It is well known that during visually guided prehension movements the
peak velocity of the arm is scaled for object distance (e.g., Gentiluc
ci et al., 1991; Jakobson & Goodale, 1991; Servos, Goodale, & Jakobson
, 1992). Those movements are being directed not to single points in sp
ace, however, but rather to objects with extent. Thus, object distance
must be computed relative to some particular point on the object. Whe
ther that point corresponds to the location of a particular edge, for
example, has not been clearly demonstrated. In the present study, subj
ects (N = 9) were presented with a series of oblong blocks positioned
at different locations. Peak velocity increased with object size for r
eaches in which different-sized objects had their near edges lined up;
in contrast, the peak velocities of reaches directed to objects of di
fferent sizes did not differ when the far edges of the objects were li
ned up. The present study, therefore, provided confirmation that subje
cts calibrate the peak velocity of their reaches relative to the far e
dge of a target object.