1. A total of ninety-six single releases of pigeons was made over distances varying from 48 to 238 miles. The birds were transported under cover which prevented them from seeing the landscape or the sun. It was generally their first release from a point out of sight of the loft.

2. The vanishing points (as observed through field-glasses) from all releases combined showed a significant tendency to be in the homeward half of the circle.

3. The direction of the bird from the release point was observed at various intervals of time from 10 sec after release. There was a statistically significant relationship between the direction at 10 sec and that at 40 sec (or 60 sec.), and the latter directions were closely related to those of disappearance.

4. To test the hypothesis that correct orientation depends on an interval of time for observing the sun's motion, some birds were allowed to see the sun for some time before flight, while others were kept in shadow. Even after the short interval of 10 sec., the no-sun birds showed no less accuracy of orientation than did the sun birds.

5. In two releases the pigeons showed a consistent choice of direction not within the home half of the circle. This suggests the need for caution in drawing conclusions based on few release points, and the desirability of using many release points and taking the average orientation at each as the statistical unit.

6. Our results strongly suggest that there is no difference between the orientation of pigeons that have and pigeons that have not seen the sun before release. This conclusion is, however, tentative until it is confirmed by further experiments using a larger number of release points.

These experiments were aided by a contract between the Office of Naval Research. Department of the Navy, Washington, and Duke University, NR 160-244 and by a grant of funds from the Rockefeller Foundation to the Parapsychology Laboratory of Duke University. The Fulbright Commission provided funds covering travel for R. H. Thouless from Cambridge University to Duke University and return. The Department of Conservation and Development of North Carolina and of Virginia co-operated by permitting the use of forestry look-out towers for releases.