344 



neglected. The specimen which Linnaeus lias written up as 

 D. sativa with his own hand is a male example of the North 

 American species named by himself D. villosa (Sjj. PI. ed. 1. 

 p. 1033. n. 7). It differs from the Amboyna, the Malabar, the 

 Ceylon, and both of the West Indian constituents of I), sativa, 

 Linn. ; it agrees as regards its foliage with the plant figured by 

 Ehret [Hort. Cliff, t. 28) which professes to be D. sativa, Linn. 

 [1], and there is nothing in its appearance or characters to 

 separate it from the plant described by Linnaeus in the Hort us 

 Cliff ortianus in 1737 as IK foliis cordatis alter nis, caule laevi. 



In 1768 the name D. sativa, Linn., was used by Miller (Gard. 

 Diet. ed. 8. No. 1) for a West Indian plant, seeds of which had 

 been sent to England by Houston before 1733. Iu the field 

 Houston had identified this with Dioscorea scandens, foliis tamni, 

 fructu racemose*. The notes of Houston and the description 

 supplied by Miller indicate that the plant intended is that 

 figured by Plumier. D. sativa, Mill., is therefore certainly = D. 

 sativa, Linn. [4], so far as Miller's description is concerned. 

 But in Miller's Herbarium as in that of Linnaeus there are two 

 specimens of D. sativa; in this case both have been written up by 

 Miller himself. One of the two, evidently taken from a plant 

 grown in the open is, as in the ease of the Linnean collection, 

 D. villosa, Linn.; the other, equally clearly derived from a stuve- 

 plant. is D. sativa, Linn. [4], the species whose seeds, Miller 

 explains (Gard. Diet. I.e.), had reached him through Houston 

 from Jamaica. In the Cambridge herbarium there is another 

 specimen of the same stove-phi nt, given by Miller to Martyn. 



In 1784 the name D. sativa, Linn., was used by Thunberg (Flor. 

 Japan, p. 151) for a cultivated Japanese Yam, termed ' kasjuo 

 kadsura.' This plant differs from all the elements included 

 by Linnaeus under D. sativa and belongs instead to one of the 

 two species included by Linnaeus under D. bulbifera, Linn. (Sp. 

 PI. ed. 1. p. 1033. n. 5). In Tlumhcrg's own* proper Herbarium, 

 however, we find that he has also written up as D. sativa a speci- 

 men which is really referable to his own species D. japonica. 



In 1788 the name D. sativa, Linn., was used by Gaertner (Fruct. 

 p. 66) in such a manner as to include, so far as citations and 

 references are concerned, all of the plants included by Linnaeus 

 under D. sativa except D. sativa, Linn. [5"], which is the West 

 Indian Dioscorea taken up from Sloane. The description, how- 

 ever, is confined exclusively to the plant which is the subject of 

 Gaertner's plate (Fruct. t. 17. fig. 4). That plate is a somewhat 

 improved version of the one used by Plumier in 1703 (Gen. t. 26) 

 to illustrate his D. scandens, foliis tamni, fructu racemoso, so 

 that, strictly interpreted, T). sativa, Gaertn. (1788) is = #. sativa, 

 Mill., as described in 1768, and = D. sativa, Linn. [4]. 



In 1789 the incidence of the name D. sativa, Linn, was critically 

 discussed by Lamarck (Encjp . Meth. vol. iii. p. 232). He arrived 

 at three definite conclusions: — 



1, that Ehret's figure, published in 1737 and acknowledged 

 for the first time by Linnaeus in 1753, must be regarded as the 

 real basis of the name D. sativa, Linn.; 



2. that the plant represented in Ehret's figure differs speci- 

 fically from all the other elements included by Linnaeus under 



