Abstract
Dips in differential cross sections of inelastic hadronic reactions may be caused by nonsense zeros or by geometrical effects. A direct way of resolving this question is to change the interaction radius in a reaction, while keeping all other parameters fixed, and to see whether the dips move. The "radius" of a virtual spacelike photon has been predicted to change with its mass. If this is verified (e.g., in ρ0 electroproduction), a study of the dip in π0 electroproduction should decide between the two schools of dip mechanisms.