224 
Psyche 
[December 
the beetles whenever the opportunity is presented, but 
with very indifferent success. When attacked the beetles 
usually wheel to present the posterior end to the ant and 
raise the apex of the abdomen. At times the beetle merely 
crouches, with the head lowered and antennas pressed back 
and in to the body, while the abdomen is raised and its apex 
directed forwards over the sloping elytra. Both maneu- 
vers serve the same purpose, the ant recoiling and behav- 
ing as though temporarily stunned. Meanwhile, the beetle 
escapes. 
Infrequently, upon meeting an ant the formicarius would 
not raise the abdomen but would dodge away and attempt 
flight, either by running or by crouching and then crawling 
away between the ant’s legs as the ant attempted to bite 
the beetle. If this retreat was blocked then the abdominal 
defense was successfully employed. 
The worker ulkei almost always attacked these beetles 
at the slender neck or peduncle between the head and 
pronotum. Once such a hold was obtained the outcome 
was more doubtful. I have witnessed such combats where 
the ant had seized a beetle by the neck and the two strug- 
gled back and forth. One of these encounters lasted twenty- 
five minutes continuously, with the beetle thrusting the 
apex of its abdomen at the ant’s head. At the end of this 
time the ant still held the beetle’s neck between her man- 
dibles, but appeared inert otherwise. The ant’s jaws were 
pried apart and the formicarius ran off, unharmed, but the 
ant proved to be dead. The same beetle was then examined 
but the integuments were not even scratched. Of course, 
the ant may have been an enfeebled individual but we 
think that it was killed by the beetle’s abdominal secre- 
tion. The protection of formicarius from ulkei, then, lies 
in (1) a hard exoskeleton, (2) agility, and (3) a defen- 
sive abdominal secretion which, under certain conditions 
at least, may be lethal. 
At times a formicarius would be killed, when placed in a 
small watch glass with five or six workers. This usually 
happened during the night, and although the beetle would 
be dismembered, it was not eaten. 
So far Megastilicus formicarius has not been observed 
to feed on dead host ants, although this is to be expected. 
