:'lk^- i^M'- fm trV^ 1 J^P", ^"i ,-:^ HARVARD UNIVERSITY. LIBRARY MUSEUM OF COMPARATIVE ZOOLOGY GIFT OF HLkA i i'jiL~IJU^9, '^'^. JUL 29 1919 I I I . r, / BEHAVIOR MONOGRAPHS V;f.:r;i;inof,r'f,'-, Edited by JOHN B. WATSON The Joiins Hopkins University Volume 3 1916-1919 Published at Cambridge, Boston, Mass. HENRY HOLT AND COMPANY 34 West 33d Street, New York G. E. Stechert & Co., London, Paris and Leipzig. Foreign Agents CONTENTS 12. The Mental Life of Monkeys and Apes: A Study of Ideational Be- havior. By Robert M. Yerkes. Pp. iv+145. $1.50. 13. A Study of Perseverance Reactions in Primates and Rodents. By G. V. Hamilton. Pp. iv-(-65. $.75. 14. An Analysis of the Learning Process in the Snail, Physa gyrina Say. By Elizabeth Lockwood Thompson. Pp. iii-(-97. $1.25. 15. The Effect of Length of Blind Alleys on Maze Learning: An Experi- ment on Twenty-four White Rats. By Joseph Peterson. Pp. iii-(-53. $.75. 16. An Experimental Studv of Transfer of Response in the White Rat. By Harry H. Wylie. Pp. iii+66. $1.00. J^PR 6 1956 Behavior Monographs Volume 3, Number I, 1916 Serial Number 12 Edited by JOHN B. WATSON The Johns Hopkins University The Mental Life of Monkeys and Apes: A Study of Ideational Behavior ROBERT M. YERKES Published at Cambridge. Boston, Mass. HENRY HOLT & COMPANY 34 West 33d Street, New York G. E. STECHERT & CO., London. Paris and Leipzig. Foreign AgenU The Journal of Animal Behavior An organ for the publication of studies concerning the instincts, habits and intelligence of organisms The Journal contains a Department of Notes in which appear brief accounts of especially interesting and valuable observations of behavior. Published bi-monthly at Cambridge, Boston, Mass., by Henry Holt and Company, New York. Each volume contains at least 450 pages with plates and text-figures. The subscription price is $3.00 per volume (foreign subscription $3.50) postpaid. Manuscripts, subscriptions, and all business correppondence should be addressed to THE JOURNAL OF ANIMAL BEHAVIOR Emerson Hall, Cambridge, Massachusetts Behavior Monographs For the publication of studies in behavior and intelligence which are too lengthy or too expensive for acceptance by The Journal of Animal Behavior Published at irregular intervals at Cambridge, Boston, Mass., in connection with the Journal of Animal Behavior, by Henry Holt and Company, New York. Each volume contains approximately 450 pages with plates and text-figures. The subscription price is $3.00 per volume (foreign subscription $3.50) postpaid. Monographs may be purchased separately at prices varying with the cost of manufacture. Manuscripts and inquiries concerning terms of publication should be addressed to the Editor of the Behavior Monographs, JOHN B. WATSON, The Johns Hopkins University, Baltimore, Md. Subscription to The Journal of Animal Behavior and the Behavior Monographs should be sent to Emerson Hall, Cambridge, Massa- chusetts. APR 6 1916 Behavior Monographs Volume 3, Number 1, 1916 Serial Number 12 Edited by JOHN B. WATSON The Johns Hopkins University The Mental Life of Monkeys and Apes A Study of Ideational Behavior ROBERT M. YERKES Harvard University WITH SIX PLATES AND FIVE TEXT FIGURES Published at Cambridge, Boston, Mass. HENRY HOLT & COMPANY 34 West 33d Street, New York G. E STECHERT & CO., London, Paris and Leipzig, Foreign Agents CONTENTS Page I. Interests, opportunity and materials 1 II. Observational problems and methods 8 III. Results of multiple-choice experiments: 1. Skirrl, Pithecus irus 21. 2. Sobke, Pithecus rhesus 43 3. Julius, Pongo pygmaiis 63 IV. Results of supplementary tests of ideational behavior: 1. Julius, Pongo pygmaus: Box "stacking experiment 88 Box and pole experiment 99 Draw-in experiment 102 Lock and key test 102 2. Skirrl, Pithecus iriis: Box stacking experiment 104 Box and pole experiment 106 Draw-in experiment 107 Hammer and nail test 108 Other activities 110 3. Sobke, Pithecus rhesus: Box stacking experiment Ill Draw-in experiment 114 Box and pole experiment 114 Other activities 115 V. Miscellaneous observations: 1. Right- and left-handedness 116 2. Instinct and emotion: Maternal instinct 118 Fear 120 Sympathy 121 VI. Historical and critical discussion of ideational behavior in monkeys and apes: 1. Evidences of ideation in monkeys 125 2. Evidences of ideation in apes 128 VII. Provision for the study of the primates and especially the monkeys and anthropoid apes 133 VIII. Bibliography 144 INTERESTS, OPPORTUNITY AND MATERIALS Two strong interests come to expression in this report: the one in the study of the adaptive or ideational behavior of the monkeys and the apes ; and the other in adequate and permanent provision for the thorough study of all aspects of the lives of these animals. The values of these interests and of the tasks which they have led me to undertake are so widely recognized by biologists that I need not pause to justify or define themi. I shall, instead, attempt to make a contribution of fact on the score of each interest. While recognizing that the task of prospecting for an anthro- poid or primate station may in its outcome prove incomparably more important for the biological and sociological sciences and for human welfare than my experimental study of ideational behavior, I give the latter first place in this report, reserving for the concluding section an account of the situation regarding our knowledge of the monkeys, apes, and other primates, and a description of a plan and program for the thorough-going and long continued study of these organisms in a permanent station or research institute. In 1915, a long desired opportunity came to me to devote myself undividedly to tasks which I have designated above as " prospecting " for an anthropoid station and experimenting with monkeys and apes. First of all, the interruption of my academic duties b3^ sabbatical leave gave me free time. But in addition to this freedom for research, I needed animals and equip- ment. These, too, happily, were most satisfactorily provided, as I shall now describe. When in 1913, while already myself engaged in seeking the establishment of an anthropoid station, I heard of the founding of such an institution at Orotava, Tenerife, the Canary Islands, I immediately made inquiries of the founder of the station. Doctor Max Rothmann of Berlin, concerning his plans (Roth- 2 ROBERT M. YERKES mann, 1912). i As a result of our correspondence, I was invited to visit and make use of the facilities of the Orotava station and to consider with its founder the possibility of cooperative work instead of the establishing of an American station. This invitation I gratefully accepted with the expectation of spend- ing the greater part of the year 1915 on the island of Tenerife. But the outbreak of the war rendered my plan impracticable, while at the same time destroying all reasonable ground for hope of profitable cooperation with the Germans in the study of the anthropoids. In August, 1915, Doctor Rothmann died. Pre- sumably, the station still exists at Orotava in the interests of certain psychological and, physiological research. So far as I know, there are as yet no published reports of studies made at this station. It seems from every point of view desirable that American psychologists should, without regard to this initial attempt of the Germans to provide for anthropoid research, further the establishment of a well equipped American station for the study not only of the anthropoid apes but of all of the lower primates. In the early months of the war while I was making every effort to obtain reliable information concerning conditions in the Canary Islands, I received an urgent invitation from my friend and former student, Doctor G. V. Hamilton, to make use of his collection of animals and laboratory at Montecito, Cali- fornia, during my leave of absence from Harvard. This invita- tion I most gladly accepted, and in February, 1915, I established myself in Santa Barbara, in convenient proximity to Doctor Hamilton's private laboratory where for more than six months I was able to work uninterruptedly under nearly ideal conditions. Doctor Hamilton without reserve placed at my disposal his entire collection of animals, laboratory, and equipment, pro- vided innumerable conveniences for my work, and in addition, bore the entire expense of my investigation. I cannot ade- quately thank him for his kindness nor make satisfactory ac- knowledgment here of his generous aid. Thanks to his sympa- thetic interest and to the courtesy of the McCormick family on whose estate the laboratory was located, my work was done under wholly delightful conditions, and with assistance from Ramon Jimenez and Frank Van Den Bergh, Jr., which was in- i See bibliography at end of report. PLATE I Figure 1. — Orang utan, Julius, in grove of live oaks'Mck of the laboratory. Figure 2. — Portrait of same subject. MENTAL LIFE OF MONKEYS AND APES 3 valuable. The former aided me most intelligently in the care of the animals and the construction of apparatus; and the latter, especiall3% was of very real service in connection with many of my experiments. The collection of animals which Doctor Hamilton placed at my disposal consisted of ten monkeys and one orang utan. The monkeys represented either Pithecus rhesus Audebert {Macacus rhesus), Pithecus irus F. Cuvier {Macaciis cynomolgos), or the hybrid of these two species (Elliot, 1913). There were two eunuchs, five males, and three females. All were thoroughly acclimated, having lived in Montecito either from birth or for several years. The orang utan was a young specimen of Pongo pygmcEus Hoppius obtained from a San Francisco dealer in October, 1914 for my use. His age at that time, as judged by his size and the presence of milk teeth, was not more than five years. So far as I could discover, he was a perfectly normal, healthy, and active individual. On June 10, 1915, his weight was thirty-four pounds, his height thirty-two inches, and his chest girt twenty-three inches. On August 18 of the same year, the three measurements were thirty-six and one-half pounds, thirty-three inches, and twenty-five inches. For the major portion of my experimental work, only three of the eleven animals were used. A growing male, P. rhesus monkey, known as Sobke; a mature male, P. irus, called Skirrl; and the young orang utan, which had been named Julius. Plates I and n present these three subjects of my experiments in char- acteristically interesting attitudes. In plate I, figure 1, Julius appears immediately behind the laboratory seated on a rock, against a background of live oaks. This figure gives one an excellent idea of the immediate environment of the laboratory. Figure 2 of the same plate is a portrait of Julius taken in the latter part of August. By reason of the heavy growth of hair, he appeared considerably older as well as larger at this time than when the photograph for figure 1 was taken. In plate II, figure 3, Julius is shown in the woods in the attitude of reaching for a banana, while in figure -i of the same plate he is represented as walking upright in one of the cages. Likenesses of Sobke are presented in figures 5 and 6 of plate II. In the latter of these figures he is shown stretching his mouth, apparently yawning but actually preparing for an attack ROBERT M. YERKES Explanation of Plate II Figure 3. — Orang utan, Julius, reaching for banana. Figure 4. — Julius walking across his cage. Figure 5. — P. rhesus, Sobke. Figure 6. — Sobke stretching his jaws (yawn?) preparatory to a fight. Figure 7. — P. irus, Skirrl. Figure 8. — Skirrl using hammer and nail. Figure 9. — Skirrl using a saw. PLATE II MENTAL LIFE OF MONKEYS AND APES 5 on another monkey behind the wire screen. Figure 7 of this plate indicates Skirrl in an interesting attitude of attention and with an obvious lack of self -consciousness. The same monkey is represented again in figures 8 and 9 of plate II, this time in the act of using hammer and saw. All of the animals except the orang utan had been used more or less for experiments on behavior by Doctor Hamilton, but this prior work in no way interfered with my own investigation. Doctor Hamilton has accumulated a large mass of the most valuable and interesting observations on the behavior of monkeys, and he more thoroughly understands them than any other ob- server of whom I have knowledge. Much to my regret and embarrassment in connection with the present report, he has thus far published only a small portion of his data (Hamilton, 1911, 1914). In his most recent paper on "A study of sexual tendencies in monkeys and baboons," he has given important information concerning several of the monkeys which I have observed. For the convenience of readers who may make use of both his reports and mine, I am designating the animals by the names previously given them by Hamilton. The avail- able and essential information concerning the individuals is presented below. List of animals in collection Skirrl. Pithecus iriis. Adult male. Sobke. P. rhesus. Young adult male. Gertie. P. irits-rhesns. Female. Born November, 1910. Maud. P. rhesus. Young adult female. Jimmy II. P. irns. Adult male. Scotty. P. irtts (?). Adult male. Tiny. P. iriis-rhesus. Female. Born August, 1913. Chatters. P. irus. Adult eunuch. Daddy. P. irus. i\dult eunuch. Mutt. P. irus. Young adult male. Born August, 1911. Julius. Pongo pygmceus. Male. i\ge, 4 years to 5 years. When I arrived in Santa Barbara, Doctor Hamilton was about to remodel, or rather reconstruct, his animal cages and laboratory. This gave us opportunity to adapt both to the special needs of my experiments. The laboratory was finalh' ROBERT M. YERKES PLATE III Figure 10. — Montecito laboratory and cages from the front. Figure 11. — ^Same from the rear. MENTAL LIFE OF MONKEYS AND APES 7 located and built in a grove of live oaks. From the front it is well shown by figure 10 of plate III, and from the rear, by figure 11. Its location was in every way satisfactory for my work, and in addition, the spot proved a delightful one in which to spend one's time. Figure 12 is a ground plan, drawn to scale, of the laboratory and the adjoining cages, showing the relations of the several rooms of the laboratory among themselves and to the nine cages. Although the construction was throughout simple, everything was convenient and so planned as to expedite my ex- perimental work. The large room A, adjoining the cages, was used exclusively for an experimental study of ideational behavior by means of my recently devised multiple-choice method. Addi- tional, and supplementary, experiments were conducted in the large cage Z. Room D served as a store-room and work-shop. The laboratory was forty feet long, twenty-two feet wide, and ten feet to the plate. Each small cage was six, b^^ six, by twelve feet deep, while the large compartment into which each of the smaller cages opened was twenty-four feet long, ten feet wide, and twelve feet deep. II OBSERVATIONAL PROBLEMS AND METHODS My chief observational task in Montecito was the study of ideational behavior, or of such adaptive behavior in monkeys and apes as corresponds to the ideational behavior of man. It was my plan to determine, so far as possible in the time at my disposal, the existence or absence of ideas and the role which they play in the solution of problems by monkeys and apes. I had in mind the behavioristic form of the perennial questions : Do these animals think, do they reason, and if so, what is the nature of these processes as indicated by the characteristics of their adaptive behavior? My work, although obviously preliminary and incomplete, differs from most of the previous studies of the complex behavior of the infrahuman primates in that I relied chiefly upon a spe- cially devised method and applied it systematically over a period of several months. The work was intensive and quantita- tive instead of more or less incidental, casual, and qualitative as has usually been the case. Naturally, during the course of my special study of ideational behavior observations were made relative to various other aspects of the life of my subjects. Such, for example, are my notes on the use of the hands, the instincts, the emotions, and the natural aptitudes of individuals. It is, indeed, impossible to observe any of the primates without noting most interesting and illuminating activities. And although the major portion of my time was spent in hard and monotonous work with my