William A. Roberta The University of Western Ontario
In 1976, Olton and Samuelson published a paper titled "Remembrance of places passed: Spatial memory in rats," in which they introduced a now piece of apparatus, the radial maze, and reported some surprising findings about rats' ability to traverse this maze. The radial maze consisted of a central elevated octagonal platform, from which eight elevated arms radiated outward for about a motor. Rats initially were trained to find food at the end of each arm and then were tested on repeated trials in which food was placed at the and of each arm and the rat was allowed to start at the center platform and roam the maze freely until all eight pieces of food were collected. The primary finding of these experiments was that rats learned to collect all available food pellets, with a very low frequency of repeating entrances into alloys. After 10 to 20 days of practice, it was common for a rat to enter eight different alleys on its first eight choices. Although rats did occasionally repeat alley entrances (defined as an error), the average number of different alloys entered on the first eight choices was between 7.5 and 8.0.
Most researchers, when initially confronted with these findings, suspect that rats' success is based on some more basic process than memory. For example, rats may be using an algorithmic strategy or fixed order of alley entrances, such as always entering the alley adjacent to the one just exited and proceeding around the maze in a clockwise or counterclockwise direction. Olton and Samuelson ( 1976) argued that this was not the case, since their rats chose alloys in random sequences. Later work has shown that, under certain circumstances, rats will choose adjacent alleys ( Roberts & Dale, 1981). However, the important point is that rats do not need to use algorithms in order to choose accurately. Numerous experiments now have shown that rats initially forced to enter a randomly chosen subset of the alloys on a radial maze and then allowed to choose freely will accurately pick out the unentered alloys. Another initially entertained explanation of the rat's prowess on the radial maze was the use of odor cues. Rats might small the presence or absence of food at the and of arms, or they might lay down an odor trail on an entered alley which would serve as a cue not to reenter that alley. Numerous control experiments have ruled out these possibilities. When alloys are rebaited after entrance, rats still choose the unentered alloys ( Olton & Samuelson, 1976; Roberts, 1979), and experiments which have either rearranged alloys