++1+++++++++ |+++ |+++ |++ 1+++++ | |++ | | |++ |++++

The following points in technique and procedure are of importance. The interval of delay was determined, from the instant of covering the food-containing box to the instant of Congo's release, by the use of an additive stop watch. The observer, except during the trials when Congo was absent from the delayed response situation or in delays of more than an hour, with few exceptions, sat continually at his post facing east with his right side toward Congo and the mooring post. Most of the time he was engaged in writing. In the early experiments when the moment for release arrived, Congo's attention was attracted by light clapping of the experimenter's hands thrice; later she gave attention the instant the experimenter moved toward the release stake. In no instance was she released until she was facing in some other direction than toward the correct box, and in almost all of the trials she directly faced the experimenter when released, and at the same time usually held the rope with one hand ready to pull on it and to move toward one of the boxes. Throughout the series of experiments the food-reward offered was relatively large. Sometimes it consisted of an entire meal, but oftener a cup containing nearly a quart of milk, or the solid food of a meal (one or two sweet potatoes and one to three baked bananas) was used separately.

Congo's initial responses surprised me greatly, for she chose correctly with evident ease and assurance, although during the period of delay she made no attempt to hold her orientation and apparently attended to the varied happenings about her as though unaware of the near presence of food. The chances of a correct choice, aside from memory, would presumably be one in four. Actually, she chose incorrectly about once in four times, and even then she almost never failed to locate the correct box on first repetition.

From the first, effort was made to prevent her from using as basis for choice odor of food, peculiarities in the appearance of the correct box aside from location and color, and cues derived from the actions of the experimenter. At no

time was attempt made to conceal the experimenter; instead he practiced rigorous uniformity of action, gazing always directly at the mooring post and observing Congo indirectly as she was released and oriented toward a box. Possible choice by odor was checked in two ways, and it was proved that she was not depending on this cue. One check was the direction of wind, for since the observations were conducted in the open it frequently happened that the wind was so steady and strong that unless it carried from the correct box toward the mooring post Congo could not possibly obtain stimulus. The second check was obtained by surreptitiously placing food with strong odor in each of the three boxes designated as incorrect. In no instance was Congo misled by this condition. Further check on the possible derivation of cues from the observer is supplied by the results of trial 38, which at this point may appropriately be described in detail.

This trial was arranged for the express purpose of discovering whether the observer could determine the choice in the absence of food. The conditions, following a trial with three-hour delay on the previous day, in which Congo chose correctly, were as follows. At 9:15 Congo was brought to the delayed response situation. Food was not placed in any box, but the experimenter selected red as the box to be chosen. After being held at the post for five minutes she was released in accordance with the usual procedure. At the moment she was facing the observer. Instantly she turned to the post, pulled on the rope, wheeled sharply to her left, and facing the white box went directly to it and opened it as though expecting food. Promptly she was returned to the post in preparation for repetition of the experiment as trial 38a. In this case the experimenter selected black as the correct box. In due course Congo was again released facing the observer, but instead of holding her orientation she wheeled to her right, grasped the rope, pulled on it, then turned sharply to her left and went directly to the green box. Once more she was returned to the post, this time in preparation for repeti

tion designated as trial 38b. The experimenter selected white as the correct box. Congo, however, by procedure similar to that in the previous trials, chose the black box. Again she was returned to the stake and this time, in trial 38c, for which the observer selected the green box as correct, she definitely, and evidently on the basis of choice made in advance of release, went directly to the red box.

In no case, despite the selection of a given box by the experimenter, was it chosen by Congo. It is noteworthy that in this series of four trials she chose in order white, green, black, red, proceeding clockwise about the circle. Whether this was accidental or evidence of method cannot be stated.

In view of the evidence from the detailed records of delayed response experiments, and especially the results of various check experiments, I state with assurance that Congo did not depend upon secondary cues for correct response. Instead, the food-box situation, as originally experienced, in some psycho-physiological manner persisted until the moment of response. Reaction may have been to the total situation, to visual configuration, or to position. In all probability the factor of position was of primary importance, and although there was no opportunity to analyze the situation in order to determine the value of different factors, it is not unlikely that the visual qualities of the several food containers were entirely neglected in favor of the positional situation.

Congo's prompt adaptation to the conditions of this experiment and her well-nigh perfect performance throughout the course of the work, doubtless will surprise the reader no less than they did the experimenter. Where discontent and impatience had been expected, she exhibited placidity and willingness to await her opportunity; and whereas attempts to destroy the mechanism and constant interference with its operation had been anticipated, these failed to appear. From the start the task seemed to appeal to her as interesting and worth while, and despite the rapidly lengthening periods of delay she worked quite as well at the end of the series of trials

1

as at the beginning. Usually the motivation was adequate; I think I might safely say optimal. In this experiment, on many of the days of its continuance, Congo obtained at least half of her daily ration and at times almost the whole.

Two trials, separated widely in extent of delay and date, as well as other conditions, will be described somewhat in detail to supplement the general statements which have been made. The first is trial 15. It extended from 12:16 to 12:32 on February 25. The green box was baited with a quart of milk and Congo was held for a fifteen-minute interval. The trial followed a shower, and although the sky had cleared, water dripped from the tree above us. She came to the mooring post from her cage with a rush, evidently keen for the experiment. At five minutes after the beginning of the trial she was noticed to look toward the green box, but most of the time she sat quietly beside the stake amusing herself as usual by playing with the mechanism, with the sand about her, or with objects which she happened to find on the ground. Apparently she was perfectly contented. Most of the time she faced toward the west away from the sun, doubtless because of the direct sunlight. When released she was facing me; a moment previously she had faced the green. As the rope was loosened she started toward the white box and continuing to it lifted the lid and looked in. Then she instantly made for the green box, but she was drawn back to the stake and held there. I thereupon re-baited the green box by taking out the cup of milk, showing it to her, and replacing it in the box. Thus preparation was made for the immediate repetition of this trial which had resulted in incorrect choice. Throughout the second period of delay Congo was quiet, apparently little if any disturbed, or at least giving no evidence of disturbance over her failure to locate the milk in her previous choice. When she was released in trial 15a she was facing the experimenter. Instantly she started toward the north but soon sharply swung around to the green box and opening it took the cup of milk. There could be no doubt about the definiteness and decisiveness of her choice.