SAPINDACE.E. 253 



Hab. Ovolau and Somu-somu, Feejee Islands : in forests. 



" Tree 20 feet high ;" the branches and foliage glabrous, except the 

 young shoots, which are minutely pubescent. Leaves alternate, desti- 

 tute of stipules, with petioles 2 to 4 inches long. Leaflets 4 to 8, com- 

 monly 6, lanceolate, with a rounded base, gradually acuminate, falcate, 

 entire, rather pale, of the same hue both sides, veiny, chartaceous in 

 texture, 3 to 6 inches long, 9 to 18 lines wide; the partial footstalks 

 2 to 4 lines long. Racemes densely-flowered and spike-like, axillary 

 and subterminal, mostly simple, short-peduncled, pubescent, about the 

 length of the petioles. Pedicels articulated with the rhachis, about 

 li lines long, subtended by a minute deciduous bract. Flowers poly- 

 gamo-dioecious. Calyx small, a line in diameter, five-cjeft ; the lobes 

 ovate-triangular, greenish and rather herbaceous in texture, minutely 

 pubescent externally, valvate or nearly so in aestivation; the bud 

 opening very early. Petals 5, very small when the calyx opens, at 

 length as long as the sepals, unappendaged within, consisting of a trans- 

 versely dilated hastate-tlvree-lobed limb, raised on a slender claw, the 

 lateral lobes involute and villous-bearded on the margin. Stamens 8, 

 glabrous, purplish, inserted within the glabrous, crenate, and com- 

 plete, fleshy disk : filaments equal, in the sterile flowers filiform and 

 thrice the length of the calyx, in the fertile flowers scarcely exserted : 

 anthers oblong. Ovary sessile, ovoid, pubescent, three-celled, with a 

 single ascending ovule in cell : style none : stigma small, three-lobed. 

 Capsule obovoid or turbinate, obscurely three-angled, glabrous, not stipi- 

 tate, nor with a tapering base, 4 or 5 lines long, very obtuse, pointless, 

 three-celled, loculicidally three-valved ; the valves coriaceo-crustaceous, 

 thick, smooth externally, densely villous inside. Seeds not seen, 

 having all fallen from the open capsules. 



This plant is a manifest congener of Cupania apetala, Labill. Sert. 

 Austro-Cal. (judging from the figure), which must belong to the genus 

 Mischocarpus of Blume ; but I know not whether this author, in his 

 Rumphia, has so referred it. From Walpers, Ann. Bot. 2, p. 216, I 

 learn that Blume has there characterized a species, Mischocarpus 

 fucescens, with minute and caducous petals. If the lobes of the calyx 

 are valvate or nearly so in aestivation in Blume's species, as they are 

 in ours, the genus Mischocarpus would appear to be sufficiently distinct 

 from Cupania. 



64 



