382 ANNUAL REPORT SMITHSONIAN INSTITUTION, 195 5 



A major behavioral and biological problem in this investigation 

 has been the nature, basis, and adaptive properties of that most 

 peculiar of insect nests, the Eciton bivouac. As mentioned before, 

 Eciton hamatwm and hurchelli may be properly termed "terrestrial" 

 species because their raiding and their bivouacking are carried out 

 mainly above ground. Without any active excavating and without 

 any manipulating of fallen materials, colonies of these species form 

 a domicile with their own bodies. A typical bivouac of Eciton 

 hamatum is shown in plate 1, upper, a cylindrical mass hanging as 

 most of them do from the underside of some projecting surface to the 

 ground. In addition to the sides or under surface of logs, other 

 typical places are the spaces between buttressed tree roots, masses of 

 brush, or even the undercut banks of stream beds or the overhanging 

 edge of a rock. In the "nomadic" phase of its activity cycle, a colony 

 in these species forms a new bivouac of this kind each evening, some- 

 where in the outlying area of the day's raid, and these tend to be open 

 clusters similar to the one represented. 



The characteristic Eciton ability to cluster their bodies, as well as 

 the manner of clustering, depends first of all upon an anatomical char- 

 acteristic, the opposed recurved hooks present on the terminal tarsal 

 segments of the worker's legs. These hooks may be seen in plate 1, 

 lower, and figure 1. The first ants to settle in a new place catch into 

 a rough or soft surface by means of their tarsal hooks, or rather are 

 pulled into this anchored position as newcomers run upon them as 

 they stand and stretch them out in a hanging position. In fact the 

 hooks are really anchored by the added weight of others that have 

 crawled down over the body of the first ant, fixing it in place and soon 

 immobilizing it. 



In the nomadic phase a new bivouac cluster is formed at the end 

 of each further day of raiding, and this is a most interesting event 

 to watch. In the advanced and most complicated stages of raiding 

 in the afternoon, caches of booty tend to be formed at each busy junc- 

 tion of raiding trails, increasing in size as more and more ants are 

 knocked around and forced out of traffic. As darkness comes and 

 raiding ceases, such clusters grow. Events become increasingly com- 

 plex at these places as an exodus develops from the old bivouac, and 

 the crowding and complications of traffic, with the tendency to stop 

 and cluster when progress is impeded in darkness, account for the 

 starting of hanging clusters from elevated ceilings near such places, 

 generally at several of them before one nears completion as the new 

 bivouac. As each new cluster begins, the initial slender hanging 

 threads may become ropes which extend to the ground, depending 

 upon a continued flow of traffic to the place. As the ropes continue 

 to grow they are joined together into a single columnar mass. At 



