1981] 
Carlin — Polymorphism in Orectognathus 
243 
(certain Camponotus species, Wilson 1971; Zacryptocerus texanus, 
Creighton and Gregg 1954). In Zacryptocerus varians, which also 
has modified mandibles useless for fighting, majors use their saucer- 
shaped heads to actively “bulldoze” invaders out (Wilson 1976). 
Major bouncers of O. versicolor are unique in using their mandibles 
to expel invaders without injury. 
To produce a caste so specialized for this form of defense, 
colonies must be under considerable pressure from ant species 
approximately the same size as Solenopsis (it would be hard to 
shoot a larger ant). When bouncing fails, majors do attack in a more 
conventional manner, as is seen in their response to successful 
invaders. (Bouncing might accidentally shoot these further into the 
nest.) It has recently been shown (Holldobler 1982) that majors also 
respond to alarm-recruitment pheromones. 
Other dacetines, including O. clarki, the monomorphic species 
most closely related to O. versicolor , often post “sentinels” at nest 
entrances (Brown 1953; he also observed occasional “retrosalience”, 
an ant striking at a hard surface and shooting itself backward — the 
same motor act as bouncing, but apparently accidental). The O. 
clarki colony, when subjected to size class polyethism analysis, 
revealed a weak division of labor very similar to that of O. 
versicolor minors and medias. It is easy to conceive of these size 
classes as the “primitive caste” (Wilson 1980) typifying the mono- 
morphic ancestor of both species, from which increasing defensive 
specialization turned the sentinels still seen in the former into the 
bouncers of the latter. 
Acknowledgments 
I am very grateful to Dr. B. Holldobler and Dr. E. O. Wilson, for 
the use of materials and the suggestion of methods, for helpful 
advice, for criticizing the manuscript, and for allowing me this entry 
into the insect societies. I would also like to thank Mark Moffett for 
suggestions, assistance, comments on the manuscript and moral 
support, Dr. R. Taylor for identifying the ants, David S. Gladstein 
for help with the repertory size estimations and polyethism curves, 
Dr. Holldobler for the photograph in figure 1, Edward Seling for 
the electron photomicrographs, and Kathleen Horton for the word 
“bouncer”. 
This work was supported in part by grants from the National 
