386 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1955 



It seems probable that the entire membership of any Eciton all- 

 worker brood is the product of fertilized eggs of essentially equivalent 

 genetic constitution and that the principal difference in the conditions 

 of their development depends upon the amount of food received. 

 Statistical evidence is available indicating a different growth rate for 

 the larvae of minim, intermediate, and major workers. It would 

 appear that the differences are established very early, and probably 

 depend first of all upon the order in which the eggs are laid, with 

 the first-laid eggs having an initial advantage which is thereafter 

 maintained to produce major-worker types, those laid later in the 

 series a disadvantage resulting in successively smaller castes. Major 

 workers are the first to mature as larvae and to spin their cocoons 

 and first to emerge from their cocoons as mature pupae, with degrees 

 of difference to the extreme of the minim workers, which are last in 

 these respects in both hainaimm, and hurchelli. In the development of 

 an Eciton all-worker brood there appears to be a trend toward an 

 extreme emphasis on underfeeding, since the typical curve of body- 

 length variations in the brood population is skewed somewhat toward 

 the smallest sizes. The conditions that influence brood development 

 normally appear to be closely similar for successive worker broods in 

 the same colony or for worker broods of different colonies, judged from 

 our population studies on broods and adult populations. 



It is very probable that the emergence of the major- worker type 

 is a strategically important factor which made possible the evolution 

 of these Eciton species as terrestrial forms. In the deep forest en- 

 vironment that evidently prevailed throughout tropical America in 

 late Jurassic times, a condition very probably congenial to the evolu- 

 tion of surface-living social insects, there also existed vertebrates like 

 the anteaters and the coatis in particular which might have curbed 

 this emergence of the Eciton^ except for the appearance of an effective 

 protective measure such as the major workers. Anteaters, and par- 

 ticularly the terrestrially and arboreally active tamanduas, are very 

 numerous in the Barro Colorado forest, yet never have I seen one of 

 them molesting an Eciton bivouac. The large and delicate brood 

 inside the exposed temporary Eciton nest would be a very palatable 

 find for the tamanduas and no doubt for the coatis also, were it not 

 for these major workers with their formidable bites and stings. 



Notwithstanding their effective equipment, the majors, because of 

 their clumsy oversized foreparts and unwieldy mandibles, play a minor 

 role in the normal work of raiding. As they trudge the trails, they 

 are characteristically shunted to the side by the more agile travelers 

 of smaller castes. These last, and particularly the intermediate 

 workers, are the normal killers of prey and carriers of booty. Far 

 from accompanying as "officers," as Belt (1874) thought, and direct- 



