380 ANNUAL REPORT SMITHSONIAN INSTITUTION, 195 5 



itself in the literature. This problem seemed to represent an ex- 

 cellent opportunity for studying "instinct" in a social insect, and I had 

 been looking for a widespread group pattern evidently influenced in 

 fundamental ways by hereditary mechanisms. 



Because Barro Colorado Island, the Smithsonian Institution's 

 Canal Zone Biological Area, is a relatively undisturbed area of 

 tropical forest in which representative Central American forms such 

 as army ants are accessible for investigation, this situation afforded 

 the best opportunity for carrying out basic phases of the research. 

 In 1932, in my first investigation (Schneirla, 1933) on the Island, the 

 phenomenon was appraised and a tentative solution was formulated. 

 This vf as strengthened into a theory in three subsequent investigations 

 on the Island (Sclmeirla, 1938). Now, after completion of eight 

 investigations on the Island with other studies elsewhere in Central 

 America (Schneirla, 1949; Schneirla and Brown, 1950, 1952), it is 

 possible to offer a systematic account and interpretation of the be- 

 havior and biological characteristics holding not only for the ter- 

 restrial species of Eciton on which these studies have concentrated but 

 probably to an appreciable extent for dorylines in general. 



Gradually, through this work, an appreciation developed of what 

 the "instinct" problem entails for one who would investigate it in the 

 natural situation. The term, after all, is just a word representing our 

 concept for a behavior phenomenon characteristic of the species; 

 "adaptive" in that it fundamentally benefits species survival, and 

 somehow basically dependent upon the genetic nature of the insects 

 themselves. But the solution of a behavior pattern such as the raid- 

 ing and nomadism of army ants as a complex system of activity, and 

 an understanding of its hereditary basis (in this case what has kept 

 the pattern predominantly in operation in this subfamily of ants 

 for more than 65 million years) is not likely to fall readily into our 

 hands. liather it requires us to investigate the entire way of life 

 of the animal, both in its behavioral or psychological and its biologi- 

 cal aspects. This is a suggestion of the problem I have pursued with 

 the army ants as material and Barro Colorado Island as base. 



Probably more than one-tenth of the more than 140 species known 

 for the three genera which constitute almost all of the tropical Ameri- 

 can dorylines are represented by colonies on the Island. Some ten of 

 these species are encountered with considerable frequency, particularly 

 the terrestrial Eciton {Eciton) species, E. hurchelU and E. hamatiun. 

 These species are correctly termed "terrestrial" in that their daily 

 raids are predominantly carried out and their temporary nests are 

 typically situated on the surface of the ground or even in the vegeta- 

 tion. Studying them, we learn something of the characteristics of less 

 accessible subterranean species as well. 



