348 





are simple but crowded, renders the identification with Mu-kelengu 

 impossible; in Rheede's plant the spikes in both sexes are always 

 solitary- When we turn to the only other known Java species 

 with transverse veins such as Bluine attributes to the leaves of 

 his plant, the difficulty disappears, for in that plant, which is 

 D. bulbifera, Linn, [lj, the leaves are always glabrous, the stem 

 is unarmed, and it is moreover the only Java species in which 

 we find a 'cluster* of simple female axillary spikes. The 

 greatest difficulties connected with Blume's account of D. sativa 

 are his statement that the male spikes are not only fascicled but 

 paniculate, and the fact that he has described 1). bulbifera as a 

 species apart. So far as the latter difficulty is concerned it 

 has, however, to be borne in mind that Roxburgh, an equally 

 competent observer, felt justified in describing as distinct species, 

 D. pulchella and D. ctispata, what we now know to be only 

 marked and readily distinguishable forms of D. bulbifera, Linn. 

 [1]. As regards the former difficulty it lias to be remembered that 

 towards the ends of the flowering twigs of D. bulbifera the leaves 

 in whose axils the fascicles of simple male-spikes arise, often are 

 much reduced and sometimes are obsolete. When this occurs the 

 postulated paniculate appearance is produced, and this condition 



is so frequent that Roxburgh, in his account of J), crispata (Flor. 

 Ind. vol. iii. p. 802), describes the male flowers as panicled. 

 Unfortunately there is no Dioscorea in the Leiden collection 

 which has been written up by Blume himself as D. sativa. This 

 defect is, however, remedied by the fact that in the Buiten/org 

 herbarium there is a specimen " ex herb. Blume " which is 

 D. bulbifera, Linn., but which has been written up as D. sativa. 

 It appears therefore as nearly certain as may be that the plant 

 which Blume believed to be identical with the Dioscorea figured 

 by Rheede as the Mu-kelengu was in reality D. bulbifera, Linn. 

 1 1], and that whatever his intention may have been Blume, 

 without being cognisant of the fact, in 1827 used the name 

 D. sativa as Thunberg had used it in 1784. 



In connection with this question of D. sativa as understood by 

 Blume, we may remark that in 18-39 Dillwyn (Rev. Ref. Hort. 

 Malab. p. 40) regarded Mu-kelengu as being D. sativa^ Linn., 

 and that in 1844 Hasskarl [Cat. alt. Hort. Bogor. p. 34) followed 

 Blume so far at least as ' form ' is concerned, though we have no 

 means of judging whether D. sativa, Hassk. (1844), and D. sativa, 

 Bl. (1827), he really the same plant. We do, however, know 



P 



Hass* 



irom its diagnosis that 1). sativa, var. ft 

 plant as D. sativa, var. ft, Bl. : 

 obvious that his D. sativa, var. 



the Mu-lelenqu of Eheede. ^ r ___ y 



karl (Hart. Malab. Clav. p. 67) restricted the name 22. sativa to 



jln-helenqu. 



In 1827 Presl applied the name D. sativa (Rel. Ha ml. fasc. 2. 

 p. 134) to a Dioscorea collected bv Haenke in the mountains of 

 Huanaco in Peru. From Presl's 'citations we find that he now 

 definitely excluded from D. sativa, Linn., the Ambovna plant of 

 Rumph, D. sativa, Linn. [6]. Tn thus following the advice 

 given by Meyer in 1818, Presl went a step further and expressed 

 the opinion that Rumph's plant is not a Dioscorea at all. 



