In 3 experiments rats were preexposed to the landmarks that surround a Morr
is pool by being placed on a submerged platform within the pool. They were
then required to escape from the pool by swimming to the platform, which wa
s in a location that had not been used during preexposure. Preexposure faci
litated subsequent escape from the pool, provided that the platform was not
moved during preexposure and the relative position of the landmarks to eac
h other remained constant throughout preexposure. In contrast, if during pr
eexposure the platform was moved from session to session (Experiment 1, or
the array of landmarks was altered unsystematically from trial to trial (Ex
periments 2 and 3), then subsequent learning to escape from the pool was di
srupted. These findings suggest that the effects of preexposure to the land
marks in a Morris pool is determined by whether or not they are of relevanc
e for identifying the location of the platform. When they are relevant, the
n subsequent learning is facilitated, but when they are irrelevant, then su
bsequent learning is disrupted.