experimental apparatus, I found time each day to get into intimate touch with the free activities of my subjects and to observe their social relations and varied expressions of individuality. As a result of my close acquaintance with this band of primates, I feel more keenly than ever before the neces- sity of taking into account, in connection with all experimental analyses of behavior, the temperamental characteristics, experi- ence, and affective peculiarities of individuals. MENTAL LIFE OF MONKEYS AND APES 9 The light which I have obtained on the general problem of ideation has come, first, through a method which I have rather inaptly named the multiple-choice method, and second, and more incidentally, through a variety of supplementary methods which are described in Section IV of this report. These supple- mentary methods are simple tests of ideation rather than sys- tematic modes of research. They differ from my chief method, among other respects, in that they have been used by various investigators during the past ten or fifteen years. It was not my aim to repeat precisely the observations made by others, but instead to verify some of them, and more especially, to throw additional light on my main problem and to further the analysis of complex behavior. What has been referred to as the multiple-choice method was devised by me three years ago as a means of obtaining strictly comparable objective data concerning the problem-solving abil- ity of various types and conditions of animals. The method was first tried with human subjects in the Psychopathic Hos- pital, Boston, with a crude keyboard apparatus which, how- ever, proved wholly satisfactory as a means of demonstrating its value. It has since been applied by means of mechanisms especially adapted to the structure and activities of the organ- isms, to the study of the behavior of the crow, pig, rat, and ring- dove (Yerkes, 1914; Coburn and Yerkes, 1915; Yerkes and Co- burn, 1915). The method has also been applied with most gratifying results to the study of the characteristics of idea- tional behavior in human defectives, — children, and adults, — and in subjects afflicted with various forms of mental disease. It is at present being tried out as a practical test in connection with vocational guidance and various forms of institutional ex- amination, such as psychopathic hospital and court examinations. As no adequate description of the method has yet been pub- lished to which I can here refer, it will be necessary to present its salient characteristics along with a description of the special form of apparatus which was found suitable for use with monkeys and apes. The method is so planned as to enable the observer to present to any type or condition of organism which he wishes to study any one or all of a series of problems ranging from the extremely simple to the complex and difficultly soluble. All of the prob- 10 ROBERT M. YERKES lems, however, are completely soluble by an organism of excellent ideational ability. For the human subject, the solution of the easiest problem of all requires almost no effort, whereas even moderately difficult problems may require many repetitions of effort and hours or days of application to the task. In each case, the solution of the problem depends upon the percept: on of a certain constant relation among a series of objects to which the subject is required to attend, and respor.d. Such relations are, for example, secondness from one end of the group, middle- ness, simple alternation of ends, or progressive n^ovement by constant steps from one end of a group to the other. It is possible to present such relational problems by means of relatively simple reaction-mechanisms. In their essential features, all of the several types of multiple-choice apparatus designed by the writer and used either by him or by his students and assistants are the same. They consist of a series of pre- cisely similar reaction-devices, any one or all of which may be used in connection with a given observation. These reaction- mechanisms are so chosen as to be suited to the structure and action-system of the animal to be studied. For the human being the mechanism consists of a simple key and the total apparatus is a bank of keys, with such electrical connections as are necessary to enable the observer to obtain satisfactory records of the subject's behavior. Let us suppose the bank of keys, as was actually the case in my first form of apparatus, to consist of twelve separate reaction -mechanism.s; and let us suppose, further, the constant relation (problem) . on the basis of which the subject is required to react to be that of middle- ness. It is evident that in successive trials or experiments the keys must be presented to the subject in odd groups, the possi- bilities being groups of 3, 5, 7, 9, or 11. If for a particular ob- servation the experimenter wishes to present the first three keys at the left end of the keyboard, he pushes back the remain- ing nine keys so that they cannot be operated and requires the subject to select from the group of three keys the one which on being pressed causes a signal to appear. It is of course the clearly understood task of the subject to learn to select the cor- rect key in the group on first trial. This becomes possible only as the subject observes the relation of the key which produces the desired effect to the other keys in the group. On the com- MENTAL LIFE OF MONKEYS AND APES 11 pletion of a subject's reaction to the group of three keys, a group of seven keys at the opposite end of the keyboard may, for example, be presented. Similarly, the subject is required to discover with the minimum number of trials the correct reaction- mechanism. Thus, time after time, the experimenter presents a different group of keys so that the subject in no two successive trials is making use of the same portion of the keyboard. It is therefore impossible for him to react to spatial relations in the ordinary sense and manner, and unless he can perceive and ap- propriately respond to the particular relation which constitutes the only constant characteristic of the correct reaction-mechan- ism for a particular problem, he cannot solve the problem, or at least cannot solve it ideationally and on the basis of a small number of observations or trials. For the various infrahuman animals whose ideational behavior has been studied by means of this method, it has been found eminently satisfactory to use as reaction-mechanisms a series of similar boxes, each with an entrance and an exit door. An incentive to the selection of the right box in a particular test is supplied by food, a small quantity of which is placed in a covered receptacle beyond the exit door of each of the boxes, Each time an animal enters a wrong box, it is punished for its mistake by being confined in that box for a certain period, rang- ing from five seconds to as much as two minutes with various individuals or types of organism. This discourages random, hasty, or careless choices. When the right box is selected, the exit door is immediately raised, thus uncovering the food, which serves as a reward. After eating the food thus provided, the animal, according to training, returns to the starting point and eagerly awaits an opportunity to attempt once more to find the reward which it has learned to expect. With this form of the apparatus, the boxes among which choice may be made are indicated by the raising (opening) of the front door. Since with various birds and mammals the box form of appa- ratus had proved most satisfactory, I planned the primate ap- paratus along similar lines, aiming simply to adapt it to the somewhat different motor equipment and destructive tendencies of the monkeys. I shall now briefly describe this apparatus as it was constructed and used in the Montecito laboratory. The apparatus was built in room A (figure 12), this room 12 ROBERT M. YERKES Explanation of Plate IV Figure 13. — Multiple-choice apparatus, showing observer's bench and writing stand. Figure 14. — Apparatus as seen from observer's bench. Figure 15. — Entrances to multiple-choice boxes as seen from the response-compartment. Figure 16. — Apparatus as seen from the rear, showing exit doors, food receptacles, and covers for same. PLATE IV MENTAL LIFE OF MONKEYS AND APES 13 having been especially planned for it with respect to lighting as well as dimensions and approaches. It was unfortunately impossible to obtain photographs showing the whole of the apparatus, but it is hoped that the four partial views of plate IV may aid the reader who is unfamiliar with previously described similar devices to grasp readily the chief points of construction. In this plate, figure 13 shows the front of the complete apparatus, with the alleyway and door by way of which the experimenter could enter. The investigator's observation-bench and record- table also appear in this figure, together with weighted cords used to operate the various doors and the vertically placed levers by means of which each pair of doors could be locked. Figure 14 is the view presented to the observer as he stood on the bench or observation stand of figure 13 and looked over the entire ap- paratus. Three of the entrance doors are shown at the right of this figure as raised, whereas the remainder of the nine entrance doors of the apparatus are closed. Figure 15 is a view of the entrance doors from below the wire roof of the apparatus. Again, two of the doors are shown as raised, and three additional ones as closed. The rear of the apparatus appears in figure 16, in which some of the exit doors are closed and others open. In the latter case, the food receptacles appear, and on the lower part of the raised doors of the corresponding boxes may be seen metal covers for the food receptacles projecting at right angles to the doors, while on the lower edge of each door is an iron staple used to receive a sliding bar which could be operated from the observer's bench as a means of locking the doors after they had been closed. The space beyond the exit doors was used as an alleyway for the return of the animals to the starting point. It will be necessary at various points in later descriptions to refer to these several figures. But further description of them will be more readily appreciated after a careful examination of the ground plan of the apparatus presented as figure 17. In accordance with the labelling of this figure, the experimenter enters the apparatus room through doorway 16, passes thence through doorways 17 and 10 to the large cage Z, from w^hich he has direct access to the animals and can bring them into the apparatus. The multiple-choice mechanism proper, consisting of nine similar boxes (nine were used instead of twelve as a matter 14 ROBERT M. YERKES of convenience of construction, not because this smaller number is otherwise preferable) is labelled F. These boxes are numbered 1 to 9, beginning at the left. This numbering was adhered to in the recording of results throughout the investigation. The other important portions of the apparatus are the runway D, from which the subject at the experimenter's pleasure could be admitted through doorway 12 to the large response-chamber E; the alleyways G, H, and I, by way of which return to the start- ing point was possible ; the observation bench C, with its approach step B; and the observer's writing table A. In the construction of this large apparatus, it was necessary to make provision for the extremely destructive tendencies of monkeys and anthropoid apes, — hence the apparent cumber- someness of certain portions. It was equally necessary to pro- vide for the protection of the observer and the prevention of escape of the subjects by completely covering the apparatus and alleyways with a heavy wire netting. Each of the eighteen doors of the multiple-choice boxes, and in addition doors 11, 12, and 15 of the runway D, were operated by the observer from his bench C by means of weighted window cords which were carried by pulleys appropriately placed above the apparatus. Each weight was so chosen as to be just suffi- cient to hold its door in position after the experimenter had raised it. For the convenience of the experimenter in the rapid operation of the twenty-one doors, the weights for the doors of runway D were painted gray, those for the entrance doors, white, and those for the exit doors, black. In each entrance door, as is shown in figure 15 of plate IV, a window was cut so that the experimenter might watch the animal after it had entered a given box, and especially note when it left the box after having received its reward. This window was covered with wire netting. No such windows were necessary in the exit doors, but to them were attached heavy galvanized iron flanges which served to cover the food receptacles. One of these flanges is labelled o in figure 17. The food receptacles were provided by boring holes in a 2 by 4 inch timber securely nailed to the floor immediately outside of the exit doors. Into these holes aluminum cups fitted snugly, and the iron flanges, when the doors were closed, fitted so closely over the cups that it was impossible for the animals to obtain food from them. MENTAL LIFE OF MONKEYS AND APES 15 A F ^ 1 F '~' 1 r m , P n _ r— 1 . . (—1 - r ^~^ \ . [— ! - . r-i_. 1 2 1 ! 3 4 i : 14 5 6 F 7 8 9 w W' Figure 17. — Ground plan of multiple-choice apparatus in experiment room A. 1 Scale — 60 A, record stand; C, bench for observer; B, step as approach to C; D, alleyway leadmg to E, response-compartment; F, one of the nine (1-9) similar multiple- choice boxes; G, H, alleyways leading from boxes to starting point at D; L alley- way used by experimenter as approach to rear of apparatus; W, W, windows; P, alleyway; Z, large cage; 16, entrance to room A; 17, entrance to apparatus and thence via 10 to cages; 18, entrance to alleywav I; 11, 15, entrances to D; 12, en- trance to E; 13, entrance door of box 5; 14, exit door of box 5; o, cover for food receptacle. 