“…After 15-30 s of constant-angular-velocity rotation, however, the semicircular canal signals have decayed toward zero, except for a small remaining gravity-sensitive modulation. Nevertheless the perception of the actual continuous rotation remains during the stimulation (4,6,35,76). Clearly this is a case where there is a conflict between semicircular canal information, which decays to its rest state and does not indicate any rotation (although strictly speaking it supports any constantvelocity hypothesis) and dynamic otolith information, which is consistent with continuous rotation.…”