ic8 
Psyche 
[March 
returned to the parent nest. During the same day four additional 
females behaved in a similar pattern. 
Another colony of R. metallica , taken at Ashton Park, Sydney, 
N. S. W., on January 8, 1964, had by May 20 produced two perfect 
females. Both emerged from the parent nest while still alate and 
perished shortly thereafter, having shown no disposition to found new 
colonies. 
This sketchy evidence of the production and behavior of perfect 
females in R. metallica surely needs to be augmented. It does suggest, 
however, that in this species the true female is indeed an evolutionary 
remnant, in which innate behavior patterns governing emergence from 
the nest, the undertaking of at least rudimentary flight, and dealation 
are retained. Behavior leading to colony foundation, however, seems 
feeble, or actually absent, and replaced by a marked tendency to rejoin 
the parent colony and assume a worker-like function within it. It is 
of course impossible yet to be certain that abnormal environmental 
conditions may not have been responsible for this aberrant behavior, 
and the isolated dealate female of R. inornata found by Wheeler may 
be suggestive. But the virtual absence of true brood queens in normal 
wild colonies surely suggests both that colony foundation by isolated 
fertile females, if it occurs at all, must be seldom successful in com- 
petition with the mode observed, and that the longevity of such 
queens, even in the parent nest, can hardly be considerable. They 
indeed appear to be of relict-like character. 
What brings about the occasional production of perfect females? 
In the two colonies where such production has been observed, it 
occurred at the height of the brood-rearing season, and at a period 
when the colony was most active in foraging and was most abundantly 
supplied with food. In each colony, the total brood of females was 
produced over a rather short interval, suggesting that trophic in- 
fluences may have been important in effecting the crossing of a 
“difficult” developmental barrier. Quantitative studies of the phe- 
nomenon, analogous to those of Brian in Myrmica, are needed. 
DISCUSSION AND SUMMARY 
The work reported has shown that workers are regularly produced 
from worker eggs in five species of Rhytidoponera in which alate 
females are rare or unknown. On the other hand, worker-laid ova 
give rise only to males (though prolifically) in R. purpurea, a species 
in which normal queens are the rule. Males are also regularly and 
copiously produced from worker eggs in artificially maintained 
colonies of R. metallica which appear to lack worker-producing 
individuals. These observations suggest that sex determination in 