16 ROBERT M. YERKES As originally constructed, no provision was made in the apparatus for locking the entrance and exit doors of the several boxes when they were closed. But as two of the subjects after a time learned to open the doors from either outside or inside the boxes, it became necessary to introduce locking devices which could be operated by the experimenter from the observation bench. This was readily accomplished by cutting holes in the floor, which permitted an iron staple, screwed to the lower edge of each door, to project through the floor. Through these staples by means of a lever for each of the nine boxes, the observer was able to slide a wooden bar, placed beneath the floor of the room, thus locking or unlocking either the entrance door, the exit door, or both, in the case of any one of the nine boxes. Since figure 17 is drawn to scale, it will be needless to give more than a few of the dimensions of the apparatus. Each of the boxes was 42 inches long, 18 inches wide, and 72 inches deep, inside measurements. The alleys D, I, and H were 24 inches, and G 30 inches wide, by 6 feet deep. The doors of the several boxes were 18 inches wide, by 5 feet high, while those in the alleyways were 24 inches wide by 6 feet high. The response- compartment E of figure 17 was 14 feet 4 inches, by 8 feet, by 6 feet in depth. In order that the apparatus might be used with adult human subjects conveniently, if such use should prove desirable, the depth throughout was made 6 feet, and it was therefore possible for the experimenter to walk about erect in it. The experimental procedure was briefly as follows: A small quantity of food having been placed in each of the food cups and covered by the metal flanges on the exit doors, the experi- menter raised door 11 of figure 17 and then opened door 10 and the door of the cage in which the desired subject was confined. After the latter, in search of food, had entered the runway D, the experimenter lowered door 11 to keep it in this runway, and immediately proceeded to set the reaction-mechanisms for an experiment (trial). Let us suppose that the first setting to be tried involved all of the nine boxes. Each of the entrance doors would therefore be raised. Let us further suppose that the right door is defined as the middle one of the group. With the appa- ratus properly set, the experimenter next raises door 12, thus admitting the animal to the response-compartment E. Any one of the nine boxes may now be entered by it. But if any MENTAL LIFE OF MONKEYS AND APES 17 except number 5, the middle member of the group, be entered, the entrance door is immediately lowered and both the exit and entrance doors locked in position so that the animal is forced to remain in the box for a stated period, say thirty seconds. At the expiration of this time the entrance door is raised and the animal allowed to retrace its steps and make another choice. When the middle box is chosen, the entrance door is lowered and the exit door immediately raised, thus uncovering the food, which the animal eats. As a rule, by my monkeys and ape the reward was eaten in the alleyway G instead of in the multiple- choice box. As soon as the food has been eaten, the exit door is lowered by the experimenter, and the animal returns by way of G and H to runway D, where it awaits its next trial. As rewards, bananas and peanuts were found very satisfac- tory, and although occasionally other foods were supplied in small quantities, they were on the whole less constantly desired than the former. Four problems which had previously been presented to other organisms were in precisely the same form presented to the three primates. These problems may be described, briefly, by defini- tion of the right reaction mechanism, thus: problem 1, the first mechanism at the subject's left; problem 2, the second mechan- ism at the subject's right (that is, from the end of the series at the subject's right); problem 3, alternately, the first mechanism at the subject's left and the first at its right; problem 4, the mid- dle mechanism of the group. It was my intention to present these four problems, in order, to each of the three animals, proceeding with them as rapidly as they were solved. But as it happened, only one of the three subjects got as far as the fourth problem. When observations had to be discontinued, Sobke was well along with the last, or fourth problem; Skirrl was at work at the third problem; and Julius had failed to solve the second problem. For each of the problems, a series of ten different settings of the doors was determined upon in advance. These settings differ from those employed in a similar investigation with the pig only in that the numbering of the doors is reversed. In the present apparatus, the boxes as viewed from the front (entrance) are numbered from the left to the right end, whereas those of the pig apparatus were numbered from the right end to the left end. 18 ROBERT M. YERKES Below are presented for each of the several problems (1) the numbers of the settings presented in series; (2) the numbers of the doors open; (3) the number of doors open in each setting and for the series of ten settings; and (4) the number of the right door. Problem 1. First mechanism at left of group Doors No. of No. of Settings open doors open right door 1 1.2.3 3 1 2 8.9 2 8 3 3.4.5.6.7 5 3 4 7.8.9 3 -.7 5 2.3.4.5.6 5 2 6 6.7.8 3 6 7 5.6.7 3 5 8 4.5.6.7.8 5 4 9 7.8.9 3 7 10 1.2.3 3 1 Total 35 Problem 2. Second mechanism from the right end of group Doors No. of No. of Settings open doors open right door 1 7.8.9 3 8 2 1.2.3.4 4 3 3 2.3.4.5.6.7 6 6 4 1.2.3.4.5.6 6 5 5 4.5.6.7.8 5 7 6 1.2.3 3 2 7 2 3 4 5 4 4 s'.['.'.['////////////.i'.2.3a'.5.6'.7'.S.9.'.'.'.'.'.'.'.'.'.9.'.'.'.'.'.'.'.'.'.'.'.'.'.'.'.'.'.S 9 1.2.3.4 4 3 10 3.4.5.6.7.8 6 7 Total 50 Problem 3. Alternately the first mechanism at the left and the first at the right end of the group Doors No. of No. of Settings open doors open right door 1 5.6.7..- 3 5 2 5.6.7 3 7 3 1.2.3.4.5.6 6 1 4 1.2.3.4.5.6 6 6 5 4.5.6.7.8 5 4 6 4.5.6.7.8 5 8 7 2.3.4.5 4 2 8 2.3.4.5 4 5 9 3.4.5.6.7.8.9 7 3 10 3.4.5.6.7.8.9 7 9 Total 50 MENTAL LIFE OF MONKEYS AND APES 19 Problem 4. Middle mechanism of the group Settings 1.. 2.. 3.. 4.. 5.. 6.. 7.. 8.. 9.. 10.. Doors open .2.3.4.... 6.7.8.9 ,2.3.4.5 ,9.... 6 2.3 ,2.3 ,3.4 ,4.5 ,7.8 8.9 No. of doors open .3. .5. .7. .3. .5. .9. .3. .5. .7. .3. Total 50 No. of right door ..3 ..7 ..4 .6 .5 .2 .4 .6 .7 It was found desirable after a problem had been solved to present a new and radically different series of settings in order to determine to what extent the subject had learned to choose the correct door by memorizing each particular setting. These supplementary observations may be known as control experi- ments, and the settings as supplementary settings. In case of these, as for the original settings, the essential facts are pre- sented in tabular arrangement. Settings for Control Experiments Problem L First at left end Doors No. of No. of Settings open doors open right door 1 2.3.4 3 2 2 6.7.8.9 4 6 3 3.4.5 3 3 4 4.5.6.7.8.9 6 4 5 6.7.8.9 4 6 6 1.2.3.4.5 5 1 7 2.3.4.5.6.7,8 7 2 8 3.4.5.6.7.8 6 3 9 5 6 7 3 5 io;;!;;;;;;;;;;;;;;;l2;3'4'5;6;7^8^9;;;:;;;;;;9;! ;!;!!!!;!;;!!;; !i Problem 2. Second from right end Doors No. of No. of Settings open doors open right door 1 5.6.7.8 4 7 2 2.3.4.5.6 5 5 3 1.2.3.4.5 6 7 8 9.... 9 8 4 5.6.7.... 3 6 5 1.2.3.4.. 4 3 6 4.5.6.... 3 5 7 2.3.4.5.. 4 4 8 1.2.3.... 3 2 9 1.2.3.4.5 6 7 7 6 10 2.3.4.5.6 7 8 9 8 8 20 ROBERT M. YERKES Problem 3. Alternate left and right ends Settings 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. Doors open .5.6... .5.6... .4.5.6. .4.5.6. .1.2.3.4.5.. .1.2.3.4.5.. .2.3.4.5.6.7 .2.3.4.5.6.7 .3.4.5.6.7.8 .3.4.5.6.7.8 8.9 8.9 No. of doors open 9 2 6 6 5 5 6 6 6 6 No. of right door .5 .6 .4 .9 .1 .5 .2 .7 .3 Problem 4. Middle Doors No. of No. of Settings open doors open right door 1 4.5.6.7.8 5 6 2 1.2.3 3 2 3 1.2.3.4,5.6.7.8.9 9 5 4 2.3.4.5.6 5 4 5 6.7.8 3 7 6 3.4.5.6.7.8.9 7 6 7 7.8.9 3 8 8 1.2.3.4.5.6.7 7 4 9 2.3.4 3 3 10 3.4.5.6.7 5 5 It was my aim so far as possible to present to a given subject each day the ten settings under a given problem in order, with- out interruption. If for any reason the series of observations had to be interrupted, it was resumed at the same point subse- quently. Occasionally it was found desirable or necessary to present only five of the series of ten settings in succession and then to interrupt observations for an interval of a few minutes or even several hours. But as a rule it was possible to present the series of ten settings. All things being considered, it proved more satisfactory to give only ten trials a day to each subject. Frequently twenty and rarely thirty trials were given on the same day. In such cases the series of settings was simply re- peated. The only pause between trials was that necessary for resetting the entrance doors and replenishing the food which served as a reward for success. Ill RESULTS OF MULTIPLE-CHOICE EXPERIMENTS 1. Skirri, Pithecus inis Problem \. First at the Left End Systematic work with the multiple-choice apparatus and method described in the previous section was undertaken early in April with Skirri, Sobke, and Julius. The results for each of them are now to be presented with such measure of detail as their importance seems to justify. Skirri had previously been used by Doctor Hamilton in an experimental study of reactive tendencies. He proved so re- markably inefficient in the work that Doctor Hamilton was led to characterize him as feeble-minded, and to recommend him to me for further study because of his mental peculiarities. With me he was from the first frank, aggressive, and inclined to be savage. It was soon possible for me to go into the large cage, Z, with him and allow him to take food from my hand. He was without fear of the experimental apparatus and it proved relatively easy to accustom him to the routine of the experi- ment. Throughout the work he was rather slow, inattentive, and erratic. Beginning on April 7, I sought to acquaint him with the multiple-choice apparatus by allowing him to make trips through the several boxes, with the reward of food each time. Thus, for example, with the entrance and exit doors of box 7 raised, the monkey was allowed to pass into the reaction-compartment E and thence through box 7 to the food cup. As soon as he had finished eating, he was called back to D by the experimenter and, after a few seconds, allowed, similarly, to make a trip by way of one of the other boxes. By reason of this preliminary training he soon came to seek eagerly for the reward of food. On April 10 the apparatus was painted white in order to in- crease the lightness and thus render it easier for the experimenter 22 ROBERT M. YERKES to observe the animal's movements, and when on April 12 Skirrl was again introduced to it for further preliminary training, he utterly refused to enter the boxes, giving every indication of extreme fear of the white floors and even of the sides of the boxes. Finally, the attempts to induce him to enter the boxes had to be given up, and he was returned to his cage unfed. The following da}^ I was equally unsuccessful in either driving or tempting him with food into the apparatus. But on April 14 he was so hungry that he was finally lured in by the use of food. He cautiously approached the boxes and attempted to climb through on the sides instead of walking on the floor. It was perfectly evident that he had an instinctive or an acquired fear of the white surfaces. As the matter was of prime importance for the success of my work, I inquired of Doctor Hamilton, and of the men in charge of the cages, for any incident which might account for this peculiar behavior, and I learned that some three months earlier, while the animal cages were being whitewashed, Skirrl had jumped at one of the laborers who was applying a brush to the framework of one of the cages and had shaken some lime into his eyes. He w^as greatly frightened and enraged. Evidently he experienced extreme discomfort, if not acute pain, and there resulted an association with whiteness which was quite sufficient to cause him to avoid the freshly painted apparatus. Having obtained an adequate explanation of this monkey's peculiar behavior, I proceeded with my efforts to induce him to work smoothly and rapidly, and on April 15, by covering the floor with sawdust, I so diminished the influence of the white- ness as to render the preliminary training fairly satisfactory. At the end of two more days everything was going so ^\'ell that it seemed desirable to begin the regular experiment. On the morning of April 19, Skirrl was introduced to the apparatus and given his first series of ten trials on problem 1. This problem demanded the selection of the first door at the left in any group of open doors. The procedure was as pre- viously described in that the experimenter raised the entrance doors of a certain group of boxes, admitted the animal to the reaction-chamber, punished incorrect choices by confining the animal for thirty seconds, and rewarded correct choices by rais- ing the exit door and thus permitting escape and the obtaining MENTAL LIFE OF MONKEYS AND APES 23 of food. The trials were given in rapid succession, and the total time required for this first series of ten trials was thirty-five min- utes. Skirrl worked faithfully throughout this interval and exhibited no marked discouragement. When confined in a box he showed uneasiness and dissatisfaction by moving about con- stantly, shaking the doors, and trying to raise them in order to escape. For the series of settings used in connection with problem 1, the reader is referred to page 18. In the first setting, the doors numbered 1,2, and 3, were opened. As it happened, the animal when admitted to the reaction-chamber immediately chose box 1. Having received the reward of food, he was called back to D, and doors 8 and 9 having been raised in preparation for the next trial, he was again adm.itted to the reaction-chamber. This time he quickly chose box 9 and was confined therein for thirt\- seconds. On being released, he chose after an interval of four minutes, box 8, thus completing the trial. As it is highly important, not only in connection with the present description of behavior, but also for subsequent com- parison of the reactions of different types of organism in this experiment, to present the detailed records for each trial, tables have been constructed which offer in brief space the essential data for every trial in connection with a given problem. Table 1 contains the results for Skirrl in problem 1. It is crnstructed as follows: the date of a series of trials appears in the first vertical column; the numbers (and number) of the trials for the series or date appear in column 2 ; the following ten columns present respectively the resiilts of the trials for each of the ten settings. Each number, in these results, designates a box en- tered. At the extreme right of the table are three columns which indicate, first, the number of trials in which the right box was chosen first, column headed R; and second, the number of trials in which at least one incorrect choice occurred, column headed W. In the last column, the daily ratio of these first choices appears. Taking the first line of table 1 below the explanatory headings, we note on April 19 ten trials, numbered 1 to 10, were given to Skirrl. In trial 1, with setting 1, he chose correctly the first time, and the record is therefore simply 1. In trial 2, setting 2, he incorrectly chose box 9, the first time. At his next oppor- 24 ROBERT M. YERKES W < Ratio of RtoW CO COOOQOC^OCOLOCO ^m^^ CO coounoLO^mn^cM'* ^r^i^^ '^s ^T;'dTiT:..'r: CO ^ c^ c^ in CO tn c£) '^ iX) CO oj CO ,-icm— i.— i -vf ci CO coLn^LO-^cD'^c^ooo cr.oocr. a-. CO O CO '~' CM CO CO ^ CO — 1 r-i -^ CM CO CM CO — 1 C^ CO ■— 1 CO --H .-1 .-1 CM ^,-(,-i,-( 1.2.3.4.5 6.7.8.9 CO . 00 CO c^oc a-. CD LC S. 8 4.5.6.7.8 ^rj<0006^^0000 00 CD irj CO od CO LO s 1 .S tr^LO t>^ t> in txJt^LOLOtrJLOtCLO'XJLnuoLO muoLOLO in 3» CM CO ;r, 00 C/2 ^ CD CD 00 00 CD t>^ 00 CD CD CD CD CD t> CD CD CO CD CD CD CD LO '^ CO CM i-H CM CO '.3 v2 S. 5 2.3.4.5.6 CM CM CMCDCMCD CM CM CM CM CM CM CM CD LO CD Ln CM CD LO CD -^ CD CD -* LO -* •^ -^ CD LO CM -^ CM CM 00 o CO 3 00 t^ t^ t^ t^ t> 00 CO ^ S. 3 3.4.5.6.7 CO cot> CO COt>-^ CO f- CO t^Lnt> CO CD CO CD LOCOCOLD-^CDCOCOCO'* COUOCO^ CO n 00 00 00 oi cioociooooocxoooooo ocoooboo G". 00 CD ,-H CO CM .-H r-l —H cvi ^ CMr-H CO •-'CO CO CM .-H CO CO -H CM -H -H CO CM .-1 ^ CM— ICO--! CO CM o ooooooocoo oooo —1 CMCO^LOCDC-OOCiO--^ CJCO--rLO .-H CM CO ^ LO CD C^ X Cr. O -H CM CO ^ o .— 1 Q 'Scr- O— iCMCO'*CDC^OOC7lOcc3^fv-i'?io 0.