ARMY ANTS — SCHNEIRLA 391 



pod bodies, and items of insect brood — is rushed along in the clutches 

 of workers in the heavily thronged, 3- to 5-cm.-wide columns. The 

 variety of captures, from tarantula femurs to ant eggs, gives evidence 

 of the wide tribute levied by the swarmers upon forest life. 



In his population studies of forest-flora life, Dr. Eliot Williams 

 (1941) was impressed by the capacity of the ecitons to deplete the 

 small life of a raided locality. One day, early in his investigation, 

 when he laid down his quadrat frame on a hill near the Barbour- 

 Lathrop trail, prepared to scoop up the soil and its faunal contents 

 for a biological census, he was perplexed at the paucity of living- 

 things discovered. Other trials in the vicinity turned out similarly. 

 The mystery was explained when we compared notes and I recalled 

 that on the previous day an Eciton praedator swarm had worked that 

 hillside. 



It is difficult to know how extensively the ecitons nmst cut into the 

 population of small forest life. A hurchelU colony is capable of heap- 

 ing more than one measured gallon of booty in just one of its forward- 

 area caches, and by conservative estimate there are 50 colonies of this 

 species on the Island. The inroads of this and the numerous other 

 Eciton species must constitute an important factor in the forest bal- 

 ance, operating against a rapid insect progeneration often cited as 

 threatening to man. 



Although hurchelU swarms typically are larger than those of the 

 subterranean Eciton praedator^ the two species, both common on Barro 

 Colorado Island, are similar in this respect : In both species the swarm 

 advances broadside with alternate "flanking movements" as a rule, 

 wheeling first to one side and then to the other. "Pseudopodic col- 

 umns" extend similarly from the front and flanks of the mass in both, 

 and in both there is a comparable fan-shaped complex of columns in 

 the rear from which a single consolidation column leads to the bivouac. 



Watching the swarm raids, first of all one gains the impression that 

 movement in the mass is utterly confused, with ants here and there 

 clustered in frenzied excitement, fighting and stinging their struggling 

 captives, elsewhere scurrying about or moving in circuitous colmnns 

 which appear to run in every direction. Nevertheless, appropriate 

 analysis soon demonstrates convincingly that the swarm is effectively 

 organized. 



In these respects Eciton behavior represents a somewhat unique prob- 

 lem in social psychology, for no animal below man is able to function 

 away from home in organized groups as sizable and at the same time 

 as regular and complex in function as these swarms. For this reason 

 I have found hurchelU swarm organization worth investigating in some 

 detail (Schneirla, 1940). It can be said that there are no leaders in 

 these swarms except in a very temporary and limited sense, and that 



370930—56 26 



