The major objective of this study was to determine the relationship between the frequency of use of the language of conditional logic by mathematics teachers and students' conditional reasoning ability. Other objectives were to determine the relationship between student mathematical ability and conditional reasoning ability and the effectiveness of quantifying teacher verbal behavior with respect to time for speaking. Twenty teachers of seventh grade mathematics were audio-taped for five classes each. Those ranked as the top five (TWHR) and the lowest five (TWLR) in the frequency of conditional moves were used to identify student subjects for the testing of hypotheses. The students were administered the "Cornell Conditional Reasoning Test" at the beginning and end of the semester. The TWHR students out-performed TWLR students on the total test, on suggestive content items, on negation items, and on the principle "p only if q, not q : not p." Mathematical ability was found to be highly correlated with the three fallacy principles of denying the antecedent, asserting the consequent, and asserting the converse, and with the total test. No interaction of mathematical ability and teacher frequency of conditional moves was found. The length of time for speaking was not closely related to the length of the transcription. Results indicate that future studies of adventitious learning could be useful to educational practices. (Author/JM)