MB. G. BENTHAM OUT BIXACEjE AND SAMTOACEJE. 79 



a distinct family, most nearly allied in his opinion to Geraniaceae 

 and to Bombaceae, from both of which it appears to us to be far 

 removed in its calyx, stamens, gynaecium, fruit and seeds, although 

 the wool of the latter as well as the foliage may remind us of Bom- 

 baceae. He hints, however, at a connexion with item as requiring 

 further investigation ; and this has proved to be so close, notwith- 

 standing the exceptional character of the seeds, that we have no 

 hesitation in bringing it into the same tribe. It agrees with JBixa 

 in habit, in the yellow juice, palmately veined leaves and large 

 flowers, in the insertion and arrangement of the deciduous sepals 

 and petals, in the large number of hypogynous stamens with pecu- 

 liar anthers, in the style, and in the valvular rather large capsules, 

 with a membranous more or less separable endocarp. It has three 

 instead of two placentas, which, although in some species they are 

 •scarcely more prominent than in Bixa, yet in other species form 

 semisepta, especially near the base, dividing the capsule into three 

 more or less complete cells ; but the most striking difference is in 

 the cochleate more or less woolly or hairy seeds, with a very much 

 curved embryo. This character, however, will not alone suffice 

 for the establishment of a distinct natural order, especially as it 

 does not exist in Amoreuxia. To the species of Cochlospermtm 

 enumerated by Planchon, we have, as yet, no additions to make. 



Amobettxia, Moc. et Sets. 



Planchon, in the above-mentioned memoir, has well shown that 

 this genus, identical with Euryanthe, Schlecht., must, in the series 

 of families, follow the fate of CochJospermum. Like the latter 

 genus, it has the flowers, valvular capsulej and soluble endocarp 

 of Bixa. The seeds, unknown to Planchon, but since received 

 amongst the rich collections of Mr. Charles Wright, and accu- 

 rately described and figured by A. Gray, are much nearer to those 

 of Bixa than of Cocklospermum, only differing from the former by 

 the cotyledons more or less bent or folded. The placentae, how- 

 ever, are completely united in the axis of the ovary and capsule, 

 dividing them into three perfect cells. But here again is a solitary 

 character, no more available for the establishment of a distinct 

 Order in this case than in that of Flacourtia. 



Bixa, Linn. 

 Of this genus there is one species, furnishing the arnotto of 

 commerce, which, by long cultivation within the tropics, in Africa 



