ARMY ANTS — SCHNEIRLA 401 



to some rhythmic environmental event such as hmar cycles at first 

 seems attractive ; however, in neither of the established Eciton species 

 does the rhythm show any identifiable coincidence with any environ- 

 mental periodicity. Moreover, numerous colonies studied simultane- 

 ously in 1946, 1948, and 1052 passed concurrently through very differ- 

 ent stages of the cycle. This turns our attention to the organic process 

 itself; however, the idea of a purely visceral rhythm or some other 

 regular organic process endogenous to the queen as pacemaker has 

 been weakened by evidence of her close dependence upon the external 

 situation of the colony. 



The nature of the controlling event has been clarified by studies of 

 brood-adult relationship in the colony in relation to changes in the 

 queen (Schneirla, 1949). T\^ien each successive brood approaches 

 larval maturity, the social-stimulative effect of this brood upon work- 

 ers nears its peak. The workers thus energize and carry out some of 

 the greatest daily raids in the nomadic phase, with their byproduct 

 larger and larger quantities of booty in the bivouac. But our his- 

 tological studies show that, at the same time, more and more of the 

 larvae (the largest first of all) soon reduce their feeding to zero as 

 they begin to spin their cocoons. Thus in the last few days of each 

 nomadic phase a food surplus inevitably arises. At this time the 

 qiieen apparently begins to feed voraciously. It is probable that the 

 queen does not overfeed automatically in the presence of plenty, but 

 that she is started and mamtained in the process by an augmented 

 stimulation from the greatly enlivened worker population. Within 

 the last few days of each nomadic phase, the queen's gaster begins to 

 swell increasingly, first of all from a recrudescence of the fat bodies, 

 then from an accelerating maturation of eggs. The overfeeding 

 evidently continues into the statary phase, when, with colony food 

 consumption greatly reduced after enclosure of the brood, smaller 

 raids evidently bring in sufficient food to support the processes until 

 the queen becomes maximally phj^sogastric. These occurrences, which 

 are regular and precise events in every Eciton colou}^, ai-e adequate 

 to prepare the queen for the massive egg-laying operation which be- 

 gins within about one week after the nomadic phase has ended. 



My theory, therefore, is tliat the queen is set off into each new 

 reproductive episode in a feedback fashion by inevitable events in the 

 colony activity cycle. It is a striking fact, however, that the instigat- 

 ing factor itself is an indirect outcome of events depending upon the 

 queen's own function at an earlier point in the cycle. 



Interestingly enough, this complicated set of events works out so 

 that when the colony is emigrating nightly over a lengthy, rough, 

 uneven route of 200, 300, or more meters in length, the queen is in a 

 contracted condition and well able to Avithstand the trials of the 



