402 ROBINSON. 



rotundatis, brevibus; corolla calyptrata; staminibus numerosis, brevibus; 

 ovario biloculari: foliis chartaceis, obovatis vel ellipticis, apice acuminatis, 

 basi acutis; venis utrinque 12-15. 



Inflorescences axillary or in the axils of fallen leaves, pyramidal-ovate, 

 3-4 cm long, 3-6 cm wide at base, usually thrice trichotomous, the 

 peduncle nearly 1 cm long or wanting, the rachis and its branches about 

 2 mm in diameter, with orbicular-ovate bracts about 1.5 mm long at 

 some or all of the articulations ; individual flowers in threes, sessile, each 

 subtended by two decussate pairs of bracteoles, 2.5 mm long, 3 mm wide, 

 broadly rounded, or one of these wanting between closely appresscd 

 flowers; calyx-tube turbinate or broadly oblanceolate, extending about 

 2.5 mm beyond the ovary, 4-5 mm long, 2.5-3 mm wide at the base of 

 the lobes; calyx-lobes 4, submembranaceous, 2-2.5 mm wide, about 0.5 

 mm long, rounded; petals falling calyptrately, the cap about 3 mm in 

 diameter, 1 mm in vertical height ; staminal disk inconspicuous ; stamens 

 numerous, their filaments 3-4 mm long, the lanceolate anthers about 0.4 

 mm long; style 5 mm long; ovary 2-celled, with numerous small ovules. 



A tall tree with a stem 60 cm in diameter, the terete or nearly terete 

 branches covered with gray bark; leaves with caniculate petioles 7-13 

 mm long, the lamina chartaceous obovatc or elliptic, 12-17 cm long, 

 6.5-9 cm wide, acute at the base, contracted at the apex into a short 

 obtuse acumen, the margins slightly revolute; primary lateral veins on 

 each side of the midrib 12-15, forming a definite submarginal vein, or 

 conspicuous on the lower surface, but less so than the midvciu. 



Luzon, Mindoro, Baco River, McGregor 8&f 



Very closely allied to the preceding, from which fuller collections may show- 

 it not to be distinct. Apparently differing by the size, shape, and texture of the 

 leaves. 



BLANCO'S SPECIES OF 1CYBTAO&S. 



Nearly all of these are referred to in the proper places in the text, but a brief 

 summary may be advisable. In no case, is his description satisfactory, and some 

 of his species may well be mixtures, and there is reason for supposing that others 

 are twice described, apart from mere changes of name. 



1. Eugenia malaccensis Blanco Fl. Pilip. (1837) 415; ed. 2 (1845) 290. 

 This has generally been reduced to E. jambos, and there is every reason to 



believe that the reduction is correct. 



2. AIybtus mananquil 1. c. 421. 

 Eugenia mananquil ed. 2 290. 



E. lobas ed. 1 857. 



E. canliflo7'a (cauliflora) ed. 2 291. 



The second and fourth names are his changes of the first and third. These 

 seem to me to be the same, at least possibly. E. mananquil has been reduced to 

 E. javanica, but he very carefully specifies that the fruit is rounded and not 

 depressed-hemispheric, which should exclude E. javanica. The species here con- 

 nected with his names is the commonest indigenous one, to which his statements 

 nearly apply. 



