15 from 18 and ran about the stake the length of her chain as if impatient with the boxes and her task. So the description continues, without indication of progress toward the solution of the problem. One additional incident is worthy of report. She succeeded in piling 18 upon 15, but they were instably placed and as she mounted the pile the boxes fell. Congo falling with them struck the ground heavily on her back. It was the first time I saw her fall in such a manner as to cause discomfort. She came down with a dull thud, apparently entirely unprepared for the mishap and unable to save herself from an uncomfortable jar. One discovers in this good reason why she should not attempt to jump. No emotional expression was observed, nor did she seem particularily disconcerted.

Obvious diminution of interest in the experiment appeared in trial 3, and neither in it nor trial 4 did indication of progress appear. In trial 5, contrary to the experimenter's intention, she succeeded in obtaining her reward by placing 15 upon 18 on center and reaching from them. She was not able to grasp the food, but by touching it with her finger tips dislodged it.

In order if possible to increase the motivation, in trial 6 a hand of ripe bananas, consisting of about a dozen, was placed one hundred and seven inches from the ground. Just beneath the hand there dangled in horizontal position a single banana. It was one hundred and four inches from the ground. The bait in this trial was unusually conspicuous because of the bright yellow color and the large quantity.

Although at first Congo feigned indifference, it was obvious that the food was almost constantly in mind and that she was searching for ways of getting at it. Two or three times she used 15 and 18, but from them she was unable to reach her objective. Then returning to her task from a brief respite, she mounted 18, pulled 15 upon it, and standing on 15 reached down for 12. From that height it was a difficult task, but grasping 12 firmly she lifted it and, standing erect on 15, raised it toward the banana instead of placing it under

her feet. In this she persisted for several seconds, but without being able to touch the food. Discouraged, she descended to the ground, dropping 12 as she went, and thereupon removed 15 from 18.

This was an excellent attempt, and although Congo failed of success it is obvious that had she placed 12 on 15 she could have obtained her reward. Especially in this trial the observer was made to feel the difficulty for her of transition from the use of two to three boxes. Her behavior suggested the question: Is it more intelligent to strive for the seemingly impossible or to abandon effort? Quite evidently she gives up when effort seems fruitless.

Because the large prospective reward appeared to increase Congo's interest and effort, it was continued in trial 7. She came to the experiment eagerly, and after several minutes of varied activity directly, and with definite intent it seemed, pulled 18 to center, drew 15 upon it, and reaching from the pile almost touched the reward. Encouraged thereby she nevertheless descended to the ground and, knocking 15 from 18, attempted from 18 to pull both 15 and 12 to her. She was able to get both in her hands simultaneously, but she failed to raise them. Her next move was to pull 15 upon 18 and then reaching down from 15 to draw up 12 and place it upon 15. The pyramid thus constituted was perfect and, instantly mounting it, she easily reached the bananas and, grasping four of them, sat down on 12 to devour them. This initial success in the use of three boxes is shown in figure 2 of plate 10.

Chiefly to test the influence of diminished quantity of reward, two bananas, instead of a hand, were offered in trial 8. They were conspicuously located about one hundred and six inches from the ground. Brought to the experiment, Congo went to work immediately. Having assembled the boxes about her, she mounted 18 and tried to pull 12 upon it. Next she turned to 15 and attempted to place the boxes in the order 18, 12, 15. The pile was instable and toppled over. Sitting upon 18 she now pulled 15 to her and after some difficulty placed

it squarely on 18. Then standing on 18 she reached down and grasping 12 pulled it up and placed it securely on 15. Once more, as in trial 7, she had achieved a perfect pyramid. With proper caution she mounted it and standing erect reached the bananas. She then sat down on 12, but finding it uncomfortably small she knocked it out of her way and sat on 15. Approximately two minutes were required for this excellent performance.

