482 Ridley.—Symbiosis of Ants and Plants. 
early stages of its development, and though the young leaves in an 
unprotected tree are not entirely destroyed, they are so injured and deformed 
that the tree has a weak and wretched appearance, and does not develop to 
such a size or robustness as a guarded tree. 
Conclusion. Among the large series of plants which have been classed 
as myrmecophilous, we find that there are all stages of development of a 
true symbiosis included among them. Thus in the case of Ficus irregularis 
there was seen a commencement of a relationship between the ants and the 
tree, due to an accidental modification which temporarily affected the tree. 
Had the swellings on the branches been of any importance to the tree and 
become a permanent feature of it, the nest stage represented by such a 
state of affairs as occurs in Pachycentria might have developed. In this the 
dilated roots have probably a function of water-storage, and form a feature 
of the plant, and owing to their peculiar growth they are apt to become 
hollowed out, and are then, occasionally at least, tenanted by the ants which 
habitually live among the roots of epiphytes. This stage of development 
of myrmecophily in which an organ modified for another purpose is adopted 
as suitable nidus for ants, while, at the same time, there does not appear to 
be any definite advantage to the plant from their presence, seems to be 
quite common. 
In Korthalsia the swollen ochrea seems at first sight to have been 
evolved for the housing of ants, but as there seems to be no advantage 
derived from them by this modification, and plants thrive as well without 
them as with them, and, further, more than one species of ant occupies the 
ochrea, such a case can hardly be considered a true symbiosis. There are 
good reasons for the development of the ochreae into their present form for a 
different purpose, and there is no special modification which is obviously 
intended to attract the ants or to induce them to remain. To the same 
case belong, I think, Dischidia ) though here the plant does derive benefit 
from the presence of ants, when they inhabit the pitcher-shaped leaves and 
convey soil to the roots within, and the same advantage is obtained by the 
nest epiphytes Thamnopteris and Platycermm. The importance of the ants 
to such plants as Clerodendron myrmecophiluin , Lecanopteris , Polypodium 
sinuosum , Myrmecodia , and Hydnophytum is less clear to me, yet all are 
habitually tenanted by ants, and seem to a greater or less extent to be 
modified to form a nidus for them. It is possible that the swollen fleshy 
stems of Lecanopteris and the Myrmecodia , resting quite exposed on the tree 
trunks, might be attacked by rats or other enemies were they not 
defended by the ants; but I have no evidence of this, and more 
observations are wanted. 
In Macaranga triloba we have the most perfect development of 
myrmecophily and a true symbiosis. The hollow stem, the retention of 
the bud-bracts for some time after their original function of protecting the 
