1982] 
Holldobler — Cerapachys 
19 
Figure 10. Schematical illustration of the experimental arrangement during trail 
tests. The colony was emigrating from nest NI to nest Nil along a natural trail a. 
During the trail tests, the whole arrangement was turned 90° (arrow). The natural 
trail a was covered by a cardboard, on which the test trail (T) and a control trail (C) 
were offered, each deviating from a in an angle of 45°. 
activity and attraction in the ants. I hypothesized therefore that the 
recruiting ant might discharge pygidial gland secretions when it 
exhibited the gaster raising behavior. The pygidial gland pheromone 
might function as an additional recruitment signal by which the 
recruiting ant keeps the raiding party stimulated when leading it to 
the prey colony. In order to test this hypothesis, I tried on four 
different occasions to close the opening of the pygidial gland by 
applying collophonium wax between the 6th and 7th tergites. Unfor- 
tunately these experiments failed; apparently the ants were too dis- 
turbed by the procedure. During two raiding expeditions of 
