidaria species from Southeast Brazil, the resemblance between Bignonia ‌triflora and Cuspidaria ‌pulchella increases. Some species are restricted to one state: Cuspidaria ‌lasiantha ‌(Bureau & K.Schum.) L.G.Lohmann (Rio de Janeiro), C. ‌floribunda ‌(DC.) A.H.Gentry and C. ‌multiflora ‌DC. (São Paulo). Meantime, others can ambiguously occur: C. ‌octoptera ‌A.H.Gentry, C. ‌pulchella, C. ‌pulchra ‌(Cham.) L.G.Lohmann (Francisco 2023; speciesLink 2023).

Almost all species mentioned above do not agree with Vellozo’s drawing, due to the terminal panicle (sometimes axillary in Cuspidaria ‌lasiantha). The shape of the leaves between them varies but is similar. Nevertheless, the only species to combine all details, including leaves, simple tendril, and axillary inflorescence reduced to a few flowers, is Cuspidaria ‌pulchella.

The only detail that made Gentry (1975a) not relate Bignonia ‌triflora to Cuspidaria ‌pulchella (treated as Arrabidaea ‌pulchella) was the less curved anthers present in the illustration. In fact, members of Cuspidaria often have anthers with forward-curved thecae (Lohmann & Taylor 2014), but this characteristic was not drawn in Bignonia ‌triflora. Nevertheless, in Lohmann & Pirani (1998), an illustration was provided with the appearance of anthers, making the similarity to Vellozo’s drawing clear. Besides, it is common for plates from Floræ Fluminensis not to include diagnostic details to support the species identification (Pastore et al. 2022).

Bignonia ‌rego has no indication of locality, vegetation, or month of the year in the original manuscript of Floræ Fluminensis. Fridericia ‌rego is a native plant endemic to Brazil. Following Lohmann & Taylor (2014) and Kaehler (2023a), this species is found in dry forests. In Brazil, it is found in the Brazilian Atlantic Forest.

A copy of the illustration of Bignonia ‌rego stored in the Manuscript Section of Torre do Tombo is accessed under the code PT-TT-MSLIV-2776_m0085 [photo!]. The search for this name on the Biblioteca Nacional do Rio de Janeiro, uses ‘Bignonia ‌sego’ with an ‘S’ instead of an ‘R’.

Candolle (1838) makes the combination: Arrabidaea ‌rego ‌(Vell.) DC. (as ‘sego’). Later, Martius (1841) changed to Vasconcellia ‌Mart. (also as ‘sego’). Kuntze (1891) transfers the species to Chasmia ‌Schott ex Spreng. Later, Gentry (1975a) re-established Arrabidaea ‌rego ‌(Vell.) DC. and considered Arrabidaea ‌agnus-castus ‌(Cham.) DC. (Bignonia ‌agnus-castus ‌Cham.) as a synonym. However, Lohmann & Taylor (2014) combined this name into Fridericia, which was followed by Zuntini (2014) and Kaehler et al. (2019). This position is reinforced here.

Bignonia ‌rego well represents Fridericia ‌rego and includes relevant diagnostic features, such as the 3-leaves with terminal foliole modified on the simple tendril, the folioles are ovate with acuminate apex, plus the shape of the inflorescence, fruits, and seeds (Reiche et al. 2020).

In addition, Vellozo (1829) commented: ‘Vulgò cognita est nomine Sipó-rego’ [Generally known by the name ‘Cipó-rego’]. A reference to the common name of the species in the region and time.

Bignonia ‌arvensis (Fig. 7) was described by Vellozo (1829) as ‘Habitat campis apricis mediterraneis trans-alpinis’, which is recognized as someplace in São Paulo state. Especially the Cerrado vegetation in the stretch of BR-459 right before the São Paulo border, in the direction of Cunha town. Anemopaegma ‌arvense is a native plant, but not endemic to Brazil. Following Lohmann & Taylor (2014) and Firetti (2023), in Brazil, this species is found in the Amazon Forest, Brazilian Atlantic Forest, and Cerrado.

A copy of the illustration of Bignonia ‌arvensis stored in the Manuscript Section of Torre do Tombo can be accessed under the code PT-TT-MSLIV-2776_m0087 [photo!].

Steudel (1840) mentioned Bignonia ‌arvensis, and later, in Sousa (1945), the combination was proposed: Anemopaegma ‌arvense. Gentry (1975a) mentioned the combination made by Sousa (1945) and suggested Bignonia ‌arvensis as a name to replace Anemopaegma ‌mirandum ‌(Cham.) DC. (Bignonia ‌miranda ‌Cham.). This position was also accepted in later treatments (Scudeller 2004; Zuntini 2014; Lohmann & Taylor 2014), and it is reinforced here.

