176 
Psyche 
[Vol. 90 
ant lion’s head was above the sand or had already been withdrawn 
below the surface, and (2) whether the beetle was retained in the ant 
lion’s hold or released. Detection of firings posed no problem since 
the discharges are accompanied by audible detonations (Eisner, 
1958). 
A total of 37 captures were witnessed. Five of these involved 
beetles that were held only momentarily by the larvae and released 
without being induced to discharge. Another three involved beetles 
that also failed to discharge, although they were held persistently 
and were eventually killed and eaten. The remaining 29 encounters 
resulted in bombardier firings (Table 1). Eighteen of these ended 
with the beetle escaping: single firings were involved in each case, 
and the ant lion’s head was in all instances exposed when the firing 
occurred. The beetles were released unharmed promptly after the 
discharge. In the other 1 1 encounters in which firings occurred, the 
ant lions had withdrawn the head beneath the sand by the time the 
beetles fired, and although there were sometimes repeated dis- 
charges, only one beetle secured its freedom. The other 10 were 
killed and eaten. It is clear that with their heads submerged, the ant 
lions are much less likely to be repelled by the spray. 
One wonders why the larvae did not consistently withdraw into 
the sand the moment they seized a beetle. We had noted that ant 
lions commonly pull their victims into the sand, but usually only 
when the prey is smaller than the predator itself. The beetles that we 
tested were roughly of the size of the ant lions or even larger, sug- 
gesting that the larvae may simply have lacked the strength to pull 
themselves under while holding such prey. That large insects are 
indeed commonly “feasted upon on the surface” had previously 
been noted (MacLachlan, 1865). 
In three instances when beetles fired at submerged ant lions, the 
latter pulled away from the site of discharge by tunneling backward 
just beneath the sand surface while keeping the beetle in tow. The 
option of burrowing without loss of prey, in a substrate where bur- 
rowing can potentially be quicker than the rate of diffusion of a 
repellent chemical, could prove helpful to ant lions also in their 
capture of chemically protected animals other than bombardier bee- 
tles. Indeed, a substantial fraction of prey items ordinarily available 
to ant lion larvae, including ants, carabid and staphylinid beetles, 
and millipeds, possess dischargeable defensive glands. Interestingly, 
