Fig. 5. Orientation of young pied flycatchers tested under a stationary planetarium sky for 11 h. The black dashed (constant compass course routes) and the black dash-dotted (great circle routes) curves represent the predicted orientations, if the birds could deduce geographical (map) information from the stars and corrected for their apparent displacement. The solid blue line represents the predicted orientations, if the birds could deduce geographical (map) information from the stars and corrected back towards the capture site. The solid green line with a -15° slope indicates the predicted orientation of the birds if they use a time-compensated star compass. The horizontal red line represents the predicted orientation of the birds if they use a time-independent star compass. The group’s mean headings during each 1 h interval are shown on the circular diagrams at the top. Each diagram corresponds to the point directly below it. The numbers in these diagrams give the number of birds that were active and oriented during each 1 h period. Directly below, the headings are plotted as circles with error bars indicating 95 % confidence intervals for the groups’ mean orientation, for easy comparison with the predictions of the different hypotheses. A solid circle indicates that the group-mean-vector was significant. Open circles indicate a non-significant group-mean-vector. The heavy solid black line indicates the best linear regression through the data points and the heavy black dashed line indicates the 95 % confidence interval for the regression. The data suggest that young pied flycatchers use a time-independent star compass, and that they are not able to deduce their geographical position from the rotational phase of the starry sky.