The illustration includes relevant diagnostic features, such as the appearance of leaves, flower position, and morphology, as well as the details presented in fruits and seeds.

Bignonia ‌pedunculata (Fig. 8) was described by Vellozo (1829) as ‘Habitat campis apricis mediterraneis’, which is recognized as someplace in the Cerrado vegetation in São Paulo state (see comments under Bignonia ‌ignea). Adenocalymma ‌pedunculatum is a native plant endemic to Brazil. Following Fonseca & Lohmann (2019) and Fonseca (2023a), this species is found in the Amazon Forest, Brazilian Atlantic Forest, and Cerrado.

A copy of the illustration of Bignonia ‌pedunculata stored in the Manuscript Section of Torre do Tombo can be accessed under the code PT-TT-MSLIV-2776_m0089 [photo!].

Miers (1863) made the combination Memora ‌pedunculata ‌(Vell.) Miers. Sandwith (1962) had already synonymized Bignonia ‌glaberrima ‌Cham. in Bignonia ‌pedunculata. Recently, other studies have followed this organization (Gentry 1975a; Zuntini 2014; Fonseca & Lohmann 2019), and it is reinforced here.

Bignonia ‌pedunculata well represents Adenocalymma ‌pedunculatum and includes relevant diagnostic features, such as the triternate leaves without tendril, ovate blade with ovate base, acute apex, entire margin, terminal inflorescence, floral morphology, and yellow color (Machado & Romero 2014). The original descriptions include features, such as ‘caule herbaceo’ (herbaceous stem) and ‘caulis cubitalis’ (stem measuring the length of a cubit, 40-50 cm).

Bignonia ‌coccinea ‌Vell. was described by Vellozo (1829) as ‘Habitat silvis mediterraneis’, which is recognized as someplace in São Paulo state (see comments under Bignonia ‌ignea).

Gentry (1975a) recognized Bignonia ‌coccinea ‌Vell., non Bignonia ‌coccinea ‌(Prush) Steud. as Dolichandra ‌cynanchoides ‌Cham. and treated the Vellozo’s name as a later homonym. However, Nascimento et al. (2023) revealed that B. ‌coccinea ‌Steudel is not validly published, and consequently, B. ‌coccinea ‌Vell. is legitimate. Thus, the combination ‘Dolichandra ‌coccinea ‌(Vell.) M.Nascim., Zuntini & J.F.B.Pastore’ was made. However, some important taxonomic details were not illustrated and undermined the stability and correct use of this name.

Fonseca (2024) identified some mistakes in the illustration, including the portrayal of simple tendrils, a bi-labiate calyx, and linear fruits. In contrast, Dolichandra ‌cynanchoides has trifid and uncinate tendrils, spathaceous calyx, and oblong fruits (Fonseca et al. 2017). Besides, Fonseca (2024) emphasized the misguided distribution presented, noting that D. ‌cynanchoides occurs exclusively in southern Brazil (Fonseca et al. 2017). However, some occurrences exist in Southeast Brazil, but they are of cultivated plants (L.H.M. Fonseca, pers. comm.).

Thus, it is evident that an opinion conflict involves this name and its application; a debate, if it remains under discussion, is entirely unfavorable to nomenclatural stability. The arguments presented by Fonseca (2024) are thoroughly addressed in Nascimento et al. (2024d), which explained all issues and showed that a discussion may persist. Nonetheless, in conclusion, Nascimento et al. (2024d) propose rejecting Bignonia ‌coccinea ‌Vell., to preserve the use of Dolichandra ‌cynanchoides and prevent this issue from resurfacing in the future.

Recently, Fonseca (2025) published a reappraisal of Nascimento et al. (2024d), explaining his reasons for maintaining the taxonomic disagreement regarding the identification of Bignonia ‌coccinea ‌Vell. as Dolichandra ‌cynanchoides. However, he acknowledges the intention of Nascimento et al. (2024d) and, for taxonomic reasons, concurs with the rejection of the name Bignonia ‌coccinea ‌Vell. In summary, although for different reasons, both authors agree that preserving the use of Dolichandra ‌cynanchoides is the best decision.

A copy of the illustration of Bignonia ‌coccinea ‌Vell. stored in the Manuscript Section of Torre do Tombo can be accessed under the code PT-TT-MSLIV-2776_m0091 [photo!] (Nascimento et al. 2023).

