A REVISION OF THE AFRICAN spe 
OF SESBANIA. 
By E. P. PuiLuips, Division of Botany, Pretoria, and J. Hurcninson (Kew). 
THE present paper is an attempt to revise the African species of the genus Sesbania, family 
Papilionaceae. Imperfect as it may prove to be, it is long overdue, chiefly because of 
the great accumulation of herbarium material since the publication of the second volume 
of the “‘ Flora of Tropical Africa `` in 1871, much of this material having remained unnamed 
or imperfectly determined. ‘The results of this investigation might. very well have been 
more satisfactory to the authors had there been more field notes available regarding the 
situation, habit, floral colouring, etc., of the specimens accumulated in the various herbaria 
which they have been able to consult.* That this information is vital in the determination 
and limitation of the species of Sesbania, at least, has been well demonstrated by Prain f 
in his critical elucidation of the Indian species. ; 
The genus Sesbania contains about fifty species which occur in the warmer parts of 
the EE mainly im or by the sides of streams, lakes, and sw amps. It appears to attain 
its greatest development in tropical Africa, a few of the species extending into South 
Africa as far as Natal, and into various parts of the Transvaal and Bechuanaland. In 
the present revision of the species from this area twenty-three are recognized to be distinct. 
They belong to two sections, nineteen to Eusesbania and four to Daubentonia, the latter 
characterized by its four- winged fruits. The third and purely American section into 
which Sesbania was divided by Bentham and Hooker (Gen. PI., I, 502) is now recognized 
by American botanists T as a distinct genus, Glottidivm, and the present authors’ views 
coincide in respect to this. They cannot, howev er, go so far as to accept the American 
view regarding the generic status of Daubentonia w hich occurs in the three widely separated 
areas, namely, the ‘south-eastern United States and Mexico, sub-tropical South America, 
and tropical East Africa. This broken distribution seems to point to a separate origin 
of the species of Daubentonia from the basal stock, Husesbania, species of which occur 
in all these areas. Glottidiwm, on the other hand, a native of Florida, is well separated 
from Sesbania by its short fruits with only two seeds and the manner of dehiscence, the 
seeds remaining inside the dry bladder-like endocarp which detaches itself as a whole from 
the outer shell. 
In the case of the African species we have found a most useful and constant character 
in the nature of the appendages on the claw of the vexillum. In the first five species 
shown in the key these are long and quite free from the vexillum in their upper half. In 
* For the privileges of examining the specimens under their charge, the authors tender their thanks 
to Sir David Prain, Director, Royal Botanical Gardens, Kew; Dr. A. B. Rendle, Natural History 
Museum, South Kensington; Dr. 1. B. Pole Evans, Director, Botanical Survey, South Africa ; Mrs. Bolus, 
Bolus Herbarium, Capetown; Dr. L. Perinquey, South African Museum; Dr. Schénland, Albany 
Museum, Grahamstown; the Director, Transvaal Museum, Pretoria; and Mr. Fred Eyles, Salisbury, 
Rhodesia. 
+ Prain in Journ. Anat. Soc., Bengal, LX VI, 366-370 (1897). 
t¢ Small, Fl. South-Eastern Unit. States, Ed. II, 615 (1913). 
