476 Ridley.—Symbiosis of Ants and Plants. 
with cellular markings, and it appears to consist of small irregular cells. 
There can be little doubt that they are derived from the bun-shaped 
bladder-glands, which are abundant on all parts of the plant where the 
food-bodies occur. They are to be seen in all stages of development, 
and seem to continue developing as long as the bud-bract exists. 
The bladder-glands are semi-transparent or quite transparent, and 
appear to be composed of about eight or nine large cells. They occa¬ 
sionally have a golden yellow tint, especially in older leaves. In young 
leaves they cover the back densely, but are more scattered in adult leaves, 
in which in some species they last for a long time. They are not con¬ 
fined to myrmecophilous plants, for they are very abundant on the leaves 
of M. trichocarpa , but here they never develop into food-bodies. In shape 
they are more or less round with the lower side flattened, the upper elevated 
into the shape of a bun. Miss Smith compares them to the lupuline glands 
of the Hop. 
The marginal nectaries. The leaves of M. triloba are in seedlings 
peltate ovate on long petioles, in older plants more or less trilobed green 
on both surfaces, or especially in young plants dark red. The adult leaf is 
quite glabrous, but when first emerging from the bud it is hairy, especially 
on the edge, and thickly covered on the back with bladder-glands. The 
margins are separated, and at the end of each tooth where a nerve ends is 
borne a green or reddish * nectary ’ or gland. These remain on the leaf till 
it is adult, when they seem to drop off. They are best developed in 
M; hypoleuca, where they appear first as short cylindric bodies of a bright 
crimson with a rounded depression at the top. They increase in length 
and become more distinctly clubbed, the top bending over slightly so that 
the depression is on one side. 
These nectaries occur in all the myrmecophilous species, but I cannot 
tell what function they perform, as the ants do not pay any attention 
to them. 
Macaranga Griffithiana, Muell. Arg. 
This tree resembles M. triloba , Muell. Arg., in its habit and myrme- 
cophily, though in specific characters it is very distinct. It is a fairly large 
tree, upwards of 40 feet tall, and has much of the habit of M. triloba , but 
its young branches and leaves are more glaucous from the deposition of 
wax. The backs of the leaves are, however, less white than those of 
M. hypoleuca. There are also differences in the flowers and fruit which, 
however, do not affect the question of the myrmecophily. This species 
seems to prefer the damper spots in and around the Botanic Gardens, even 
localities which are liable to be flooded from time to time, while M. triloba 
prefers the drier borders of woods. 
Like M. triloba its stems are hollow and tenanted by ants which also 
