1. The purpose of this study was to examine the effect of changing ini
tial stance conditions on the postural response of the cat to horizont
al plane translations of the support surface. Cats were trained to sta
nd, unrestrained, on a moveable force platform. The platform was trans
lated linearly in each of 16 directions in the horizontal plane, with
a ramp-and-hold displacement. The animal's response was quantified in
terms of the forces exerted at the ground. The trajectory of the cente
r of mass (CoM) was computed from the forces. 2. Stance length was var
ied along the longitudinal (sagittal) axis by adjusting the distance b
etween the forepaw and hindpaw force plates. Translation perturbations
of the platform were recorded at stance distances varying from 66 to
110% of the preferred stance distance. 3. Changing stance distance had
a significant effect on the amplitude and direction of the active for
ces exerted by the cat both during quiet stance and during the respons
e to platform translation. At long stance distances, each limb exerted
a force outward, along the diagonals during quiet stance. The respons
e to translation was characterized by an invariance in the direction o
f force exerted against the ground, a strategy that was described prev
iously. At short stance distances, quiet stance forces were more later
ally directed. The force constraint strategy was usually not observed
for the response to translation. Nevertheless, the cats were equally e
ffective at all stance distances in restoring the position of the cent
er of mass after translation of the support surface. 4. There was no d
iscrete boundary between the presence and absence of the force constra
int, suggesting that the strategy for exerting forces against the supp
ort surface is characterized by a continuum of response, from a bimoda
l, or anisotropic distribution of force vectors on the one extreme, to
a uniform, or isotropic distribution on the other. Arguments are deve
loped to suggest that the force constraint strategy may be useful in s
tabilizing the vertebral column during the response to platform transl
ation, to allow linear translation of the CoM rather than bending of t
he trunk.