The centripetal force generated by a rotating space vehicle is a poten
tial source of artificial gravity. Minimizing the cost of such a vehic
le dictates using the smallest radius and highest rotation rate possib
le, but head movements made at high rotation rates generate disorienti
ng, nauseogenic cross-coupled semicircular canal stimulation. Early st
udies suggested 3 or 4 rpm as the highest rate at which humans could a
dapt to this vestibular stimulus. These studies neglected the concomit
ant Coriolis force actions on the head/neck system. We assessed non-ve
stibular Coriolis effects by measuring arm and leg movements made in t
he center of a rotating room turning at 10 rpm and found that movement
endpoints and trajectories are initially deviated; however, subjects
readily adapt with 10-20 additional movements, even without seeing the
ir errors. Equilibrium point theories of motor control errantly predic
t that Coriolis forces will not cause movement endpoint errors so that
subjects will not have to adapt their reaching movements during rotat
ion. Adaptation of movement trajectory acquired during Coriolis force
perturbations of one arm transfers to the unexposed arm but there is n
o intermanual transfer of endpoint adaptation indicating that neuromot
or representations of movement endpoint and trajectory are separable a
nd can adapt independently, also contradictory to equilibrium point th
eories. Touching a surface at the end of reaching movements is require
d for complete endpoint adaptation in darkness but trajectory adapts c
ompletely with or without terminal contact. We have also made the firs
t kinematic measurements of unconstrained head movements during rotati
on, these movements show rapid adaptation to Coriolis force perturbati
ons. Our results point to methods for achieving full compensation for
rotation up to 10 rpm. (C) 1998 Published by Elsevier Science B.V. All
rights reserved.