To examine the assumption of Chomsky and others about a Language Acquisition Device, the number and the Irregular Forms of exemplars were controlled by using the rule adopted by Palermo & Eberhart (1968). The results showed that 1) as the number of exemplars presented increased, the mean number of trials to reach a criterion of rule acquisition decreased significantly, although no difference was found between the mean number of exemplars, 2) this relationship remained consistent regardless of the insertion of Irregular Forms among the Regular Forms, 3) rule acquisition was not possible on a `basis of insufficient information about the rule structure. Implications of these results were discussed from the viewpoint that a constant amount of experience together with a certain kind of information is required for rule acquisition.