458 Ridley.—Symbiosis of Ants and Plants. 
duces disastrous results ; such, as I hope to show, are the Macarangas. In 
these the symbiosis may be said to be complete. In others we find that 
though the plant possesses modifications utilized often by ants, such modi¬ 
fications have another function of importance to the plant, and there are no 
modifications which can only serve to attract the ants. Furthermore, in 
some of these cases the ants have apparently no function to perform that is 
of any use to the plant. In these cases there is no true symbiosis, even 
though ants commonly occupy a position on the plant. 
Then again, we have a series of plants—epiphytes—which, though 
undoubtedly deriving advantages from the ants, have no special modi¬ 
fication to induce them to occupy the plant, the modification by which the 
plants obtain the services of the ants being obviously intended originally 
for a different purpose. 
Ants are extraordinarily numerous and varied in habits in the tropics, 
and are very quick in finding out the most suitable locality to put their 
nests. Any kind of hole, e. g. in a stick if large enough, is adopted. 
A cool place under or between leaves, a deserted termites’ nest, anywhere in 
fact where they can put the eggs and larvae away from the action of sun 
and rain, is speedily utilized, and even if they commonly adopt a particular 
part of a plant, such as the swollen sheath of Korthalsia or the spathes 
of Dctemonorops , or even a hollow stem, it does not necessarily follow that 
this is a case of true symbiosis. When Professor Schimper was stopping in 
Singapore, as he and I were looking into various oecological questions, we 
found on a tree of Ficus inaequalis , a native of Celebes cultivated in the 
Botanic Gardens, a number of branches which were swollen just below the 
nodes, and partly split and hollowed out. These were tenanted by ants. 
A branch of this fig-tree is figured in the Pflanzengeographie, PI. 83, as an 
example of myrmecophily. I have since constantly examined this tree 
and others in the Gardens, and find that these swellings and the ants have 
quite disappeared. The branches were, I believe, swollen by some disease 
or abnormality of growth, and in course of development split and the holes 
produced were occupied by ants. The appearance of the twigs closely 
resembled that of Humboldtia laurifolia (figured on the same plate), a more 
truly myrmecophilous plant, but in the case of the Ficus the swelling and 
presence of the ants appear to have been abnormal and accidental. 
A somewhat similar case occurs in Pachycentria Uiberculata , an 
epiphytic plant of the order Melastomaceae, in which the roots, embedded 
in a mass of vegetable debris, ants’ nests, &c., are frequently swollen at 
intervals into fusiform dilatations. In most of these I find no opening or 
hollow space, the tubers being quite solid ; occasionally there is, however, 
an opening and a hollow space within (sometimes after the style of the 
tuberous stem of Myrmecodia ), and in this hollow ants have placed their 
nest. Beccari describes and figures this plant and allied species, but says 