Subsequent repetitions of this experiment, trials 9 to 12, were made for purpose of photographic record (see plate 10), demonstration, or to indicate further the motivating value of large versus small quantities of food as prospective reward. The general outcome of the observations was that Congo's ability to use three boxes when she so desired, to obtain suspended food, was convincingly demonstrated. The motivation appeared to be extremely variable. In general, a large quantity of food seemed more likely to attract attention and stimulate effort, but as indicated by the description of trial 8, the customary quantity might yield as prompt and signal success as an extraordinarily large quantity. Although no considerable skill was attained by Congo in the manipulation of three boxes to construct a pyramid, she exhibited tendency to place the boxes appropriately near center and in such relation as to constitute a stable structure. To the experimenter her difficulty in making the transition from solution of the two-box to that of the three-box problem is puzzling. The preliminary trial with three boxes clearly enough indicated tendency to try to use all of the boxes, but in the regular trials it seemed at first as though she were restricted to the use of two boxes, and that the necessity for employment of three constituted a veritably new type of problem. This suspicion is further confirmed by the results of the experiment with four boxes, now to be described.

Problem 19. Box stacking with four boxes. In the situation described for the previous experiment, the box stacking observations were continued with placement of the reward

which necessitated the use of four boxes. A hand of bananas one hundred and fifteen inches above the ground, unless otherwise specified, was offered as reward, and as means of obtaining it, the four cubical boxes previously described were placed six feet off center as follows: 9 to the south, 12 to the north, 15 to the west, and 18 to the east.

In the very first trial Congo eagerly began work, with every indication of expected success. Gathering the boxes about her she began to pile them and in the course of a few minutes she had constructed and proved the uselessness of 18, 9, 12; 18, 12, 9, and 15, 18, 12. In case of the last of these structures she had placed the basal box practically on center, but the pile was instable and promptly fell. Several times she succeeded in arranging the boxes in the order 15, 18, 12, but this construction of course did not suffice to bring her within reach of the reward. The really essential point in the record of this trial is that in the course of the regular thirty-minute interval of work Congo did not attempt to use the fourth box as part of the pyramid, nor indeed was there indication of attention to it.

The second trial exhibited marked diminution of interest and expectation of success. She amused herself by playing with the boxes, throwing them about in the air as though they were toys. Only a small part of her time was given to manipulation of the boxes in relation to the food. Again she showed entire willingness to work with two or three of the boxes in combination, but it seemed not to occur to her that the fourth might be added to the structure.

As my available time for the continuation of this experiment was limited, I decided at this point to attempt to facilitate solution of the four-box problem by eliminating certain of the initial steps in pyramid construction. To this end for trial 3 I pre-arranged the situation by placing 18 on center, 15 upon 18, 12 six feet west of center, and 9 six feet east of center. A hand of bananas was suspended one hundred and fifteen inches from the ground. When brought into this situation Congo paid

J

little attention to boxes or food. Deliberately she knocked 15 from 18; then she proceeded to play with 12, throwing it into the air and moving it about rapidly with hands and feet. Apparently she had definitely given up the problem and was not to be tempted to undertake its solution. Again as I watched her, I wondered whether perhaps in the long run it is good. judgment to refuse to attempt the seemingly impossible. Here Congo, as in certain other experiments, after sizing up the situation, definitely refused to work.

In the next trial, 4, she was extremely rambunctious, but only slightly interested in the problematic situation. It was her delight apparently to knock the boxes about as though irritated by them and impatient of their presence.

The fifth trial found her in optimistic mood. She had just succeeded easily in the slot box experiment and, it seemed, was greatly encouraged by the experience. In preparation for her I had arranged the boxes as follows: 18, 15, 12 were pyramided on center and 9 was six feet to the west of center. Instantly, when given opportunity, she mounted the pyramid, and then lifting 12 she moved it about and readjusted it to suit herself. Then again she climbed upon it and looked toward the food without reaching for it. Next she got down and removed 12 and 15 from 18. Presently she took up 9 and carried it toward the other boxes. Thus she approximated all four boxes to the stake. They were within easy reach and 18 was excellently placed as base of the pyramid. But instead of attempting to construct it she lay down beside the boxes and amused herself with them, her chain, and acorns in the sand about her. The concluding paragraph of my notes on this trial reads: "Evidently box stacking does not generalize in the sense that when the gorilla has learned to use two boxes together she can by virtue of this acquired ability use additional boxes as necessity requires. Apparently there is marked difference between the spread or transfer of training from box to box and from stick to straw, cloth, or any other functional substitute for a stick. It is clear that Congo is not likely to suc