HEMIGYROSA DEFICIENS. (Nat. order_Sapindacese.) 



lor Gen. Char, see letter press to Plate cli. 



HEMIGYROSA DEFICIENS. (WA.) A middling sized or small tree, leaves 1-2 feet long abruptly pinnate, petiole 

 terete smooth, leaflets 5-7 pair opposite sub-opposite or alternate, lanceolate to linear-oblong obtusely acuminated, entire, quite gla- 

 brous ou both sides, 5-9 inches long by lj,-]f broad, petioles 2-4 lines long, racemes paniculate solitary or congested on the old axils or 

 from nodes on the boughs, densely villous very variable in length from 2 inches to nearly 2 feet long, calyx leathery and densely villous, 

 5 parted much imbricate the 2 outer segments generally much smaller than the others, petals 4 crenulate or sub entire equal or sub- 

 equal oblong to obovate with a short claw glabrous in the upper half densely villous towards the base at the back and furnished Inside 

 with a large entire hairy scale which is -| or more than \ as long as the petal. Male flowers : disk incomplete 3 angled unilateral, stamens 

 8-10 rarely only 6, unilateral slightly exserted, filaments very hairy no rudimeut of an ovary. Hermalhrodite flowers, disk incomplete 3 

 angled unilateral, ovary hairy slightly 3-lobed 3-celIed, stamens 8 regularly arranged round the ovary included, filaments hairy, anther3 

 fertile but smaller than in the male, style short thick hairy, stigma large umbilicate nearly entire papillose, ovules solitary attached to 

 the centre of the axis, fruit unknown. — Sapindus ? deficiens, WA. Prod, page iii. Anornosantb.es, Beiith. and Hook. Gen. PI. 1 p. 402, 

 and Bedd- vide Manual p. lxxii. 



Tinnevelly ghats, common 2-4000 feet elevation. Anamallays (head of the ghat from Palghat up to the Nelliampatty coffee estates) 

 2500 feet elevation, and toioards the higher ranges at 5000 feet elevation. I have only lately met with this tree in the localities above mentioned, 

 and there can be no doubt as to its being a true Hemigyroca, the male trees are far more abundant than the fertile ones, it appears to be in flower 

 all the year round as I found it in flower in Tinnevelly in September, and on the Anamallays in March, but of very many trees examined only 

 one had hermathrodite flowers and there was no fruit, the male racemes are sometimes of great length though generally only a few inches long, the 

 genus differs from Sapindus in its ii regular flowers and unilateral disk and in its inflorescence being lateral instead of terminal. 



Analysis. 



1. A small portion of the paniculate raceme (from a raceme 18 inches long.) 



2. Highly magnified portion of the same. 



3. A male flower bud. 



3ft. Full male flower with 8 stamens. 



36. A male flower, sepals removed showing only 6 stamens. 



4. Petal, outside view. 



5. The same inside view, showing the large entire scale. 



6. A male flower, sepals and petals removed showing the unilateral 3 lobed-disk and 8 stamens. 



7. The same showing 10 stamens. 



8. Anther?, front and back view. 



9. Hermathrodite flower. 



10. The same, sepals removed. 



11. The same, sepals and petals lemoved showing the unilateral 3-lobed disk and 8 stamens regularly arranged round the ovary. 



12. The same seen from the side opposite the disk, 



13. A petal, inside view. 



14. The anthers, front and back view. 



15. The slightly lobed hairy ovary and style and the oblique entire papillose stigma. 



16. Ovary cut vertically the cells with a solitary ovule attached to the centre of the axis. 



17. The same cut vertically. 



N. B. — As the only supposed species of Anomosantb.es turns out to be a Hemigyrosa, the genus Anomosantb.es must be expunged from the 

 Manual. 



231 



