13 
Has.—Natal, Inanda 1,800 feet alt., Wood, 616, 3843. 
Caulis cue pedalis, Large florifera 1-1} ped. longa. Folia 2-3 poll. 
longa, 2-11 poll. lata. Racemi, 14-2} poll. longi. Bractee 1} lin. 
longe, já -1} lin. latæ. Pedicelli 13-2 lin. longi, Calyx 3 lin. longus. 
Corolla 7 lin. longa. 
Var. longipes, V.E. Brown. Pedicellis quam bracteis 14-3 plo 
longioribus; corolla aurantiaca di morphs. aliæ profunde faleato-cy ced 
formes, alic rectæ minus alte cymbifor 
Han.—Tropieal Africa: ae Hills, 1,000-2 ,000 feet alt., 
Meller ; Maravi Country, West of Lake Nyassa, Kirk ; lower plateau 
of Lake Nyassa, Thomson; Buchanan, 885; Blantyre, Scott ; valley 
of Umzingwani River, Baines; Angola, Cunene, Johnston, 
Caulis 8-12 pedalis. Bractee 1-2 lin. long». Pedicelli 2-5 lin. 
longi. Corolla 5-9 lin. longa. 
his species is closely allied to P. esculentus, but differs in its more 
erect habit, taller stems, and closely sessile leaves, with broader rounded 
bases, rather rougher surface, and more prominent reticulation. The 
tropical variety longipes may prove to be a distinct species when the 
leaves are known, but in DUE except the length of the pedicels it 
agrees with the Natal plan 
As in the case of veni other Labiatze, the corolla shows consider- 
able variation ; two specimens collected by Meller at the same place and 
time have very different corollas; in one it is only five lines long, with 
the lower lip straight and very shallowly boat-shaped, in the other it is 
8-9 lines long with the lower lip falcate and deeply boat-shaped ; possibly 
they are sexual forms. 
‘The three species of Plectranthus described above, together with 
P. defoliatus, Hochst., form a small group differing from all the rest of 
the genus us by the peculiar ity of flowering after the leaves have all 
fallen away. Possibly they are all tuberous rooted species, but P. escu- 
lentus is the only one which affords evidence of this. The leaves of all 
but the Natal plants are at present unknown. Although all have the 
same habit, it is noteworthy that, whilst P. densus and P. defoliatus 
have all the calyx-teeth equal, and belon z tothe section Jsodon, the 
others have the large elliptic upper calyx-tooth of the section Coleoides, 
with which they also agree in the form of their corolla. 
Plectranthus incanus, Link Enum. II., p.120; Hook. fil, Fl. Brit. 
Ind., IV., p. 621. Plectranthus cordifolius, Don Prod. Fl. Nep., p. 
116; and Benth. in DC. Prod. x ii, p. 66. This Indian plant is 
stated by Bentham to have a tuberous root, but this is certainly an error. 
None of the other authors who have dispen the plant under various 
names say that it is tuberous. Link describes it as a perennial, and Aiton 
chara an irse and this last appears from the specimens to be its true 
c 
riim anthus madagascari ensis, Benth. Labiate, p. 38, and in DC. 
Prod. 68; Baker, Fl. Maurit., p. 258. . mauritianus, 
Bojer, Hon. Maurit., p. 254. This plant is mentioned by several 
authors (Naudin and Mueller, Manel de PAcclimateur, p. 428; 
Mueller, Select Extra-tropical PL, ed. 8, p. 371; and Pollless ct 
Bois in Hev. Scien. Nat. Appliq., 101, p. 686), as a tubereus-rooted 
Labiate cultivated in Madagasear, Mauritius, and the East Coast of 
Africa, under the name of Oumime or Hounine. But in all probability 
his is an error, and the plant intended is most likelv aee Eos 