From the description at locality origin in Vellozo (1829) ‘Campis apricis mediterraneis habitat’, which is recognized as someplace in São Paulo state (see comments under Bignonia ‌ignea). Jacaranda ‌caroba (Fig. 9) is a native plant endemic to Brazil. Following Gentry (1992b) and Farias-Singer (2023), this species is found in the Brazilian Atlantic Forest and Cerrado.

Morawetz (1982) chose the collection Pohl 223 (F No. 875015 [photo!], W No. 0057739 [photo!]) as Neotype for Jacaranda ‌caroba. Consequently, this typification is also superseded here due to the presence of the original illustrations of Floræ Fluminensis (Art. 9.19 of the ICN, Turland et al. 2025).

A copy of the illustration of Bignonia ‌caroba ‌Vell. stored in the Manuscript Section of Torre do Tombo can be accessed under the code PT-TT-MSLIV-2776_m0093 [photo!].

Candolle (1845) observed the species and arranged the combination: Jacaranda ‌caroba. Gentry (1975a) also recognized Jacaranda ‌caroba and listed a substantial number of illustrated characters. Later, other studies also accepted this organization (Gentry 1992b; Zuntini 2014), and it is reinforced here.

The illustration includes relevant diagnostic features, such as the bi-pinnate leaves with acute apex, terminal inflorescence, and color of flowers (Gentry 1992b). It is also commented ‘Vulgo dicitur Carôba. Habet vim antiveneream. Ideo rusticani eam magni faciunt. Qualitas ‌amara.’ [Generally, it is called ‘Carôba’. It has anti-venomous power. Because of this, the peasants consider it of great value. The quality is bitter, a reference to the common name, flavor, and possible uses by regional people, which indicates Vellozo or someone in the entourage knew about this species.

Bignonia ‌elliptica was described by Vellozo (1829) as ‘Habitat silvis mediterraneis trans-alpinis propre prædium dictum Boavista’, which is recognized as someplace in São Paulo state (see comments under Bignonia ‌ignea), especially near the Chapel of São José da Boa Vista in Cunha/SP (Pastore et al. 2021). Jacaranda ‌macrantha is a native plant endemic to Brazil. Following Gentry (1992b) and Farias-Singer (2023), this species is found in the Brazilian Atlantic Forest, and Cerrado.

Bignonia ‌elliptica ‌Vell. is a later homonym of Bignonia ‌elliptica ‌Thunb. (1821), therefore it is illegitimate (Art. 53.1 of the ICN, Turland et al. 2025). A copy of the illustration of Bignonia ‌elliptica ‌Vell. stored in the Manuscript Section of Torre do Tombo can be accessed under the code PT-TT-MSLIV-2776_m0095 [photo!].

Gentry (1975a) associated Bignonia ‌elliptica ‌Vell. with Jacaranda ‌macrantha ‌Cham. and explained that the substitute name Jacaranda ‌elliptica ‌Steud. (1840) is legitimate, but it is a synonym of J. ‌macrantha. In Farias-Singer (2023), Bignonia ‌elliptica ‌Vell. does not appear in the search or as a synonym for Jacaranda ‌macrantha. However, Bignonia ‌elliptica ‌Vell. (as a synonym of Jacaranda ‌macrantha) is found in other studies (Gentry 1992b), and it is reinforced here.

Bignonia ‌elliptica ‌Vell. well represents Jacaranda ‌macrantha and includes relevant diagnostic features, such as the bipinnate leaves, ovate, acuminate apex, cuneate base, entire margin, inflorescence, flowers, fruits, seeds morphology, and corolla color (Gentry 1992b). Although the line drawing lacks the majority of leaf venation, this detail does not harm the recognition of the species.

Bignonia ‌obovata ‌Vell. was described by Vellozo (1829) as ‘Habitat silvis maritimis, campisque’, which is recognized as a place in Rio de Janeiro state (see comments under Bignonia ‌squalus). Jacaranda ‌puberula is a native plant, but not endemic to Brazil. Following Gentry (1992b) and Farias-Singer (2023), this species is found in coastal evergreen forests, semideciduous forests, or montane forests. In Brazil, it is found in the Brazilian Atlantic Forest and Cerrado.

Bignonia ‌obovata ‌Vell. is a later homonym of Bignonia ‌obovata ‌Spreng. (1825), therefore it is illegitimate (Art. 53.1 of the ICN, Turland et al. 2025). A copy of the illustration of Bignonia ‌obovata ‌Vell. stored in the Manuscript Section of Torre do Tombo can be accessed under the code PT-TT-MSLIV-2776_m0097 [photo!].

