480 
Ridley.—Symbiosis of Ants and Plants. 
Of the myrmecophilous species, we have two series. In both the 
stem at first is solid and woody, but as it develops the internodes dilate, and 
the pith, which is proportionately large, disappears and the stem becomes 
hollow. The bud-bracts are large and more persistent. The leaves bear 
on their backs numerous bladder-glands, and some of them develop into 
food-bodies which are used by the ants inhabiting the hollow stem as food 
for their young. 
In most cases the stems and leaves are more or less glaucous or white 
from the excretion of wax. The young leaves possess nectaries on the 
edges. 
The differences between the two series are as follows :—In one 
the bud-bracts are lanceolate and leaf-like, and remain sub-erect; they 
bear neither bladder-glands nor food-bodies. The food-bodies and glands 
are borne on the under side of the young leaf, which for some time remains 
with its lobes deflexed. In the other the bud-bracts bear on their outer 
surface bladder-glands, especially at the base. They are long-persistent, 
and soon after opening are deflexed, till meeting the stem they form 
a chamber, in a ring shape, within which food-bodies are developed. The 
young leaf opens flat and entire, and is not deflexed and bears no food- 
bodies. 
That these modifications are not caused by any action of the ants 
is shown by the fact that all occur in plants in which ants are absent. It is 
clear, however, that they cannot be of any use to the plant except for the 
purpose of attracting ants, and causing them to permanently inhabit the 
plant. Observations show that plants of these species in which ants are 
absent suffer seriously from the attacks of caterpillars. Two of these 
modifications, namely, the hollowing of the stem and the development 
of some of the bladder-glands into food-bodies, may be said to be comple¬ 
mentary of each other, that is to say, neither alone would serve the desired 
purpose. Food-bodies alone would only attract ants temporarily, as they 
would take them to their nests as speedily as they could be produced. The 
additional advantage of the hollow stem induces them to remain as perma- 
ment guardians of the plant. Hollow stems which furnish abodes for ants, 
unaccompanied by any other inducement to them to occupy the plant, 
do occur in other plants, e. g. Clerodendron myrmecophilum , Ridley, but 
further investigation of these plants is required. In the case of such plants 
as Hydnophytum , the stem is naturally hollowed out with an external 
opening before the incursion of the ants. But this is not the case in 
Macaranga nor in Clerodendron , where the opening to the hollow chamber 
is made by the ants themselves. 
The modifications which bring about the occupation of the plant 
by ants are for the most part simple developments of organs originally 
possessing other functions. Bud-bracts occur in most other species of 
