144 
Psyche 
[Vol. 95 
nest, because in this initial phase almost each worker returning from 
the feeding site to the nest appeared to lay a trail. 
The anatomical sources of the trail pheromones 
The fact that recruiting Meranoplus workers frequently tap with 
their abdominal tip the ground, suggests that either one of the major 
sting glands (Poison gland and/or Dufour’s gland) or the rectal 
bladder are involved in trail communication. Each of the three 
organs were dissected out of freshly killed workers, crushed on 
hardwood applicators and rubbed along a line of 25 cm length. Each 
test trail was offered to the ants simultaneously with a control trail, 
which consisted either of water, or of the secretions of one of the 
other organs. Both trails originated in the same circle of 0.5 cm in 
diameter directly at the nest entrance, but diverged in an angle of 
40° . The sides of control and test trails were changed in irregular 
sequence. 
All Meranoplus species investigated have unusually large Dufour’s 
glands. In the species Meranoplus sp. 11 the Dufour’s gland 
stretches almost the whole length of the gaster, whereas the poison 
gland is relatively small, with a pair of bulging glandular filaments 
(Fig. 3). In a series of pilot tests it quickly became apparent that 
secretions from both the poison and Dufour’s glands elicit trail 
following behavior in Meranoplus workers, whereas the contents of 
the rectal bladder had no effect as a trail substance. There was, 
however, a significant difference between the ants’ responses toward 
the two trail substances: Trails drawn with crushed poison glands 
elicited a much stronger initial response than those drawn with 
Dufour’s glands. The poison gland trails, however, were only effec- 
tive for a few minutes, whereas Dufour’s gland trails had an orient- 
ing effect for at least 7 hours and possibly even longer. Table 2 
shows the quantitative results of the trail tests. It is interesting to 
note that considerably more ants followed the Dufour’s gland trail 
when it was presented simultaneously with a poison gland trail, 
probably due to the stimulating effect of the poison gland secretions 
(independent samples t-test: p = 0.067). This is further supported by 
the following results. 
Dufour’s gland secretions alone did not elicit a substantial trail 
following response in Meranoplus workers. But when a crushed 
poison gland on the tip of an applicator stick was offered simul- 
taneously at the nest entrance, a significant number of ants emerged 