Gentry (1975a) treated Bignonia ‌obovata ‌Vell. as a later homonym, and it was synonymized in Jacaranda ‌puberula. This position was also accepted in later treatments (Gentry 1992b; Zuntini 2014), but in Farias-Singer (2023), this name does not appear as a synonym of J. ‌puberula. However, our study agrees with Gentry (1992b) and considers Bignonia ‌obovata ‌Vell. as a synonym of J. ‌puberula.

Bignonia ‌obovata ‌Vell. well represents Jacaranda ‌puberula and includes relevant diagnostic features, such as the bipinnate leaves, tubular-campanulate corolla and color (purple), oblong-obovate fruit, and winged seeds (Gentry 1992b). Although there is a lack of venation in the majority of the leaves and the margin is slightly serrate (a common characteristic in Rio de Janeiro collections), this detail can be variable, and its absence does not harm the recognition of the species.

In addition, Vellozo (1829) commented: ‘Carôba etiam dicitur’ [Also known as ‘Carôba’], as a reference to the common name of the species.

Bignonia ‌tuberculosa (Fig. 10) was described by Vellozo (1829) as ‘Habitat silvis maritimis pharmacopolitanis’, which is recognized as someplace in Rio de Janeiro state (see comments under Bignonia ‌triphylla). Zeyheria ‌tuberculosa is a native plant, but not endemic to Brazil. Following Gentry (1992b) and Lohmann (2023h), this species is found in the Brazilian Atlantic Forest, Caatinga, and Cerrado.

A copy of the illustration of Bignonia ‌tuberculosa stored in the Manuscript Section of Torre do Tombo can be accessed under the code PT-TT-MSLIV-2776_m0099 [photo!].

Steudel (1840) treated Bignonia ‌tuberculosa and made a combination: Jacaranda ‌tuberculosa. Later, Bureau (1868) revised this name and proposed the combination in another generic delimitation: Zeyheria ‌tuberculata ‌(Vell.) Bureau. Gentry (1975a) and other studies also accepted it (Gentry 1992b; Zuntini 2014), and it is reinforced here.

Bignonia ‌tuberculosa well represents Zeyheria ‌tuberculosa and includes relevant diagnostic features, such as the 5-leaves, more or less oblong, acuminate apex, rounded base, entire margin, fruit with obovoid capsule, and winged seeds. These characteristics agree with Zeyheria ‌tuberculosa ‌(Gentry, 1992b).

In addition, Vellozo (1829) commented: ‘Florem non vidi’ [Unseen flower], an indication that at the moment of Vellozo and his entourage’s passage, specimens of Bignonia ‌tuberculosa were not found with flowers.

Bignonia ‌digitalis was described by Vellozo (1829) as ‘Habitat campis apricis mediterraneis. Trans-alpinis’, which is recognized as someplace in São Paulo state. Especially the Cerrado vegetation in the stretch of BR-459 (see comments under Bignonia ‌squalus). Zeyheria ‌montana is a native plant endemic to Brazil. Following Gentry (1992b) and Lohmann (2023h), this species is found in the Amazon Forest, the Brazilian Atlantic Forest, Caatinga, and Cerrado.

Gentry (1975a) accepting Zeyheria ‌digitalis ‌(Vell.) Hoehne does not consider the priority of Z. ‌montana and mentions the posterior homonym Zeyheria ‌digitalis ‌(Vell.) L.B.Sm. & Sandwith (Sandwith 1953). However, Sandwith (1953) considered the year 1825 as the publication year of Floræ Fluminensis; this detail led the author to attribute priority to Bignonia ‌digitalis. Nevertheless, 1825 represents the initial date of printing, which ended in 1829 (Carauta 1973). Bignonia ‌digitalis has had priority over Zeyheria ‌montana since 1826, the year of effective publication (Martius 1826). Later, other studies proposed the appropriate organization (Gentry 1992b; Zuntini 2014), and it is reinforced here.

A copy of the illustration of Bignonia ‌digitalis stored in the Manuscript Section of Torre do Tombo can be accessed under the code PT-TT-MSLIV-2776_m0101 [photo!].

