imitation experiment on February 19, trial 11, is extremely interesting and important. Because of a sudden shower she was in her cage just prior to this trial. Apparently she was quite ready for something to do. I let her see me place a large orange in the box and then proceeded to demonstrate the removal of the snap as usual. When the mechanism was presented to her for work she began instantly and for four minutes labored almost continuously. She poked at the snap, turned, twisted, and pulled it; only once did she go away and then for but a moment. Once I heard the spring bolt of the snap click; a second later again, and in the next two minutes I heard it several times. She was working with extreme concentration, intense interest, and apparently with approaching localization of attack on the spring bolt. As I watched her fumble with the snap I wondered what she would do if the mechanism were two or three times as large, and therefore easier of manipulation with her short, thick fingers. In my notes on these observations I commented: "Why this great accession of interest in an experiment which at various times in the past fortnight she has virtually given up? Is it the cloudy day, with somewhat lower temperature, coupled with higher humidity and possibly a more stimulating atmosphere? Hunger certainly does not account for her unusual interest and activity." Previously she had succeeded, to my great surprise, in the wound chain experiment and had received her entire breakfast as reward. She had then spent an hour in her cage without attention from the experimenter. Possibly this, combined with her unwonted success in the wound chain experiment and the lure of the large orange, conditioned her favorably for the spring snap problem. Still I am puzzled, for I had no reason to expect other than the usual futile fumbling activity; and whereas previously I have felt reasonably certain that success in removing the snap was almost entirely accidental, I am now warned by the clicking of the spring bolt that the mechanism is being attacked in a manner which promises early success. In the very next trial, no. 12 of February 20, she watched

the setting of the apparatus and my demonstration with evident impatience, meanwhile trying, even by reaching out with straws, to get hold of the box. This most unusual behavior undoubtedly resulted from her previous near success. She made it perfectly clear that she wished to work for herself instead of watching me operate the mechanism. With the box within reach she began to work feverishly, pulling, turning, and twisting the snap as in the preceding trial. In fifty-five seconds she removed the snap; precisely how I am unable to say, for her hands entirely obscured the mechanism from view.

From this time on the spring-snap-imitation experiment was repeated to enable me to discover her method of operating the mechanism and the rapidity of acquisition of skill. With reference to the first item, I discovered that she did not specifically attack the movable part of the spring snap but instead rubbed it vigorously with the knuckles of her fingers and thus sooner or later, forcing the bolt back, gave the snap an opportunity to spring out of the staple. This manipulation is little better than fumbling and it offers no evidence of insight. It is amazingly different from that of the chimpanzee, which in my experience either of its own initiative or quickly by aid of imitative copy, learns to operate this type of spring snap directly and efficiently by specific attack on the spring bolt, which is drawn back and the open hook of the snap thereupon skillfully removed from the staple of the hasp by a single movement of the animal's wrist. More striking contrast can scarcely be imagined than between Congo's method of removing the spring snap and that of a comparable specimen of the chim

panzee.

From the record of trial 17 given on February 26 are taken the following typical supporting observations. Congo began work instantly, eagerly, and with evident expectation of success. With the backs of her fingers and chiefly the knuckles of her right hand she pushed the snap about. Quickly she got it off (fifty-six seconds), but it seemingly was rather good luck than insight, for what happened was that she pressed hard

against the snap again and again and finally got it into such
a position that as her downward moving fingers forced the
spring bolt back the snap either sprang or fell out of the staple.
She worked without observation of the relation of bolt to re-
mainder of snap or of the snap itself to the staple. There can
be no doubt that her solution is an original one, and that my
imitative copy instead of influencing her actions specifically
merely stirred her to effort.

