On factors influencing the Seward-type latent learning in guppies, Lebistes reticulatis

Abstract

A series of latent learning experiments was performed through the procedure similar to the SEWARD'S (9) and the MAIRE'S ones (4) by making use of guppies, a kind of fishes, and some principles were discussed to explain the latent learning phenomena.
In Experiment I, it was found that 20 trials of one-directional exploration from S to E₁ or E₂ of the maze (see Fig. 1) during 5 days were not sufficient for these subjects to cognize the structure of the maze, while 40 trials were quite sufficient.
In Experiment II, it was postulated that there might be found two types of learning, i. e., conditioning accompanied by reinforcement and cognizing which was independent of reinforcement. The results of this experiment seem to be unexplainable in terms of any ready-made principles except the postulated principle of the learning of cognizing. In brief, cognizing was postulated to be a learning related to the effects of starting-point, while conditioning was to be a learning related to the effects of goal : Cognizing as such may be characterized by the start-gradient.
In Experiment III, the effects of hunger drive on latent learning, cognizing in our term were dealt with. Any indications of the Seward-type of latent learning could not be found under the condition of weak hunger drive.
In Experiment IV, the problem of water temperature as a factor influencing latent learning was investigated : The results of this experiment indicate that it was difficult for guppies to learn latently the maze when the temperature of water was 15°C or 35°C, while the optimum water temperature for the subjects was 25°C. The negative effect of changed water temperature, I suppose, might last at least for 24 hours after water temperature had been changed to the optimum one.