DESCRIPTION:--An arborescent branched shrub. _Leaves_ up to 0·5 m. long, about 4 cm. broad at the base, gradually tapering to the apex, almost flat on the upper surface, convex on the lower surface, with ovate somewhat incurved teeth 1 cm. apart below, about 1·3 cm. apart above, glabrous. _Inflorescence_ solitary or 2-3 from each rosette of leaves, unbranched. _Peduncle_ 7 mm. in diameter, cylindric with scattered membranous bracts. _Floral-bracts_ membranous, 1·3 cm. long, 9 mm. broad, oblong, with straight margins, distinctly veined. _Pedicels_ 2 cm. long, elongating in the old flowers. _Raceme_ dense, up to 22 cm. long. _Young flowers_ erect-spreading, cylindric, with a subacute apex; old flowers pendulous. _Perianth_ (in open flowers) 3·2 cm. long; outer segments reddish, 5 mm. broad, obtuse; inner segments white, 9 mm. broad, with a distinct mid-rib, reddish below, greenish above, obtuse. _Stamens_ projecting. _Ovary_ 8 mm. long, cylindric; style 3·2 cm. long, cylindric; stigma minute. (National Herb., Pretoria, No. 2904.) PLATE 187.--Fig. 1, young and adult flowers; Fig. 2, median longitudinal section of a flower; Fig. 3, pistil; Fig. 4, bract. F.P.S.A., 1925. [Illustration: 188.] PLATE 188. DERMATOBOTRYS SAUNDERSII. _Cape Province, Natal._ SCROPHULARIACEAE. DERMATOBOTRYS, _Bolus in Hook. Ic._ Pl. 1940. * * * * * =Dermatobotrys Saundersii=, _Bolus in Hook. Ic._ Pl. Zwilt’s 1940; _Fl. Cap._ vol. iv. sect. ii. p. 206. * * * * * This interesting member of the _Scrophulariaceae_ is peculiar in that it grows as an epiphyte on other trees. Miss Pegler records it from the Kentani District as growing on the Cape Chestnut (_Calodendron capense_). It ranges from Kentani to Zululand. It differs also from most members of the family in having 5 well-developed stamens, and for this reason was first referred to the family _Solanaceae_, but the straight or almost straight embryo indicates an affinity with the family _Scrophulariaceae_. Mr. Saunders, who first collected the plant, described it as a parasite which killed the host on which it grew, but the late Dr. Medley Wood states that the plant has a tendency to fix itself to trees already dead. We are indebted to Mr. K. B. Jameson, Scottsville, Maritzburg, for the specimen from which our illustration was prepared.