Yet another psychologically important result of this exper-
iment is Congo's failure to carry over her interest in the spring
snap or her ability to manipulate it to other situations than
those of the experiment just described. For the purpose of
motion picture photography the spring snap box was arranged
on the platform outside the grill with the snap facing away
from the grill instead of toward it. It was with extreme diffi-
culty that I succeeded in holding Congo's attention on the
mechanism in this position or in getting her to work at it,
and not once in several trials was I able to induce her to remove
the snap. In almost all of the experiments of this report in
which she was moored to a stake during the progress of
observation, a spring snap of the type described was used at
one or both ends of the leash. Yet only once or twice, even
after the series of experiments which I have described, did
I observe her manipulating one of these snaps. Twice she
succeeded by accident in freeing herself from a stake during
an experiment, but it was by a sudden jerk on the chain and
not by manipulation of the snap itself. It therefore appears
that transfer of ability to operate a spring snap did not occur.
This is contrary to the result of the box and pole and pipe and
rod experiment, but in general it is consistent with Congo's
behavior in other experiments. Indeed her behavior in the
spring-snap-imitation experiment gave me the impression that
her ability to solve the problem was very closely bound up with
the particular situation in which she attained success. Obvi-
ously her success was the result of crude, ill-directed efforts,
and lacked accompanying insight. Whether the mechanism

1

was operated in a few seconds, a few minutes, or not at all, depended rather on good luck than good management.

From the observational results presented in the foregoing pages it would appear that my earlier tentative conclusions relative to the narrowly limited mechanical ability or aptitude of Congo are wholly confirmed. Although in the winter of 1927 she succeeded in solving simple mechanical problems in which she had previously failed, this indicates, I believe, rather the combined results of general adaptation to experimental situations and maturation than increased aptitude for dealing with mechanisms or greater insight into their essential characteristics and relations. Also supporting the conclusion that she is mechanically inapt in a truly remarkable degree is the repeatedly confirmed observation that she gave almost no attention to such mechanisms as the locks, hooks, snaps, and like devices which were used about her cage, in the construction of experimental apparatus, and for mooring her to trees. A notable illustrative instance is the following. On the inside of the door leading into her cage was a light hook such as is used on screen doors, which served to hold the door closed after one had entered the cage. It might naturally be supposed that Congo could, if she had so desired, either thrust the door open by force or lift the hook out of its eye, but during my months of observation of the animal and daily use of this device to hold the door shut, I never once saw her manipulate it. Were this not a typical observation, I should report others. It is a clear case of mechanical inapitude. The chimpanzee is highly gifted in this respect, as compared with Congo. I think the same statement might be applied to the orang-outan, but as I have not had opportunity to make strictly comparable experiments and should have to depend upon intimate experimental acquaintance with only one specimen, I prefer to limit the comparison as indicated above.

SUMMARY OF SECTION VI

In this section, six experiments with their results have been summarily described. Of these, five were repetitions of experi

17

ments presented in essentially the same form in 1926, two were presented also in novel variations, and one experiment was first given in 1927.

The situations and Congo's response are as follows:

(1) Diagonal rope experiment, in which Congo promptly succeeded in 1926 and to which on repetition in 1927 she instantly reacted with perfect adaptation. Her response may cause the reader to suspect that this problem is too easy of solution to yield significant results. This inference, according to H. C. Bingham, is not justified by the responses of young children.

6

(2) Hooked rope experiment. When it was initially presented, Congo reacted with gradual adaptation which afforded scant evidence of insight. Finally, she mastered the problem. On repetition the following winter she reacted successfully, although imperfectly, but after a few trials she exhibited a degree of adaptation or perfection of response not previously attained. It would thus seem that either the ripening of her specific experience or her general maturation during the interval between observations facilitated adaptation.

(3) Milk bottle experiment. In its original form, with the bottle containing solid food, this experiment yielded success. initially and also in its repetitions after an interval of a year. The results of repetition were as usual superior to the original performances. Variation A of this experiment, which involved. use of the bottle as milk container, was mastered very slowly and awkwardly by the gorilla. On first presentation, she was entirely incapable of adaptation, and it was only after daily practice for a fortnight that she achieved a fair amount of proficiency in drinking from the bottle. By contrast, the chimpanzee of comparable age and experience, as far as known, adapts not only much more quickly than did Congo but also more perfectly.

(4) Wound chain experiment. To this situation Congo in 1926 responded negatively, although adaptation might reason

"Unpublished observations.