--H CMCMCMCMCMCMCMCMCMCO^'^-' ^"•' CO c MENTAL LIFE OF MONKEYS AND APES 25 tunity, he chose box 8, which was the right one. The record therefore reads 9.8. In trial 3, setting 3, he chose incorrectly twice before finally selecting the right box. The record reads 6.7.3, and so on throughout the ten trials which constitute a series. The summary for this series indicates three right and seven wrong first choices, that is, three cases in which the right box was entered first. The ratio of right to wrong first choices is therefore 1 to 2.33. Since the total number of doors open in the ten settings is thirty-five, and since in each of the ten settings one door is describable as the right door, the probable ratio, apart from the effects of training, of right to wTong first choices is 1 to 2.50. It is evident, therefore, that Skirrl in his first series of trials closely approximated expectation in the num- ber of mistakes. By reading downward in any particular column of results, one obtains a description of the changes in the animal's reaction to a particular setting of the doors. Thus, for instance, in the case of setting 1, which was presented to the animal in trials numbered 1, 11, 21, and so on to 141, it is clear from the records that no definite improvement occurred. But oddly enough, in the case of setting 10, which presented the same group of open doors, almost all of the reactions are right in the lower half of the column. For setting 2, it is evident that mistakes soon disappeared. Comparison of the data of table 1 indicates that the number of correct first choices is inversely proportional to the number of doors in use, while the number of choices made in a given trial is directly proportional to the number of doors in use. During the first week of work on this problem, Skirrl im- proved markedly. His performance was somewhat irregular and unpredictable, but on the whole the experiment seemed fairty satisfactor3\ Cold, cloudy, or rainy days tended to dimin- ish steadiness and to increase the number of mistakes. Simi- larly, absence of hunger was unfavorable to continuous effort to find the right box. The period of confinement, as punishment for wTong choices, was increased from thirty seconds to sixty seconds on April 26. But there is no satisfactory evidence that this favored the solu- tion of the problem. Work on May 4 was interrupted by a severe storm, the noise of which so distracted the monkey that 26 ROBERT M. YERKES he ceased to work. Consequently, observations were inter- rupted on the completion of trial 132, and on May 5, the series was begun with setting 3. On this date, eighteen trials were given in succession, and in only one of them did a mistake occur. Since the ten trials numbered 133 to 142 were correct, Skirrl was considered to have solved problem 1. and systematic train- ing was discontinued. On the following day, as a measure of the extent to which the animal had learned to select the first door at the left no matter what its position or the number of doors in the group presented, a control series was given in which the settings differed from th§ regular series of settings. These supplementary settings are presented at the bottom of table 1 together with the records of reaction in ten trials. Since in only six of these ten control settings was the first choice correct, it is scarcely fair to insist that the animal was reacting on the basis of an ideational solution of the problem. Rather, it would seem that he had learned to react to particular settings. A careful study of all of the data of response, together with notes on the varied behavior of the animal during the experiments, justifies the statement that Skirrl's solution of problem 1 was incomplete and unreliable. It was highly de- pendent upon the particular situation, or even the particular door at the left end of the group, and slightly if at all dependent upon anything comparable to the human idea of first at the left of the group. This particular series of observations has been described and discussed in some detail in order to make the chief points of method clear. It will be needless, hereafter, to refer explicitly to many of the characteristics of reaction or to the important points in the construction of tables which have been mentioned. A graphic representation of Skirrl's learning process in problem 1 is presented in figure 18. The irregularities are most strik- ing, and fairly indicate the erraticness of the animal. The curve is based upon the data in next to the last column of table 1, that is, the column presenting the errors or wrong first choices in each series of trials. Unquestionably, the form of such a curve of learning should be considered in connection with the method or methods of selecting the right box employed by the animal during the course MENTAL LIFE OF MONKEYS AND APES 27 Trials Figure 18. — Error curves of learning for the solution of problem 1 (first box at left end). of experimentation. It appears from an analysis of the behavior of Skirrl in problem 1 that there developed a single definite and persistent method, namely, that of going to one box in the group, and in case it happened to be a wrong one, of choosing, on emergence from it, the next toward the right end of the group, and so on down the line. Having reached the extreme right end, the tendency was to follow the side of the reaction-chamber around to the opposite end and to enter the first box at the left end of the group, which was, of course, the right one. This method appears, with certain slight variations, in approxi- mately ninety per cent of the trials which involved incorrect choices. Thus, in the case of trials 121 to 130, of which eight 28 ROBERT M. YERKES exhibit right first choices, the remaining two exhibit the method described above except that the final member at the right end of the group was in each case omitted. On the whole, Skirrl's behavior in connection with this prob- lem appears to indicate a low order of intelligence. He per- sisted in such stupid acts as that of turning, after emergence from the right box, toward the right and passing into the blind alley I, instead of toward the left, through G and H, to D. In contrast with the other animals, he spent much time before the closed doors of the boxes, instead of going directly to the open doors, some one of which marked the box in which the reward of food could be obtained. It is, moreover, obvious that his re- sponses, as they appear in table 1, are extremely different from those of a human being who is capable of bringing the idea of first at the left end to bear upon the problem in question. Problem 2. Second Jrom the Right End Following the series of control trials of problem 1 given to Skirrl on May 6, a period of four days was allowed during which the animal was merely fed in the boxes each day. This was done in order that he should partially lose the effects of his previous training to choose the first box at the left before being presented with the second problem, the second box from the right. On May 11 regular experimentation was begun with problem 2. Naturally the situation presented unusual difficulties to the monkey because of his previously acquired habit, and on the first day it was possible to give only five trials, in all except the first of which Skirrl had to be aided by the experimenter to find the right box. He persistently, as appears in the first line of records of table 2, entered the first box at the left. The series was continued on May 13, but with very unsatisfactory results, since he apparently had been greatly discouraged by the unusual difficulties previously met. Only four trials could be given, and in these the showing made was very poor. It is noteworth3^ however, that in trials 6, 7, and 8, May 13, there was no marked tendency to choose the first box at the left. Thus quickly had the force of the previous habit been broken. For problem 2, the total number of open doors in the ten set- tings is fifty, as appears from the data on page 18, and as ten of MENTAL LIFE OF MONKEYS AND APES 29 these fifty open doors may be defined as right ones, the expected ratio of right to wrong first choices in the absence of previous training is 1 to 4. The actual ratio for the first series given in problem 2 is 1 to 8, while in the second series it is 0 to 10. On the morning of May 13, work was interrupted in the ninth trial by what seemed at the moment a peculiarly unfortunate accident, but in the light of later developments, an incident most fruitful of valuable results. Skirrl, in trial 9, directly entered box 1. Since this was not the right box, he was punished by being confined in it for ten seconds. While in the box he howled and when the entrance door was raised for him to retrace his steps, he came out with a rush, showing extreme excitement and either rage or fear, I could not be sure which. At intervals he uttered loud cries, which I am now able to identify as cries of alarm. Repeatedly he went to the open door of box 1 and peered in, or peered down through the hole in the floor which received the staple on the door. He refused to enter any one of the open boxes and con- tinued, at intervals of every half minute or so, his cries. For thirty minutes I waited, hoping to be able to induce him to complete the series of trials, but in vain. Although it was obvious that he was eager to escape from the apparatus, he would not enter any of the boxes even when the exit doors were raised. Instead, he gnawed at the door (12 in fig. 17) to the alleyway D and attempted to force his way through, instead of taking the easy and clear route to the alleys, through one of the boxes. His behavior was most surprising and puzzling. Finally, I gave up the attempt to complete the series and returned him to his cage by way of the entrance door to the response-compart- ment E. I then entered the apparatus to seek some explanation of the animal's behavior, and my search was rewarded by the finding of two sharp pointed nails which protruded for an inch or more in the middle of the floor of box 1. My assistant, who had been charged with the task of installing the locks for the several doors, had used nails instead of screws for attaching staples underneath the floor and had neglected to clinch the nails. Skirrl, in the dim light of the box, doubtless stepped upon one of the nails and inflicted a painful, although not serious, injury upon him- self. It was impossible for him to see clearly the source of his 30 ROBERT M. YERKES injury. He was greatly frightened and expressed the emotion most vigorously. His behavior strongly suggested a super- stitious dread of some unseen danger. It may be that the in- stinctive fear of snakes, so strong in monkeys, was partly respon- sible for his response. The first result of this accident was that more than two weeks were lost, for it was impossible, during the next few days, to induce the animal to enter any of the multiple-choice boxes voluntarily. From May 14 to May 24, I labored daily to over- come his newly acquired fear. The usual procedure was to coax him through one box after another by standing at the exit door with some tempting morsel of food. After several days of this treatment, he again trusted himself to the boxes, although very circumspectly and only when both entrance and exit doors were raised. Not until May 24 was it possible to resume regular experimentation, and on that day it was found necessary to indicate the right box by raising the exit door slightly and then immediately lowering it. Trials in which this form of aid was given are indicated in table 2 by a star following the last choice. Gradually, Skirrl regained his confidence in the apparatus and began to work more naturally. For a long time he would not stand punishment, and it was necessary for the experimenter to be very careful in locking the doors, since the sound of the bar sliding beneath the floor often frightened and caused him to quit work. Day after day the tendency to peer through the holes in the floor at the entrance to the boxes rendered it clear that the animal feared some danger from beneath the floor. This behavior was so persistent that much time was wasted in the experiments. On the last day of May, punishment by confinement for ten seconds in wrong boxes was introduced, but since this tended to discourage the monkey, there was substituted for it on June 1 the punishment of forcing him to work his way out of each wrong box by raising the entrance door which had been closed behind him. This he could fairly readily do, and his stay in a box rarely measured more than ten seconds. As a variation in the mode of procedure, confinement for thirty seconds was tried on June 5, but it worked unsatisfactorily and had to be abandoned. During this series, the animal was MENTAL LIFE OF MONKEYS AND APES 31 startled by the sound from one of the sliding bars under the floor, and in the sixth trial he refused to work. As improvement was very slow, varied modes of rewarding and punishing the animal were tried in the hope of discovering a means of facilitating the work. Among the former are the use of banana, grapes, peanuts, and other eagerly sought foods in varying quantities, and in the latter are included periods of confinement ranging from ten seconds to sixty seconds. In the end, confinement of about thirty seconds, combined with a small quantity of food which was much to the monkey's taste, gave most favorable results. All this time Skirrl's attention to the task in hand was seldom good. He was easily diverted and even when extremely hungry, often stopped work in the middle of an early trial, yawned re- peatedly and finally sat down to wait for release from the ap- paratus. The results obtained during the long continued trials with this animal in problem 2 are presented in table 2, which differs from the previously described table, first, in that several of the trials are followed by an asterisk to indicate that aid was given by the experimenter, and second, in that two additional columns, headed, respectively, R and W, are presented. These give the right and wrong first choices for each day, whereas the two columns preceding them give the same data for each series of ten trials. Similarly, the ratio of right to wrong choices is pre- sented for each day in table 2, instead of for each series of ten trials as in table 1. From the results of table 2, several peculiarly interesting facts appear. In the first place the influence of the habit of choosing the first box at the left disappears with surprising suddenness, and in the second place, there are remarkable con- trasts in the results for different settings as they appear in their respective vertical columns. Thus, in the case of setting 1, after the first trial mistakes became relatively infrequent, whereas in setting 6, which involved the same number of doors, mistakes continued to be the rule until nearly a thousand trials had been given. The most likely explanation of this difference is that for some reason the animal avoided box 9. The reactive tendencies, or better, the methods of reaction which manifested themselves during this long series of observations 32 ROBERT M. YERKES Pi ^ o o o oco CO OO CD CO 00 _o O o q pco CO LOp 00 CO 00 *- ^ o 00 c5 ^ '^CM CM .-H-:iJ ^ CM CM rt « O +-> Di '-' (6 ^ a;^' ^ ^^ ^ ^ ^ ^ 00 o 00 00t> C^ CD 00 CO I— ( I— 1 '* T— { Pi I— 1 o CM CM CO CO M CD CD ^ 00 o 00 ooo c^ CD 00 LO 00 CD 00 CD ooc^ Di - o CM CM CO CO -*CM LO CM"* CM '* CM 00 00 * C^ LO t^ tr^ 00 o CO CD 00 CD 00 00 CO Si "O LO t>CDt> CO 06 c-- cbLoo6 c- c/j 00 LO ^ co^ CO ^ t>- 060606 o6'*CO CD in CD LOLO LOCDOCD t> CD CD CD CD 10 CD t> LO ^ 00 05 fC * CO CO CM CO CO — ICO uh cvi ^' -* coco CO COCO CO COCO'-icM CM CMr-i ^ ^' -* ^^ ^ -^^ ^ "^ "^ "^ "^ ■* ^'* 00 CO 00 00 "^o. 00 CO 00 ^^t- l> 00 ^Zj. * t^ ■ • ^^ 00 00 CO ^LOCOOO^CDCO '* 00 c/) dS -^ CD CD 00LO'-i^CD^t>CD 00 CD I>CO '^ •rf LO LO CD CM LO vjOcoCM 00 oot>ai 00 coco CM^ LO f 00 CO CO"* CO o •^ ^OJ -* ^oa CMLO'*Cvi'* c/j CO CO • -^^ -^Tf ^ CO'^ CO ■ cm"^ CO CM CO COLO ^ lOCOTt LOLO '^ LOLO LO LOLO LOCO LO LO LO LO LO CM CM CM CM CM CD CD t^ t> cbcDCD t> c/) lO ^ CD O t^OO CD CO OOLO^'oO c^ C^ ■ CO LOI> CD .-HLC ^ ^ in '-'CD CO CDLO ■^ '^ ^^ * LO . * CMLO LO CO CM c/i CO •— ( ^cDc^i cbcDLO COLO CM CD LOLO '^cjcvi cbcD^LO LO CD CD LO LOLO CD CO LO ^CD t^ LOCD CD CD >o C^LC t> CM c^ LO d^^ CD LO CO CA) CO Cvj CO COLO CD CD LO CO C^CD cvioci'* LO CD CD I> CD CD CD C^CDC^ CO LOt> ^co -^ ^^ CO CO IM PC ♦ CO * CO CO CO ^ '* CO CAJ c^i ^ -* CO'* '^ CO'* CO CM 00 CO '* coco rHc^io^ ^ ^-^ Tp ■^'* ^ •<*'*^ ^ Tt ^ 00 t^ .— 1 90 CJ 00 CAJ t> l>^ * * t> tr- 00 00 00 00 00 00 00 00 CT3 00 00 00 00 00 en o o oo Q oo o 000 0 00 CM CO '^LO CD C-QO o^ 0 ♦""* OJ CO Tj'LO i2 f-H T-H T—H T—K r—\ d t; "3 Z £ ,-H r-H ^ T^ 1-^ rM r-H r-H .— 1 r-H r-H .— I ,—1 ^H 1—1 r— * CM CO -^ LO CDt^ 00 OlO »-H C^l OT* ' ' *""* 1 1 T-H CO OJ >.^^ 0) r3 s2 ^ LO CM coc^ CM CM 00 C^J CM CO P 1— > I— 1 3 CM ' CO 3 T^i MENTAL LIFE OF MONKEYS AND APES 33 '*N y-t ^ ^ ^ ^ ^^ .— I .—1 .—1 ^ ^ -^ ^ 2 O) ^ ^«J 00 c^ lO CD in CO CO n< ^ cOqq 05 c s ■fc p 00 CO LO C£> COlM CVJCO in ^ in •* C^ CO CO t>c^ .—1 .—1 lOOOincOCOOOCDOOOiCXS 00 I> LC a5i> 00 t> OOOOCOt>Ot>OOCOOOtOOO TfLO •a incq.-H.-c^M'^c^^iM IMCO LO ^co (M CO CMCJxj- CO t> 00 t> CO'* * . . CDt^ t> t> CO t>t> 00COt> t> LOCO odt>odt>o6c^t> 00 co' co-^ooc^ co' c^ CO tx:^ i> t> 00 1> c^ c^ o t>t> coco ^ CO CO CO CO CO CO coco CO CO CO CO CO CO CO CO ^ CO Tjl-^-rJ LO o ^- oot>t:^oo . . . * t>- .... COOSCOOO 00 CO LO CO CO C^ to 0000 t>^-CDt> oooooooooot>coc^oot>cdLncTioot>^cbLrit>Lo'>ooooooooaii>~oococbt>cDoooo '^ Tj-LO Tj< CO -* -* '^ -* '^ LO LO LO LO LO in Ln inio Lfiio in" "^ '^ '^ ^^ ^* '^ ^* ^* '^ '^ ■^ ^^ T^ inLriinLriLnLnininLriLriin in in Di cvico6o6o6cDcd in in in CO CO 00 CO 00 l>- in CO o i>-Ln CO ■<* inoi>- i o c C^00t>00CDt>t>I>CO0000 in CD in in in CO coin in in in in in co in in co in '^ in in in in in \s>\si cbcb CD coco CD CO CD CD CD CD CD CD CD CD CO CD CO COCO in CO in in -^ in in in CO CO CO CO CO CO coco COC^I> CD CO CO CD CO t>C^COI>t>t> CO C^CD CO CO ■<*co CO CO C-^ CO CO coco CO CD CO CD CO CO CO CO CD t>i t> oi CD t> t> t> t> t> C^ t> t>i COC-^ COCOrf •^-^'csi CO CO CO CO CO CO ■^ "^ "^ ■^ Tf ■'^j^ coco CO '*COCO CO Csico TfT^'"*' COCOCOCOCOCOCOCOCO '^ 00 CCOOOO OOOOOOOOoioO bo 00 00 00 t^oo 00 C^ 00 t>- 00 00 00 00 00 00 00 CO 0000 b OOCDOOOOOCO CDC^r^OOCTlO'— ICMCO"* ^.