Bignonia ‌digitalis well represents Zeyheria ‌montana and includes relevant diagnostic features, such as the 5-leaves, elliptic, acute apex, cuneate base, entire margin, terminal inflorescence, irregularly 2-dentate calyx, tubular-cylindrical corolla, fruit, and seeds morphology. These characteristics agree with Zeyheria ‌montana ‌(Gentry 1992b). Although there are leaves without venation, the lack of this characteristic does not harm the recognition of the species.

In addition, Vellozo (1829) commented: ‘Tota ‌planta ‌tomento subtilissimo rufo tecta’ [The entire plant is covered by a very fine and reddish tomentum], a reference to the indument present on the plant.

Bignonia ‌heptaphylla (Fig. 11) was described by Vellozo (1829) as ‘Habitat silvis maritimis’, which is recognized as someplace in Rio de Janeiro state (see comments under Bignonia ‌squalus). Handroanthus ‌heptaphyllus is a native plant, but not endemic to Brazil. Following Gentry (1992b) and Lohmann (2023c), this species, in Brazil, is found in the Brazilian Atlantic Forest, Cerrado, and Pampa.

A copy of the illustration of Bignonia ‌heptaphylla stored in the Manuscript Section of Torre do Tombo can be accessed under the code PT-TT-MSLIV-2776_m0103 [photo!].

Bignonia ‌heptaphylla was first treated by Martius (1841), who made the combination Tecoma ‌heptaphylla ‌(Vell.) Mart. Then Toledo (1952) transferred it to Tabebuia ‌Gomes ex DC. In Mattos (1970), the species was combined in Handroanthus.

Gentry (1974) initially considered Bignonia ‌heptaphylla as ‘nomen confusum’, but the next year, after analyzing many collections near Rio de Janeiro, Gentry (1975a) affirmed that Tabebuia ‌heptaphylla should be adopted. After the re-establishment of Handroanthus by Grose & Olmstead (2007), H. ‌heptaphyllus ‌(Vell.) Mattos became the most widely used name. This position was also accepted in later treatments (Gentry 1975a; Zuntini 2014), and it is reinforced here.

The illustration includes relevant diagnostic features, such as the number of leaflets (6-7), more or less oblong, acuminate apex, rounded base, serrated margin, terminal inflorescence, cupular calyx, elongate fruit, and seed details, that agree with H. ‌heptaphyllus ‌(Gentry 1992b). Although the leaf venation was not represented in the illustration, the lack of this characteristic does not harm the recognition of the species. The original description includes features, such as ‘Caulis ‌arboreus’ [tree stem] and ‘Corolla violacea’ [purple corolla].

Bignonia ‌leucantha was described by Vellozo (1829) as ‘Habitat silvis maritimis’, which is recognized as someplace in Rio de Janeiro state (see comments under Bignonia ‌squalus). Sparattosperma ‌leucanthum is a native plant, but not endemic to Brazil. Following Gentry (1992b) and Lohmann (2023f), this species, in Brazil, is found in the Amazon Forest, Brazilian Atlantic Forest, Caatinga, Cerrado, and Pantanal.

A copy of the illustration of Bignonia ‌leucantha stored in the Manuscript Section of Torre do Tombo can be accessed under the code PT-TT-MSLIV-2776_m0105 [photo!].

Schumann (1894) treated Bignonia ‌leucantha and made the combination: Sparattosperma ‌leucanthum. Later, Gentry (1975a) recognizes and agrees with Schumann (1894). This position was also accepted in later treatments (Gentry 1992b; Zuntini 2014), and it is reinforced here.

Bignonia ‌leucantha well represents Sparattosperma ‌leucanthum and includes relevant diagnostic features, such as the 5-leaves, oblong-ovate folioles, acuminate apex, cuneate base, entire margin, terminal inflorescence, irregular calyx, tubular-campanulate corolla, and color, elongated fruit, and winged seeds (Gentry 1992b). The original description includes features, such as ‘Caulis ‌arboreus’ [tree stem].

Bignonia ‌quinquefolia was described by Vellozo (1829) as ‘Habitat silvis pharmacopolitanis’, which is recognized as someplace in Rio de Janeiro state (see comments under Bignonia ‌triphylla). Cybistax ‌antisyphilitica is a native plant, but not endemic to Brazil. Following Gentry (1992b) and Lohmann (2023b), this species, in Brazil, is found in the Amazon Forest, Brazilian Atlantic Forest, Caatinga, Cerrado, Pampa, and Pantanal.

A copy of the illustration of Bignonia ‌quinquefolia stored in the Manuscript Section of Torre do Tombo can be accessed under the code PT-TT-MSLIV-2776_m0107 [photo!].

