The study reported is directly related to the work of Eastman (1975), which was a continuation of work begun by Carry (1968) and Webb (1971). Eastman obtained a significant interaction when the criterion measure of transfer was regressed on factors of general reasoning and spatial visualization. The Necessary Arithmetic Operations test and the Abstract Reasoning test of the Differential Aptitude Test Battery were used to measure these aptitudes, respectively. Two treatments, one graphical and the other analytical, were used to present methods for finding solution sets of quadratic inequalities. In the graphical treatment, concepts were presented in a symbolic-deductive mode. Eastman observed that factor analytic studies indicated that the measure for spatial visualization used by Carry and Webb (paper folding) had factor loadings on deduction. Conjecturing that the problem of deductive versus inductive presentation of the learning material might have been a confounding variable in the studies by Carry and by Webb, Eastman selected a different measure for spatial visualization, revised the instructional treatments to reflect a figural-inductive mode and a symbolic-deductive mode of presentation, and was successful in isolating an interaction. The present study was an attempt to generalize the results of Eastman to another mathematical content area. Two instructional treatments were written by the authors to present basic logical inference patterns and common logical fallacies. Three inference patterns were presented-modus ponens, modus tollens, and the law of hypothetical syllogism. The two treatments can be characterized as symbolic-deductive and figural-inductive. In the figural-inductive treatment the concept of Euler diagrams was introduced and used to determine whether a conclusion was possible according to one of the three inference patterns. The use of Euler diagrams reflects the figural nature of the treatment. The inductive nature of the treatment was reflected by the fact that the subjects learned to analyze premises by numerous examples without being given-either symbolically or verbally-a statement of the inference patterns. By contrast, the deductive nature of the other treatment was reflected by the fact that the inference patterns were given to subjects as rules prior to any exemplification of the rules, and were given in symbolic form. The symbolic nature of the treat