Sound Shadows

Hagstrum knew that homing pigeons could hear sounds as low as 0.05 hertz, low enough to pick up infrasounds that were down around 0.1 or 0.2 hertz. So he decided to map out what these low-frequency sound waves would have looked like on an average day, and on the days when the pigeons could home correctly from Jersey Hill.

He found that due to atmospheric conditions and local terrain, Jersey Hill normally sits in a sound shadow in relation to the Cornell loft. Little to none of the infrasounds from the area around the loft reached Jersey Hill except on one day when changing wind patterns and temperature inversions permitted.

That happened to match a day when the Cornell pigeons had no problem returning home.

"I could see how the topography was affecting the sound and how the weather was affecting the sound [transmission]," Hagstrum said. "It started to explain all these mysteries."

The terrain between the loft and Jersey Hill, combined with normal atmospheric conditions, bounced infrasounds up and over these areas.

Some infrasound would still reach Castor Hill, but due to nearby hills and valleys, the sound waves approached from the west and southwest, even though the Cornell loft is situated south-southwest of Castor Hill.

Records show that younger, inexperienced pigeons released at Castor Hill would sometimes fly west while older birds headed southwest, presumably following infrasounds from their loft.

Hagstrum's model found that infrasound normally arrived at the Weedsport site from the south. But one day of abnormal weather conditions, combined with a local river valley, resulted in infrasound that arrived at Weedsport from the Cornell loft from the southeast.