NOTES ON MIMOSEiE. 81 



last part of Pceppig's work, whose description perfectly 

 agrees with the plant before me, except that his measure- 

 ments of the leaflets and flowers are rather smaller. 



S. polyphyllum, p. 342.— Mart. Herb. Fl. Bras. n. 1102. 



Adenanthera, Hook. Journ. 4, p. 343. 

 A. pavonina, p. 343. — Philippine Islands, Cuming, n. 1691. 



Elephantorrhiza, Hook. Journ. 4, p. 344. 



Burke's collection from the Macalisberg district contains 

 specimens in flower and fruit of two forms, which appear to 

 he distinct species although it is difficult to assign positive 

 characters to them. In the one E. Burchellii, Benth. 1. c, 

 the leaflets are generally three or more pair to each pinna, 

 and when full grown scarcely more than 3 or 4 lines long, 

 the flower spikes from 2 to 2£ inches. In the other 



• Burkei, sp. n., the leaflets are much more than twenty pair 

 to each pinna, and near 6 lines long, and the flower spikes 

 from 4 to 5 inches. 



Prosopis, Hook. Journ. A, p. 346. 



Engelmann and Gray, PI. Lindheim, p. 34, describe the 



rurt of P. glandulosa as having the inner lining of the carpel 



ardened into distinct almost bony husks round each seed, 



and they state that there is the same structure, except that 



these husks are thin and paper-like in P. dulcis, whilst in the 



section Strombocarpa this lining is continuous through the 



Pod, and they suggest the adoption as distinct genera of the 



r ce sections Adenopis or true Prosopis, Algarobia> and 



Strombocarpa ; but striking as some of these differences in 



e Pod are, their adoption as generic characters would (as 



ln the case of the Acacia gummifera) make it necessary to 



stabhsh a separate genus for almost every species of which 



he pod is known to us, and as a whole Prosopis may be 



dls tinguished from the allied genera by the pod which is 



^dehiscent (or nearly so), with a membranous or crustaceous 



