Fig. 2. Experiments 1, 2 and 4. (A) Filled circle, nest entrance; open circle, position of fictive nest when path integrator has not been reset; square, position of feeder or fictive feeder when path integrator has been reset. Black lines, trajectories of ants released close to the nest (N=15); red lines, trajectories of ants reset into the nest (N=17); blue lines, trajectories of ants released in cages into the nest entrance (N=16); grey lines, trajectories of control ants, heading to the feeder (N=40). All runs were cut to the first transgression of a 6 m circle around the nest entrance. (B) First intersections of trajectories with a 6 m circle around the nest entrance. Open circles, first intersections; black, Experiment 1; red, Experiment 2; blue, Experiment 4; grey, control; arrows give the mean direction and the directionality with the radius of the large circle meaning a maximum concentration with r=1 and the marker denoting the border of significance with P=0.01.