280 
Psyche 
[Vol. 93 
under intensive study, these were often marked workers, i.e. active 
on the surface). Males were held by their legs, wings or antennae, 
and resisted fiercely; some managed to struggle free. After releasing 
the males in the vicinity of the nests (30 cm -1 m away), the workers 
ran back into the entrance holes. The uninjured males cleaned their 
antennae and then immediately attempted to enter the nests again. 
On some occasions a number of workers cooperated in the eviction 
of foreign males, and some workers also chased males when they 
came across them outside the nests. Eviction did not always follow 
a male’s entrance, and some marked males remained underground 
for at least 15 minutes. 
Discussion 
In Ophthalmopone berthoudi copulation was never observed 
above ground, and it is inferred that it occurs exclusively inside 
foreign conspecific nests. This is an unusual situation in ants, who 
usually mate some distance from the nests. However, copulation can 
take place in the immediate vicinity of nests in queenright and 
queenless ponerines, and in socially parasitic myrmicines (e.g. Har- 
pagoxenus ; Buschinger and Alio way, 1979). In Rhytidoponera 
chalybaea, in which colonies have either a queen or gamergates, 
large numbers of workers and males mill around nest entrances, 
and males make repeated attempts to mate with workers (Ward, 
1981). However, males also enter nests and may mate with workers 
there. In the queenless R. metallica, workers attract males by the 
release of a pygidial gland pheromone; this distinct behavior (‘sex- 
ual calling’) occurs outside the nest entrances (Holldobler and Has- 
kins, 1977). The pygidial gland has been found in O. berthoudi 
(Villet et al., 1984), and we speculate that if young workers release 
this sex pheromone, they only do so inside the nests and hence 
encounter males underground. Sexual calling was never observed inf 
the field or in the laboratory. 
Direct data are not available on the activities of males inside 
foreign nests, and the occurrence of mating is inferred from the large 
proportion of inseminated workers in nests excavated after the 
period of male activity (Peeters and Crewe, 1985). The existence of 
many gamergates in some nests (up to 108) strongly suggests that 
males copulate more than once; otherwise, such nests would need to 
be visited by larger numbers of males than we observed entering any 
