ARMY ANTS — SCHNEIRLA 397 



movement to a site in Lutz Creek, and a gi-eat heap of empty cocoons 

 around the vacated spot indicated that the colony had opened most 

 of its brood at this place. Another clue to a long occupancy of the 

 site was a whitened circular area on the rock exactly where the bivouac 

 base had rested, evidently caused by the protracted action of Eciton 

 chemicals. 



Next day the colony developed three extensive trail systems in a 

 large raid in Lutz Valley. In the evening it emigrated over the 

 principal trail of the central system and before 11 p. m. had mainly 

 completed its new bivouac across Donato trail. This pattern of large 

 daily raids and nightly emigrations was repeated during the next 

 17 days (a nomadic phase of 18 days), in which the colony moved in 

 an irregular course to a site near Barbour-Lathrop 12. In the mean- 

 time a large new brood attained its full-term development. At the 

 last nomadic site the spinning of cocoons by these mature larvae was 

 in full swing, a process greatly assisted by workers laying the larvae 

 on wood detritus and shifting them about at intervals (Beebe, 1919). 

 Next day the first processes of enclosure seemed to have been com- 

 pleted for most of this brood. Now, clustered deeply under a large 

 log on the hill east of Barbour-Lathrop 12, the colony remained for 

 19 days, a "statary" phase which coincided with the pupation of the 

 current brood. Between August 24 and September 11 the bivouac 

 shifted only about 25 centimeters farther beneath the great log. 

 Daily raiding was always much less vigorous than in the nomadic 

 phase, indicated by the use of only one trail system each day rather 

 than two or or three as in the active phase (fig. 3). Each day after 

 dusk the ants were found retreating into their bivouac, generally 

 issuing in some other direction on the following day. In this way 

 the colony "boxed the compass" in its successive raids from the 

 statary site and managed to strike most sections of a circular area 

 approximating 300 meters in radius. 



These facts, first of all, oppose the hypothesis that colonies of this 

 doryline species are forced to emigrate when they have exhausted 

 the prey in a given area. For example, two days after this colony 

 had left its sit« near Barbour-Lathrop 3 after a one-day stand, 

 another hamatuvi colony stopped nearby and remained there 20 days, 

 raiding in the area as it passed through a statary phase. If there was 

 booty enough in this area to hold the second colony for that number 

 of days, it is difficult to understand the moves of our record colony 

 to its last two nomadic sites as based on scarcity of prey. It is simi- 

 larly difficult to believe that the booty supply in the vicinity of the 

 statary site was more plentiful than in other localities at which the 

 latter colony remained only for a one-day stand in its nomadic 

 phase — certainly not seventeen times more plentiful. 



