NEW AFRICAN PLANTS. 181 at the mouth) ; the short invohicre-segments are subquadrate in shape and fimbriate at the top, the glands larger and more fleshy than in the bisexual cups ; in both cases the glands bear on the inside a shallow transverse keel. The filaments and pedicels are glabrous ; the fimbriate scales equal the latter in length. The lateral cups (1 line long) bear each at the base a pair of smalb adnate bracts ; the short involucral segments are broader than in the male cup, and rounded with a slightly fimbriate margin ; the pedicel of the female flower is ^ line long. Near E. triacantha Ehrenb. apud Boissier, but is at once dis- tinguished by the four-thorned pulvini. Kniphofia insignis, sp. nov. Foliis pluribus bipedalibus herbaceis linearibus superne sensim angustatis, nervo mediano plicatis, margine minute scabridula, racemum speciosum late cylin- dricum baud attingentibus ; bracteis albidis acuminatis pedicellos perbreves plus triplo excedentibus ; perianthio tubuliforme saepe curvato super ovarium constricto, dentibus brevibus ovatis ; sta- minibus semper inclusis, stylo exserto. Hab. Sheik-Mahomet, Nov. 13th, 1894, Donaldson Smith. The fleshy stalk is 2 ft. 4 in. long below the raceme, with a diameter at the base of about ^ in. The long narrow thin leaves are 2 ft. in length, and about 6-7 lines broad at the base. The raceme, which becomes dense-flowered above, is 10 in. long. The pointed scarious bracts are 3-4 lines long, the flower-pedicels 1 line. The perianth-tube is 15 lines long, the segments 1^ lines ; the diameter of the tube in the dry state is 3 lines. The stamens are about three-fourths the length of the perianth ; the style becomes exserted nearly 2 lines beyond the mouth. Is near K. Schimperi Baker, which it closely resembles in leaf- characters, but is distinguished by its much larger flowers and denser raceme. It may be worth calling attention to the fact that Kniphofia elegans Engl, in Hochgebirgsflora, p. 162, is a synonym of K. Schimperi Baker, published in this Journal nearly twenty years before (Journ. Bot. 1874, p. 4), and based on Schimper's Abyssinian plant No. 1200 ("1863-8"), which Engler also quotes as authori- tative for his own species. By a strange coincidence K. Schimperi Baker is omitted from the list of Kniphofias in Schinz & Durand's Conspectus, though the other species described by Baker in the same paper are included; Schimper's No. 1200 is referred to K. elegans Engl. As a set-off Mr. Baker is credited with a species K. unifolia "in Bntt. Journ. of Bot. (1885), p. 298." No such species occurs on the page referred to, nor have I been able to discover it elsewhere. Is K. ensifolia Baker intended? This is described on the page in question, and included in the Conspectus in its proper place, with a correct reference. Albuca Donaldsoni, sp. nov. Glabra, bulbo . . . . , foliis basi membranaceis lanceolatis, superne lineari-acuminatis, siccis involutis et reflexis ; scapo valido cum racemo denso ovato, folia excedente, bracteis longissimis supra basin triangularem fili- K %
