ANTS AND PLANT-STRUCTURE 389 
which is adjacent to and inundated by the rivers 
and lakes, but sometimes deep in the virgin forest, 
wherever the land is so low that the water of rains 
may accumulate thereon to a slight depth. All the 
species have the unmistakable aspect of their order 
— the ribbed opposite leaves, the polypetalous flowers 
with beaked porose anthers, etc. ; but they are dis- 
tinguished at sight from most others of the order 
by the large, thin, lanceolate or ovate acuminate 
leaves, very sparsely set with long hairs, and having 
a hollow sac or a pair of sacs at the base either of 
all the leaves, or (more frequently) of only one of 
each pair when that one is much larger than the 
other. The leaves in the majority of the species 
have but three ribs ; a few species, however, have 
five- or even seven-ribbed leaves ; but, in all, the 
origin of the innermost pair of ribs is an inch or so 
up the midrib from the base of the leaf; and it is 
this portion of the leaf, from the insertion of the 
inner ribs downwards, which is occupied by the sac. 
The latter sometimes takes up only a part of the 
breadth of the leaf, when it is technically considered 
to be seated on the leaf (Epiphysca) ; in other 
cases the sac in its lower half absorbs the whole 
breadth of the leaf, when it seems to be seated half 
on the leaf, half on the petiole (Anaphysca) ; or, 
lastly, throughout its length it absorbs the whole 
breadth of the leaf, and then seems seated entirely 
on the petiole (Hypophysca). That it is really 
formed in all cases at the expense of the lamina, 
and not of the petiole, is proved by the occasional 
occurrence of imperfectly-developed sacs in the 
hypophyscous form, bordered by a narrow wing con- 
tinuous with the leaf, and giving to the latter a 