— ir- 1,— I^CMCVICMCMCM oo o oo LOCO t>- OOO) CM C>J C^J CM CM oocoooooooo oo cMco^incot>-oocxiO'— icM co^ CO CO CO CO CO CO CO CO '^ '^ ^ '^^ I'^ina t^s OOa 02- 2 '^'^ '^ ^£5 - ^2 - 34 ROBERT M. YERKES ci .9 Pi 1: 9.00 1: 2.33 1: 2.33 1: 2.33 1: 3.00 .—J 1.86 1.50 1.86 00 CO CM 1: 4.00 1: 4.00 1: 9.00 ^ ^ ^ '-' ^ a> ^t>i> i£5 CD CO CM CO t> 2 ^o. Pi -H CDCOCO LO ^ C^ 00 t> CO '^ CO— 1 ^ C; t^ t> t> t^ t> 00 CDC^ CDt>LOt>CD t:^ !>• oi 00 (X t> cr- t> Di •-I CO CO CO CO CO CM ^CO ■^colocotj^ 00 CO -H CM -H 00^ 00 o 00 l> CD LO CO t> t> 00 00 00 t> 00 l>l> o in t>LOt>I>l>- I> 060606060606 t-- c/5 CO CO oococo coco^ -* '^ '^ CO 'd' -^ '^ CO CO-* CO CO CO CO -<:)' ■^' 00 '^ '^ 00 CO CO CO CO CO ^" ^' '^ "^ '^ CO 00 in 00 "* .00 90 CD 00 oo t> 00 00 O"! 00 00 00 LO CD 00 lo'co COLO 00 ooi>^cd'ooooooooi>^ 00 CO '^ LO C£5 00 LO 00 00 CD C--^ 00 00 oi 00 t>^ t>co cot> ^Tf 00 Lnin-rt* CD COLO in c^ urj -^ Tf -^ rf ^ -* '^-^ r^ Tji -^ Tf "* ^ CO LO^ ^-*-^^ CMLn ■<# LO immn in LO in LOLO LOLO LO in LO CO LO LO 10 10 -^ CO CM in CD CO CO CM Co' CM CM CM CM CM CM CM CM CM CM CM CM CO CM CM CM CM CM CM CM CM CO CO CO 00 CO CO CO CO CO CO CO CO CO CO COCO CO CO CO CO 00 CO LO CO 00 t> CD LO LO !>• CD tr- c^ t> t> c^ 00 I>-C- L c C^ L t^C^OCDt>C 0'*^ O^LT DLoa 6 i i 5_t>- 06 06 06 06 CO t^ ' CO CD in "^ CO cj I— I LO LO LO LO LO LO LO LOLO LO LO LO LO LO m CD LO in in in in in LO CD CD CD CD CD CD CD coco CO CO CD CD CD CD CD CD CD CD CO CD CD CO LO -^ CO eg CO CD CD CD CO t^ CD CD C^ t> t> t> t> I>^ t> coco CD t> CD CO CO CD CD CO CD CO CO t>^ CD tD 0^ t>^ t> t> t>^ CD CM CJ CO CO "^ CO CO '^r CO CO CO COCO CO CO 'l^ CO CO CO CO CO cocococo^co CO ^ --r ^ -rr "* ^ ■* 'i"^ ^ ^ ^ Tl< -^ '^ ^ TJ'Tf^'*^ '^ I-H 00 00 0000 c^ t> I> 00 00 00 l>^ 00 00 00 00 oooot>I oot>t>c^oo 00 00 00 O0t>00 00 00t^ 00 No. of trials OOOOOQO LO CD C^ 00 CTi q T-H ■^ -^ -^ -^ T)> LO LO ':3< LO CD i>- oc cr. o ■^ tt -^ Tf -^ '-r LO LOLO ^CM LOLO ooooo ^LOCDC^OO LO LOLO LOLO CO -^ LO CD t^ LO LOLO LOLO in 00 in 000000 0 •-' CM CO f in CD CO CD CO CD CD - T3Ln^c^i-0'X)cotr)f^t>LO^Dt>.t>«r>cocM'*LOrj-'sOoot> 00 c^ ^00 t> '-0 00 00 tr> yi t^ c^ CO t^ c^ o t> c^ t> c^ ;d t>- 1^ 00 00 c^ 00 00 00 00 o 00 1> 00 00 1^ c^ t>. f^ t^ c^ CO COCOCO CMCO OTCOCMCO CO 'sT CO COCO CO -^ CO CO -^ '^ -^ CO CO CO CM CM CO CO CO CO -^ CO "tf CO CO CM •* CO CO CO CO CO CO 00 Tj" 00 '^ lt: 00 00 lOt>- ITi CO CC'X> t>- ^O 00C£)C£) OOCO tr>OCO0OCt>00l:^[>-»OO0OOO<£JoO0OOO0OOOO00O0OOO0O0OOOOO0O 00 c> co^ ■^ Tf -^ rf -^ >* LOCO -^Ti< Tj. UD ■^O ^00 t> t^ T:t" tr^ LOLOLOCX5 o LO c^c^t- ir^'*i:^LOLo't> oococc)t>c£)^c^t>t>t:^c^o6t>t>tr^!^c^c^t>c^i>ooocooi>^f^t:^t>.oct:^ioo6cO!£)o6 t> LOLO LO LOLOLOLOLO LO LOLOLOLOLO LOLOLO LO LOCDCO LOCO LOCC!Cr>COCOLOCOLOLOLOLOCOUOLOLOCOCOCOCOCOLOLOCOCO'.OLOLO CO LO CO CO CO CO CO CO CO CO ^CO COCOCOCOCOCOCOCO COCOCO CO c^cot> CO tr^ CO CO ^~ CO c^ CO t^- ^-^ CO CO i> CO LO t>^ CO !>• t> i> c^ i>- 1^ t~- !>• CO t>^ [>- -^ co t> CO CM CO CO CO CO CO COCOCO coco COCOCOCOCO^CMCOCOCMCOCOCOCO-^-^'^COCO-rfCOC'tCOCOCOCOCO CO CO CO 000000X00 00 00 oot^oo 00 00 coi>-oot:^c^ooxoot>coc^ooooooxc^xcoococoot^oooooooooo co cc PSS o^ ooooooooooooooooooooooooooo o o KrfSS or^ cMcO'^i'Locot^xciO^Dco^LocDc^xcr. o^c^jco^uocotr^oo a--, o COCOCO i>t^ t>i>-i>.t>t>c^c-^i>.xxxxxxxxxxo^. cr. c^. aioi CT-CT. CTiC) ai o JS?^°2 cr^p •— icMcoTiHLOcct^XG%0'-''Mco'*Locoi:^xcT, o-HC^ico^Locor^ x a-> COCOCO co^ t:^t>t>ot>t>i:^c^c^xxxxxxxxxxaia;c>cnc^a5C)cji o^ ct> OCMCO ^-^ ^L0,C0^t:^-CX>--O^^,C^1- — i'-^— I - -H ~- r-i - ^- -^ - -H - " CM- CM" CM - ^ CO ^ ^ . 36 ROBERT M. YERKES ffl < o ^ CM CDC^ c-co o oc- CO 00 00 CD CD^ o oco -* rt o2 CD ^<5 CDO d r-ld CD K (v; ^ ;-;^ ^^ i-H ^^ ^ ^ 03 22^ ^CO O lO-^ CO Di ;i3 oco CDt> O LOCO c- ^ CO CD 00 LO^ ■*co o LOT:t< co Di t^TtCM LOCO cot- o LOCO c- 00 o- -^ s S CO ode- LOCr. ^ CD 00 '^ 00 « 3 CO t>.o6LOt> coc^ CM CD CO LOt^OO 00 00 CD 00 t> c- 00 00 00 CD ^ Ol co CM . ^LO C^ <^ w CO-^CO xt* c^i (73 =^ r-H CM CJ CO CM LOCO CO CM CM -^ CM CM coco co c- 1> CD ^ — '^- -' 00 ^O. 1> 00^00 CD CO •<^^ LO * ■O CO CMLO c/5 fo CM T?*^ LO ^ CO cm' ^ nJlocolo ^Lnin CO'* LO ^ ^ LO^COCMLOCM 1 t-4 Ol, C «o ^D 1^ CO S r4 in .>- C/) ^ CM CM ^ LO a^ CM CM 00 CM CM CO CM CM LOCO LO ^j °9 LO '^ 'i coc^ CDC- CM CO a t>C^CDLri t^oo LOOO t>CD^t> CM CO CO (/) ^^ -^^ 1 CD CD LOLO CD CD LO ^ ^ CO CO CM CD C— cvi CMLO ^CO CMLO LO LOCO CD CD CM ^'-* LO CO CD CD t> CD vO t>^ CD LO f^ LO CO CO T^ C75 C/) ^ CO CDCD^ ^ cooc CMC-; 00 CJ C^LOLO CO CD COLO CD r-;cD 00 00 c- '^ CD LO LO ■^^ ro COLO CO CA) <^j CO CO ojco CM CO CM CO COCO coco CO CM coco CM 00 —1 '^ 00 (> c- ^ c^ 00 00 CD LO LO OOOOC^ t-oo OOM 00 CDC- C- ooo oo OO o oo o ^CMCO "^LO CDC- 00 — ICM 00 Cfi OOO oo OO o ©"rjlS z°| .-H 1-H .-H rH r-H J^ t-H jLi r-H ri ^ O »—' C^3 co^ LOCO C- f-H CM ooo OO CO o r-( .— 1 .— 1 ^H ■— 1 '"""' '"' JJ • « oj - CO c^a LO CO CO '^ CO ^ CO cr. t> 00 LO i>- I> CO c^ Di Cv] - CO c^a LO co CO T7*1 CO 00 (A C/2 00 CM CO LOO . * CD GOO) coo-. 00 CO CD"*' t>cb 00 t>^ coco CO CO 00 . * . . * COLOCT. OOCTlCOCOaiCTJ t>I>C~-CDt^cbLCCCo6 a-. 00 00 00 CO '^ 00 i>i CO CO CO CO C2 CO t^co CO ir^ CO t>- 00 -^t> 00-* oo Oi LO 00 O^ LO c. 00 00 00 00 CO CD OCOC . . * t^^CDOC^ LO CD 00 LO LO 00 CD r^ LO CD CD C^ t> t^ •* t>oooot>c^ 00 CO 00 LO CD'* LOCO CDLO COr-i Tt'cDr-HcblO'-HLOCD.-iCOC^LO'-i LOc^icb^cDLOc-icoco^cDcvico CDOOCOLOxfCMCDLO'^CDLOcb'* i^i r-l CM CO CO •^ MENTAL LIFE OF MONKEYS AND APES 43 2. Sobke, Pithecus rhesus Problem 1. First at the Left End Sobke was somewhat afraid of the experimenter when the investigation was undertaken, and instead of willingly coming out of his cage when the door was raised, he often had to be coaxed out and lured into the apparatus with food. Whereas Skirrl was frank and rather aggressive, Sobke was stealthy in his movements, furtive, and evidently suspicious of the experi- menter as well as of the apparatus. He was perfectly safe to approach, but would not permit am^one to touch him. After a few days, he began to take food from the hands of the experi- menter. Preliminary work to acquaint this monkey with the routine of the experiment was begun on April 13. As in the case of Skirrl, he was lured into the apparatus and was taught the route through the boxes to the starting point by being allowed to obtain food once each day in each of the nine boxes. The procedure was simple. The entrance door and the exit door of a particular box were raised and the animal admitted to the reaction-compartment and permitted to pass through the box whose doors stood open, take its food, and return to the starting point. Sobke verv' quickh^ learned the route perfectly and came to work steadily and rapidly. After five days of prelim- inary work of this sort, he was so thoroughly accustomed to the apparatus that it was evidently desirable to begin with regu- lar training experiments. The first series of trials was given on April 19. Both punish- ment and reward were employed from the first. The punish- ment consisted of confinement for thirty seconds in each wrong box, and the reward of a small piece of banana, usually not more than a tenth of a medium sized banana for each correct choice. The total time for the first series of trials was fourteen minutes. This indicates that Sobke worked rapidly. My notes record that he worked quickly though shyly, wasted almost no time, made few errors of choice, and waited quietly during confinement in the boxes. In this, also, he differed radically from Skirrl who was restless and alwa^^s tried to escape from confinement. Throughout the work on problem 1, punishment and reward 44 ROBERT M. YERKES were kept constant. Everything progressed smoothly; there were no such irregularities of behavior as appeared in the case of Skirrl, and consequently the description of results is a relatively simple matter. Sobke invariably chose the end boxes. His performance was in every way superior to that of Skirrl. As previously, the detailed results are presented in tabular form (table 4). From this table it appears that, whereas the expected ratio of right to wrong first choices for this problem is 1 to 2.5, the actual ratio for Sobke's first series was 1 to .67. This surprisingly good showing is unquestionably due to his marked tendency to choose the end box of a group; and this tendency, in turn, may in part be the result of the preliminary training, for during that only one box w^as open each time. But, if the preliminary training were responsible for Sobke's tendency, it should be noted that it had very different effect upon Skirrl, and, as will be seen later, upon Julius. The results for the ten different settings of the doors for prob- lem 1 as they appear in table 4 are of interest for a number of reasons. Inthefirst place, the setting 1. 2. 3 appearing twice, — at the beginning of the series and again at the end — yielded markedly different results in the two positions. For whereas no mistakes were made in the case of setting 1, there were fifty .per cent of incorrect first choices for setting 10. Again, satis- factory explanation is impossible. It is conceivable that fatigue or approaching satiety may have had something to do with the failures at the end of the series, but as a rule, as is indicated by settings 1,2, and 6, if correct choices were made at the beginning, they continued throughout the day's work. In this problem, Sobke's improvement was steady and fairly rapid, and in the eighth series, trials 71 to 80, only correct first choices appear. Consequently, seventy trials were required for the solution of the problem. This number is in marked contrast with Skirrl 's one hundred and thirty-tw^o trials. Immediately following the first perfect series, Sobke was given two series of control tests on April 28. Conditions were un- favorable, since the day was stormy and the rain pattering on the sheet-iron roof made a great din. Nevertheless, he worked steadily and well up to the sixth trial, which was preceded by a slight delay because of the necessity of refilling some of the food boxes. After this interruption, wrong choices occurred in MENTAL LIFE OF MONKEYS AND APES 45 w >« H-l m a. < H QJ XI o C/J Pi ^ C^COCOCOLOLO^O 00 ca o o CO "* ^ "* C^3 C^J ^ o oooooooo o Pi Pi ^ ^ ^COOOCOCMCM— lO CO Pi cotxr^o-ooooo^S [>• ^ '^COCOCOC^ICM'-HO CM-H c^ cot-t^t^ooooOiS 00 CX) LO • IC LOLO 00 00 00 I> CO C/} 'S*^ LO 1< -roooo^'*^^'* PO COCO 00 t>- CD C/3 »o LOLO LO OC^LOCOLOLOLOIO ^s CMC^] LO to 00 -* c^ CO CO sO CM ^.-l cocococococococo "^ LOCM CD o^ LO lO 00 C/2 CO CM CM COC-ICMCOCMCMCMCM coco Oi 00 -^ c^ t^ ai 00 c- i>- tr^ CO LO CTi c^ c. c^ t> cr. t:^ i> ■* ^^ ^^ CO C£> in C/) ^ CO PC fC COCO^COCOCOCOCO coco — 1.— 1 y—{ 1—1 .-1 c^i CO -* LO CO t> QJ _ feOiO^cMco^co:^ 00- C-r-i CM CM CM CM CM CM CM CM " < 46 ROBERT M. YERKES trial 6. And again after trial 9, there was brief interruption, followed by wrong choices in trial 10. The ratio of right to wrong choices for this first control series was therefore 1 to .25. Six minutes after completion of the first control series, a sec- ond was given under slightly more favorable conditions, and in this only a single wrong choice occurred, in that box 2 was first chosen in trial 6 instead of box 1. From the results of these two control series, it is evident that Sobke's solution of problem 1 is reasonably adequate. He is easily diverted or disturbed in his work by any unusual circumstances, but so long as every- thing goes smoothly, he chooses with ease and certainty. Whether it is fair to describe the behavior as involving an idea of the relation of the right box to the other members of the group would be difficult to decide. I hesitate to infer definite ideation from the available evidence, but I strongly suspect the presence of images and relatively ineffective or inadequate ideation. It is perfectly evident that Sobke is much more intelligent than Skirrl. In practically every respect, he adapted himself more quickly to the experimental procedure and progressed more steadily toward the solution of the problem than did Skirrl. The contrast in the learning processes of the two monkeys could scarcely be better exhibited than by the curves of learning which are presented in figure 18. The first, that for Sobke, is sur- prisingly regular; the second, that for Skirrl, is quite as sur- prisingly irregular. These results correlate perfectly with the steadiness and predictability of the former's responses and the irregularity and erraticness of the latter's. Problem 2. Second from the Right End On the completion of problem 1 Sobke was in perfect condi- tion, as to health and training, for experimental work. He had come to work quietly, fairly deliberately, and very steadily. His timidity had diminished and he would readily come to the experimenter for food, although still he was somewhat distrustful at times and became timid when anything unusual occurred in the apparatus. As preparation for problem 2, a break in regular experimen- tation covering four days followed the control series of problem 1. On each of these four days the monkey was allowed to get MENTAL LIFE OF MONKEYS AND APES 47 food once from each of the nine boxes, both doors of a given box being open for the trial and all other doors closed. For this feeding experiment, the doors were opened in irregular order, and this order was changed from day to day. Systematic work with problem 2 began on May 3, with pun- ishment of thirty seconds for mistakes and a liberal reward of food for each success. Early in the series of trials it was dis- covered that Sobke was likely to become discouraged and waste a great deal of time unless certain aid were given by the experi- menter. On this account, after the first two trials, the method was adopted of punishing the animal by confinement for the first ten mistakes in a trial., and of then, if need be, indicating the right box by slightly and momentarily raising the exit door. Every trial in w^hich aid was thus given by the experimenter is indicated in table 5 by an asterisk following the last choice. In the first series of trials for this problem, aid had to be given in seven of the ten trials, and even so the series occupied seventy- one minutes. It is possible that had no aid been given, the work might have been continued successfully with a smaller number of trials than ten per day. But under the circumstances it seemed wiser to avoid the risk of discouraging and thus spoil- ing the animal for use in the experiment. It should be stated, also, that it proved impossible to adhere to the period of thirty seconds as punishment in this series. For the majority of the wrong choices confinement of not more than ten seconds was used. For the second series, given on May 4, the conditions were unfavorable in that it was dark and rainy, and the noise of the rain on the roof frightened Sobke. He refused to work after the fourth trial, and the series had to be completed on the fol- lowing day. The total time required for this series was seventy- eight minutes. The work on May 6 was distinctly better, and the animal's behavior indicated, in a number of trials, definite recognition of the right door. He might, for example, make a number of incorrect choices, then pause for a few seconds to look steadily at the doors, and having apparently found some cue, run directly to the right box. No aid from the experimenter was needed in this series. On the following day improvement continued and the ani- mal's method of choosing became definite and fairly precise. He 48 ROBERT M. YERKES was deliberate, quiet, and extremely business-like. The time for the series was thirty-one minutes. The period of punishment was increased on May 12 to thirty seconds. Previously, for the greater number of the trials, it had been ten to fifteen seconds. This increase apparently did not disturb the monkey, for he continued to work perfectly throughout the series, although making many mistakes in spite of deliberate choices and the refusal of certain boxes in each trial. An interesting and significant incident occurred on May 13 when at the conclusion of trial 5, Doctor Hamilton came into the experiment room for a few minutes. Sobke immediately stopped working, and he could not be induced to make any choices until Doctor Hamilton had left the room. This well indicates his sensitiveness to his surroundings, and his inclina- tion to timidity or nervousness even in the presence of conditions not in themselves startling. Work was continued thus steadily until May 28 when, because of the failure of the animal to improve, it seemed wise to increase the period of confinement