Abstract
The author sets up conditions for measuring the "distinctive initial period of generalized learning" which he maintains precedes the responses to localized cues. Formica ants were made to traverse a 6-unit maze in returning to the nest with food. Changes in the chemical stimulations from the maze pathway were made as a means of altering the specific cues. 3 groups of 4 animals each were run 32 trials. The author thinks that the rapid initial reduction in errors and retracings found in the 1st 8 runs indicates a generalized learning not involving specific responses to the individual alleys. Specific avoidance responses to blinds are not acquired in this stage. Rather lengthy discussions of various factors substantiating the theory of a generalized learning phase are given. Tests of statistical significance of the differences mentioned are not utilized.