AUNOTT ON SOUTH AFRICAN PLANTS, 255 



character, the ovarium is saitl to be always trilocular, where- 

 as, in several of the Cape species, it is decidedly 2-celled, so 

 that it is doubtful what is intended to be done with these. 



I shall now take the species in the order in which Drege 

 distributed them. — I. Celastrus obtusiis, laiirinus, and No. 

 1925, have the ovary immersed in the disk, a bifid style, 

 stigmas flat, oval and spreading, and belong to Scytophylliim, 

 E. and Z. — 2. No. 6*727, b.; here are five petals and 

 stamens, the latter inserted between the lobes of the disk ; 

 style one, cylindrical; stigmas three, patent; ovary nearly 

 quite immersed in a fleshy 5-lobed and crenulated disk, 3- 

 celled ; ovules 2, collateral in each cell : this I refer to 

 Pterocelastrus — 3. No. 6725. Petals five, patent ; style one, 

 short and thick; stigmas three, short ; ovary immersed in the 

 disk, 3-celled; ovvdes 2, collateral, in each cell; the other 

 characters nearly as in No. 6727, whence I refer this also to 

 Pterocelastrus. — 4. C. lanceolatus; this belonsfs to Celastrus 

 E. and Z., and is perhaps C. stenophyllus, E. and Z,; the 

 style is thick, stigmas two, erect, ovary seated on a 5-lobed 

 fieshy disk, and 2-celled ; the last character separates it from 

 Catha^ Endl., but it is nearer that than Celastrus^ Endi. — 5. 

 C. linearis, Th. seems correctly named ; it is very closely 

 allied to the last species, and exhibits the same structure of 

 flower and ovary; in both, the ovules are in pairs in each cell, 

 and collateral. — 6. C. refracta, E. M. ; petals erect, oblong, 

 and stamens five; ovary scarcely half immersed (nearly sessile) 

 in a crenulated disk ; style short, thick ; stigmas two, emar- 

 ginately 2-lobed, erect ; ovary 2-celled, 4-ovuled. The leaves 

 are opposite, and the branches acutely 4-6 angled or almost 

 winged ; hence I infer that it is Cassine scandens, E. and Z. ; 

 but it cannot be a true Cassine, if Endlicher be correct in re- 

 ferrinor that genus to IlicinecE, although I have reason to en- 

 tertain doubts about this — 7. C. buxifolius : this differs from 

 the cultivated plant by the inflorescence much shorter than 

 the leaves, and is perhaps rather C. patens, E. and Z., stamens 

 five; ovary seated on a crenulated fleshy disk, 9-furrowed, 

 globose-ovate, 3-cel!ed, 6-ovuled ; ovules collateral ; style 



