Abstract
21 college students served as Ss in an experiment concerned with speed of adding simple numbers. The results partially confirmed the hypothesis of Groen (1967) and Restle (1970) of a mental analog mechanism in which speed of adding two numbers is a linear function of the smaller number but independent of the larger number. However, in contrast to that hypothesis, the results indicate that speed of addition is a linear function of both the larger and smaller numbers in a problem. Finally, significant departures from this general finding were also noted, indicating that on certain well-learned problems the answer may be obtained by simple memory search or associations rather than by an analog computational process.