474 Ridley.—Symbiosis of Ants and Plants. 
down, and in an unoccupied internode small ragged bits only remain. 
In a plant occupied by ants these are removed, and no trace of the pith is 
left. The walls of the tunnel are then dark brown from the excreta of the 
ants. As the stem grows the ants carry their young into the upper inter¬ 
nodes by perforating the septa of the nodes, so that only the terminal 
internode, still green and sappy, is not inhabited, for into this the ants 
do not attempt to tunnel. In each internode there are from one to three 
perforations made by the ants to communicate with the exterior. Usually, 
there is one to each internode, and it is not rare to find an internode 
not perforated. When more than one is made they are generally in a line 
close together. They are sometimes made in the upper part, sometimes in 
the lower part of the internode, and usually in a shallow inconspicuous 
groove on the side further from the leaf. 
It is clear, I think, that their external openings are made from the out¬ 
side, at least in the case of the first ones on the stem, but the ones made in 
the upper internodes in some cases, at least after the ants have occupied 
the lower ones, appear to be made from the inside. 
Within the hollows of the internodes are placed the young larvae. The 
queen ant is usually to be found in the lowest occupied internode. 
There is no nest properly speaking, and no material brought from 
outside. The larvae lie quite loose and motionless within the hollow. 
They possess on the body a number of very short hairs by which, 
I imagine, they retain their position on the vertical walls of the tunnel. 
Around the head of each I have found four or five * food-bodies * placed 
near the mouths, presumably for them to eat. The £ food-bodies ’, which 
will be described later, are brought into the hollow stem from the bud- 
bracts through the lateral perforation. I have seen the ants carrying 
them about in their mouths, and even bearing them off when the stem 
has been split and they commence to remove the larvae into a safer spot. 
Besides the larvae and food-bodies, there is nothing else in the hollow 
except occasionally some Coccidae are brought in. This, however, is more 
common in M. Grijjithiana than in M. triloba. These coccids are orbi¬ 
cular or elliptic in outline, of a pink colour, and covered with elevated 
round basses on the raised back. They are usually put in internodes 
not occupied by the larvae, and must be brought in quite young, as when 
adult or nearly so they are too large to pass through the perforations. 
I have only seen these coccids in the hollowed stems, and have failed 
to find them on the leaves or outer parts of the plants. 
As the tree develops and becomes thicker and more woody, the per¬ 
forations to the surface disappear, but the hollow tunnel still remains, 
never entirely closing up, but is now unoccupied by the ants. The ants 
in fact only remain in the trunk of the tree as long as the external per¬ 
forations remain. I have seen them still inhabiting the base of a young 
