XV, 2 
Merrill: Melodorum of Loureiro 
129 
lacca, Burkill 2510; Perak, Scortecliini 1916 (all these distributed 
a§ Polyalthia aberrans Maing., flowering specimens) : Java, cult. 
Hort. Bogor XI- A- 1 1-7 1 (four sheets, Polyalthia siarnensis 
Boerl., flowering specimens from the type plant) , XI-A-63 ( Po- 
lyalthia affinis Teysm. & Binn., flowering and fruiting specimens) . 
As several detailed descriptions of this species, as well as no 
less than four illustrations of it, with details of the flowers and 
fruits, have been published, it would seem that a further de- 
scription is unnecessary. The genus Melodorum Lour., as I 
understand it, contains a single definitely known species, which 
while well defined and characteristic as a species presents a 
combination of characters which render it somewhat difficult to 
separate Melodorum from several not closely allied genera. It is 
clearly no Unona (that is, Desmos ) ; and it is equally aberrant 
in Polyalthia and in Popowia, the two other genera in w T hich it 
has been placed. I am personally of the opinion that Melodorum 
as originally described by Loureiro and typified by Melodorum 
fruticosum Lour, (not of modern authors) is a valid genus, 
more closely allied to Popowia than to Polyalthia, and that it 
belongs in the tribe Miirephorae. Pierre in his critical discussion 
of Unona mesneyi notes that Maingay described the inner petals 
as imbricate in bud, but that he found them to be perfectly 
valvate in Wallich’s specimen. Both series of petals touch by 
their thickened margins, and those of the inner series remain 
in this position long after anthesis. I have seen no fresh 
material of Melodorum fruticosum, but the figures given by 
Pierre, Boerlage, and King present the outer series of petals as 
more or less spreading in anthesis ; in all the herbarium specimens 
examined by me none of the petals are spreading. The persistent 
valvate position of the inner petals is a character by which the 
genus can be readily distinguished from Polyalthia and all the 
other genera in the tribe Unonae; while in the Mitrephorae it 
is readily distinguished from Fissistigma Griff. ( Melodorum 
auct., non Lour.) by its globose buds and more or less spreading 
( ?) outer petals ; from Popowia Endl., which seems to be its true 
alliance, it differs entirely not only in its facies, but also in its 
larger, long-pedicelled flowers; in its outer petals, which are 
much larger and entirely different from the sepals; and by the 
inner petals being valvate by their much thickened margins but 
not connivent. 
Melodorum arboreum Lour. 9 = Unona sylvatica Dunal, 10 the 
9 Fl. Cochinch. (1790) 351. 
10 Monog. Anon. (1817) 91. 
