L. COMBRETACEtE. 
563 
many as calyx-lobes, usually small, imbricate or valvate. Stamens as many or 
twice as many as calyx-lobes, rarely indefinite, inserted on the calyx ; anthers 
opening in longitudinal slits or (in Gyrocarpea ) in 2 valves. Ovary inferior, 
1 -celled, with 2 or more pendulous ovules, or (in Gyrocarpcat) with 1 only ; style 
filiform or scarcely any, with an entire terminal stigma. Fruit coriaceous, char- 
taceous or drupaceous, indehiscent (except in a few species not Australian). Seed 
solitary, pendulous, without albumen ; cotyledons convolute or folded, very rarely 
flat inside and furrowed outside ; radicle short, superior. — Trees shrubs or woody 
climbers. Leaves alternate or opposite, entire, without stipules. Flowers in 
axillary or terminal racemes spikes or heads, or (in Gyrocarpecc) in cymes. Bracts 
usually small ; bracteoles sometimes larger, often wanting. 
The Order is distributed over the tropical regions of the New and the Old World, a very few 
species extending beyond the tropics in S. Africa or in N. India. Of the four Australian genera, 
three are common to America, Africa, and Asia, one of them restricted to seacoasts, the fourth is 
endemic. — Benth. 
Suborder I. Combreteae. — Calyx-lobes valvate. Stamens without g lands or staminodes at 
their base ; anthers dehiscing by a longitudinal slit. Ovules 2 to 7, suspended by long funicles. 
Flowers racemose or spicate. 
Anthers opening in slits. Ovules 2 or more. Flowers in racemes, spikes, 
or heads. 
Calyx-tube not produced above the ovary. Petals none. Stamens 10 . 1. Terminalia. 
Calyx-tube produced above the ovary. Petals 5. Stamens 10 or fewer. 
Bracteoles small. Ovules 2 to 5. Maritime shrubs 2. Lumnitzera. 
Bracteoles enlarged and forming wings to the fruiting-calyx. Ovules 
10 to 12. Silky or tomentose shrubs 3. Macropteranthes. 
Suborder II. G-yrocarpeae. — Calyx-lobes valvate or imbricate. Stamens with glands or 
staminodes at their base ; anthers opening by recurved lateral valves. Ovule 1, suspended by a 
short funicle. Flowers cymose. 
Anthers opening in 2 valves. Ovules solitary. Flowers small, in cymes. 
Petals none 4. Gyrocarpus. 
1. TERM IN ALIA, Linn. 
(Leaves usually at the end of the branches.) 
(Chuncoa, Ruiz and Pav.) 
Calyx-tube not produced above the ovary ; limb campanulate or urceolate, 
5-cleft. Petals none. Stamens 10, longer than the calyx. Style filiform. 
Ovules 2, rarely 3. Fruit ovoid, terete, angular, compressed or with 2 or (in 
species not Australian) 3 to 5 longitudinal wings. Cotyledons convolute. — Trees 
or erect shrubs. Leaves alternate or rarely opposite, usually marked with minute 
pellucid dots, often only visible under a strong lens. Flowers hermaphrodite or 
polygamous, small, green, white or rarely coloured, sessile in loose spikes, rarely 
contracted into dense heads, either axillary or clustered on the old nodes. Calyx- 
tube usually small and narrow, the limb much broader. 
The genus extends over nearly the whole range of the Order, but is most abundant in Africa 
and Asia. The Australian species appear to be all endemic, with the exception of T. Catappa 
and T. microcarpa. Several of them, however, are as yet insufficiently known. They are often 
not to be distinguished without the fruit, which, when succulent and not winged, is rarely 
perfect in herbarium specimens, and we do not as yet know how far the fruit may vary in the 
same species. Some with broadly winged fruits have precisely the foliage and flowers of others 
which have wingless fruits. The circumscription of species here given may therefore require 
much revision when more perfect materials are obtained. — Benth. 
The subdivision of the genus into sections, or with some botanists into distinct genera, has 
been founded on the fruit alone, and although the line of demarcation is often very indefinite, 
no better character has as yet been found. The Australian species are included in Chuncoa, 
with 2 or 3 distinct wings to the fruit, Catappa with 2 wings, confluent above and below so as 
completely to encircle the drupe, and Myrobalanus without wings ; but in T. volucris the wings 
are often slightly confluent so as to do away with all real distinction between Chuncoa and 