Gentry (1975a) discussed the stabilization of Cybistax ‌antisyphilitica and listed Bignonia ‌quinquefolia as a synonym of Cybistax ‌antisyphilitica. This position was also accepted in later treatments (Gentry 1992b; Zuntini 2014), and it is reinforced here.

Bignonia ‌quinquefolia well represents Cybistax ‌antisyphilitica and includes relevant diagnostic features, such as the 5-leaves, elliptic, acuminate apex, cuneate base, entire margin, terminal inflorescence, calyx long, acuminate teeth, tubular-campanulate corolla, the oblong capsule fruit, and winged seeds. These characteristics agree with Cybistax ‌antisyphilitica ‌(Gentry 1992b). Although the leaf venation was not represented in the illustration, the lack of this characteristic does not harm the recognition of the species.

Bignonia ‌flavescens has no indication of locality, vegetation, or month of the year in the original manuscript of Floræ Fluminensis. Handroanthus ‌serratifolius is a native plant, but not endemic to Brazil. Following Gentry (1992b) and Lohmann (2023c), this species is found in seasonal forests. In Brazil, it is found in the Amazon Forest, Brazilian Atlantic Forest, Caatinga, Cerrado, and Pantanal.

Bignonia ‌serratifolia was originally described with reference only to the voucher Ryan s.n., without indicating the precise specimen or herbarium, and therefore the name lacks a designated holotype (McNeill 2014). Thus, a lectotype has been here designated from the main Vahl herbarium, which includes both vegetative and reproductive characteristics in the same specimen.

A copy of the illustration of Bignonia ‌flavescens stored in the Manuscript Section of Torre do Tombo can be accessed under the code PT-TT-MSLIV-2776_m0109 [photo!].

Mattos (1970) recognized Bignonia ‌flavescens and made the combination in Handroanthus ‌flavescens ‌(Vell.) Mattos. Later, Gentry (1975a) treated B. flavescens as a variation of Tabebuia ‌serratifolia and suggested the synonymization. After the reestablishment of Handroanthus by Grose & Olmstead (2007), the name was synonymized with Handroanthus. This name, as a synonym of H. ‌serratifolius, was also accepted in later treatments (Gentry 1992b; Zuntini 2014), and it is reinforced here.

Bignonia ‌flavescens well represents Handroanthus ‌serratifolius and includes relevant diagnostic features, such as 5-leaves, elliptic, acuminate apex, rounded base, slightly serrated margin, terminal inflorescence, and infundibuliform corolla. These characteristics agree with Handroanthus ‌serratifolius ‌(Gentry 1992b).

In addition, Vellozo (1829) commented: ‘Vulgo dicitur Ipéuva, vel Ipê do Campo’ [Generally known as ‘Ipéuva,’ or ‘Ipê do Campo’], as a reference to the common name of the species.

Bignonia ‌longiflora was described by Vellozo (1829) as ‘Habitat silvis maritimis pharmacopolitanis’, which is recognized as someplace in Rio de Janeiro state (see comments under Bignonia ‌triphylla). Handroanthus ‌speciosus is a native plant endemic to Brazil. Following Gentry (1992b) and Lohmann (2023c), this species is found in montane forests. In Brazil, it is found in the Brazilian Atlantic Forest.

Several names, with the same basionym, were proposed as correct to Bignonia ‌longiflora ‌Vellozo. In chronological order, these include: Tecoma ‌speciosa ‌DC. ex Mart. (Gelseminum ‌speciosum ‌(DC. ex Mart.) Kuntze, Tabebuia ‌vellosoi ‌Toledo (Handroanthus ‌vellosoi ‌(Toledo) Mattos), Tecoma ‌longiflora ‌(Vell.) Bureau & K. K.Schum., and Handroanthus ‌longiflorus ‌(Vell.) Mattos. It should be highlighted that the earlier specific epithet speciosum/speciosa/speciosus could not be legitimately combined with the genus Tabebuia ‌Gomes ex DC., because Tabebuia ‌speciosa ‌Standl. (Callichlamys ‌latifolia ‌(Rich.) K.Schum.) was already established. Therefore, when Tabebuia was generally accepted as the correct genus, the appropriate name was Tabebuia ‌vellosoi. However, Handroanthus ‌Mattos came to be regarded as the valid genus for many species, following the generic delimitation provided by Mattos (1970) and later supported by Grose & Olmstead (2007). Mattos (1970) treated the taxon as Handroanthus ‌vellosoi yet seems to have overlooked the oldest epithet (i.e., “speciosa”), which continues to be available within the genus.

