1962 年 33 巻 1 号 p. 8-24
The present experiment was designed to investigate the effect of motivating conditions on the estimation of time in rats with a single T-maze of corridor type. A diagram of the apparatus used is shown in Fig. 1. The animal was put on the starting platform, and the door 1 and door 2 were opened at the same time. When the animal ran past, door 2 was closed. A compartment was thus formed between door 2 and door 3 in which the animal was confined for the desired length of time. If this interval was 50 seconds the animal was to learn to make a right (or a left) turn which leads to the food at the goal. If the shorter interval of 10 seconds was used the animal was to turn reversely to the left (or the right). Thus the animal was required to stay in the detention compartment for the desired length of time before he was permitted to react to the bifurcation, where he was expected to learn an alternative choice of pathway to the goal according to the length of interval. Non-correction method was used. The time of food intake at the goal was about 25 seconds.
In Experiment I it was shown that the rat could react differentially to different time intervals (10sec vs. 50sec) presented before the reaction was made. After each animal had learned the 10sec vs. 50sec discrimination, the differential between the time intervals was reduced by increasing the shorter interval and decreasing the longer one. Thus, it was shown that the animal could also discriminate 15 seconds from 45 seconds (Table 1 and 2). These results well agreed with that of Heron.
As for a “timing mechanism” which afforded such a time discrimination, the hypothesis was presented that excitatory potential as well as inhibitory one might be summated at the same time in the organism during the time interval spent in the detention compartment. The stimulus change by opening the door (d3) would lead to reduce the inhibitory potential and result in the evocation of reaction. The level of tension thus summated at the time of reaction would afford the internal cue to select a definite pathway to the goal.
In Experiment II, six conditions of food deprivation (1-, 12-, 24-, 36-, 48-, and 60-hr) which would be supposed to affect the level of tension, were chosen. In the test trials, the animal was confined for 30 seconds in the detention compartment. The ratio of frequencies of responses to the goals associated with the shorter and the longer time interval showed that the animal estimated 30 seconds much longer when he was tested after each of the 24-, 48-, and 60-hr food deprivations in comparison with the 1-, 12-, and 30-hr deprivations (Table 3 and Fig. 2). It is necessary to note that test trials were run every day at the same time (i.e. midnight) under all conditions of deprivation throughout the experiments except Experiment VI.
These results in Experiment II did not, however, seem to be a monotonic function of the Food Deprivation Hours (FDH). Then, another way of explanation of the result required to assume the factor of the Maintenace Schedule Cycle (MSC) in addition to the factor of FDH. It is matter of course that the drive would be determined by this MSC which was learned to be a daily cycle with reinforcement at a fixed time for a rather long period.
In Experiment III the change of distribution in frequencies of responses to the right goal as well as to the left one was asked for by decreasing the time interval in the detention compartment under the conditions of 12-, 24-, and 48-hr food deprivation. If the time interval which evoked the equal frequenceis of responses to both goals was to be found it might be defined to be the psychological middle time interval between the shorter and the longer time interval used in the training of the time discrimination. Then it would be supposed that such an estimated middle time interval might shift to the shorter time under the lower drive conditons. The results shown in Table 4 and