THE MONTANE FLORA OF FIJI. 147 
MELASTOMACE, 
MELASTOMA DENTICULATUM, Labill. Sert. Austr. Caled. 65, t. 64. 
Nadarivatu, 2700, in dry open spaces. Fl. Aug. . 593. 
Distrib. Fiji (Viti Levu, Vanua Levu, Ovalau, Totoya, Ngau). New 
Hebrides, New Caledonia, Tahiti, Samoa, Rarotonga, Solomon Islands. 
This shrub only occurred sparingly at Nadarivatu, but in many parts of 
the island where land has been cleared and neglected it has taken entire 
possession of the soil, becoming a perfect pest. 
MEDINILLA HETEROPHYLLA, А. Gray, Bot. U.S. Expl. Exp. 598, t. 75. 
Vatavula, Nabucara Hill, 1000, epiphytic in mixed forest, general. FI. 
Aug. 539. 
Distrib. Fiji (Viti Levu, Ovalau). 
MEDINILLA RHODOCHLENA, A. Gray, Bot. US. Epl, Exp. 600. 
Nadrau Valley, 2000, epiphytic in forest. Fr. Sept. 145. 
Distrib. Fiji (Viti Levu, Ovalau). 
MEDINILLA LONGICYMOSA, sp. nov. (PI. 14. figs. 21-23.) 
Glabra, foliis longe petiolatis, oppositis, ovatis, obtusis, integerrimis, 
3-nerviis, costa media subtus prominente ; cymis lateralibus 1-3-floris, longe 
peduncuiatis, eymarum bracteis minutis, pulverulentis, bracteis floralibus 
conspicuis, petiolatis, roseis, lanceolato-acutis, ante anthesin caducis ; 
floribus albis, pedicellatis, 4-meris; calycis tubo cylindraceo, dentibus 4, 
minimis haud conspicuis, rufo-pulverulentis ; petalis late ovatis, acutis, 
carnosulis; staminibus 8, inter se similibus, antheris elongatis, recurvis, 
apieibus haud evolutis, connectivo bilobo, antieis unicalearatis, posticis 
bicalcaratis ; ovario toto calycis tubo adhærente, 4-loculari ; stylo columnari, 
basi incrassato, stigmate punctiforme. 
Hab. Nadarivatu, 3500, on wooded ridge. 884. 
An interesting epiphyte, of bushy growth, very conspicuous from the 
foliose floral braets, which are 1:5 em. long and 8 mm. wide, of a lovely pink 
colour, the same shade as in some Bougainvillea sp. They fall before the 
flowers expand. The young inflorescences, minute bracts, and the calyces 
are pulverulent. My specimens show only buds, which are sufficiently 
advanced to prove that the plant is a true Medinilla, and as the habit differs 
so much from anything yet known in the genus, I have thought it advisable 
to figure it. The specimen 547, of exactly the same habit, has much larger 
bracts, and longer, flexuous pendent peduncles. It is probably a different 
species. 
