On Macaranga triloba . 81 
which does not mix with glycerine (see figs. 5, 8 and 9). It will be 
remembered that Francis Darwin ( 4 ) showed that the Beltian cor¬ 
puscles contain oil—insoluble in glycerine but soluble in absolute 
alcohol. Associated with these “ food-bodies ” are glands resem¬ 
bling the Lupulin-glands of the female inflorescence of the Hop, 
which De Bary calls “ Bladder-glands.” The bladder is formed 
by the cuticle, which is separated from the epidermis by the 
secretion. In Macaranga there can be seen, in surface view, 
a plate of 8 or 9 cells, with radial and periclinal walls, below the 
bladder. If the multicellular food-bodies arise from these glands it 
must be by further division by anticlinal walls so that the circular 
plate of cells becomes a sphere. Satisfactory intermediate stages 
are wanting, but in some of the glands hexagonal markings are seen 
on the cuticle similar to those on the food-bodies (see figs. 4 and 6 ). 
The cuticle of the food-bodies is very tough ; it resists crushing and 
few re-agents can penetrate it. 
Nectaries are borne at the serrated tips of the leaves. They 
are identical with those described by Beccari in Macaranga 
caladiifolia , which species differs from the one under discussion in the 
shape and position of the stipules and in the absence of food-bodies 
in them. The nectaries in M. caladiifolia are cup-shaped glands 
found at the extremity of veins near the base of the peltate leaf 
(see figs. 7, 12 and 13). 
Finally there were found in the internodes many ants, not only 
adults, but also pupae and larvae. Their presence, which is 
apparently an unusual feature, is probably due to the fact that 
the material was put into spirit within a few hours of gathering, so 
that the ants had no time to escape, or would not leave their pupae 
in the internodes and could not drag them through the holes. The 
ants have been identified by Colonel Bingham as belonging to 
an apparently undescribed species of Creinastogaster near C. daisyi, 
Forel. 
This is the case for Macaranga triloba as a claimant for 
Myrmecophily. Of course one cannot but agree with K. Schumann 
when he says: “ Der Beweis dass ein Gewachs als Ameisenpflanze 
zu betrachten ist, kann im strengsten Sinne des Wortes nur an 
den Localitaten gefiihrt werden, wo dasselbe seine Heimath hat. 
Vor allem muss durch das Experiment gezeigt werden, dass die 
Entfernung der Insecten auf das Gewachs einen bemerkbar 
schadigenden Einfluss ausiibt.” Such an experiment has yet to be 
tried on Macaranga triloba in its native habitat, but the striking 
