Sapindus. octandria monogynia. 281 



of most parts of India. Flowering time the beginning of 

 the hot season. 



Trunk straight ; branches also nearly erect, and few of 

 them. Bark smooth, and ash-coloured ; height of the tree 

 generally about twenty feet. Leaves alternate, about the 

 ends of the branchlets alternately pinnate ; from six to 

 twelve inches long. Leaflets subalternate, from eight to 

 twelve in number, or from four to six pair, entire, ob- 

 liquely lanceolate, oblong, smooth on both sides, and a- 

 bout four inches long. Petioles common, round, flexuose, 

 smooth. Panicles terminal, and from the exterior axils, 

 diffuse, composed of diverging, compound ramifications. 

 Calyx five-leaved. Petals five, equal, and regular. Nec- 

 tary, two woolly scales near the base of each petal. Sta- 

 mens six or eight ; filaments woolly. Germ three-sided, 

 sitting, with the stamens, on a large glandular recepta- 

 cle. Style single and short. Drupes generally solitary, 

 seldom more than one coming to maturity, one- celled, 

 subglobular, very smooth, and yellow, with a pretty large 

 ridge round the base on the outside, the inside mark- 

 ed with the two abortive lobes of the germ. Nut solitary, 

 round, and smooth, aftixed to the inside of its cells, where 

 a considerable quantity of woolly fibres intervene. 



With the pulp of the fruit the Hindoos wash linen, &c. 



In January, 1808, a healthy young tree of about twen- 

 ty feet in height, reared from seed, received from North 

 America, under the name Sapindus Saponaria, flower- 

 ed abundantly, and ripened many seeds. It differs from 

 my detergens ; 1st. in being a larger tree, and more ra- 

 mous. 2ad, In the leaflets being acute, and lanceolar, 

 that is taper at each end. 3rd. In the calyx, and corol 

 consisting of six parts each, which are rjund, and shorter 

 than the germ ; and in the petals being without the two 

 woolly scales, so conspicuous in detergens. I therefore 

 conclude they are distinct species, and doubt if the Ame- 



Jj 