Nascimento et al. (2024a) proposed the new combination, accompanied by taxonomic notes, including a comparative table highlighting the morphological differences among species related to Handroanthus ‌speciosus. In addition, the study presents distribution maps, further nomenclatural considerations (e.g., the suppression of the neotype), as well as typifications, habitat descriptions, distribution data, and phenological details.

Bignonia ‌tababuya was described by Vellozo (1829) as ‘Habitat maritimis au loca inundata, vel a mari, vel a fluviis’, which is recognized as someplace in Rio de Janeiro state (see comments under Bignonia ‌squalus). Tabebuia ‌cassinoides is a native plant endemic to Brazil. Following Gentry (1992b) and Lohmann (2023g), this species is found in the Brazilian Atlantic Forest.

Bignonia ‌cassinoides was originally described with only a brief mention of the collector and collection locality: ‘[...] découvert par M. Commerʃon, aux environs de Rio-Janeiro, au Bréʃil’. As a result, the name lacks a holotype and requires typification (see comments on Bignonia ‌triphylla ‌Vell.). Gentry (1992b) treated Tabebuia ‌cassinoides and indicated the specimen in P as the ‘type’ – a valid indication under (Art. 7.11 of the ICN, Turland et al. 2025). Although Gentry was probably only aware of the specimen in P, we interpret his reference as a lectotypification of that material. Additionally, we report the discovery of an isolectotype in C.

A copy of the illustration of Bignonia ‌tababuya stored in the Manuscript Section of Torre do Tombo can be accessed under the code PT-TT-MSLIV-2776_m0113 [photo!]. The spelling in the text is ‘Bignonia ‌tabebuya’, maybe he mistakenly used an ‘E’ instead of an ‘A’, because the original text, illustration, and its copies are spelled with an ‘A’.

Gentry (1975a) related Bignonia ‌tababuya with Tabebuia ‌cassinoides. Later, this position was also accepted in later treatments (Gentry 1992b; Zuntini 2014), and it is reinforced here.

Bignonia ‌tababuya well represents Tabebuia ‌cassinoides and includes relevant diagnostic features, such as the simple leaves, elliptic, cuneate base, entire margin, terminal inflorescence, infundibuliform corolla, oblong fruit, persistent calyx, and winged seeds (Gentry, 1992b). Although the leaf venation was not represented in the illustration, the lack of this characteristic does not harm the recognition of the species. It is also commented ‘Radix ‌eosdem ‌usus habet, quos suberis cortex’ [The root has the same uses as the cork oak bark], regarding the common uses of the species, which shows that regional people, Vellozo, or someone in the entourage knew this plant.

Bignonia ‌leucoxylla (Fig. 12) was described by Vellozo (1829) as ‘Habitat tum maritimis, cum mediterraneis ad cacumina, et declivia montium’, which recognized double occurrence in Rio de Janeiro and São Paulo states. Especially at the stretch of BR-459 right after the São Paulo border in the direction of Cunha town. Tabebuia ‌obtusifolia is a native plant endemic to Brazil. Following Gentry (1992b) and Lohmann (2023g), this species is found in the Brazilian Atlantic Forest and Cerrado.

Gentry (1975a) considered Bignonia ‌leucoxylla as ‘orthographic variations’ of Bignonia ‌leucoxylon ‌L. So, a binding decision was organized with the names and sent for review. However, parallel usages of the two epithets in the period between 1894 and 1953 were not documented, concluding that, in the interest of nomenclatural stability, the request was unnecessary (John McNeill, pers. comm.). Thus, Bignonia ‌leucoxylla needs to be continually treated as a later homonym of Bignonia ‌leucoxylon (Art. 53.2 of the ICN, Turland et al. 2025).

A copy of the illustration of Bignonia ‌leucoxylla stored in the Manuscript Section of Torre do Tombo can be accessed under the code PT-TT-MSLIV-2776_m0115 [photo!].

Bignonia ‌leucoxylla well represents Tabebuia ‌obtusifolia and includes relevant diagnostic features, such as the marks presented in the young branches (dots or lines), simple leaves, elliptic, rounded apex, cuneate base, entire margin, terminal inflorescence, flowers with large and toothed calyx, and infundibuliform corolla. These characteristics agree with Tabebuia ‌obtusifolia ‌(Gentry 1992b).

