Notes on the nomenclature and typification of eight names in Bignoniaceae

Lectotype is designated for two names, Bignonia xylocarpa Roxb., Radermachera tonkinensis Dop and second - step lectotype is also designated for six names, Radermachera corymbosa Steenis, R. elegans Steenis, R. pentandra Hemsl., R. pierrei Dop, R. poilanei Dop and Stereospermum seemannii Rolfe


Materials and Methods
An extensive literature survey was carried out and the protologue of each name was studied. The types were studied from online resources of A, E, HK, K, L, NY, MO, P, U, US (13), JSTOR Global Plants and the Director of BO was also consulted for study of type images. Roxburgh icons at CAL and K were also studied.

Notes
Bignonia xylocarpa Roxb. was introduced into Calcutta Botanic Garden from Soonda by Dr. Andrew Berry (14). Roxburgh's understanding of this species was based on either live wild plants or the specimens grown in the Garden (14). We could not trace any specimens of Roxburgh but could trace Roxburgh's drawing number 2076 at CAL and K. The drawing at K is depicted with leaves tri-pinnate; panicles terminal; calyx 5-toothed; corolla large, white, tinged yellow, 5-6-lobed, lobes sub-orbicular and much curled along margins; stamens 5-6, stigma 2-lobed; fruits linear, woody, rough, tubercled; seeds winged, all matching well with characters specified in the protologue. Additionally, the Kew drawing is clear, partially coloured and comparatively better than the CAL drawing. Therefore, we designate Roxburgh Icon No. 2076 (K) as the lectotype.

Notes
The protologue cites two gatherings, Achmad 164, and Achmad 181 (15). We have been able to trace two specimens (BO, L) belonging to Achmad 164 and five specimens (BO, L) belonging to Achmad 811 [181 as error in the protologue]. The protologue associated the vernacular name koedo koedo paya with collection number 181. The collection number 181 in the protologue is erroneous because all five specimens (BO and L) with vernacular name koedo koedo paya bear the collection number 811. Among all specimens, two sheets belonging to Achmad 811 at BO were annotated as "Type specimen!". Moreover, 'type specimen' was also cited with Achmad 811 (BO) in the protologue, therefore, these two specimens as well as three specimens at L can be considered as syntypes. van Steenis (7) cited "Type: Achmad 181 (BO, L). N. Sumatra, Simalur I" and his indication of type should be considered as inadvertent first-step lectotypification because specifying herbaria was not necessary before 1990 under Art. 9.22 of the Code and Achmad 811 (181 as error) is a single gathering. The specimen BO-1297607 is in brittle condition and bears three fragments devoid of leaves, flowers, fruits, two detached leaves and van Steenis' annotation 'Type specimen!'. We designate Achmad 811 (BO-1297608) ( Fig. 1) as the second-step lectotype because the specimen is in good condition and bears a single fragment with leaves and the annotation 'Type specimen!' in van Steenis' handwriting.

Notes
Radermachera elegans Steenis was described based on the gatherings Koorders 16253β, 16254β, 16255β, 16256β, and 16257β collected from Sulawesi (Celebes) which can be considered as syntypes. We could trace three specimens belonging to Koorders 16253β, 16254β, 16255β at L and there are no extant specimens at BO. Furthermore, Koorders 16256β (BO, L) was cited as the "type specimen" in the protologue and we have been able to trace two specimens of this collection at BO and L. van Steenis (7) cited "Type: Koorders 16256 (BO, L), NE Celebes, Minahasa" and it should be considered as inadvertent first-step lectotypification following Art. 7.11 and 9.22. Therefore, we designate here Koorders 16256β (BO-1897992) (Fig. 2) as second-step lectotype because it bears the annotation "type specimen!" in van Steenis' handwriting.

Notes
A single gathering, Pierre 570 was cited in the protologue (16). We have been able to trace nine specimens at A, K, L, NY and P and all are syntypes under Art. 40 Note 1 of the Code and McNeill (17). van Steenis (7) cited "Type: Pierre 570 (A, K, P), Cambodia, Mt. Schral". Although three herbaria were cited by van Steenis (7), but before 1990 speci-fying herbaria was not necessary for effective lectotypificaton under Art. 9.22 of the Code. As Pierre 570 is single gathering, therefore, citation of type should be accepted as inadvertent first-step lectotypification under Art. 9.17. Therefore, second-step lectotype designation is required and we designate Pierre 570 (P00609721) as second-step lectotype because it bears Dop's annotation and a packet bearing flowers and seeds are also attached.  Henry 10909 (A, K), Yunnan, Mengtze" and his indication of "type" must be accepted as inadvertent first-step lectotypification following Art. 9.17 and Art. 9.22 because A. Henry 10909 is a single gathering. However, two herbaria were mentioned by van Steenis but specifying herbaria were not necessary before 1990 for effective lectotype designation under Art. 9.22. Therefore, the name would undergo second-step lectotype designation and we designate here the specimen K000779284 as second-step lectotype because Hemsley was based at Kew.

Notes
Dop (19) described Radermachera poilanei based on two gatherings, Poilane 5953, and Poilane 8759. We have been able to trace five specimens belonging to Poilane 5953 at A, NY, P and three specimens belonging to Poilane 8759 at K, P. Santisuk (4) cited "Type-Annam: Ca-na, Phanrang, 1-4-23, Poilane 5953 (K, P)" and this citation of type should be accepted as effective lectotypification under Art. 7.11 of the Code because specifying herbaria were not necessary before 1990 under Art. 9.22. As Poilane 5953 is a single gathering deposited at A, NY, P, a second-step lectotypification is required under Art. 9.17. Thus, we designate here Poilane 5953 (P00609723) as second-step lectotype because the specimen bears Dop's annotation.

Notes
Rolfe (20) cited a single gathering, Cuming 996 from Philippines in the protologue of Stereospermum seemannii.
There are two specimens at K and one specimen at P which can be considered as syntypes following McNeill (17) and Art. 40 Note 1 of the Code. van Steenis (7) cited "Type: Cuming 996 (BM, K), Philippines" and his indication of "type" should be considered as inadvertent lectotypification following Art. 7.11 and 9.22 because specifying herbaria were not necessary before 1990. The specimen K000779259 consists of a flowering fragment without leaves, and is mounted together with another specimen, with barcode K000779260, bearing the annotation "This lower specimen is Stereospermum chelonoides DC., and is probably not from the Philippines at all" by an unknown hand. We have chosen to designate the specimen K000779261 as second-step lectotype because it is unambiguously a part of the type gathering and bears Rolfe's annotation "Stereospermum seemannii Rolfe!" in his own handwriting. .