Gigasiphon humblotianum (Leguminosae: Caesalpinioideae: Bauhinieae) as a Drift Seed

Abstract

During a meeting of the Conchological Society of Great Britain and Ireland F. J. M. Pinn asked me if I could name various drift fruits and seeds he had discovered on the Kenya Coast at Msumarimi, 6-4 km N of Kikambala, whilst hunting for shells. I attempted to do this with only moderate success. One seed, obviously leguminous, was curious owing to the very characteristic crescentic funicular-lobe scars, restricted to the tribe Bauhineae and reaching a point of maximum development in the seed under notice, which proved to belong to the genus Gigasiphon (Baill.) Drake. I should have recognised the genus at once, having collected the very rare East African species, G. macrosiphon (Harms) Brenan, for the first time in Kenya and also having had to deal with the New Guinea species G. schlechteri (Harms) De Wit quite recently, (Plate 23 top and middle rows left, bottom row right.) My attempts to name the seed led me to the collection of drift seeds preserved in the Museum at Kew where indeed I did manage to match it with several seeds wrongly named 'Mucuna sp.' collected on the beaches of the Aldabra group of islands. It also matched a drawing in C. R. Gunn & I. V. Dennis's book on drift seeds (A World Guide to Tropical Drift Seeds and Fruits. New York, 1976, fig. 66/A-D), also incorrectly captioned Mucuna sp. Once it was evident that all these seeds belonged to Gigasiphon it became clear that they had to belong to G. humblotianum (Baill.) Drake, the Madagascan species of the genus of which only a photograph of the type was represented in the Kew Herbarium. Baillon described it in a short note entitled 'La plus longue fleur de la famille des Legumineuses' under the genus Bauhinia mentioning, however, that it deserved a section which he called Gigasiphon. He gave no locality but added 'sa gousse t later in the same bulletin he mentions the plant again giving the vague locality 'Humblot, Madag. bor.-or. absque loco (1880)'. Drake treated Gigasiphon as a genus also mentioning that Humblot has said 'gousse tres grande' and adding the same vague indication of locality 'r6gion Nord-Est'. Wondering about the present status of this plant I wrote to Prof. J.-F. Leroy at the Paris Herbarium and he kindly arranged for two more recent sheets to be sent to Kew on loan. One of these has a fruit with a portion removed to reveal one of the seeds. Unfortunately this material still does not absolutely confirm my conclusion since the seeds in the pod are small, only 2-3 cm wide; but despite the fact they are loose within the pod I feel certain that they are quite undeveloped particularly bearing in mind the great width of the pod. The other particulars fit well; G. humblotianum grows by the sea and there are ocean