In addition, Vellozo (1829) commented: ‘Vulgò dicitur Caixeta’ [Commonly called ‘Caixeta’] as a reference to the common name of the species; ‘Præbet materiariis lignum tenue’ [Provides a thin wood for carpentry], a reference to potential uses of the species; ‘Hunc oʃʃendi apud scaturiginem fluvii Tuguahy’ [This is observed at the source of the Taguay River], a comment about the probable collection place of the plant.

Bignonia ‌curialis was described by Vellozo (1829) as ‘Habitat silvis maritimis, et mediterraneis’, which is recognized as an ambiguous occurrence in Rio de Janeiro and São Paulo states (see comments under Bignonia ‌leucoxylla). Jacaranda ‌jasminoides is a native plant endemic to Brazil. Following Gentry (1992b) and Farias-Singer (2023), this species is found in the Brazilian Atlantic Forest, Caatinga, and Cerrado.

A copy of the illustration of Bignonia ‌curialis stored in the Manuscript Section of Torre do Tombo can be accessed under the code PT-TT-MSLIV-2776_m0117 [photo!].

Candolle (1845) treated Bignonia ‌curialis and made the combination: Jacaranda ‌curialis. Gentry (1975a) suggested that J. ‌curialis (based on B. curialis) should be treated as a synonym of Jacaranda ‌jasminoides. This position was also accepted in later treatments (Gentry 1992b; Zuntini 2014), and it is reinforced here.

Bignonia ‌curialis well represents Jacaranda ‌jasminoides and includes relevant diagnostic features, such as the bi-pinnate leaves, elliptic, acute apex, cuneate base, terminal panicle, tubular corolla, and cupular calyx. These characteristics agree with Jacaranda ‌jasminoides ‌(Gentry 1992b). Although the leaf’s venation was not represented in the illustration, the lack of this characteristic does not harm the recognition of the species.

In addition, Vellozo (1829) commented: ‘Etiam vulgô dicitur Ipê roxo’ [Also widely known as ‘Ipê roxo’], as a reference to the common name of the species; ‘Duritiem, incorruptibibilitatem hujus ligni, cæterorumque, quæ ad hanc familiam pertinent, quis est, qui ignoret?’ [Who is ignorant of the hardness, incorruptibility of this wood, and of the others that pertain to this family?], as comments about the quality of the species wood.

Crescentia ‌cuyeté ‌Vell. was described by Vellozo (1829) as ‘Habitat tum maritimis, tum mediterraneis’, which is recognized as an ambiguous occurrence in Rio de Janeiro and São Paulo states (see comments under Bignonia ‌leucoxylla). Crescentia ‌cujete L. is probably a native plant to at least Northern Central America and Mexico (Gentry, 1980). Following Gentry (1980) and Lohmann (2023a), this species occurs in the Amazon Forest and the Brazilian Atlantic Forest.

Crescentia ‌cuyeté ‌Vell. is here interpreted as a citation of Crescentia ‌cujete L. (1753). Previous analyses also seem to reach the same conclusion, as Gentry (1975a) did not include Crescentia ‌cuyeté ‌Vell. in his study. Thus, in this case, there would be no need to designate a ‘type material’, but rather to mention the ‘reference material’ deposited in the BN (mss1198655_107 [photo!]). A copy of the illustration of Crescentia ‌cuyeté ‌Vell. stored in the Manuscript Section of Torre do Tombo can be accessed under the code PT-TT-MSLIV-2776_m0213 [photo!]. This decision follows the ICN, as we assume that Crescentia ‌cujete L. and Crescentia ‌cuyeté ‌Vell. share the same type, allowing both names to be inserted as orthographic variants (Art. 61.2 of the ICN, Turland et al. 2025).

Gentry (1980) suggested a typification for this species, selecting the specimen LINN 779.1 in the ‘Linnaean Herbarium’. Nevertheless, Wijnands (1983) noticed that this specimen was purchased by Linneaus in July 1758, so it must be rejected. In addition, Wijnands (1983) analyzed another specimen of Linnaean Herbarium (LINN 779.2) and also perceived that it could not be the type of Crescentia ‌cujete ‌L. Thus, Wijnands (1983) designated (with Gentry approval) a lectotype for Crescentia ‌cujete L. based on Plukenet’s illustration.

Crescentia ‌cuyeté ‌Vell. well represents Crescentia ‌cujete L. and includes relevant diagnostic features, such as the simple leaves, cauliflorous inflorescence, bilabiate calyx, tubular-campanulate corolla, and calabash spherical fruit. These characteristics agree with Crescentia ‌cujete L. (Gentry 1980). Although the leaf venation was not represented in the illustration, the lack of this characteristic does not harm the recognition of the species.