as punishment to sixty seconds. In the meantime, it had sometimes been evident that Sobke was near to the solution of his problem. He would often make cor- rect choices in three or four trials in succession and then appar- ently lose his cue and fail utterly for a number of trials. After June 1, in order to hasten the solution of the problem, two series per day were given. In some instances the second series was given almost immediately after the first, while in others an interval of an hour or more intervened. It was fur- ther found desirable to give Sobke all of his food in the appa- ratus. When the rewards obtained in the several trials did not satisfy his hunger, additional food was presented, on the com- pletion of the series of experiments, in one or more of the food cups. On days marked by unwillingness or refusal to work, very little food was given. Thus, the eagerness of the monkey to locate the right box was increased and, as a matter of obser- vation, his deliberateness and care in choice increased corre- spondingly. Sixty seconds punishment was found satisfactory, and it was therefore continued throughout the work on this problem. It was evident, on June 9, from the behavior of the monkey as well as from the score, that the perfect solution of the prob- MENTAL LIFE OF MONKEYS AND APES 49 lem was near at hand. This fact the experimenter recorded in his daily notes, and sure enough, on the following day Sobke chose correctly throughout the series of ten trials. The time for this series was only ten minutes. The choices were made deliberately and readily. An analysis of the data of table 5 reveals five methods or reactive tendencies which appeared more or less definitely in the following order: (a) Choice of first box at the left, because of experience in problem 1. This tendency w^as very quickly suppressed by the requirements in connection with problem 2. Indeed one of the most significant differences which I have dis- covered between the behavior of the primates and that of other mammals is the time required for the suppression of such an acquired tendency. The monkey seems to learn almost imme- diately that it is not worth while to persist in a tendency which although previously profitable no longer yields satisfaction, whereas in the crow, pig, rat, and ring dove, the unprofitable mode of response tends to persist during a relatively large number of trials, (b) The tendency to choose, first, a box near the left end of the group, to go from that to the box at the extreme right end of the group, thence to the one next in order, which was, of course, the right box. This tendency appears fairly clearly from May 7th on. (c) The box at the extreme right was first chosen and then the one next to it. For example, in setting 2, box 4 would be chosen first, then box 3. Or, if this did not occur, the method previously described under (b) was likely to be employed, as for example, in setting 8, where such choices as 7.6.5.1.8 appear, (d) In certain series there appeared a marked preference for a particular box, usually box 3 (see results for May 24). This was doubtless due in a measure, if not wholly, to the fact that box 3 was the right box twice in each series of ten settings. But it should be added that the same is true of box 7, for which no preference was mani- fested at any time, (e) Direct choice of the right box. The five reactive methods or tendencies enumerated above roughly appeared in the order named, but there were certain irregularities and the order as well as the time of appearance varied somewhat from setting to setting. In general, method c was the most frequently used prior to the development of method e, the direct choice of the right box. 50 ROBERT M. YERKES .2 rt o 05 o d d d i-H ciJ d CD OO OO oS 8 •^ai o in coo coo CMCji — ( —1 CO CO CM o CO d ^ o o o —I OiOC a^ en OOCTl CD OCT) o '^ ci o o o — ICM - - CM— 1 Tj< CO— 1 co CO ^ o o o CTlOO cr. a> 00 t^ t> o C/5 LO -^ CO coco CO coo 00 00 CO 00 CO °0c^ CO LOOj^oo o6o-<*cbo6f^ 00 ^ t>co LOOOLO 00 CO ooo o CO CO 00 00 00 CO 00 in in CO CO CO in 00 t--. 00 00 00 o CM— 1 xf r-iCV! . * ^^co ^co-^ ^ CO CM CM CO O^ « •^ -^ CM c^ '^ ^ — iCM-<#CO C/2 ^,->^ '^^ r-iCMCOCMCO CM 00 CO CM'* — ^' CO — < — i -^ ^ — ( »^CMCO CO ^^,-1 -rf Tl< Tt T^f* ^ •* CO —1 00 00 00 00 in • CO 00 00 Cr. CD CDOC LOCOOO'* t> LO —i LO cr- a •-< CM ^00 00 CO CM LO CJi LO CO 00 LO CO CM lO CD 00 Cji CO 05 cnt>cocoi>oot>cr-. LO CD 00 t> CJl ooo o o C<3 ^ ■* CO CO LOCO s 0) c^ in ^ LOCO CO . * LOin-<^io LO^ COLO ^^lO LOCO"* in CM in in in in Tj* c^i CM CM in in •S c/) CO C^i LO LO LO LO CO ^ CM CO roLOio-* ^lO ■* r} 00 00 00 LO cot^t> t> o ■^ lO to CO ■^^t>t> c^-* 00 t> 00 00 00 ^ LO (/) C/D LO 00 OCMOOOO 00 LO '^OO^ ^'^ t^ in ^t- a -^ LO 00 00 -^ CD ^CO CD^LOLOLOOO CD inoo o o. en Tf U3 COr-H . * r-(,-HLO LO CDLO r-icb LO CO C-JLO ,-Hcr> CM CM en CO cvi CO CD CO CD COCMLO Tt LO LOLO — ICOLO ^^ — 1 -^ CD CO COLO CO CD CD CO LO in CM -i^iO in t> CO CO CO CD vO t> i>i CD cvi CO 'X t^ t> c^ t^ CO lO 5 CO CD CM CM CM CD CD c« CO OJ t> C^ t> f- CM CVl C^JCD CO CO CO i>- c^co CO coco t>o CD CM COt>CMj^CVlC^C^ CO -*CO '-f ,-H— 1 . CO —I'* M fo CM^CO ^ CO-* CM CO c/) "^ ^ CO ^CMCM ^ CO CO CO — 1— ITf -- '^ — 1^^ CO I-H co^ CO CO"* '^ CO c-cr-^ avoooo ac t>I>-00 a;a-. o 00 00 00 C/) aicx 00 00 crJoo 00 00 00 C^CJ>I> C^-OiOi I>C-- t^ t> tr-t- c^ ot- 00 00 o o C^l O CO CM oo ^LO CO -* 8 LO CO OO 00 cr- t^OO 8 C7) CO -^C^1 O— 1 C^J o CO a. LO i" c re CO c^oo o ^ CM CO •* LOO 00 cr. MENTAL LIFE OF MONKEYS AND APES 51 2.33 1.50 1.50 o oooooo LO O CD LC LO LO CD r-H -^ ^ ,-i ^ T-^ '^ ° 1.22 1.86 0.82 0.54 0.54 1.22 0.25 0.00 oco o^ ^•^^ ^ ^^^^^^ -^ 1— ( .— ( t^ ;£)«£> CD OOLCCDCDCDOO CM ;i3 S2 <^ t^ t> ^il '^o LOCO CO "^ "^ ■<# (MLn-^-*-*CM 00 oi t- ^; J2 2 o^ S2 lOC^ I>.C£>C£> CD OOLOCDCDCDOO CDCD lOCDCDC^CDCOCO'*-<3*COCDlOCO'-iO lOCO COTf<'* ■^ CMLO'^^'^CM T^r^ LO^^CO^OC^CDCDt>TjCJi2 LOI> t^ C^ 00 00 LOCD CO t> CO c^ '^ 00 ^ t>tr^i>- oot> o LO t^OO C-- coc> CM CD t>C£3t> t> ooooooococo 00 1> CDI>t>00C0Ot>t>I>-t-t>t>00t>t> 00 00 CM '*' CO 00 CO 2.1.4 4.4.3 3 CM . CD CM I>CD CO CO CO ^■<:l<-^^COCOCOCO CO^CMCOCOCOCOCOCO'^fOCOCOCOCOCOCOCO t>t^ 00 00 00 00 5.4.3. 6.5.9.8 6.5.8 3.8 6.5.9 6.5.8 8 00 CO 'S* 6.5.4.3 8 6.5,9 6.5.8 8 CO CM C£)t>t> 00 ^0000t>t>C5 00 lO OOC^I>t>t>OOOOOOOOOOC7300000000 CO CM ■^ ^'Tj* LO ifi'^iri^^ LO ■^ LO LO -^ -^ -^ -^ •* cMfoCMtnm -^ -^ Tt"# CM Tj< Tf '^^ LOCMCMLOLOLOLO'*^^'* CO CM ■^ coLOLOCMco loiOLnLncoLO LO-^ COLO'*COCOCOCOCOCOCOCOCOCOCO'* cotj< COCO CM CD in CM CM CM •- < CO CM CM CM CM CM CM CM CM CM CM CM 0-] CO CM T)< LO COCOCOCO'-ICOCOCOCOCOCMCO coco COCOCOCMCMCMCMCO— ICMCMCOCMCMCM LO-* 00 '^ t> ■* fc -* t^ t>I>l>l>- t> C^IO t^t^t> t>. t^t^C^ C^5 LOt>I> 00 00000000C>00 00 CD C^t^00O0O0l>-C^C^0O00O00Ot>t>-C^ COCO LO LO LO LO LO CD LO LO LOCDLO CD COLO CD ^LO CM LO CD CD LO LO CD CO CD LO -^ ■^'—1 in CM CDCD COLOCD CM CM CD CD CO '^ C^ CD CM LO ^ CM LO COIOCO CO CO LO LO LO -nP CD COLO COCOCO^COCOTi'COLOLOCOCO'^LOLO CDCD CD CD cDt-: LO CD C^ t>CDCO C^CDCOCD CD t^CD CD CD CD C^ t> CD LO -^ CD cow CMt^CD CMCOt> CMO^ C- ^1> C^CDCDCD^ '^CDCDC^CD CMl> CO LOO COt^COCDCOCOLOCO LOCO CDCOt>COt>lOCDt>t>t>'*t>-CDCDCD .-hCD 00 00 CD ifi CO LO CM CD LOCD CO CO CO CO COCOCOfOCOCO coco COCOCOCOCOCOCOCOCOOOCOCOCOCOCO CM COCO 00 Cji 00 ^00 00 oc t^ 00 00 00 05 0> 00 t> c:^ 00000000 00 LO l> t>I>.t^ 00 C^ 00 t>- 00 f- t> ooo I>t>I:^O2000000000000t>000C0000 CDI>- ooo LOCDC^ o oooooo CO -00050'-i CMCMCMCOCOCOCOCOCOCOCOCOCO^^ oo •-HCM -^LTjCD C^ 00 O. O ^ CM CO — 1 --i^CMCMCMCM 'S'LO CM CM CDt:^000)0'-^CMCO'*LO'DI>-COCT50 CMCMCMCMCOCOCOCOCOCOCOCOCOCO'* 1— 1 f— 1 O^CM CMC^OCM '^ LOCD t^OCOJ^ C_^ CM CMCMCMCMCMCO 3 "^ - •—5 CMS COi: ^3 L03 t>3 00s CDS g ^CM 52 ROBERT M. YERKES Examination of table 5 indicates that some of the settings proved very easy for Sobke; others, extremely difficult. Con- sequently, the number of methods which were tried and re- jected for a given setting varies from two to five. Setting 2 proved a fairly simple one, and after the inhibition of the tendency to choose the first box at the left, the only definite tendency to appear was that of choosing the first box at the right, and then the one next to it. After one hundred and thirty trials, this method suddenly gave place to direct choice of the right box, and during the following twenty-eight series, no error occurred for this setting. Setting 4, on the contrary, proved extremely difficult, and a variety of methods is more or less definitely indicated by the records. It is needless to lengthen the description by analyzing the data for each setting, since the reader by carefully scanning the columns of data in table 5 may observe for himself the various tendencies and their mutual relations. Sobke's curve of learning (figure 19) in problem 2, is extremely irregular, as was that of Skirrl. Similar irregularities appear in the daily ratios of right to wrong first choices presented in the last column of table 5. Most of these irregularities were due, I have discovered, to unfavorable external conditions. Thus, dark rainy days and disturbing noises outside the labora- tory were obviously conditions of poor work. On the day following the final and correct series for problem 2, a control series was given. In this Sobke seemed greatly surprised by the new situations which presented themselves. Repeatedly he exhibited impulses to enter the box which would have been the correct one in the regular series of settings. He frequently inhibited such impulses and chose correctly, but at other times he reacted quickly and made mistakes. It was evident from his behavior that he was not guided by anything hke a definite idea of the relation of the right box to the other members of the group. In a second control series given on the following day, June 12, confusion appeared, but less markedly. For the first setting, a correct choice was made with deliberation. For the second setting, box 3 was immediately chosen, as should have been the case in the regular series of settings. Sobke seemed confused when he emerged from this box and had difficulty in locating MENTAL LIFE OF MONKEYS AND APES 53 the right one. Then followed direct correct choices for settings 3, 4, and 5. For setting 6, there is recorded a deliberately made wrong choice, and so on throughout the series, the choices being characterized by deliberateness and definite search for the right box. Uncertainty was plainly indicated, and in this the be- havior of the animal differed markedly from that in the conclud- ing series of the regular experiment. It seems safe to conclude from the results of these control series that Sobke has no free idea of the relation of secondness from the right and is chiefly dependent upon memory of the particular settings for cues which lead to correct choice. Problem 3. Alternately First at Left and First at Right For four successive days after the last control series in con- nection with problem 2, Sobke was merely fed in the apparatus according to previous description (p. 43). He exhibited a won- derfulh' keen appetite and was well fed during this interval between problems. The method of experimentation chosen for problem 3 in the light of previous experience was that of confining the monkey for a short time, ten to fifteen seconds, in the wrong box, in each of the first ten mistakes for a given trial, and of then aiding him to find the right box by the slight and momentary raising of the exit door. Aid proved necessary in a few of the trials during the first four days. Then he worked independently. As work progressed it w^as found possible and also desirable to increase the period of confinement, and in the end, sixty sec- onds proved satisfactory. It was also thought desirable to in- crease the number of trials per day from a single series during the early days to two or even three series from June 29 on. Often three series could be given in succession without difficulty. During the early trials on this problem Sobke worked remark- ably well, but later his willingness diminished, evidently because of his failure readily to solve the problem, and it became extremely difficult to coax him into the apparatus. On days when he entered only reluctantly and as it seemed against his will, he was likely to be nervous, erratic, and often slow in making his choices, but above all he tended to waste time by not returning to the starting point, preferring rather to loiter in the alleyways or run back and forth. 54 ROBERT M. YERKES W < « _o ^ 1 o 8 8 CO CO 8 O in g COCOQ CO COO O O C CM m 00 CMO t^ ooo CD 03 O T*! en C-J V. ^ ^ CMCM'* ^ 'Pv! 'P. .-H I— 1 1—1 ^ 00 CJi tr^ 00 CD CD t>t>-00 in ^ o. aim 00 Qi CM - CO CM -* ■* CO CO CM in 00 ^^ ^^ ^ ^ 00 en t> 00 CO CO C^I>00 in CD com'* '*mmmoo ca CM - co CM -* ^ CO CO CM in^ •^inco commm t^ Oi 2S 00 Oi iy> m ^ci: l> 00 CO CD 00 o^ O) O^ Gl 05 cjicoc:^ G^ cn cncj>co 00 05 00 CT3 OiO CO c^ c^ CO CO co^ 00 CO CO CD en 00 CJiOl c^cbco 00 cncoooj oocooi . . . . « ... r^^oooicoo^cji-^ 00 CO CO coc^cx <^dco^ r^ • CD oc3°°^t^ " cn ^ 00 '* t> oot-r- 00 ooincnincnooinoocot>oooo 00 en t> 00 CO 00 00 CO CO CO 00 OiOOOO CTi en cDCTicr. OJO^cDcnt ■^ d Gl CTl C:^ CTl Ol CTl Oi Ol OlCJiCTiCJi^ in in m Wi ^ CM CM in in CM CM oc ^ in O-J ^ CM CM ^in CMLO m'* m CO CO C^j CM ^in CM in in CO '^ Mc^iin CMCM CMCM cMmmm CO co^ CO CM CM CM CM CMCOC^l CO'* co'*m CO CM CO com LO CM CO t- ^ CM CM '^ C^] C/3 CO CM CO CM ^ in ^ CM CO '^ ^ CM in '*CMin CMCM C-JCMCM C^l CM CM CM CM 00 CD c/i o CD iri -^ CO 00 00 00 00 00 00 00 03 00 00 00 00 00 00 00 00 00 00 ^ in 00 t> CO 0000 inoo t>oo . . » . . * t~_t>-*c>c^^t>oo 0000 00 00 CO'* 00^ * 'S' '* 00^ ^ ^ c/) LO oooot>i>-oooooooo^oooot>ooooin t>in'* '*^ m olm C^'*'* ^CD •* O 00 t^CO 00 t> 00 t>. 00^ t> 00^ tr^ 00 00 00 00 00 00 00 00 00 00 00 00 00 00 o CD CD CD •^ in 1—1 >-H . . * CM CM CD CO CO CO CD CO ,-( =M ^ CDCO CO --icDJ^ .-( CD -^ in CO --H CD CD CD 00 CD C/3 CO i-H^-"* r-i CD in ^ CMCM,-; ^' -< CM i-l i-t --H r-i CO cDmcDmco CO'* in CO CO c^a CM 1— 1 CO <^j'CD rHlOrHr-HCMCMCOCM r-ICMCM--!'* CD ^ LO inT-i in'* -*CD co^ CO '^ COCMCMCD"-! incMi-i ■<*Ln r-l CD c^ ^^'^mcocDinLn ^ in CO ^ c^ CO c^ J CO in CO CO'* --I i-H CO'-H'-l CM CO CO t> CM CD u-:i in o- o in ■^CD CO LO t> t^ CD in in t--in C^ mio in in o in CD CD t>LncD t^in l>l>.t>C^t> CD CO in LO ■-^ t> CD in in CD mm CO tfi CD in CD in in CD com c^ CD CD m mm CD CO CO CO CD c- in t>coc^ t^t- cDmm mmcocom No. of trials o S o 00 C^l o CO O in in ooo C^OOOi cot-- 00 oo So ooo CMCO^ T— 1 1— 1 l-H .—1 .—1 .—1 r-HCMCO ooooo mcDt^oocn '*mcDc^oo 1— 1 1— 1 1— I .— 1 1— ( C3 Q 00 Oi CM CM CM CO CM '*incD CM CM CM 00 cn CMCM ., o^ " CO >> 3r-is CMCO- MENTAL LIFE OF MONKEYS AND APES 55 od TttO^ LOLOCO CO LOLOLO OJ CO CO CO CVl CO CO '* CO CO LO CO ' ICVICOCOO --H co-^'* LOLOt> t> LOLCLO oot>i>t>-c»c^t>-Loc^CT!cr, oot>i>2<^ cocx> 00'* Tf in 00 C:^ t£>C75 ■^ odco in CO t^ (T. (Ji (Ji cn CT5 coco CO '^COOl •^ t^ LO t^ C^ LO CTiO^LOoocT. cDcn (ji ooa^t>Tt*LocoCT)a". C75 COCO CO CO Lo'c^co cococo LO cococo cOLo'cocococo 00 cococo coco cr. oioj cTiooci Oi odaioo CTiaiaiaiodaicoaicooioot^cococo^cTicoco LO LC LO LO LO CO CO oj CO LO cci IfiTtiT}" CO-^LO LD LOLOC^] LOLOUOLnLnLOLnCOLOLOLOLnLOLnLOLCLnLOLO cvi CM CO CM CM LO CM CMCMCM CM CM^CM '^CMCM CM LOLOCM CMlOCMCMCMLOLOLOCMCMCOCMCMCMCMCMCMCMCM 000000 OOt^-00 00 00 00 •^ -^ 00^ 00I>00 00 00 1>- 00 00 00 00 00 00 00 00 00 00 00 00 '^ LO 00 00 -* ^ 00 00 ^ ■<*tO^D VO ■* -^ •^ Tf LO ■* -^ ^ "* t^ Tji -^ ■rj' Tjl -^ TJI -^ -^ Tt -^ ododoo 060606 06 060606 06 06 06 06 06 06 06 '^ 06 06 06 06 06 ^ 00 Tf 'st* ^ 06 <0 «£) ^Dtr> cOtO CM to ^o CTiCMCM LOCM'X) ^O tDCOCI3 ^D^t£>LO^'r!<£)CriCOCOtriC£;^0<£i^COt£>C£) LncdtTi'H c£) COCM CM CO ^ ,-t ^O «X5 CO ^ ^ --H ^(X)^ r-I^CMCM^DLO'-HCM^ ^^r-t^^r-<^^,-t^^^^^^r-lr-l^^ i>t>-io' C^t>C^I>C^C^LOt>C^t^LOC^[>t>t>I>t>t>t^ ^C LO LO LO LO LO LO t^ LO LO LO LO LO LO LO LO LO LO LO LO LO LO LO LO LO LO LO LO LO 800 •-ICM CM CM CM 000 CO '^LO CM CM CM 000 t^ 00 c^. CMCMC^3 0000000000000000000 O'-HC-jcoM'Loc^t^ooaiO'— icMco^LocDr^oo cococococococoxicocO'^'-rr'^'vf^^^'*-^ L03 CO 3 t>3 00 3 CJi 3 3 i:::;^ ^3 O- CM- CO, ^, LO- — 1 "—I ^ o o o Ol en 00 00 COCO ^co LOCO ftj o o o -- - C.3 CM ^'a' CDC- LO •^ ^ o o o 05 en 00 00 COCO ^00 inco Di o o o - - CM CM ^Tl< cot- LO-* CDir^ o 00 00 CO CD 00 c^ (/) CO COCO l>- CO C^ 00 00 co CO t> t- C^ CD l>t> I>C- c-t> CD CT5 in C2 00 COCOOOLO 0000 '^CO ■^COCOCOCO^LOCDoi'^CO eo 00 CO CO CD C/) t- COOiCTlCO-^OCOO^LOCO CO CD C^ co^ ^CD coco a-iCoco^coco'*coLnt^t>- LO t> LO O-LC LO t- oc- CO CM 00 c/5 CO CO cocM c^ico^ CO CO CO CM^i- CO CM CMLO CD ^ LO LO'* COCO LOLO LO LO ^ lOLO CM CO CO CO COCO LO CO CO COCO CD LO ^ LO-* coco CO CO 1— I CM CM CM 1— I CM CM CO CO CO CO CO CO CM CM CM CMCJ CM CM LO IX> 05 00 COCiLOCO— ILO CO COLO LO ^00 C^LOt>-^ LOCOLO . * .... ... COCOLOCMCOCOi-iLOCDOCO CD t— en c^ CD a5Cvi.