14 
Psyche 
[March-June 
threads coated with mineral oil were partially effective as guards 
against Monomorium } but induced in the ant simple avoidance 
rather than active withdrawal; a single artificial “repagulum” 
proved totally inadequate for protecting an egg-clump. 
Our field data is most complete and convincing for the ground- 
and low-bush-foraging species Dorymyrmex pyramicus (3 to 4 mm 
long). This is fortunate in the sense that pyramicus appears to be 
quite abundant in areas normally chosen by Ululodes for ovi- 
position; however, Formica sp., Crematogaster sp., and several 
other undetermined species of bush-foraging ants are also potentially 
important egg predators and should be more extensively tested. 
The ant-fences of Ululodes are equally effective against ant 
predators after hatching has occurred and the larvae are sitting in 
a group on their eggshells; in fact, repagula remain full and fluid- 
covered even three weeks after eclosion, while the larvae apparently 
rest on the twigs for no more than ten days. Insect predators or 
other disturbances introduced artificially elicit no defensive reactions 
from the larvae during their gregarious phase. Instead, upset larvae 
simply drop to the ground, suspended at first on silken threads two 
to three centimeters long. 
B. Ascaloptynx furciger 
The repagula of Ascaloptynx furciger seemingly serve no dis- 
cernable protective function. No species of ant tested (see Table 1) 
displayed any pronounced or consistent negative reaction to the 
repagula. However, despite the accessibility of the main egg-mass 
to ant predators, no eggs were broken or carried off in the course of 
these experiments. Ants were observed to probe and bite the eggs, 
but even robust specimens of Formica sp. and Pogonomyrmex sp.* 
seemed unable to pierce the shells. Repagula., too, for the most part 
remained unharmed; in the one case where a Formica sp. crushed 
a repagulum in its jaws, the ant showed no further interest in eating 
the structure or retrieving it for the colony. Admittedly, our sample- 
sizes are quite small and incomplete so far as ant species are con- 
cerned, but the contrast between the reactions of ants to Ululodes 
eggs and repagula compared with their reactions to those of Asca- 
loptynx is striking. 
The repagula of recently hatched Ascaloptynx egg-masses always 
appear to be shriveled or collapsed, almost as if the abortive eggs 
had been consumed by predators. Observations of the activities of 
*Pogonomyrmex is not normally a twig-forager: the egg-mass was thrown 
directly on to the nest mound, at midday. 
