180 CHINA. [Leguminosm. 



considerably swelled, in which we have only been able to trace one ovulum suspended from a funiculus 

 that rises from the very base of the cell. If there be really only one ovulum in the germen, this plant must not 

 only be removed from Connarus, but from the tribe, and be placed near to Pistacia ; in which case it may, 

 with the P. oleosa of Loureiro and De Candolle, be referred to Dr. Hamilton's genus, Cussambium. (Wern. 

 Trans, v. 5. p. 256.) Against this, however, the structure of the advanced germen seems considerably to 

 militate, being compressed, and evincing a tendency to split readily, nor does the epicarp separate, as in the 

 case of Pistacia, and other young drupes. We are therefore inclined to suppose, that there may actually 

 be two ovula present in the unimpregnated germen, and we place it in Connarus. 



Ord. XXIX. LEGUMINOSM. Juss. 



1. Crotalaria retusa. Linn. Sp. PL p. 1004. — Lupinus Cochinchensis. Lour. Cochin. 2. 

 p. 520 ? 



2. Crotalaria Vachellii ; stipulis subulatis aculeiformibus minutis recurvis, foliis trifolio- 

 latis, foliolis oblongis subretusis cum mucronulo subtus adpresse pubescentibus subpellucido- 

 punctatis petiolum duplo longioribus, racemis patentibus subelongatis oppositifoliis, legu- 

 minibus globosis breviter pedicellatis hirsutis styli basi indurato uncinatis. 



This species is exceedingly common at Canton, but we have not been able to discover that it is noticed 

 by Loureiro. It approaches very close to C. virgata, Keen, and Roxb. (n. 373 of his drawings at the India 

 House,) or C. divaricata, Graham, in Wallich's Catalogue of East India Plants. The stems are herbaceous, 

 but what the duration of the plant may be we are ignorant. It belongs to a small groupe, proposed by Mr. 

 Brown to constitute a genus, Cyrtolobus ; hut as no character of it is yet published, we have no means of 

 judging of its validity, and therefore leave it in Crotalaria. Clavulium of Desvaux, (Ann. Sc. Nat. v. 9. p. 

 407,) another genus, consisting of species removed from Crotalaria, is not distinct from it, or it must com- 

 prehend many more species than the two pointed out by the Author. 



1. Melilotus leucantha. Koch. 



Dr. Graham (Wallich's List of E. I. Plants, n. 5942,) appears perfectly correct in uniting this with M. 

 altissima, Thuill. Perhaps, therefore, the latter name ought to be retained as the oldest; but M. leucantha 

 ' is not only more expressive, but more generally adopted by Botanists. In like manner M. Indica is also 

 superseded ; it being no way distinct from M. parviflora, and is generally diffused throughout the world. 



1. Indigofera hirsuta ; caule suffruticoso erecto, ramis teretibus, petiolis pedunculisque 

 hirsutis, foliis pinnatim 3-5-jugis, foliolis obovato-oblongis obtusis utrinque villosis, racemis 

 folio longioribus, leguminibus tetragonis 6-spermis pendulis hirsutis. De Cand. — Linn. Sp. 

 PL p. 1062. De Cand. Prodr. v. 2. p. 288. Burnt. Zeyl. t. 14. Lam. III. t. 626. 



1. Lourea vespertilionis ; foliolis lateralibus nullis aut minimis, terminali transversim et 

 falcatim oblongo subemarginato longitudine decies latiore. De Cand. — Desv. Journ. Bot. 

 3. p. 122. — Hedysarum vespertilionis. Linn. 



1. Desmodium polycarpum. De Cand. Prodr. v. 2. p. 334. (non Wall.) — D. angulatum. 

 De Cand. I. c. p. 335. {sec. specimen e Wallich, et in " List of E. Ind. Plants," n. 5729.) 

 D. purpureum. Hook, et Am. supra, p. 62. — D. Hippocrepis. De Cand. 1. c. p. 338. — 

 Hedysarum purpureum. lioxb. Hort. Bengh. p. 57. — H. polycarpum. Poir. in Encycl. 

 Meth. 6. p. 41. —Hippocrepis barbata. Lour. Cochin. 2. p. 553. Burm. Zeyl. t. 53. / 2. 

 Lam. III. t. 628. / 4. 



It is unnecessary to repeat here the character we have already given of this species ; but we may remark 



