68 MR. bentham's synopsis or legnotide^, 



and sometimes on the same specimen. The leaves are generally 

 quite entire; occasionally some may be found with a few very 

 small teeth towards the extremity, but never toothed all round, as 

 in the G. lancecdfolia. I have therefore no hesitation in con- 

 sidering this as the G. integerrima, DC, to which I would refer 

 the G. zeylcmica, corymhosa and sinensis of Amott, G. timorensis, 

 BL, G. octopetala, F. Muell. (the parts of the flower are often in 

 eights in the Indian specimens), and Pootia cereopsifoUa, Miq. 



O. laneecefoUa, E-oxb., is only known to me from specimens 

 from the Calcutta G-arden and one or two from Sylhet. It 

 appears well distinguished by the leaves being regularly serrulate 

 all round, and, as far as I have seen, by a more pear-shaped fruit, 

 which appears to ripen generally more than one seed. The leaves 

 are usually considerably broader than those figured by Wight 

 from Roxburgh's drawing. The G. conjlnis, BL, is evidently 

 closely allied to this species, if not identical with it. G. sym- 

 metria, BL, appears to be a form of 0. integerrima with the leaves 

 slightly toothed at the top. 



G. lucida, Boxb., appears to me to be in some respects a made- 

 up species, resulting perhaps in a confusion between G. integerrima 

 and G. lancecufolia. The specimens originally sent from the Cal- 

 cutta Garden as G. lucida belong to the broader-leaved forms of 

 G. lancecefolia, although I have not seen any quite so broad as 

 those figured in Eoxburgh's ' Coromandel Plants ' and in Wight's 

 * Icones.' Both these figures are taken from Roxburgh's drawings, 

 the work probably of a native artist ; and, as is so frequently the 

 case, more or less made up, as is evidenced by the venation of the 

 leaves, so difierent from that of any Garallia. I should there- 

 fore be disposed either entirely to reject G. lucida, or to consider 

 it merely as an ill-represented variety of G. lancecefolia. 



G. celeUca, BL, judging from Borneo specimens which agree 

 with his description, is a distinct species with looser inflorescence, 

 much larger flowers, &c. 



G. cuspidata, BL, and G. multijlora, BL, both from Borneo, but 

 unknown to me, appear from his description to be both of them 

 good species. 



Baraldeia, Dup. Thou., is considered by B. Brown, and after him 

 by other botanists, as a species of Garallia. It is unknown to me : 

 the Hookerian herbarium, although now rich in Madagascar plants, 

 contains, as far as I can find, no Legnotideous species from that 

 island. 



