AFRICAN SPECIES OF CROTALARIA. 945 
enumerated the 34 species then known from that region, many of them 
described by Dr. De Wildeman in the ‘Annales’ of the Congo Museum ; 
but the number of species now known is somewhat larger, thanks principally 
to the labours of T. Küssner, M. Bequaert, and the Rev. F. Rogers. 
The plant enumerated in the * Sylloge? as C. spartea, R. Br., is C. glauca, 
Willd., but by a curious coincidence the true C. spartea has recently 
been collected in the Congo Region ; C. lanceolata, E. Meyer, is in part 
C. intermedia, Kotschy ; C. Descampsü, Micheli, is a very close ally 
of C. nytkensis, Baker, and C. dubia, De Wild., appears hardly to be 
specifically distinct from C. prolongata, Baker. 
There are, in addition to the above, a number of other descriptions of 
plants of this genus, principally by Kotschy, Dr. Schweinfurth, Dr. Taubert, 
Dr. Harms, and my father, to which I have referred in the body of the 
paper. 
EXTERNAL MORPHOLOGY AND CLASSIFICATION. 
The Genus Crotalaria belongs to the Tribe Genisteae of the Papilionacem. 
The flowers are zygomorphic and of the ordinary Papilionaceous type, the 
posterior petal being the Standard or Vexillum, the lateral petals the Wings 
or Al», and the two anterior being eonnate to form the Keel or Carina. 
There is considerable diversity of Aabit within the limits of the genus. 
Some species, such as C. Küssneri, Bak. fil, and C. parvula, Welw., are 
dwarf, erect, or suberect annuals ; some, like C. Baumii, Harms, are pros- 
trate annuals ; some, like C. psammophila, Harms, are annuals about 1 metre 
high. Numerous species are suffruticose. C. dumosa, Franchet, is a little 
shrub. Many of the species are tall shrubs such as C. agatiflora, Schweinf., 
C. imperialis, Taubert, and C. Engleri, Harms, reaching 3-4 metres high 
or more. Some of the African species are spinescent. 
The stipules are organs of great importance for taxonomic purposes. 
In some species they are very small or completely obsolete, e. g., in 
C. striata, DC., in others they are always setaceous; sometimes they are 
lanceolate, while in some species they are foliaceous and may be either 
sessile or petiolulate; sometimes they are semilunate, as for instance in 
C. verrucosa, Linn. 
The leaves are either simple or trifoliolate, or rarely 3-4-foliolate, or 
5-foliolate as in C. cleomifolia, Welw. and C. Burkeana, Benth. ; but I have 
not retained the Group Multifoliolatz, since in C. cleomifolia for instance the 
leaves are also sometimes trifoliolate, and €. Burkeana, Benth., is so clearly 
an ally of the 3-foliolate C. polysperma, Kotschy, it seems advisable that 
in the sequence these species should closely approximate to one another. 
In the calyx the lobes are generally free and either equal in length to, 
or longer than, the calyx-tube ; they are generally either lanceolate, or 
triangular, or ovate. In C. platysepala, Harvey, they are broad and suddenly 
