ARMY ANTS — SCHNEIRLA 395 



birds flitting about and calling from underbrush perches; the other 

 is that moment near the end of a nighttime emigration when a pre- 

 viously monotonously plodding column begins to show an excited 

 rushing about of workers, indicating that at last the queen is about to 

 pass. The problem of Eciton nomadism has been a fascinating one 

 to study. 



That the dorylines are characterized by a nomadic existence has 

 been known to science for more than a century, particularly from the 

 observations of Savage (1847) on African driver ants, and Bates 

 (1863), Sumichrast (1868), and others on the American forms. But 

 the frequency of occurrence of the emigrations and their causes or, 

 conversely, the cause of failures to move were quite obscure in the 

 earlier literature. The dominant hypothesis was Vosseler's (1905) 

 assumption that a colony moves on when it has exhausted the supply 

 of prey around its temporary nesting place. The suggestion of 

 Miiller (1886) that a colony needs more food when it has a growing 

 brood, and hence moves about more, although teleological, was much 

 closer to what now appears to be the actual explanation. But nearly 

 50 years later it was possible for Ileape (1931) and Fraenkel (1932) 

 to conclude from the current literature that the doryline ants move on 

 according to food scarcity and hence are a case of irregular "emigTa- 

 tion." From what wo now know, the dorylines emerge as a special 

 pattern of true migration. In my first few weeks of work on the 

 Island in 1932, findings indicated the existence of specific "nomadic" 

 and "statary" conditions in the terrestrial ecitons, related to active 

 and inactive stages respectively in brood development (Sclmeirla, 

 1933). Laboratory and field results increasingly supported a theory 

 of cyclic nomad-statary phases, alternation of these opposite phases 

 depending upon energizing of the worker population by successive 

 broods in their active and passive stages of development (Schneirla, 

 1938). 



Emphasis was first placed upon cross-sectional surveys and upon 

 the study of the major turning points in the postulated cycle. A 

 record of three complete cycles, accomplished by a colony of 

 E. hamatum in somewhat more than four months, is sketched in 

 figure 4. The first complete record of a cycle was obtained in 1936. 

 Soon after my arrival early in the evening on August 5, I was for- 

 tunate enough to find a hamatum colony in bivouac just behind the 

 laboratory, near Snyder-Molino 2, then in deep forest. From their 

 almost concealed cluster located far beneath a large log, the ants were 

 found at dusk pouring out in a single heavily thronged column in 

 which were seen large numbers of newly emerged callow workers. 

 Masses of callow workers were seen in the bivouac walls, and a con- 

 siderable number of cocoons remained unopened. Almost all the 

 cocoons had been opened that night before the colony completed its 



