

CHRYSOBALANACE^l. 491 



of glands attached to the base of the midrib : they are glabrous, 6 to 

 8 inches long and 2 or 3 inches wide, rather coriaceous and lucid, 

 especially the upper surface, copiously feather-veined; the veins 10 

 to 12 pairs, divergent, and connected by an elaborate network of 

 veinlets. Petiole only 2 or 3 lines in length. Stipules linear, 5 or 6 

 lines long, very caducous. Flowers crowded in axillary or terminal, 

 short-peduncled, sometimes geminate racemes, which are shorter than 

 the leaves, more or less spreading. Bracts oblong-linear, coriaceous, 

 caducous. Pedicels articulated with the axis, lh to 2i lines long, 

 bibracteolate near the summit; the bractlets oblong-linear, longer 

 than the pedicel, shorter than the calyx-tube, deciduous. Calyx 

 externally silky-canescent ; the tube funnelform or club-shaped, a little 

 oblique, 3 or 4 lines long, not thickened nor fleshy ; the orifice equal, 

 the interior villous with reflexed hairs, in the manner of the genus ; 

 the lobes obovate or oblong, obtuse, shorter than the obovate-orbicular 

 and deciduous petals; the latter nearly equal, convolute in aestivation, 

 except the posterior, which is wholly external in most cases. Stamens 

 monadelphous at the very base (or raised on a narrow annular disk), 

 inserted on the throat of the calyx; the 14 or 15 anterior fertile, with 

 elongated and filiform filaments (involute in aestivation) ; the posterior 

 7 to 10, reduced to as many rigid and short teeth. Ovary two-celled, 

 with a single erect ovule in each cell, very villous, its long stipe 

 laterally adnate to the anterior side of the calyx-tube quite up to 

 the throat; the free side densely bearded with deflexed villous hairs. 

 Style basal, filiform, as long as the stamens, excessively villous at the 

 base, the upper part glabrous, the apex tipped with a simple stigma. 

 Fruit not seen. 



This species is manifestly related to P. JacJdanum, Benth. (Petro- 

 carya excelsa, Jack), and to P. Griffitkianum, Benth.; but it has not 

 the fleshy calyx which Bentham assigns to his section Sarcostegia. 

 Having a racemose inflorescence and short pedicels, this species is con- 

 venient for determining the position of the several floral organs in 

 relation to the bract and axis. We find the odd sepal to be anterior, 

 with a slight obliquity: it nearly, but not exactly, subtends the ovary, 

 as in Leguminosce, being between that and the bract, but a little to the 

 right of the median line (vide Fig. 4) : thus essentially confirming the 

 view indicated by Mr. Bentham, in Hooker's Journal of Botany, 2, 

 p. 211. 





