361 



vol. iii. p. 138), by Presl in 1827 (Bel Haenk. fase, 2. p. 134), and 



by Grisebach in 1842 (Flor. Bras. vol. iii. p. 27). So far these 



J references had regard only to Plunder's illustration. Ivunth's 



treatment of this plate under D. sativa has already been adverted 

 to; in another place (Enum. vol. v. p. 330) he did expressly 

 exclude Plunder's plant from D. piperi folia, but he did not even 

 there definitely identify it with a known species. It is to Grisebach 

 that we owe the reference of the plant figured by Phimier to a 

 properly characterised species, though his treatment still leaves 

 something to be desired. In 1864 (Flor. Brit. West Ind. p. 587 

 he identified it with a Martinique plant which Presl had named 

 D. multi-flora in 1844 (Bemerk. p. 116), overlooking when he did 

 so that he himself had alreadv published a different J), mult i flora, 

 Mart., in 1842 (Flor. Bras. vol. iii. p. 35)/ 



The bibliographical history of this Martinique plant has been 

 rather chequered. The plant was collected by Kohnut on behalf 

 of Sieber and was issued by the latter, at first without a specific 

 name (Herb. Mart. n. 27), in 1821. A specimen which reached 

 Sprengel was described by that author in 1822 as D. martini - 

 censis (Neue Entdeck. vol. iii. p. 17). t Shortly afterwards 

 Sieber himself issued a list in which it was identified with D. 

 altissima, Lamk (Herb. Mart. Sup pi. n. 27). In 1825 Sprengel 

 accepted this determination and treated his T). martinicensis as 

 identical with J), altissima, Lamk (Syst. vol. ii. p. 153). Presl 

 in 1844 recognised the error, but unfortunately overlooked the 

 existence of the name D. martinicensis , Spreng. (1822) for the 



species he was describing and of the name D. multiflora, Mart. 



(1842) for another species. Kunth in 1848 while overlooking the 

 former inadvertence detected the latter; to remedy it he suggested 

 that the Martinique plant be termed D. Sieberi (Act. Acad. Berol. 



* The difficulty caused by the existence of the name D. midtiflora in 

 connection with two distinct species was pointed out in 1916 by Hauman 

 (Anal. Mils. JSfac. Hist. Nat. Buenos Aires, vol. xxvii. p. 458, footnote). 



f That the Martinique plant of Phimier is the same as the Martinique 

 plant of Sieber is evident from a comparison of Kohaut's specimens with 

 Plumier's plate and is confirmed by the comparison of Plumier's original 

 description, here reproduced, with those of Sprengel and Presl and Kurith. 

 Plumier's account is as follows : — 



Dioscorea scandens foliis tamni fructu racemoso Phimier MSS. torn. ii. 

 fig. 144. 



Huiusce planta radix radicis Bryoniae vulgaris in modum carnosa 

 non napiformis quidem sed in raraos et in appendices distenta, 

 tenera Candida multoque sapore conspicuo donata. Cauliculum 

 unicum producit tenuissiraum sed lentum alteque supra vicinas 

 arbores scandentem et ramos plurimos emittentem etiam loneris- 

 simos et tenuissimos quibus inhaerent ampla folia cordata tamni 

 vulgaris foliis perquam conformia sed ampliora, nervosiora, obscuriiis 

 aliquantulum virentia et amariuscula longis anneetuutur pediculi- 

 longus appendet Horibus onustus exiguis candidis monopetalis 

 eampaniformibus pat^ntibus et in sex segmenta divisis, uc senideni 

 tribus staminulis bicipitibus instructis. Ex calyce autem perquam 

 exiguo surgit pistillum quod infimam floris partem perfodit abitque 

 deinde in fructum trigonum et in tria loculamenta multum compressa 

 et fere semicircularia divisum, seminibus foeta planis orbicularis et 



marginatis et amarissimis. 

 Mulcoties plantam reperi variis in loci? insulae martinicanae potissimum 

 in ilia regione quae vulgo le Mont Hoel dicitur ad partes arcis Divo 

 Petro dicatae. 



