470 Ridley.—Symbiosis of Ants and Plants. 
so large. As they wither they curl in and make a large cabbage-like ball 
in which the vast myriads of ants live. Platycerium , when removed from its 
tree to a plant-house, seldom thrives well unless ants have an access to it from 
the ground, as during the removal the original inhabitants usually leave the 
nest. It should, however, be mentioned also that in a plant-house or such 
place it has not the advantage of the fall of leaves in it, at least to the 
ordinary extent, and also the water-supply from above is perhaps not what 
it ordinarily gets. But considering the smaller catch of leaves which this 
bulkier plant gets and the large amount of soil usually found in the centre, 
I have little doubt that the main feeding of this plant is effected by the 
Dolichoderus. and that the adaption of the erect fronds, though primarily for 
the catching of fallen leaves and for covering the roots when they are dead 
to prevent a loss of water, yet has a secondary function in encouraging the 
presence of ants, which thus supply an additional source of nutriment 
without which the plant would not thrive. 
Clerodendron myrmecophilum, Ridl. 
This plant and Cl. brevifios , Ridl., are both inhabitants of very wet spots 
in Singapore, Johore, and other parts of the Peninsula, and both have 
hollowed stems in which live ants, much in the same way as they do in 
Macaranga. The stems are usually quite simple and unbranched, bare of 
leaves at the base. There do not appear to be any further modifications to 
attract ants, and it is difficult to see any advantage their presence can be 
to the plant. Under cultivation, as the plant is propagated by cuttings, the 
ants soon disappear and the plant seems none the worse, except that it is 
very liable to attacks of coccids and aphids which are often to be found 
beneath the young leaves, which they spoil. Ants, however, protect these 
insects and often cover them in with walls and masses of sand beneath the 
leaf. It is quite conceivable that the removal of coccids by ants from the 
young foliage, which they seriously injure or destroy, to the interior of a 
hollow stem where it appears from their constant presence in Macaranga 
they are harmless, may be advantageous to the plant, but I have no 
evidence to offer for this in Cterodendron , and the blights I have found on 
the young leaves are not the kinds transferred to their nests by ants. 
Beccari, in describing C. fistulosum of Borneo, suggests that the tunnelling of 
the stem by the ants strengthens it. I hardly think this would account for 
the ants, as the same structure occurs in Macarangas even when no ants 
are present. 
Macaranga. 
The genus Macaranga includes upwards of ioo species of trees and 
shrubs distributed over the tropical regions of Africa and its islands, Indo- 
malaya, and the Australian and Polynesian regions, but as those defined 
