114 
the same author, in the same publication, described the very distinct R. spectabilis. I am 
doubtful whether R. cinnamomea is distinct from Dolicholus venulosus, Hiern. (R. venulosus, 
K. Schum.), and I think R. Rehmanni is synonymous with Fagelia bitumanosa, D.C. In 1906 
Dr. Bolus, n.” Trans. $. Afr. Phil. Soe.” XVI. 383, described the curious and aberrant 
Rk. pauciflora. 1 hardly think R. uncinata, Schlechier, from Packhuis Berg, is separable 
from R. bullata, Benth. In 1918, in “ Bull. Soc. Bot. Fr.,” LX, 461, Gandoger quite correctly 
segregated the R. memnonia of Harvey, calling the silvery plant with a rigid stem R. 
albissima. 
Several plants have been issued with ms. names to which I must allude. Wilms No. 375 
J would place as synonymous with R&. monophylla, Schlechter; Wilms No. 387 appears to 
me to be a form with more angled leaves than usual of R. sigmodes, Benth. I have retained 
R. graciliflora, Harms, founded on Wilms No. 367, as a variety of R. Totta, D.C.; it differs 
in the shape of the leaflets from the typical form. Quite recently in the “ Kew Bulletin ”’ 
Mr. Burtt-Davy described R. wnifoliolata. It is also a close ally of the true R. Totta, D.C., 
but is sometimes unifoliolate, and has shorter petioles. A litle later in the same publication 
he also describes R. confusa, R. Burkei, R. Harmsiana Schl., var. Burchell, and R. Pentheri 
Schl., var. Hutchinsoniana. With R. Burkei and R. Pentheri, S chl., var. Hutchinsoniana, he 
has associated my name. 
NOTES ON TAXONOMY. 
Harvey, in the “ Flora Capensis,” arranges the species of Rhynchosia then known to 
him in four sections: (1) Chrysoscias ; (ii) Polytropia : (im) Copisma ; and (iv) Orthodanum. 
The first of these is well marked by the character of the inflorescence, which is generally 
umbellate, more rarely the flowers are solitary. The leaflets are narrow and either tomentose 
or villose below. T his section I retain as originé ally delimited. 
The section Polytropia is at once recognized by having pinnate or bipinnate leaves. 
But there is a difficulty with regard to the sections Copisma and Orthodanum. Certain 
plants in these sections are very distinct froni one another ; take, for instance, RF adenodes, 
Ecklon and Zeyher, and R. minima, D.C., with a distinctly voluble stem, which is the main 
character of Harvev’s Copisma, and R. Orthodanum, Be nth., which is erect, rigid, and suffru- 
ticose, and which is the type of Harvey’s section Orthodanum. But there are certain 
intermediate plants which are difficult to place. Take, for instance, R. viscidula, Steud. : 
the plant from Little Namaqualand has a somewhat rigid stem and cannot be correctly 
placed in Copisma. | have, therefore, thought it advisable to unite the sections Copisma 
and Orthodanum, and am calling this section Eurhynchosia. 
I have introduced the section Cyanospermum. Benth., for R. cyanosperma, Benth. 
It may be easily recognized by the robust voluble stem, the racemose inflorescence, the 
broad calycine segments, whic 'h are densely matted on the back with a grey tomentum, 
and the dark blue seeds. I am also introducing the section Arcyphyllum Torrey and Gray 
for two species which occur south of the Tropic. The stem in this section is voluble, the 
flowers are in dense, oblong, sessile, subsessile, or pedunculate clusters, and with the calycine 
teeth much narrower than in section Cyanospermum. 
There are one or two anomalous species which require consideration. One of these 
is R. pauciflora, Bolus, with solitary pedicellate flowers and broad calycine lobes. It comes 
from the Transvaal and Swaziland. R. ficifolia, Benth., I am excluding from the genus. 
It is a Pueraria. Some species from Great Namaqualand such as #. namaensis, Schinz, 
I have not included. as it is recorded from Rehoboth, which is just north of the Tropic. 
DESCRIPTION OF SPECIES. 
SECT. I.—CYANOSPERMUM, Benth., in Benth. and Hook “Gen. Plant ” (1865), 543. 
Stem voluble, robust. Inflorescence racemose. Calyx as long as or slightly longer than ~ 
the corolla, segments rather broad, subequal in length, _Uppermost connate. Poes] 
dark blue. 
> 
