162 D. Prain — Little Andaman and the Nicohars. [Dec. 



obviously rests with those who would unite, not with those who can 

 distinguish, the trees ; till the point is settled the writer feels compelled 

 to follow Blume in treating them as distinct. In the Andamans Carapa 

 moluccensis flowers in November and December, 6'. obovata flowers in 

 March and April. 



LEGUMINOS.^. 



Desmodium umbellatum do. 



Abrus precatortus Linn. 



Erythrina indica Lamk. 



Canavalia turgid a Grab, in Wall. Cat. 



This is the common sea-shore Ganavalia of the Andaman, Nicobar 

 and Malayan coasts and is undoubtedly the plant intended by Graham 

 as 0. turgida Wall. Oat. n. 5534 A, a plant collected by Wallich in 

 Penang. Cat. n. 5534 B, from Siam (Herb. Finlayson), is not re- 

 presented at Calcutta but most probably is, since Graham and Wallich 

 thought so, the same plant. G. turgida is certainly not identical with G. 

 eiisifoQ'mis, even if we admit that the Ganavalia gladiata, cultivated in the 

 Eastern Hemisphere, is conspecific with the American cultivated plant ; 

 nor is it the same as G. virosa W. & A., with which Mr. Kurz has 

 identified it {Journ. As. Sue. Beng. xlv, pt. 2, p. 127) and which the 

 writer agrees with Mr. Baker in considering the wild form of Ganavalia 

 ensiformis {G. gladiata). The interior of the pod, even more than the 

 different shape, makes the proposal to treat G. turgida and G. virosa 

 as conspecific quite impossible. Perhaps the confusion of G. turgida with 

 G. virosa may have arisen from the fact that Wall. Oat. 5534 0, from Ava, 

 is true G. virosa. A reference, however, to the Lithographed Catalogue 

 itself shews that Dr. Wallich only doubtfully refers the Ava plant to 

 Graham's species. Mr. Baker doubtfully refers G. Stocksii Dalz. & 

 Gibs., Bomb. Fl. 69, to G. turgida ; this is, in the writer's opinion, highly 

 improbable because 1., G. turgida seems always strictly confined to 

 sea-shores and to the banks of muddy estuaries and never has been 

 collected inland ; and 2, though it extends from the Salt-lakes near 

 Calcutta and from the Sunderbuns at the top of the Bay of Bengal to 

 the Indo-Chinese and Malayan Coasts generally, it has not yet been 

 found anywhere on the coasts of India proper, of Ceylon, or of the 

 Laccadives. 



Though a characteristic sea-shore species, G. turgida is not con- 

 specific with the G. ohtusifolia of the coasts of India proper, which ap- 

 parently does not occur in the Andamans, the Nicobars, or the Malay 

 Peninsula — on the shores of the Andaman sea, though it does occur 

 in Java (as pointed out by Prof. Miquel) to which island G. turgida also 

 extends. 



