&ladru8.~] XLI. CELASTRACE.f; (oliver). 363 



exceeding or equalling the leaves, dichotomous, 1-4 or 5 from each axil, first 

 fork often helow the middle. Calyx 5-partite ; segments rotundate. Ovary 

 ovoid, inserted by a broad base in the centre of a fleshy disk more or less 

 ' free at the margin, 3-celled. Style deeplv 3-fid ; lobes at length spreading 

 or recurved and 2-fid (at least in 'the Natal plant). Fruit I have not seen.— 

 C rubra (Herb. Gerrard). 



Mozamb. Distr. Rovuma river, Drs. Kirk ! and Meller ! Inhambane (Klotzsch). 

 Also south of the tropic. 



'• C Atkaio, Rich. Fl. Abyss, i. 132. Armed ; with terete pulverulent 

 ranches; spines long, straight, slender. Leaves coriaceous, shortly petio- 

 le, elliptic-oblong, usually obtuse, crenate-dentate, glabrous, l-l| in. long, 

 W n ' ac '- Cymes axillary, divaricately forked from the base, equalling 

 * leaves. Peduncles glandular. 



Wile Land. P rov . Choa, Abyssinia, PjiUi. 

 nave not seen a specimen. The description is from Richard, who says the plant is near 

 » -iw/o/HM, Linn. 



8 - C. arbutifolius, Ilochst. ; Rich. Fl. Abyss, i. 133. A stiff, much- 

 ranched, usually spinose shrub. Extremities glabrous or puberulous. Spines 

 j raight, \~i i n . Leaves small, very coriaceous, elliptical or obovate-ellip- 

 2, Ver y °l>tuse or even emarginate, rounded or cuneate at the base, obso- 

 «av serrulate or subentire, glabrous, ^-li in. long, f-f in. broad ; petiole 

 s ual| . y distinct, 1-2 lines. Peduncles axillary, usually 2 or 3 from the 

 atneaxil, $-$. i n> bearing a few-flowered, often umbellate cyme of shortly 

 Fuicellate flowers. Ovary nearly quite free, raised upon a fleshy disk, 3- 

 IJef i; style with 3 spreading or recurved linear lobes. Capsules size of a 

 g*P*. usually 3-valved.— C. parviflorus, Ferr. et Galin. Voy. Abyss. Atlas 



|J ,le Land - Abyssinia^/J^^/ Dr. Roth ! 

 torrT rS ' Ferret and G a'in»er (Voy." Abvss. iii. 107) identify this plant with Celastms 

 Ibecal Vahl (° atha spinosa, Forsk.).* They may be correct, but as Forskal describes 

 psnle as 2-locular, I prefer to leave it under Hochstetter's name at present. 



s J, C * hteolus, Delile in Ann. Sc. Nat. Ser. 2. xx. 90. A glabrous 

 ,. °_, unarmed in our specimens. Leaves very coriaceous, pale green, ovate- 

 ■WH or varying from oval to oblanceolate, subacute or obtuse, more or 

 ct "ieate at base, usually coarselv dentate-serrate or sinuate-dentate, with 



sm , or a cute teeth, glabrous, l£-4 in. long, £-l| in- broad > occasionally 

 fi) , r; P e tiole 1-3 lines. Pedicels simple, in axillary 2-6-8-flowered 

 base I / . articulat ed near the base, in fruit 2-4 lines long. Ovary broad- 

 2>"»mersed in a fleshy disk, 3-celled. Stigma 3-lobulate. Capsule 

 «w } Size of a pea.— Ferr. et Galin. Voy. Abyss. Atlas Hot. t. 8.-C. 

 ^■dentatns, Hochst. in PL Schimp. Abyss. 



BXoL I *t nd " Ab yssinia, Schimper ! ,. . . 



j f anab - I>istr. ? Zanzibar, Br. Kirk ! In fruit only. Perhaps distinct. 

 "*o in fr .? icbard - w ho is probably right, in his identification of Delile's plant. A spec, - 

 %I £ D ° f ' "PParently, an allied plant collected at Cape Coast {Bra») is in the herba- 



10 of 'he British Museum. 