-H •^ Oi Oi •^ o t> f^t>.CD^OOt>CDCOC^C^C^ C^l ■"^ ^ CTJ as •-H 00 -— 1 ^CM CM CM CMC nco^-*-<*i>-rJCMcs ^LOLO ^in lO 00 00 LO^ Loooco CO CO CO CD CO c^co CA) LO ^COOO'*'* t>. t- I>- t> -* ^ CD ^ OO^^^OOLO "* ^ ^ CO LO LOLO coco CDt> ^ ot5 . . * . . * en t> 00 t> t:^ 00 t-00 t> c-^oo 00 00 00 C/) t^ t^ CT> t^Ci CTi t^ cnc^ t>t> t> 00 c^ 00 c- 00 00 00 00 (Ji C^ t> C^ t^ Oi t>-c^ c^t^ c^ i> c^ c^c^ c-t- t>.I> ^ CO cot> . * -* ■* '^ ^ '^ ^ LO CO'* I>li0 en LO I>-C^C^ CM CO I> t> CO c^ LO LO"* CO"* LOCO .— 1 1— 1 i-H .-It^ --' .— ( 1— 1 CM CM CO CM CM CM CDt> c^ !M 00 LOCO L6LOt> LOC^ LOLO CD LO o6 c^ 1> t>- t> c- cn CO cna.00 COlO CD CD CD LO LO LO ID LO LOlOlO LOOJ LO LO LO CO CO I>CO C-C- c-co CO en o4 CM CO CO CO CO CM CM CM CM CO ■^ CO COCO COCO coco d "o O s o CO c^:i o CO o CD LO o S8 c^oo go cno oo CM CO -HCM >— I >— ( •Is CM CM CM CO CM C^J CO CM CM ooo-j CM CM o^ coco to < MENTAL LIFE OF MONKEYS AND APES 61 oo o oc^o o Qor;-ot>.t-- ^ "— ' ^D^ LO lO-^tO C£> LO LO -^ LO '^r -^ CT) ^^ LC LOCC^ "^ LO LO to LC CO cxi to 00 Ln^o to«o Ln m^ iolo^lO'*^ ^ -:j- (M un^ ^-* Lo LOto^ '^ LoiooLOco^ ^ ~ CO CO C^ 00 00 ai C050 CO ^ ^ coco coco CO t>^l>^CO t> CO -^ CO CO LO CO I>it> I>t> I> o6o6t> 00 t> C^ t>^ t>i !>• 00 -^ -^ -^ '^ -^ ■<*' -^ -^ "^ "^ LOlO LOLO LO LO'^'* Ln LO ^LOLn-^ coco coco CO COCOcb CO CO-^COCOCOCO MM MM CM M CV] CM (M CM CM CM COincOLO CMLOCM LO OCO^CO'lO t-lCOCOLO LO t> LO 00 1> CO t>- CO ^t>^coLnLC co oo COCMo6t>CMt>LnCCOCMLOI>LOC7) CO CD t>co t>CD 00 t> CO COCOI>- LOCO Tf' t>! CO CO CO CO c-^o6cot>t>c-^c>co 00 00 00 00 00 00 00 00 00 00 00 t> 00 t> I> 00 1> 00 t> [>• t>l>00 t> t>t>00 00 00 00 ■^ CO I> 't ^ -^ t> CD LO ^ "^LO '^ ■* '^ >* ^TtlO-r- LO t>-*CO LO LOCO 'i'LO in ^^^ LO lO^iniocoLO CD CMIOCM CD CDt> LOCM CO CM CM in CD CD CO CM CO CM CM C^ LO t^ CO t> c^ r>.t- t>I> t> c^o t^ c^ ^.l>t>I>t>t> Tj-CM CO^CM Tf T? -^ Tj<" CM '^ CO ^ CO . . * .... . . T^TtCO-^-^CMrfCOCM-* CM CM -^ CM '^ ^ -* '^ '^ CM CO CO CO COCOCO CO COCOCOCOCOCO o o o oo oo o LO CO t>- 00 a; o— < cm ^ ^ ,-, ^^ CMCM CM '^ LO CO t^OO OlO -^ ^ ,-1 — I ,-1,-H ,-(CM CM ooo co^ in CMCM CM oooooo t- 00 CI O -H CM 0:1 C^l CM CO CO CO 62 ROBERT M. YERKES laboratory to take charge, I found that Sobke was no longer trying to solve the problem as when I had gone away. His attitude had changed in that he had become indifferent, care- less, and obviously discouraged with his task. I immediately set about reinstating the former attitude by lessening the number of trials and the punishment, and by in- creasing the value of the reward, but my best efforts, continuing up to August 28, failed markedly to improve the condition. The number of correct choices did somewhat increase, but at no time did the animal attain the degree of success which he had achieved as early as July 31 in the eleventh series of trials. During the last two weeks of experimentation, all possible efforts were put forth to discover the best combination of re- wards and punishments. Punishment was varied from 0 to confinement of sixty seconds, and man}^ kinds of food in different amounts were tried as rewards, but in spite of everything Sobke failed to improve markedly. From time to time, notably on August 12 and 21, he exhibited peculiarly strong resentment toward me and repeatedly attempted to attack me. The outcome of my experiments with problem 4 is peculiarly interesting in that it indicates the importance of a favorable attitude toward the work and the extreme risk from disturbing or discouraging conditions. It seems not improbable that had the work progressed without change in experimenter, or method of procedure, and above all without the disturbance of the paint- ing, Sobke might have solved problem 4 within a few days. This is by no means certain, however, for in problems 2 and 3 the ratio of right to wrong choices instead of increasing steadily increased very irregularly. The detailed results for this problem are given in table 7. Reactive tendencies which appear are : (a) persistent choice of the end boxes followed, subsequently, by (b) the tendency to locate the middle box directly. This proved fairly easy when the number of boxes involved was only three as in settings 1, 4, 7, and 10. Setting 4 was most difficult of all, because box 9 was avoided or ignored. When the number of open boxes was as great as five, as in settings 2 and 8, the task was obviously- more difficult, but whereas success in setting 2 appeared early, in setting 8 it failed to appear during the course of experimentation. For the settings 3, 6, and 9, involving either seven or nine open boxes, MENTAL LIFE OF MONKEYS AND APES 63 the direct choice of the middle box was next to impossible, and Sobke tended to choose, first of all, a particular box toward one end of the series, for example, box 2, in setting 3, and box 7 in setting 9. To the experimenter, as he watched the animal's behavior, it looked as though effort each time were being made to locate the middle member of the group. This appeared relatively easy for groups of three boxes, extremely difficult for as many as five boxes, and almost impossible for seven or nine. 3. Julius, Pongo pygmcEUS Problem 1. First at the Left End The orang utan, Julius, was gentle, docile, and friendly with the experimenter throughout the period of investigation. He at no time showed inclination to bite and could be handled safely. As contrasted with Skirrl and even with Sobke, he adapted himself to the multiple-choice apparatus very promptly, and only slight effort on the part of the observer was necessary to prepare him, by preliminary' trials, for the regular experiments. But in order to facilitate work, he was familiarized with the apparatus by means of regular route training and feeding in the several boxes from April 5 to April 9. On April 10 the apparatus was painted white as has been stated previously, and on the following Monday, April 12, Julius when again introduced to it gave no indications of fear, uneasiness, or dislike, but worked as formerly, making his round trips quickly and eagerly entering any box which happened to be open, in order to obtain the reward of food. The regular experimentation was undertaken on April 13, and the results of the first series of trials with Julius are sharply contrasted with those obtained with the monkeys in that fewer choices were necessary. Instead of the expected ratio of right to wrong first choices, 1 to 2.5, the orang utan gave a ratio of 1 to 1. An additional markedly different result from that obtained with the monkeys is indicated below in the total time required for a series of trials. As examples, the data for the first, second, fifth, and tenth series are presented. Time for Series of Trials 1st series 2nd series 5th series 10th series Skirrl ' 35 min. 20 min. 14 min. 10 min. Sobke 14 " 17 " 10 " 9 " {8th series) Julius 12 " 11 " 14 " 9 " 64 ROBERT M. YERKES It is also noteworthy that JuKus in the presence of visitors or under other unusual conditions worked steadily and well, whereas the monkeys, and especially Sobke, tended to be dis- tracted and often refused to work at all. Almost from the beginning of his work on problem 1, Julius began to develop the tendency to enter immediately the open door nearest the starting point. In case the group of open doors lay to the right of the middle of the apparatus, this method naturally yielded success; whereas if the group included doors to the left of the middle, it resulted in failure. Obviously it was a most unsatisfactory method, and although it enabled him to make more right than wrong first choices, it prevented him from increasing the number of right choices, and as table 1 indicates, it maintained the ratio of 1 right to .67 wrong first choices for eight successive days. On April 23 a break occurred in which the number of correct choices was reduced from six to five. Julius worked very rapidly and with almost no hesitation in choosing. My notes record " he seems to miss the point wholly. It is doubtful whether the punishment is sufficiently severe." At this time he was being punished by thirty seconds confinement in each wrong box, the interval having been held fairly steadily from the first series of experiments. On April 26 it was increased to sixty seconds, in an effort to break him of the habit of choosing the " nearest " door. But he became extremely restless under the longer confinement and tried his best to raise the entrance and exit doors. Since there was at this time no mechanism for locking them when closed, it was difficult for the experimenter to prevent him from escaping by way of the entrance door or from raising the exit door sufficiently to obtain the food. In- deed, the longer confinement worked so unsatisfactorily that on the following day I substituted for it the punishment of forcing him to raise the entrance door of the wrong box in order to escape for a new choice. He was rewarded with food in the alleyway H, beside door LS (figure 17), only when he chose correctly on first attempt. This method discouraged him extremely and proved wasteful of time. Consequently, in a second series on the same date return was made to the former method, and he was rewarded with food whenever he found the right box. But on April 28, MENTAL LIFE OF MONKEYS AND APES 65 the two methods were again employed, the first in the initial series and the second in a final series of trials. The animal's persistent attempts to raise the doors gave the experimenter so much trouble that on April 29 barbed wire was nailed over the windows of the entrance doors with the hope that it might prevent him from working at them. But he quickly learned to place his fingers between the barbs and raise the doors as effec- tively as ever. On April 30 the rew^ard of food was given only when the first choice was that of the right box and in that event it was placed in the alleyway H as stated above. As it seemed absolutel^^ essential to break the unprofitable habit of choosing the nearest door, on May 3 a new series of settings was presented, in which only the doors to the left of the middle of the row of nine boxes were used as right doors. That is, in this new series, doors 1 to 4 occur as right doors; 5 to 9 do not. As punishment for wTong choices on this date, Julius was confined in the wrong box from one to five minutes. It was difficult to keep him in, but by means of cords which had been attached to the doors, this was successfully accomplished. Yet another and slightly different series of settings was employed on May 4, and this, proving satisfactory, was continued in use until the end of the experiment, with punishment ranging from sixty to one hundred and twenty seconds for each mistake. Naturally the modification of settings introduced May 3 greatly increased the proportion of wrong first choices. Indeed, as appears in table 8, the ratio of right to wrong immediately changed from 1 :.67 to 1 :4.00. Between May 3 and May 10, no steady and consistent improvem.ent in method or in the number of correct first choices occurred, and on the last named date, Julius chose correctly only three times in his ten trials. At this time there was, as my notes record, no satisfactory indi- cation of progress, and the status of the experiment seemed ex- tremely unsatisfactory in as much as in spite of the experimenter's best efforts to break up the habit of choosing the nearest door, the orang utan still persisted, to a considerable extent, in the use of this method. The only encouraging feature of the results was an evident tendency to choose somewhat nearer the left end of a group than previously. A series of correct first choices was obtained on May 11, 66 ROBERT M. YERKES Di o ^ Di oi O CO c/3 -^ 00 C/2 C/3 CO CO in CO C/3 C/) 00 c\i g^l lOOOOOOOO^O- in Tj< Tfrti -^ Tj< Tj< t;3< T* LO -^ LO 2 '^ °0 LO(£)t£>LnU3lO 2 S 2 ^ ^ LO •>* -^ Tti Tt Tf -^ Tf -^ lO -^ LO CD -^ CO CO -^ -^ -^ -rt -^ -^ in CO CO CO CO CO CO CO CD LO C£> LO -^ CD t-- t> CD CC CO CD COCO ^ .-H ^ CVl 1-1 ^ ^ Oa CM C^] ^ »-H 00 C^ •-H N CJ CCl >-< CM .-H .— I CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO 00 CO CO 00 coco t>t>Ot>t>-t>l:^t^t>t>I>-t>t:^t:^t>l>t:^l:^I>- CD ^ -rtOO immnio LO in in in in in lo LO CO Lo in LC in Lrno in imn C^ CD CD CO CO CO CD CD CD CD CO CO CO CO CO CD CD CD CD CD COCO CM CM '^OJCM CVIin CM CO CM CM CM CVI CM CM CM CM CO Tji CO cvj '^ in CM CM c^ CO ^ CO CO CO CO CO CO '^ oj 00 CM in CO in cmcm ■^ -^ -^ -^ -^ ■^ in in CD -^ -^ ■* -^ -^ in CO -^ -^ -^ in -rf oo-^ t>t>t^t>-t^t>t>t^t>t:^t:^t^C^t>t>.i:^t>.t^t> CO CO CO CO CO CO CO CO CO CO CO CO 00 CO oococo T^Tji^^^^T^j^^'^iniriTf^^rtcoco^^'^ coco 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 CO 00 00 00 00 0000 .-( CM CM r-H CM rH .— I >— I r-l CM CM --i CM ■— I CM I— I .— I CM t-l I-H .-HCM 00 CO CO CO CO 00 CM CO 00 CO CO CO 00 CO CO CO CO CO CO CO coco OOOOOOOOOQOOOOOOOOO o oo ^cMco^incot-ooaio^c^ico^LncDt^ooo'; o i-hcm T-icMco-^incot^ooaiO'— icMoo^tncDc^oo d o^ ,_, ^ ^ ^ _i _ ,-1 ,-1 ,-, ^ CMCM bsc0'*incot>cr-0'— icMco^coo^ oo- a~. - o » «_,- D.rt .-I rH —I .-H rt CM CM CM CM CM CM CM ' CM~*CM~"CO -^•^- MENTAL LIFE OF MONKEYS AND APES 67 8 oco oooooo c CO c LO q CO q -^' c LO CD 00 t> O 1 o c^ CM CO LO -<*< OJ CO 2 O 1 1 1 1 00 C^ LO CD 00 O O O 1 ^ 1 CSl CO LO ^ CM CO 2 1 1 2 2.3.4.5 6.7.8.9 4.3.2 CO c^ CMCM — ^.— 1^ CO CO CM CM CM CM —1 1 CO 1 CM 4.5.6.7.8.9 4 LOCT; -^00 COI> rice a-. 1 CM CM CM C- i CM C^l t-^ CO CO CO rr CO CO CM t> 1.2.3 3.1 00 t> CD LO ■«1< . 00 C£f L.O 2.3.4.5 4.3.2 CO t>. .-H^ CD (M CM ^ — ' ^ ^ ir! CO CO CM C^l CO CM ^ LO 4.5.6.7.8.9 4 CD LO PC 00 CO CO C2 CO sO Tt •rj' CO CO '^ ^ CO CD CO LO CO LO -^ CO ri CM CO CM CM CM C^3 CM "* CO CO CO CO CO CM CD LC' CO r4 CM CO CO CDCTl LOod -^^-*^ ^ ^ >* 00 00 3s CO "s CO • C^] Ol CM CM CO CM •^ CO CM CO -^ -^ CM CD LO PC LO CD LO CO ^ CO coco '^r CO CO CO ^ -r CO 00 CO CO CO C^J — ' — — ^ CO CM CM C^l CV) ^ ^ CO 1 1 c CO C^l ■^ LO ® t^ w a-, o CM CM CM CM CM C^l CO CO -^r LO S c^ 00 5^ CM CM C^l C^J C^:i C^l C^] o CO r— 1 CO 1 ^ LO CD t^ 00 2 ;il! CM 68 ROBERT M. YERKES greatly to the surprise of the experimenter, for no indication had previously appeared of this approaching solution of the problem. It seemed possible, however, that the successes were accidental, and it was anticipated that in a control series Julius would again make mistakes. But on the following day. May 12, the presen- tation of the original series of ten settings, which, of course, differed radicahy from the settings used from May 4 to May 11 was responded to promptly, readily, and without a single mistake. Julius had solved his problem suddenly and, in all probability, ideationally. Only three reactive tendencies or methods appeared during Julius's work on this problem: (a) choice of the open door near- est to the starting point (sometimes the adjacent boxes were entered); (b) a tendency to avoid the " nearest " door and select instead one further toward the left end of the group; (c) direct choice of the first door on the left. The curve of learning plotted from the daily wrong choices and presented in figure 18, had it been obtained with a human subject, would undoubtedly be described as an ideational, and possibly even as a rational curve ; for its sudden drop from near the maximum to the base line strongly suggests, if it does not actually prove, insight. Never before has a curve of learning like this been obtained from an infrahuman animal. I feel wholly justified in concluding from the evidences at hand, which have been presented as ade- quately as is possible without going into minutely detailed description, that the orang utan solved this simple problem idea- tionally. As a matter of fact, for the solution he required about four times the number of trials which Sobke required and twice as many as were necessary for Skirrl. Were we to measure the intelligence of these three animals by the number of trials needed in problem 1, Sobke clearly would rank first, Skirrl second, and Juhus last of all. But other facts clearly indicate that Julius is far superior to the monkeys in intelligence. We