20 
Psyche 
[Vol. 89 
Cerapachys we succeeded, however, in diverting individual ants 
from the raiding column over a distance of at least several centi- 
meters by presenting two applicators in front of them, one contami- 
nated with pygidial gland secretions and the other with water. Both 
applicators were slowly moved away from the columns in opposing 
directions. Of a total of 10 ants tested, 4 responded by following for 
a few centimeters behind the applicator with the pygidial gland 
secretions; no ant followed the control applicator. Although these 
results can be considered only preliminary, they do suggest that 
pygidial gland secretions might be involved in the recruitment pro- 
cess of Cerapachys. This suggestion was further supported by the 
results of a series of experiments in which I offered artificial trails 
drawn with crushed poison glands. I compared the trail following 
response of Cerapachys (within the first two minutes) successively 
either to trails drawn with poison gland secretions only or to poison 
gland trails offered simultaneously with pygidial gland secretions. 
For each kind a total of 6 experiments was carried out. Between 
each test at least one day had elapsed. The following response 
appeared to be stronger to poison gland trails when offered together 
with pygidial gland secretions (5.5 ± 2.9) than to those offered with- 
out pygidial gland secretions (3.0 ± 1.4) (0.1 > p > 0.05; Students 
t-test). Because of lack of material this series could not be extended, 
and thus the results remain only suggestive. 
The two final experiments demonstrated that a trail (10cm long) 
drawn with one crushed poison gland, was still effective as an 
orientation cue several hours after it had been drawn. Using the 
same experimental arrangement described above (Fig. 10), I was 
able to show that emigrating Cerapachys would follow poison gland 
trails, 2 and 6 hours old, when they were offered after the natural 
trail had been covered. On the other hand, crushed poison glands 
introduced into the nest tube after 2 and 6 hours, or poison gland 
trails offered 2 and 6 hours after they had been drawn, did not elicit 
excitement or spontaneous trail following behavior. From these 
results it appears that the poison gland material might contain a 
short lasting stimulating component as well as a longer lasting 
orienting component. 
