56 
Forest Flora of Portuguese East Africa. 
A. Catechu, Willd. Vcm. nanus — io, Moroga and Nematonga. A large or medium-sized tree, having short, stout, recurved thorns usually present in pairs. Leaves 
12-18 c.m. long, the rachis somewhat pubescent, with a sessile gland near the base and several between the pinnae ; pinnae many ; leaflets many, smaller than those 
of A. caffra. Inflorescence .spicate; spikes axillary, 5-10 c.m. long, petals much longer than the calyx. Pod flat, linear, about 6-seeded, 10 c.m. long, 2 c.m. wide. 
Zambesia, scarce. Mr. Burtt-Davy, speaking of this species in the Transvaal says : — “Our largest species of Acacia, forming a handsome and striking tree on the 
river banks of the eastern middle veld, between 2000 and 2500 ft. alt called ‘ White Thorn ’ . . . . tall trees with trunks up to and 2 ft. diam.” 
A. caffra, Willd., var. rupestris. (New variety.)' Vcm. names — 4, 13, Tole; 15, Mutholwane. Small tree with leaf 15 c.m. long; about 14 pairs of pinnae; about 30 
pairs of leaflets to each ; rachis with gland near base and between one or more of the upper pairs. Stem somewhat angular, with recurved prickles 3-7 m.m. long, 
scattered on its surface. Inflorescence spicate. Spike 7-10 c.m. long, seldom panicled. Pod 10-12 c.m. long, i‘5 c.m. wide, pointed, tapering to both ends, flat. 
1 he whole tree pubescent. Probably corresponds with A. eriadenia, Benth., but is not specifically distinct from A. caffra. The type of A. caffra was not noticed in 
the Province ; it is a larger tree, usually near water, less pubescent and more panicled. The present form is frequent on and below the Lebombos, in Maputa and 
Marracuene, less common nearer Lourenzo Marques, and not seen in other districts. It occurs in the Transvaal, Natal and Cape Colony. Sim. 6235. 
Plate XXXIX. B. x, Leaf and spike ; 2, Pod. 
A Burkei, Benth. Unknown to me; described by Harvey, as: — “Young twigs and petioles patently hispid; prickles in pairs just below the nodes, recurved; petiole 
unarmed or armed, and with a petiolar gland below the pairs ; pinnae 3-6 jugate ; leaflets 5-8 jugate, obliquely oblong or obovate, obtuse or mucronulate, the 
younger villous at the margin ; spikes tufted, loose, peduncled, rather longer than the leaf. Legume ? Bark dark ashey-brown, rugged. Leaflets drying pale 
4 lines long, 1^-2 lines wide, the uppermost largest and obovate. Spikes 2-3 inches long.” 3 
Mr. Burtt-Davy, who calls this Aapjies-doorn cites, along with Transvaal localities, “Manganja hills, Zambesia, fine tree, 30-40 feet high; wood not hard but 
tenaceous, used by natives for hoe handles, &c\, Melkr 9 in Herb Kew.” 
A. delagoensis, Harms. Concerning this species I know nothing beyond Mr. Burtt-Davy’s record , } viz. “Umbolosi, Schlechter, 11,718, in Herb Lurich. Probably 
extends into the low veld of the Transvaal, apparently near to A. Burkei.” y 
A. lasiopetala, Ohv. Unknown to me, described by Prof. Oliver as " Extremities uniformly softly hoary-tomentose. Stipular spines, straight or subrecurved 
pubescent nearly to the apex. Young leaves, especially on the rachis, silky tomentose ; pinnae in 14-22 pairs; leaflets 20-30 pairs, probably often more numerous’ 
not wholly developed in our specimens ; gland sessile near the base of the rachis. Peduncles in pairs or solitary from the lower axils, pubescent, 1-2 inches long’ 
Bracts obsolete. Flowers capitate. Calyx pilose pubescent, lobes ovate obtuse. Petals externally silky, cohering | or § their length, about half as long again as 
the calyx. Ovary subsessile, glabrous, subtruncate above, with an obliquely inserted filiform style. Fruit unknown. Foot of Impemba Peak, Shire River Dr. 
A i*L’ . Mr : Bur “-Davy, who has compared Transvaal specimens with those from the Zambesi (Buchanan’s), says:— “In foliage and pubescence it is much like 
A. Rehmanmana, Sch.nz, but the legumes are large and woody, in shape a little approaching those of A. hebeclada, but longer, glabrous and shining. The bark is 
quite diflerent from that of A. Rehmanmana, being pale yellow, soft and papery, at length flaking off.” 
A. pennata, Willd. Vent. name — 4 Inekwa. Scandent by nature, erect when support is absent. Prickles scattered round the stem and between the nodes often 
numerous, not stipular, recurved, 1 c.m. long. Leaf 12-18 c.m. long ; the rachis often bearing prickles on its lower surface, and a gland on its upper • pinn^io-xc 
pairs or not regularly opposite, leaflets ,5-30 pairs, 7 m.m. long, 2 m.m. wide. Inflorescence capitate in axillary and terminal panicles. Axillary abordve branches 
set with numerous small recurved prickles, and with tendril nature assist the scandent habit, but where there is no support the plant assumes erect habit 4-xo feet 
high much branched, and with the inflorescence in lax terminal panicles, with numerous capitate heads and eventually 3 o or more pods in a panicle Pod 10 ic 
c.m. long, 2-2 5 c.m. wide, thinly coriaceous, flat, many-seeded. Seeds oblong, 7 m.m. long, dark brown. A bad forest weed. Umbalusi; Lourenzo Marques &c 
Plate XXXVIII. B. 
• Acacia caffra, Willd., var. rupestris. (Var. nov.) Arbor parva pubescens ; cauli paullo angulato aculeis recurvis 3-7 m.m. loneis soarsis arraato ■ Mil* ic n m t„ r 
“ c “ 3 °J"f 5 * racl “ f-W. P'°P» I»»i» into pinnies pnrfila J imWccn.ia spies. 7 ..o c.m. long., nuo panicuhla ■ Luminc io-,i c m loll ,'-Tc T,’ “’'T* 
Maputa et Marracuene, minus (requens versus Lourenzo Marques. Verisimiliter A. eriadenia Benth Sim 6235. Tab. XXXIX. B. & 0 12 c.m. longo, 1 5 c.m. lato, acuto, piano. Frequens in 
3 “ Flora of Tropical Africa," II., 282. * Kew Bulletin No. 4. 1908, 157. « “ Flora of Tropical Africa," II., 347. 
