OF THE DIPTEROCARPACE E. 11 
The underside of leaves is in some species clethed with minute 
and very thin round scales. As a rule the pubescence of leaves, 
petioles, stipules, branchlets, and inflorescence consists of stellate 
or fasciculate hairs, which are nearly always one-celled. An 
exceptional case of pluricellular hairs is on the inside of the calyx 
of Vatica lanceefolia. The ramifications of the inflorescence, the 
calyx, and the outside of petals are, as a rule, hairy. Dryoba- 
lanops and Doona (most, but not all species), and some species of 
Hopea, form the most notable exceptions. 
The inflorescence is, so far as my researches have gone, always 
racemose, never cymose. Racemes or spikes, very often uni- 
lateral, are arranged in axillary or terminal panicles, the branches 
of which are frequently distichous and regularly alternating. I 
have called these panicles racemose when the branches are not 
further ramified (Engler, Nat. Pflanzenfam. iii. 6, fig. 123 F, 
Shorea Warburgii, Gilg). 
There are, however, numerous and most remarkable instances 
of flowers, which at first sight appear to be terminal, standing in 
a bifurcation of the inflorescence (Plate III. fig. 18, Vatica 
perakensis, King, subgen. Synaptea). A careful examination, 
however, shows that one of the apparently equivalent branches 
of the bifurcation is the main axis, and that the flower which is 
apparently terminal is in reality the lowest flower on either the 
main axis or the branch. In some cases, particularly in Vatica 
lanceefolia, Blume (Retinodendron), there are 2 pseudo-terminal 
flowers, one the lowest flower of the main axis, the other of the 
branch (Plate IIT. fig. 4). Other species of Vatica with pseudo- 
terminal flowers are V. pallida, Dyer, and sarawakensis, Heim, 
both of subgenus Retinodendron. Most deceptive is Vatica 
Mangachapot, Blanco (subgen. Synaptea). The specimens collected 
by Dr. Warburg of this species at first sight give the impression of 
a cymose inflorescence. A similar inflorescence I have found in 
other genera. A drawing of Pachynocarpus umbonatus, Hook. f. 
(Plate III. fig. 25), clearly shows that the apparently terminal 
flower belongs to the main axis on the left. Anisoptera oblonga, 
Dyer, and Dipterocarpus intricatus, Dyer, have a similar inflor- 
escence. That of the last-named species is well represented on 
tab. 215 of Pierre’s ‘Flore Forestière de la Cochinchine’ In 
this species the bifureation of the panicle with an apparently 
terminal flower seems to be the rule. In order finally to settle 
the question, whether the interpretation here given is correct, 
