CONTRIBUTIONS TO SOUTH AFRICAN ASOLEPIADOLOGY. 357 



purple; the corona, which is rather membranous for the genus* 

 is white. The pollinia are a Httle compressed, but not so much 

 as is usually the case in Gomphocarpus. The umbels appear to be 

 generally three-flowered. I have to thank Mr. J. R. Sim for the 

 only specimen I have seen of this species. 



37. Asclepias pseudo-crispa Scliltr., n. sp. Herba erecta, 

 scabra, e basi ramosa, 15-20 cm. alta ; ramis subteretibus bifariam 

 puberulis satis dense foliatis ; foliis lineari-oblongis linearibusve 

 obtusiusculis, supra sparsim pilosulis, subtus glabrescentibus, mar 

 ginibus interdum reflexis, 2-3-5 cm. longis, medio 0"4-0-6 cm 

 longis ; floribus in umbellis terminalibus multifloris erectis, pe 

 dunculo bifariam piloso, tereti, folia paulo excedente, pedicellis 

 gracilibus filiformibus pilosis, subfequilongis (1-1-5 cm. longis 

 calycis segmentis lanceolatis acutis pilosis, quam corolla subduplo 

 brevioribus ; coroll^e lobis reflexis ovatis obtusis, margine tenuiter 

 ciUatis ; corona foliolis gyuostegium paulo excedentibus breviter 

 unguiculatis, lamina cucullata obtusa, margine utrinque medio in 

 dentem linearem porrectum productis ; antheris oblongis, mar- 

 ginibus cartilagineis alseformibus, loculos sequantibus, appendice 

 hyalino traasverso apice inflexo emarginato, marginibus loculorum 

 alte excisis ; polliniis oblique pyriformibus paulo compressis, basi 

 attenuatis, caudiculis divaricatis brevissimis, glandula oblonga basi 

 afifixis. 



In collibus prope King Williamstown (Kaffrari*), alt. 1400 ped., 

 Dec. 1898, J. R. Sim, No. 1633. Loco incerto, probabiliter in 

 Transvaalia, Mrs. Saunders (J. M. Wood, No. 5380 in Herb. Schltr.). 



One of the many allies of G. crispus R. Br., from which it is 

 easily distinguished by the narrow straight leaves and conspicuously 

 smaller flowers, besides several other differences mentioned in the 

 diagnosis. The flowers seem to be brownish. The locahty of 

 Mrs. Saunders's specimen, which Mr. J. M. Wood kindly forwarded 

 to me, is unknown ; Mr. Wood suggests that it may have been 

 collected in the Transvaal, but this is as yet doubtful. 



38. Asclepias kigida Schltr. (Gomjjhocarjms rigidus Dene.). 

 Herba valida, erecta, simplex ; caule subtereti scabro dense foliato ; 

 foliis erecto-pateutibus ovatis vel ovato-ellipticis undulatis, apice 

 subacutis scabre pilosulis, prresertim costa margineque, 3-6 cm. 

 longis, supra basin 2-3-5 cm. latis, pedunculo brevissimo ; floribus 

 in umbellis extra-axillaribus alternantibus subquadrifloris, folia baud 

 fequantibus, pedunculo pedicellos aequantibus bifariam puberulo, 

 pedicellis pilosis, 0-7 cm. longis ; calycis segmentis lanceolatis 

 acutis pilosis, vix 1 cm. longis, infra medium 0-4 cm. latis; 

 corollse subcampanulatte lobis ovatis obtusis glabris, marginibus 

 apice s^pius subreflexis, calycis segmenta paulo excedentibus ; 

 coronse foliolis depressis, e basi lineari-cuneata in laminam con- 

 cavam rotundatam obscure 5-lobatam dilatatis; antherarum ob- 

 longarum marginibus cartilagineis basi oblique truncatis, loculis 

 subfequilongis, appendice hyalino ovato obtusissimo, marginibus 

 loculorum profundms rotundato-emarginatis; polliniis more generis 

 valde compressis oblique oblongis basin versus attenuatis, caudiculis 

 patulis paulo compressis glandula oblonga obtusa supra basin adnatis. 



