1981] Holldobler — Communication in Orectognathus 
249 
were designed to localize the anatomical source of a possible trail 
pheromone in O. versicolor. 
Close-up cinematography and photography revealed that many 
Orectognathus workers, when moving back and forth between the 
displaced colony and the nest tube, touched their abdominal tips 
intermittently to the ground, presumably depositing droplets of trail 
pheromone. Three major exocrine glands open at or near the 
abdominal tip of O. versicolor workers: the poison gland and 
Dufour’s gland, both of normal size, and a relatively large pygidial 
gland, which opens between the 6th and 7th abdominal tergites (Fig. 
3). 
Most myrmicine ants have a more or less well developed pygidial 
gland (Holldobler et al. 1976, Holldobler and Engel 1978; Kugler 
1978), but in O. versicolor this gland is more complex than usually 
found in Myrmicinae. It more closely resembles the pygidial gland 
of some ponerine species, for example Pachycondyla laevigata , in 
which it serves as the source of a trail pheromone. The paired 
reservoir sacs (invaginations of the intersegmental membrane be- 
tween the 6th and 7th tergites) are filled with a clear, lightly 
brownish liquid. Several ducts lead from paired clusters of glandular 
cells into the reservoir, penetrating the intersegmental membrane 
(Fig. 4). The cuticle of the 7th tergite has a grooved structure (Fig. 
5), underneath of which is a large glandular epithelium (Fig. 4). 
Orectognathus versicolor workers, when engaged in trail laying 
behavior, usually hold their gaster in an almost vertical position. 
This brings the opening of the pygidial gland very close to the floor 
so that part of the grooved structure on the 7th tergite can be easily 
put in contact with the surface of the ground. 
In the next series of experiments we tested the trail following 
response of O. versicolor to artificial trails drawn with glandular 
secretions of the poison gland, Dufour’s gland and pygidial gland. 
The glands were dissected out of freshly killed workers and for each 
trail test one gland of a kind was crushed on the tip of a hardwook 
applicator stick and smeared once along a 20-cm-long pencil line. 
The trails were made to originate either from the entrance of the 
nest tube or from the periphery of the clustered colony, which had 
previously been shaken out of their nest tube into the test arena. As 
a control a second trail was offered simultaneously which was 
derived either from a droplet of water or from one of the other 
