Dr. Arnott on the Genus Rhizophora. 369 



celled, while Blume attributes to the whole genus a three- 

 celled ovarium. 



7. B. maldbarica (Arn.) ; foliis elliptico-oblongis utrinque acuminatis, pe- 

 dunculis petiolum sequantibus apice trifloris, floribus arete sessilibus, 

 calycis laciniis obtusiusculis, ovario biloculari, tigello cylindrico demum 

 supra medium paullo attenuato obtuso. Wight. Cat. n. 2452. — Rhizo- 

 phora cylindrica, Lin. sp. p. 635 (ex parte). Rheed. Mai. vi. t. 33. 



Hab. ad oras Malabariae, Rheede, Wight. 



Both this and the preceding have the germinating tigellus 

 tapering slightly and obscurely angled towards the point. I 

 as yet know of no character to separate the two except the 

 narrower and more pointed leaves, and longer peduncles of 

 the Malabar plant. I have never, in the present one, observed 

 more or less than three flowers on each peduncle, but the two 

 lateral ones often drop off before expansion. 



8. B. parviflora (Arn.) ; foliis elliptico-oblongis seu lato-lanceolatis, pe- 

 dunculis semel bisve trichotomis multifloris, ovario 3-loculari, tigello 

 subcylindrico, Roxb. — Rhizophora parviflora, Roxb. Ft. Ind. ii. p. 461. 

 — R. cylindrica Roxb. Hort. Beng.p. 36. Wall. Cat. n. 4877. 



Hab. Ad ostia Gangis. 



My character of this is taken from Roxburgh, the only spe- 

 cimen so named, which I have seen, being from Penang 

 (Wall. Cat. 4877* c.) and having neither flower nor fruit; its 

 leaves are slightly acuminated at both ends, so that the species 

 may be R. cylindrica of Blume. 



If the number of cells of the ovary and flowers on the pe- 

 duncle be liable to variation, the three species of this subsec- 

 tion may be combined, unless some of them be found to have 

 the flowers pedicellate, respecting which Jack, Blume, and 

 Roxburgh are silent. 



§. Vice not a. 



9. B. sexangula (Spr.) ; foliis ovato-lanceolatis, peduneulis unifloris, ca- 

 lyce 10-fido magno laciniis acutis patentibus, petalis ciliatis, tigello 

 prismatico aequaliter hexagono obtuso. — Rhizophora sexangula, Lour. 

 Cock. (ed. Willd.) p. 363. 



Hab. In Cochinchina, Loureiro. 



This obviously approaches very closely to B. Australis, and 

 by the tigellus to B. gymnorrhiza, Bl., but from the imperfect 

 description of the petals, it is uncertain whether it be a form 

 of either, or a distinct species. Loureiro adds, " filamenta 20, 



Ann, Nat. Hist. Vol. 1. No. 5. July 1838. 2 b 



