Flora Zambesiaca

Taxon Detail

FZ volume:2 part:1 (1963) Geraniaceae by T. Müller

(Search other Kew databases for: Pelargonium asperum)

Pelargonium graveolens L’Hérit.

in Ait., Hort. Kew. 2: 423 (1789); Geran.: t. 11 (1791–92). — Harv. in Harv. & Sond., F.C. 1: 306 (1860). — Knuth in Engl., Pflanzenr. IV, 129: 428 (1912). — Burtt Davy, F.P.F.T. 1: 190 (1926). TAB. 22. Type a plant from S. Africa cultivated in Europe.


Geraniospermum terebinthenaceum (Cav.) Kuntze

Rev. Gen. Pl. 1: 94 (1891). Type as for Geranium terebenthinaceum.

Geranium graveolens (L’Hérit.) Thunb.

Prodr. Pl. Cap. 2: 115 (1800). Type as for Pelargonium graveolens.

Geranium radula Roth

Bot. Abh.: 51, t. 10 and t. 12 fig. 9–16 (1787), non Cav. (1787). Type a plant from S. Africa cultivated in Europe.

Geranium terebenthinaceum Cav.

Diss. Bot. 4: 250, t. 114 fig. 1 (1787), non Murr. (1785). Type a plant from S. Africa cultivated in Europe.

Pelargonium asperum Ehrh. ex Willd.

in L., Sp. Pl. ed. 4, 3: 678 (1800). Type as for Geranium radula.



E: Melsetter, Chimanimani Mts., N. of upper Bundi Plain, fl. & fr. 29.xii.1959, Goodier & Phipps 335 (BM; COI; K; LISC; SRGH).


Cape Prov;  Transvaal; 


Among rocks 1500–2200 m.


Suffrutex up to 1·3 m. tall, strongly aromatic; internodes 1–8 cm. (or 0·2–1 cm. in E.M.&W. 348) long; vegetative parts, peduncles and pedicels glandular and pubescent, tomentose when young, hairs patent or ± oppressed on the leaves, glands short-stalked to ± sessile. Leaf-lamina 2–7 × 2·5–8 cm. (2·3 × 2·5–4 cm. in E.M.&W. 348), broadly ovate to depressed-ovate, pinnatipartite or sometimes almost palmatipartite, cordate at the base; two basal segments much larger than the rest and often further subdivided with segments often pinnatilobed and irregularly dentate, margin revolute; petiole l·5–4 (8) cm. long (0·2–1·7 cm. in E.M. & W. 348); stipules 4–9 × 3–7 mm., deltate to broadly ovate, acute, often bifid, membranous. Inflorescence a pseudumbel of (1) 2–5 (7) flowers; peduncle (0·2) 1–4 cm. long, leaf-opposed or in the axil of the smaller of 2 apparently opposite leaves; bracts 3–8, 5–8 × 2–4 mm., ovate, membranous; free part of pedicel (2) 4–8 (10) mm. long. Spur 4–8 (10) mm. long. Sepals 7·5–11 × 2–5 mm., narrowly lanceolate to narrowly ovate or lorate to narrowly oblong, acute, patent-pubescent and glandular. Petals pink with darker veins; 2 posterior ones (14) 17–20 × 5–5·6 (6) mm., spathulate; 3 anterior ones (11) 13–15 × 2–3·5 (4) mm., oblanceolate to narrowly obovate, unguiculate. Stamens with fertile filaments 7, 10–16 mm. long; sterile filaments (staminodes) 3, 5–8 mm. long, all 10 connate at the base for 1·5–5 mm.; anthers 2–2·8 × 0·9–1 ·2 (1 ·5) mm. Ovary with basal part tomentose; rostrum pubescent; style 5–8 mm. long; stigmas 2–2·8 (3·2) mm. long. Fruit 1·8–2·4 cm. long; cocci 5–5·5 × 1·4–1·7 mm. Seeds 3·2–3·6 × 1·4–1·8 mm., obovate, pale brown, minutely reticulate.


Some specimens from Umtali Distr. (e.g. E.M. & W. 348 referred to in the description) have much shorter internodes and peduncles, and slightly smaller leaves, but they agree otherwise and are probably habitat forms. Specimens from the Transvaal have leaves with more numerous and narrower segments, and the pedicels are shorter (2–4 mm.). The few specimens I have seen from the Cape seem to agree more with the material from our area. The type of Pelargonium graveolens is a cultivated plant and differs from the wild specimens, besides slight differences in general appearance, in having very short filaments (up to 6 mm. long), a feature which I have noticed in all the cultivated plants I have studied, but never on a wild plant. It is possible that these cultivated plants are of hybrid origin, in which case the wild form should be given a new name. I prefer, however, at least for the present, to maintain the established name P. graveolens for the wild plants, as the solution of this problem is beyond the scope of this Flora. A new description of the wild material would involve the exact establishment of the somewhat obscure relationship between it and the very closely related P. radens H. E. Moore (and also some undescribed material).It would be inadvisable to transfer the name Geranium radula Roth to Pelargonium as the same name was also used for P. radens H. E. Moore in the same year (Cav., Diss. Bot. 4: 262, t. 101 fig. 1) and at present it is not possible to decide which publication has priority. Moore has used the name P. asperum Ehrh. ex Willd. for the hybrid P. graveolens × P. radens (in Baileya, 3: 22 (1955)). I feel certain, however, that L’Héritier’s plate and the one of Roth to which Willdenow refers are identical.