=Roodia=, _N. E. Brown_ (genus novum). _Herba_ succulenta, nana, caespitosa. _Proles_ bifoliatae. _Pedunculus_ bracteatus, uniflorus. _Calyx_ in tubum distinctum supra ovarium productus 6-lobus. _Petala_ numerosa, libera. _Stamina_ numerosissima, pluriseriata, omnia a basi in tubum abrupte inflexa, exterioribus brevioribus. _Ovarium_ multiloculare. _Stigma_ parvum, inconspicuum, sessile, integrum. _Fructus_ multilocularis; loculi alis binis subchartaceis tecti et dorsum versus tuberculo clausi. =Roodia digitifolia=, _N. E. Brown_. _Planta_ 8-11 cm. alta, caespitosa, glabra. _Proles_ bifoliatae, confertae. _Folia_ erecta, 8-11 cm. longa, 10-15 mm. crassa, digitiformia, subteretia faciebus interioribus planis, apice obtusissima, inferne viridia, superne purpureo-tincta. _Pedunculus_ 3-6 cm. longus, apice 4-5 mm. crassus, erectus, uniflorus, infra medium bibracteatus, purpureus. _Calycis tubus_ cum ovario 10 mm. longus, 10-11 mm. diametro; lobi inaequales, 5-8 mm. longi, 5-7 mm. lati, ovati, obtusi vel subacuti. _Corolla_ circiter 4 cm. diametro; petala numerosa, subquadriseriata, exteriora circiter 15 mm. longa et 1 mm. lata, interiora breviora, linearia, apice obtusa vel dentata, pulchre purpurea. _Stamina_ numerosissima, in tubum calycis abrupte et arcte inflexa. _Ovarium_ supra concavum. _Stigma_ sessile, parvum, inconspicuum. _Fructus_ 12-14-locularis. Cape Province: Van Rhynsdorp Division, near Van Rhynsdorp (_Mrs. E. Rood_). Growing at the Division of Botany, Pretoria (Garden No. 403). * * * * * This interesting plant is so like some species of _Mesembryanthemum_ in general appearance, that most would unhesitatingly place it in that genus. For a long time past, however, as I have elsewhere stated, it has been dawning upon me that _Mesembryanthemum_, as at present understood, is rather of the nature of a natural order than of a single genus. In vegetative characters it presents a very great amount of variation, and in most cases any particular kind of variation is found to be common to several species, so as to form of them a group, indicating a generic difference from the others. About a hundred years ago Haworth recognised this and gave generic names to some of the groups he had formed, which have neither been accepted nor taken notice of by subsequent authors.