72 MEMOIRS OF THE NEW YORK BOTANICAL GARDEN [Vol. 9. No. 1 



glandularum filiformium 0.75 mm. longarum apice curvato-incrassatarum praedita; 

 petala 5, 1 mm. longa, plana, infra medium in unguem 0.5 mm. latum abrupte 

 augustata, supra medium ovato-triangularia et 1.5 mm. lata, subintegra, quoque 

 basi glandula unica subquadrata 0.3 mm. alta et lata praedito; stamina 5, inferne 

 in columellam tenuem 1.8 mm. altam connata; apice columellae filamenta libera 

 5, horizontaliter radiatim patentia, 0.75 mm. longa; antherae 0.5 mm. longae, 

 siccitate albidae; ovarium rudimentarium in apicem columellae insidens, 0.75 mm. 

 altum, 0.5 mm. latum, canalem fumarium simulans. Flores foeminei nondum visi. 

 Fructus pedicellis 4.5-5 mm. longis suffultis. Perianthium fructiferum rigidum, 

 incrassatum; sepala 5, obovato-elliptica, apice rotundata, 2.5 mm. longa, 1.5 mm. 

 lata, basi glandulis binis praedita; petala 5, sepalis subaequalia, obovata, basim 

 versus magis augustata, basi ut videtur eglandulosa. Capsula subglobosa, 4-5 

 mm. diametro, matura brunnea, glabra, siccitate rugulosa, in tres valvas acute 

 bifidas findens; pericarpium 0.9-1 mm. crassum, lignosum. Semina plus minusve 

 ellipsoidea, 3-3.2 mm. longa, 2.3-2.6 mm. lata, testa nigra nitida, sub lente 

 valido subtilissime punctata; caruncula subquadrata 1.5 mm. lata, siccitate 

 aurantiaca. 



Mlanje District: Mlanje Mountain; west slope, in V accinium savannah, shrub 

 1.5 m. high, flowers green, 1830 m., June 21, 1946, 16398; ibid., frequent in 

 grassland, usually among rocks, shrub about 1 m. high, dioecious, leaves glau- 

 cous, flowers cream, 1850 m., July 18, 1946, 16866 (TYPUS in Herb. Kew.); ibid., 

 ^ plant of 16866, flowers cream, fruits brown, 1850 m., July 18, 1946, 16861, 



This is a fine discovery by Mr. Brass of a new species of unusual distinction 

 that seems altogether to have escaped those who have visited Mlanje Mountain 

 in the past. 



The arrangement of the glands within the cf flower is that of § P auciglandw 

 losae Pax & K. Hoffm., and here the only species to which C. brassii is at all 

 akin is C. whytei Hutch. But C. brassii is most distinct in its cordate-based 

 sessile leaves, few-flowered inflorescences (or the % flowers solitary), the quite 

 glabrous pedicels, the capsules without pale tubercles outside etc. 



With the naked eye there seems nothing unusual about the stem. Under a x 10 

 lens the surface of the stem is however seen to be covered with low tubercles, 

 elongate longitudinally, pointed at each end, and often confluent into irregular 

 ridges. Closer inspection will reveal that each tubercle has, approximately in its 

 middle, a tiny impressed dot. Lenticels are the obvious guess, but these tubercles 

 occur on the very youngest green stems where lenticels would presumably be 

 rather superfluous. The tuberculate surface of the stem persists until the latter 

 has attained a diameter of at least 4 mm. (I have not seen older stems); the cork 

 cambium is thus presumably late in starting. The appearance of the impressed 

 dots on the tubercles recalls that of the gland on the back of cypress leaves. 

 This odd stem structure is constant and characteristic in C. brassii, and quite 

 absent in C. whytei. 



I have sent some leaves and portions of stems of C. brassii to Dr. C. R. 

 Metcalfe, Keeper of the Jddrell Laboratory, where he very kindly had them ana- 

 tomically examined by Dr. M. Y. Stant. 



Cavities occur in the cortex of the stem beneath the ridges; these spaces are 

 most probably the remains of secretory glands, although there was no trace of 

 the actual secretion. This, however, may be seasonal, as with the glands of 

 Osmanthus. Sections showed evidence that a definite layer of cells was origi- 

 nally present as a lining to each cavity, although many of the constituent cells 

 of this layer had become disorganised when the specimen was dried. 



