346 



satica, Linn. [3], and the other is D. sativa, Linn. [4]; Poiret 

 was quite entitled to accept, if he chose, the considered 

 judgment of Willdenow, expressed in 1806, that the two plants 

 concerned ought both to be included in T). sativa. But Poiret' s 

 preliminary statement, which is that this plate (III. t. 818) 



n 



ft 



[11, and is in opposition 



to Lamarck's explicit intimation that the two species there 

 depicted, which are D. sativa, Linn. [3] and D. sativa, Linn. 

 [4], should be excluded. Poiret's action, the responsibility for 

 which lias at times been attributed to Lamarck himself (conf. 

 Flor. Brit. Ind. vol. vi. p. 296), is so unwarranted that it can 

 only be explained as the result of inadvertence. 



lu 1796 the name J), sativa was applied by Eodschied (Bemerk. 

 Essequeb. p. 57) to a Dioscorea found by him in Dutch Guiana, 

 His specimens, from the words of Arowabisch, in which the plant 

 is wild, were studied by G. W. F. Meyer in 1818, and were found 

 to represent a species different from any of the elements included 

 by Linnaeus under D. sativa. This species was described as 

 D. lutea, Meyer [Prim. Fl.Esseq. p. 282), and when characterising 

 it Meyer took occasion to endorse the view expressed by Lamarck 

 in 1789, insisting in particular that the Amboyna plant of Rum- 



species intended by Linnaeus. 



["61, must 



In 1806, Willdenow in re-editing the Species Plantarum made 

 very little modification in the conception of D. sativa enunciated 

 by Linnaeus. The West Indian species taken up from Sloane, 

 D. sativa, Linn. [51, which Gaertner had not cited in 1788, was 

 again included and the onlv Linnean element left out is the 

 oilier "West Indian species taken up from Plumier, D. sativa, 

 Linn. [4]. With regard to this particular plant the argument 

 advanced by Lamarck in 1 789 could not be disregarded ; it was 

 accordingly transferred by Willdenow (Sp. PI. ed. Willd. vol. iv. 



95) to D. piperif 



There are, how- 



ever, indications that Willdenow must have felt the difficulty 

 with regard to the identity of D. sativa that had been experienced 

 by Lamarck, by Miller and by Linnaeus himself. The sheet in 

 Willdenow* s own herbarium which has been written up as D. 

 sativa is stated both bv Grisebach (Mart. Flor. Bras. vol. iii. 

 p. 43) and by Kunth (Enuw. vol. v. p. 340) to bear the number 

 18422. According to Grisebach the specimen so numbered is the 

 Bast Indian D. deltoidea, Wall. (Cat. Lith. n. 5110); according 

 to Kunth this specimen is a male example, collected by Hunne- 

 mann, of that form of the North American D. villosa, Linn., 

 which was described bv Walter in 1788 as J), quaternata (Flor. 

 Carol, p. 246). Unless there be two sheets in Willdenow's 



* This involved the introduction of a new confusion owing to the 

 circumstance, of which Willdenow does not appear to have been aware, that 

 Plnmier's plant, D. sativa, Linn. [4], twines to the left, has only three fertile 

 stamens, and has seeds which are winged all round, whereas D. piperifoUa, 

 Hurnb. and Bonpl., twines to the right, has six fertile stamens and has seeds 

 which are winged only at the top. In 180o, however, the taxonomic value 

 of these morphological differences were not so fully appreciated as they are 

 now. 



