OF THE DIPTEROCARPACEA, 55 
of the next higher internode. The case seems to be analogous 
to what has above been noticed in regard to Dryobanalops 
aromatica and oblongifolia, where the main resin-duct sometimes 
branches below the base of the internode. Of the 3 corner 
ducts one is destined for the apical, the other 2 send branches 
into the lateral leaf-traces of the internode. These 4 resin- 
ducts run through the pith in the same manner for more than 
half the internode, and Plate I. fig. 7 represents a section made 
at about 3 of its length. Higher up (fig. 8) the 2 lateral ducts 
bifureate, and the branches, which are intended for the lateral 
leaf-traces, pass through the wood into the bark, where we find 
them in a section made immediately under the petiole (fig. 9). 
At the same time the apical resin-duet has subdivided into 5 
branch-ducts, the 2 lower and larger of which enter the axillary 
bud, while the 3 others enter the apical leaf-trace. Burck (‘Sur 
les Diptérocarpées,’ p. 161) states that the lateral cortical leaf- 
traces do not enter the petiole, but lose themselves in the inter- 
cellular spaces of the stipular cushion. In the case of Hopea 
cernua, 1 have not yet been able to make the examination 
necessary to verify this statement; but in the case of two nearly 
allied species, H. Mengarawan, Miq., and H. Pierrei, Hance, I 
am certain that the lateral leaf-traces enter the petiole. The 
stipules in these species are exceedingly small and early deciduous. 
The two species last mentioned have a structure similar to that of 
H. cernua, but there are important differences. H. Pierrei has 
5, not 4, resin-ducts at the base of the internode, and H. Men- 
garawan has 8-10. They agree, however, with H. cernua in this 
point, that the apical duct subdivides into a large number, up to 
8, some of which enter the petiole, while the others are destined 
for the axillary bud. They also agree in this point, that the two 
lateral leaf-traces do not run through the bark for any distance ; 
they separate from the central cylinder immediately under the 
insertion of the petiole. Burck claims for H. cernua and the 
other species of section Dryobalanoides two main resin-ducts, 
which run through the entire axis and send out branches into the 
apical and lateral leaf-traces. Matters are apparently more com- 
plicated and more varied in the different species of this section. 
The petiole of Hopea cernua has at its base 3 distinct vascular 
bundles, one on the underside, 2 lateral, each containing ore 
resin-duct (Plate I. fig. 13). The central mass consists of a 
four-sided ring of xylem and phloém, enclosing a semicircular 
