Abstract
The hypothesis that ergatoid reproductives of ponerine ants are derived from the worker caste is rejected on morphological and theoretical grounds. A syndrome by which ergatoid morphology could arise in the queen caste of some species is then outlined, implicating a combination of dietary specialization on prey that cannot be handled by a single ant, and an inability of foundresses to initiate new colonies alone. Secondary aspects of ergatoidy which may facilitate its development, and the occurrence of ergatoidy in other ponerine species, are explored. It is suggested that this syndrome gave rise to legionary behaviour in the army ants of the Dorylinae and Ecitoninae.