352 



is, nor does it seriously impair the value of Rheede's figure, which 

 has been cited bv Hooker (Flor. Brit. Ind. vol. vi. p. 289) under 

 D. daemona, Roxb. The less serious inadvertence of adding 

 the aerial tubers of D. bnlbifera to the figures of the two sexes 

 of another Dioscorea cannot prevent the citation of these figures 

 under D. Wallichii, Hook, f., the species which, notwithstanding 

 this error, they manifestly represent.* Since, however, one of 

 these two figures is the basis of D. aculeata, Linn., published in 

 1753, it follows that the name D. WaUichii, which was not pub- 

 lished until 1892, must be subordinated to D. aculeata. For the 

 moment, however, our interest is confined to the circumstance, 

 already made clear by Hooker (Flor. Brit. Ind. vol. vi. p. 296) 

 that neither Katu-katsjil nor Kattu-kalenyu can be included in 

 the aggregate T). sativa, Linn., as built up during the period 



17*7-67. 



In 1842 Grisebach described as D. sativa, Linn. {Flor. Bras. 

 vol. iii. p. 43), a plant from Brazil which also differs from any 

 of the elements included by Linnaeus in this species. The plant 

 in question had been figured by Velloso in 1827 as D. heptaneura 

 (Flor. Flu m. vol. x. t. 124). Grisebach, unlike any of his 

 predecessors from Willdenow onwards, was guided in his action 

 by Lamarck; he excluded from D. sativa, Linn., everything save 



[i] 



ft 



* It may appear singular that an inadvertence so trivial as the addition to 

 the figures ot* Kattu-Kelengu and Katu-Kat&jil of the aerial bulbils charac- 

 teristic of D. bulb if era should have obscured the identity of these plants 

 longer than the more serious error of adding the fruits of a Clematid to the 

 stem of a Dioscorea concealed the identity of Podava-Kelengu. Linnaeus be- 

 came acquainted with that species through Rumph's figure of Ubium sylvestre 

 (Herb. Amboyn. vol. v. t. 128) which he misidentified in 1764 (Amoen. Acad. 

 vol. iv. p. 131) with Rheede's T?jageri-Nuren, his own D. triphiflla. But in 

 none of his w r orks did Linnaeus account for Podava-Kelengu. Barman 

 (Hort. Malab. index) made no attempt to identify Podava-Kelengu in 1768. 

 Roxburgh, who had received a plant from the Moluccas through 0. Smith 

 about the close of the eighteenth century, identified it with the Ubium 

 sylvestre of Rumph, but failed to recognise its identity with Podava-Kelengu. 

 In 1789 Lamarck (Encyc. Meth. vol. iii. p. 347) mistaking the Naravelia fruits 

 for flowers, conjectured that it might be a Trichosanthes. Buchanan 

 (afterwards Hamilton) who seems to have become acquainted with the plant 

 itself during his Mysore journey in 1802-3, evidently without finding its 

 fruit, thought it might be a Smilax. The plant is poisonous, hence 

 Hamilton's names S. virosa and 8. narcotica, first made public in 1832. In 

 1818, however, Dennstedt (Schl. Hort Malab. pp. 15, 20, 33) recognised that 

 Podava-Kelengu is a Dioscorea and named it D. hispida, Demist. As this 

 author has been careful at p. 33 — the passage usually cited — to associate 

 this name only with the first plate (Hort. Malab. vol. vii. t. 51) where no 

 Naravelia fruits are shown, and as these fruits are not referred to in 

 Rheede's text, D. hispida has to be regarded as the earliest unimpeachable 

 one for the species, taking precedence over the name D. hirsuta employed by 

 Blume in 1827 (Enum. PI. Jav. p. 21) for Ubium sylvestre, Rumph. But 

 Blume, like Dennstedt, recognised that Podava-Kelengu is a Dioscorea, which 

 he named D. hirsuta iriflorescentia monstruosa. This failure to recognise the 

 true explanation of Rheede's second figure (Hort. Malab. vol. vii. t. 52) does 

 not detract from the merit of Blume as the first author to recognise that 

 Ubium sylvestre and Podava-Kelengu are conspecific. Since then no one save 

 Kunth in 185u {Enum. vol. v. p. 438, footnote) has expressed any doubt on 

 the subject. In 1841, by some inadvertence, Steudel (Nomencl. ed. 2, vol. i. 

 p. 511) reduced D. hispida, Dennst., to D. cylindrica, Burra. (Flor. Ind. 215 

 [315]), which is in reality Ipomoea Batatas. This error has somehow crept 

 into the Index Kewensis. 



