Kanahia.'] lxxxv. ASCLEPiADEiE (brown). 207 



vii. o7. Asclepias laniflora, Forsk. Fl. ^-Egypt-Arab. 01 ; Vahl, Symb. 

 i. 28, t. 7 ; Deiile, Cent. PI. Voy. Meroe, 49, t. 3, fig. 3. 



Wile Iiand. Eritrea: (iiiula Valley, 3100 it., Schweinfurth Sf i?ua, 2148 ! 

 Ariot Valley, 4400 ft., Schiceinfitrth Sf Riva, 677 ! bed of the River Lava, at 

 Arbaseiko, P^'Hcir/; wcav Kevcw, Penzig. Galabat: re<;ion of Mataiinna, Schwein- 

 fiD'th, 216! Sennar, Kotschy, 177! Blue Nile, Muriel, S, 14! Kotschy, 4721 

 475! Fazokl ; Mount Akaro, Cailliaud {c\ Deiile); near Kliartooni, 5r/iM?et«/«r///, 

 856 ! Upper Nile, Freeman Sf Lucas, 86 ! Abyssinia : near Adowa, Sfeudner, 743 ! 

 Schimper, 120 ! 939 ! by streams near Jelajeranne, Schimper, 1909 ! Tacazze 

 Valley, Steudner, 742 (ex Engler) ; Somak Efat, Roth, 122 (413) ! West shore of 

 Lake Zouai (Zwai), Wellby ! Somaliland : 'by the Dagatto Torrent, Ghika (i-x 

 Schweinfurth) ; near Sagado, Riva, 962 (ex Schumann). 



XMCozaxnb. Dist. German East Africa : west of the Pare Kaugc, 600-2300 ft 

 Rohnel (ex Schiceinfurth). 



Also in Arabia. 



According to Decaisne this is not the true Asclepias lanijlora of Forskal, but on 

 this point he is certainly mistaken, as there is a specimen, received from Korskal, 

 in the British Museum Herbarium, which I have examined and found to be identical 

 with the plant here described. It was upon tliis specimen that Robert Brown 

 founded the genu? Kanahia, although his description of the coronal structure is not 

 very exact. 1 have not seen any exan)ple from Africa of the plant which Decaisni- 

 has described under the name of Kanahia Forskalii, which has shortly bitid coronal- 

 lobes, something like those of K. consimilis, N. E. Br., but not so deei)ly bitid, and 

 the pollen-masses are quite different. 



The solid structure of tlie coronal-lobes of K. lanijlora is evidently only a varia- 

 tion of that which usually obtains in Asclepias, and is formed by the soldering 

 together of the intlexed sides of an ordinary complicate coronal-lobe ; indeed, this is 

 completely demonstrated in a specimen collected in Eritrea at Ginda (Schweinfurfh 

 Sf Rim, 2148), in which the top of the coronal-lobes is deeply furrowed all along, 

 forming an intermediate stage between the entire coronal-lobes of typic.il K. lam- 

 flora and the deeply cleft lobes of K. glaherrima. 'J'he pollen-masses, however, are 

 short as in K. lanijlora. 



2. K. glaberrima, N. E. Br. Stems 4-0 ft. or more high, simple 

 or branched, glabrous, except some minute bristles in the axils of the 

 leaves and along the rudimentary stipular line. Leaves ascending; 

 petiole 1-4 lin. long; blade 2\-Q in. long, 2-7 lin. broad, broadly linear 

 or linear-lanceolate, tapering to an acute point, acutely narrowed into 

 the petiole, slightly thickened along the margins, ({uite glabrous. 

 Flowers in pedunculate several -flowered bracteate umbel-like racemes 

 (the axis of the umbel shortly elongating), lateral between the bases of 

 the leaves; peduncles U-2J in. long, glabrous ; outer bracts 1-0 lin. 

 long, the succeeding ones gradually smaller, deciduous, glabrous ; pedicels 

 G-12 lin. long, glabrous. Sepals JJr-4 lin. long, varying from ovate- 

 lanceolate to linear-lanceolate. Corolla deeply r)-lobed, somewhat I'^dex^ 

 (or, in the dried state, often more or less campaiiulate), whitish ; lobes 

 4-5 lin. long, bordered inside with woolly hairs, glabrous on the 

 back. Coronal-lobes arising J- 1 lin. above the base of the staminal- 

 column, with the tips reaching to its summit, Ij-U lin. long 

 and about 1 lin. broad, complicate, obliquely deltoid-ovate viewed 

 sideways, with the dorsal side much the shorter, the inflexed 

 sides produced into acute or linear-acuminate teeth sometimes in- 



