MaieriaU for a Flora of the Malayan Peninsula. 365 



lobes. Petals 5, much longer than the calyx, thick, spx'eading, orbicular. 

 Disc broad, convex, fleshy. Stamens 3, inserted near the apex of the 

 disc ; the filaments broadly triangalar, compressed ; anthers small. 

 Ovary flat, buried in the disc, the short conical thick style protruding ; 

 ovules 3 iu eacli cell. Fruit unknown. 



Perak : Scottechini, No. 1848. 



This is known only by the late F. Scortechini*s scanty specimens. 

 It is a very distinct species. 



7. Salacia grandiflora, Kurz in Journ. As. Soc. Beng., 1872, 

 pt. 2, p. 300. A shrub or small ti^ee ; young branches rather slender, 

 their bark pale when dry. Leaves coriaceous, oblong-elliptic or elliptic, 

 shortly cuspidate, the edges entire, the base sliglitly narrowed or 

 rounded ; both surfaces glabrous, the upper very shining and reticulate ; 

 main nerves 7 to 9 pairs, ascending, curved, sliglitly prominent ; length 

 5 or 6 in., breadth 2 25 to 3*25 in., petiole '5 in. Floioers glabrous, '25 to 

 "45 in. in diam., in groups of 3 to 6 from very short axillary or exti\i- 

 axillary bracteolate tubercles ; their pedicels '2 in. long. Calyx cupular, 

 •deeply divided into 5 sub-orbicular fleshy segments. Petals 5, orbicu- 

 lar or obovate-orbicular, concave, spreading. Disc yej-'j convex, fleshy, 

 •glabrous. Stamens 3, inserted towards the apex of the disc ; the filaments 

 broad, triangular, recurved. Ovary bui-ied in the disc, the 3-angled 

 style alone protruding, stigma small. Fruit globular or ovoid, glabi^ous, 

 1 to 1"25 in. diam., the calyx and corolla persistent at the base while young 

 about 1"5 in. long when ripe. Kurz For. Flora Burma, I, 259 ; Laws in 

 Hook. fil. Fl. Br. Ind. I, 626. 



Malacca : Griffith. Perak : King's Collector, Nos. 5924 and 7579. 

 DisTRiB. — Tenassejim (Kew Distrib.), No. 891. 



Kui'z describes his S. grandiflora (Journ. As. Soc. Beng., pt. 2, p. 

 •300), as scandent, and as having its flowers on minutely bracteolate 

 axillary or extra-axillary tubercles. He has with his own hand written 

 the name S. grandiflora on various specimens in the Calcutta Herbarium 

 which, although they have all axillary tubercles, sheW lio evidence of 

 being scandent. These sheets are as follows :^H elf er, Tenasserim or 

 Andamans (Kew Distrib.), Nos. 898 and 891 ; Grifl&th, Malacca (with- 

 out number); and Wall. Cat. No. 2812 from Penang. These in turn 

 agree with v.arious specimens from Penang, Perak, Singapore and 

 Malacca which are described by their collectors as small erect shrubs, 

 and not scandent. The chaiacter of being scandent must therefore 

 be eliminated from Kurz's diagnosis of (S. grandiflora. In this view 

 Mr. Lawson appeal's to agree, for he describes both S. grandiflora and 

 *S. longifolia Hook. fil. (which I reduce to a variety of 8. grandiflora^ 

 as small trees. There- ts however a scandent species vevy closely re-s 



651 



