Abstract
The object of the study was to determine whether larval amblystoma are able to establish associations with non-essential signals, i. e., are capable of true learning. This necessarily rules out such effects of training as facilitation of primary reflexes. It was found that the larvae of Amblystoma paroticum could form associations between food taking and the movement of an object in the visual field. In response to movement above them, as, passing the hand back and forth, they would rise from the floor to the top of the aquarium. In the dark room, they did not form associations with the turning on of a constant light, but did learn to respond to intermittent flashes, 52 to the minute. They also learned a simple T-maze of which one end was closed, the other open, allowing the animal to slip down into a dish of water. This constituted the reward. Five animals were given in all 537 trials in 9 experimental series. Each animal in at least one series learned to take the “correct” turn. The results of each series were subjected to statistical analysis, which showed that of the 9 experiments, 2 fall within the probability of the results being due to chance alone, the remaining 7 unquestionably show the effects of learning. The third type of reaction studied was the snap reflex. It was attempted to bring about inhibition of this reflex. For this purpose an earthworm was put into a large glass tube and, in a test, the preparation was held in front of the salamander for 5 minutes and the observer counted the number of strikes made during the period. Two tests a day were given.
Get full access to this article
View all access options for this article.