therefore must conclude that where very different metlwds of learning ap- pear, the number of trials is not a safe criterion of intelligence. The importance of this conclusion for comparative and genetic psychology needs no emphasis. MENTAL LIFE OF MONKEYS AND APES 69 Problem 2. Second j'rom the Right End Julius was given four days' rest before being presented with problem 2. He was occasionally fed in the apparatus, but regular continuation of training was not necessary to keep him in good form. During this rest interval, locks were attached to the doors of the apparatus so that the experimenter by moving a lever directly in front of him could fasten either one or both of the doors of a given box by a single movement. On May 13 Julius was given opportunity to obtain food from each of the boxes in turn, and trial of the locks was made in order to familiarize him with the new situation. He very quickly discovered that the doors could not be raised when closed, and after two days of preliminary work, he practically abandoned his formerly per- sistent efforts to open them. The locks worked satisfactorily from a mechanical point of view as well as from that of the adapta- tion of the animal to the modified situation. Problem 2 was regularly presented for the first time on May 17, on which day a single series was given. The period of pun- ishment adopted was twenty seconds, and for each successful choice a small piece of banana was given as a reward. After the first trial in this series, in which Julius repeatedly entered the first box at the left, that is box 7, there was but slight ten- dency to reenter the first box at the left of the group. Instead, Julius developed the method of moving box by box toward the right end of the group. The choices were made promptly, and their systematic character enabled the animal to obtain his reward fairly quickly, in spite of the large number of mistakes. In the second series, the orang utan developed the interesting trick of quickly dodging out of the wrong box before the experi- menter could lower the door behind him. This he did only after having been punished for many wrong choices to the point of discouragement. The trick was easily broken up by the sudden lowering of the entrance door as soon as he had passed under it. There appeared on May 21 an unfavorable physical condition which manifested itself, first of all through the eyes which ap- peared dull and bloodshot. On the following day they were inflamed and the lids nearly closed. Julius refused to eat, and experimentation was impossible. Until June 2 careful treat- ment and regulation of diet was necessary. He passed through 70 ROBERT M. YERKES what at the time seemed a rather startHng condition, but rapidly regained his usual good health, and on June 3, although some- what weak and listless, he again worked fairly steadily. Since it was now possible to lock the doors and confine the animal for any desired period, on June 5 the interval of punish- ment was made sixty seconds, and a liberal quantity of banana, beet, or carrot was offered as reward. No increase in the num- ber of successful choices appeared, and Julius showed discourage- ment. Sawdust had been strewn on the floor, and in the inter- vals between trials as well as during confinement in wTong boxes, he took to playing with the sawdust. He would take it up in one hand and pour it from hand to hand until all had slipped through his fingers, then he would scrape together another handful and go through the same process. Often he became so intent on this form of amusement that even when the exit door was raised, he would not immediately go to get the food. The reactive tendencies which appeared in the work on prob- lem 2 will now be presented in order, since I shall have to refer to them repeatedly, and the list will be more useful to the reader at this point than at the conclusion of the presentation of daily results. The following is not an exhaustive list but includes only the most important and conspicuous tendencies or methods together with the dates on which they were most apparent. (a) May 17, choice of first box at left of group or near it, then the next in order, and so on, until the second from the right was reached. This method with irregularities and certain definite skipping was used at various times, sometimes over periods of several days, during the course of the work. (b) June 3, preference for number 3 and number 4 developed immediately after the orang utan's illness and when he was work- ing rather listlessly. On June 9 and 10, the original tendency (a) reappeared and persisted for a number of series. (c) June 14, a tendency to choose the box at or near the right end of a group, and then the one next to it. In connection with this tendency, which of course required only two choices in any given trial, interest in playing with the sawdust on the floor developed. Again on June 21, the animal returned to the use of tendencv (a). MENTAL LIFE OF MONKEYS AND APES 71 (d) June 29, movement to box at right end of group, hesita- tion before it, and turning through a complete circle so that the second box from the right was faced. This, the correct box, was often promptly entered. This method, if persisted in, would obviously have yielded solution of the problem. (e) July 5, approach to and pretense to enter the box next to the right end (right one), and then choice of some other box. This feint is pecuharly interesting, and its origin and persistence are difficult to account for. (f) In connection with the tendency to pretend that he was going to enter the second box from the right end, Julius developed also the tendency to turn around in front of the box at the right end, starting sometimes to back into it, and then to enter, in- stead, the box second from the end. (g) July 6 and 7, a fairly definite tendency to take the one next in order or, instead, to go directly to the right box. (h) July 10, direct first choices without approach to other boxes appeared for the first time on this date. For this problem, it proved impossible to establish and main- tain uniform conditions of experimentation. Instead, because of the failure of the animal to improve and the tendency to discouragement, both punishment and reward had to be altered from time to time, and other and more radical changes were occasionally made in the experimental procedure. Below for the sake of condensed and consecutive presentation, the most important conditions from day to day are arranged in tabular form: Conditions of Experiment from Day to Day for Problem 2 Date Punishment Reward May 17 20 sec. confinement Food in right box for each (Aid after 10 trials) trial " 18 to 21 30 sec. confinement Food (banana) in right box for each trial " 22 to June 2.. Illness, no experiments , June 3 15 sec. confinement Food (banana) in right box for each trial " 4 30 " " Food (banana) in right box for each trial " 5-10 60 " " Beet, carrot and loquat, in addition to banana " 11 10 to 30 sec. confinement Beet, carrot and loquat, in addition to banana " 12 to 15 60 sec. confinement Beet, carrot and loquat, in addition to banana " 16 60 " " Banana and sweet com — for- mer preferred 72 ROBERT M. YERKES Conditions of Experiment from Day to Day for Problem 2 — Continued Date Punishment Reward June 17 (1st series). 60 sec. confinement Food (banana, as in early series) " 17 (2nd series). No confinement in wrong box; Food only for correct first but instead, return to start- choices ing point by way of alleys " 18 to 22 No confinement in wrong box; Food only for correct first but instead, return to start- choices ing point by way of alleys " 22 (2nd series). No punishment; allowed to en- Food for each trial ter boxes until right one was found " 23 Return to starting point. After five wrong choices of a given box the animal was held for 60 sees, in one of the boxes and was then released by way of the exit door and rewarded when the right one was chosen " 23 (2nd series). No punishment Reward for each trial " 24 (1st series). Return to starting point Food only for correct first choices " 24 (2nd series). No punishment Reward for each trial " 25-30 Same as on 24th July 1 (1st series). No punishment " " " " " 1 (2nd series). Return to starting point Reward only for correct first choices " 2-8 Same as on July 1 " 8 (2nd series). No punishment Reward for each trial " 8 (3rd series). Return to starting point Reward only for correct first choices " 9-10 Same as for July 8 (3rd series) " 10 (2nd series). Momentary confinement in Reward for each correct choice wronj. boxes " 12 Return to starting point Reward for correct first choice " 12 (2nd series). 30 sec. confinement Reward for each correct choice " 12 (3rd series). 5 " " " " " "13 30 " " " " " " 14-17 Return to starting point Reward for correct first choices " 17 (2nd series). 60 sec. confinement Reward for each correct choice "19 30 " " " " " " 20-26 10 " " " " " " " " 27-30 Right box indicated by slight Reward in each right box raising of exit door momen- tarily. No punishment " 30 (2nd series). Return to starting point Reward for correct first choices "31 " " " " " " " " " " 31 (2nd series) to Aug. 10. . 10 to 60 sec. confinement Reward for each correct choice Aug. 10 (2nd series). Threatened with whip " " " " " " 11 (1st series). " " " " " " " " " 11 (2nd series). 10 sec. confinement " " " " " " 12 Threatened with whip " " " " « " 12 (2nd series). 10 sec. confinement " " " " " « ig 10 " " " " " " " " 19 (2nd series). Threatened with whip..'... .. " " " MENTAL LIFE OF M(JXKEYS AND APES 73 With the above reactive tendencies and modifications of method in mind we may continue our description of results. On June 9 there developed a tendency to increase the magnitude of the original error by choosing nearer the left end of the groups. This is odd, since one would naturally suppose that an animal as intelligent as the orang utan would tend to avoid the general region in which success was never obtained and to focus atten- tion on the right, as contrasted with the wrong end of each group. It obviously contradicts the law of the gradual elimination of use- less activities. In other words, it is wholly at variance with the principle of trial and error exhibited by many infrahuman organ- isms. Julius, although making many mistakes, worked dili- gently and, for the most part, fairly rapidly. The day's work proved most important because of the change in method and also because of the appearance of hesitation, the rejection of cer- tain boxes, and the definite choice of others. My notes record "this is a most important day for Julius in problem 2;" but subsequent results do not clearly justify this prophecy. The method of choosing the first box at the left and then of moving down the line until the right one was reached was so consistently followed that during a number of days it was pos- sible for me to predict almost ever\^ choice. Indeed, to satisfy my curiosity in this m.atter during a number of series I guessed in advance the box which would be chosen. The percentages of correct guesses ranged from ninety to one hundred. June 10, for example, yielded two series for which the ratio of right to wrong first choices was 0 to 10, and in which the method described above was used consistently throughout. It was inevitable that punishment by confinement and the discouragement resulting therefrom should interfere with the regularity of work and make it extremely difficult to obtain strictly comparable results from series to series and from day to day. The data for this problem, as presented in table 9, have values quite different from those for the monkeys, chiefly because of the more variable conditions of observation. It was occasionally noted that the disintegration of a definite method and the disappearance of the tendency on which it depended occurred rather suddenly. Frequenth^ it happened that having used an inadequate method fairly persistently on a given day, the animal would on the following day exhibit a 74 ROBERT M. YERKES o ^ o o o o o o o LO CO CO § CO oo CO oo O O § § Pi ^2 PJ o 6 c^ "* ;; CM '^ cci coo-. CO G-j o 6 ^ o Ol 00 CO C^ 00 c- Sot LO 00 o Di o - oq 'st" CO CM CO LO^ LO CM o ^ o cr, 00 ■ CD c^ 00 c^t>,xcr. i>ooai a-. 2 o Pi o - CM '^ CO C^l CO CO CM -H COCM-H -HO o 00 c^ c^ f-t- c^ o cr> 00 CO C^ LO LO c^ t^ r> Ot^C^CDC^t^C^CO CD CD CO ^o in to OC !£! -* rf to 1>- CD CD CCCOCOLOCOCDCOLO LOLO LO cr> CO LO 00 LO CO 00 LT ^00^ LOCO c^ SCO ^LO LO LO LOLOLO-^LOLOLO-rf -Nf-Nt"^COo6xf*CO coco CO CO' ^ CM '^'^CD ■ ■ ■ CO CD ■ °^'-^^Lr;LO^ C^C^LO ^00-^00 00 -^00 "^00 tr- CD COCD^0000C0t>C0t> cot>- coc^ 00 .^ LO t^C^LO^^'Tj'OOLOLOOOC^C^CMCDCMCD cvic£) CM CO LO LO ^ CO cn '^ <^ f^ CO t>-^^ "vfCMcoco— iLO'-HLO '^ S o LO '^ '^ ^ CO c^i -*^ '^ O) CO cvi CO CO CO '^ ^ ^ '^ ■TfLO'^f coco CO -o oa CM C^l "* CO CO ■^^CO'^ CO LOCO CM CM M CLh ^ CO CM CM rd CO ^~' CM CM .-H CO —1 CM CMC^1 CM C 5 CO CM -^ CM CM CO coco CM -^ CM CM CM r-^rH ^ c^ 6 «2 00 LO LO CO LO oo CO CO CD LO CO LO CO coco LO t^ UO LO cbo-cb LOCDLO CD CO LOLO CD LO 03 1 •^ ^ ^ ^ LO ^ M< '*'*-;)< "* -*LO^ ^ -^ '^ !X> LO LO LO cc; IC LO Tf "sf ^ CO CO CO LO LO LO LO COLO LOCOT:tCMCO CO coco LOLO LO CO LO CO LO C-t^-^CD LO CO COLO LOCO CD CO CD •^'^ "^ (r> -* CO Cvj "* -* -rf 00 ■"^ I>I>00 l>-00 t>- C/3 !>; c^t-t^ 00 00 cr)t>-oo 00 ■ 00 cr. 00 00 00X00 00 00 00 tr^ir^i> c^ t> t--a5i>- t>- t^t^t^t^ t> t^o OD^ c- O) o <—< s O CO o s o CO oooo O00CT-. O OOO —•CM CO ^LO s o 2 °1 -^ - CM CO ■^ lO CDt>00CT: O^CM co'-f LO 00 I— 1 o CM CM §co —> ■^LO s t> 003 CTi = 2 3 MENTAL LIFE OF MOXKEYS AXD APES 75 CO ^c-. if3 CO c^ LC X t^ C5 ^ c^ to .—1 L': ri- CO L.T CM CO - <£) CO 00 t> C X t^ t> CT. X cr. C7-. • o X X a; cr. 2 X •^x c^. w cc'-'ooc^n -^]-- ^ CVl (Nl - -- c c^:i -^-C^l - t> t^ LCO t> LO t> X toc^ to cc I> LOX "* CDC^CD t>- CO t^ t>COXI> in in c^o^o LOLf: ^ LOLOCn co cox c^ to to to to 00 t>Ti'[>i:^ O- X '.i: UO UO LO LO CCLOCn LO to-^to LOtOXLO CO CO CO ^ ^ 00 '* CM CO c^ COC^]C0 ^ CO ^coco CM CM Ti< ^ I— I ro^cMcoco CO — CO ^ '^^rf C^JCO CM 00^ ^ CO ^ -^1"^'^'* '*^^C0 c^ ^ ■^• ^TT -*^ CMCM^ CO 'Tf -r- X X — t^LOOX. X X X XXt> X 4.5.6.7 6.7.8 7.8 5.6.7 LOX uocx> ! X X 'O LO ^ CM CO X t> c-:XLO t^ to CO CM CM C^J to O x^ t>-tox <£> to ^ >— 1 CO '-H to CO OLC t^t>to (N t^toxoi CO LOCO ^o -^LO nL0^C0lO^t>C0 inc^ LOC^ l>- r^ t> "^ LO CC L.O ^ ^ UO ^ LO to «3«n tOLO t> X c- tDt>t:^t> ccc^ in^'^LCiccLOLOLOLOLn c^ to to to to o to t>t> t^^ iD t^LOXXX LOXtr-cDu' ; ^ LO '-0 'X! en ^ LO lo LO LOtC tOLO T* to toto to 2 in to" in-^LCLo T^ to to to LO LO LO LO LO to to to '.O CO tO'^tOtO to to to to C^l LO CO t^ c^ to t-c^t>t> rtCO Ttr-H CM'* CO CO ^.-ICOCM OOCM CO coco LO CM LO O'OtO ^to LO I>T!>LO to LOLO LO LOC^LO LO LO c^:i tj to to to -* T# r:J< ,-( CO CM ^ -*CO^^ 00 ^ c^x X t> c~- X ci c~. 00 ai cc CO X ai t> crl X i>^ t> ai t> ai ci o ooooo oooo C^ X C". O — Cv] CO '* LO to •—I ^^0]C^5CM CMCMCMCM Tf ^CO '^ ^ ^ T* -* -^' -rfi X C^X t> c. t^ t^ c^t> t>xc^cixc^ C^I> t> t^ C^ CI C^ 1>- ci c-~ t> r> c-~ o"i t-~ t> CO ^ ox t> t>X XX oo "^LO 00 CO "-I ^ CM-r}< . 76 ROBERT M. YERKES 1 -<-' ^ CO CM Q u^ ,_l Q CO o _o CO CM o t> CO o CO LO "oS CM '-' '=* CM ^ CO CM '-^ « »— ( i-H .— I >— 1 '^" •^ •^ •^ ^ ^ ^ CO CM I> lO '^ (Si '~' '"' '"' CM •—t ■-^ '-' Di CO C5 T}< 00 CO LO CO 00 ^ 00 COCO LOOl I>00 CO 00 LO cococ^ 00 t>I> -*00 Di CM ^-* LO— 1 COCM ^ CM LO ^^co CM coco COCM 00 c^ c^ o l> 00 CO LO CO t- 00 t^ LO COO O) LO CO CO tr^co t>t> 00 CO 00 LO lO !>[>• 00 CO 00 00 f^ t> t> t>t^I> 00 t>-o6 c^od-* '^ "* CO CO CO 05 fC CO ^00 -^ CO-* ^ c/i CM '^ CO rfr^f co^ CO CO CO CO -* "^ CO-* CO ^ ^co CO^^'*'* -# ^ ^ ^-*-* CO ^^ CO-* 00 00 ^o-. LO '^LO -* ^ 00 00 t>i jSr-ioO t>00 LO C/3 CO 00 t>^ OOCDCMCT) 00 00 CJCD CO 00 t>00 00 l> OOCTJ I> t^cot^oo t>^t>00 cjicrioooo 00 c^ -* LO ^ LO -^ -*^-* CM I> ■^ LO LO-* LO cmlolo 0) c/j CO CM •* ^co '^ LO LO T^H LO -* -* "* -* -* -* 1 LO LOLO LO LO LO LO LO LO LO LO colOlo LO LO^ ^co CM C 5 C/3 CM CO COCM CM CO CO CM --H CM CM CO COCM CM CO CO CM CM CO CO CM CO COCM COCOCO COCOCM ■— ICOCO CM COCM CM CO 3 COO o^coco coco CO CO CO CO CO CO CO CO CM CO CO CO coco CM CO C c^ S 00 c^t>- LO O <2 LO CO LO 00 00 LOOO C-- t> CO CD t> CO 00 CD t^co C/3 TP t> c-^o6 CO 00 t>t> o 00 00 ooood 00 0C3t> 00 CO 1 CD "i LO COLO CD COLT LO LO ■^ '^ -r}< coco COLT CO CD CO LOCO LO LO LO C/3 CO CM CD CD'^ COCOLOCDCOLO LO CO CD CO CD LOTf -* CO LococomcocDCDCDcbcd CO CO LOCO CD CO coco COLO !>. CO vO CO t^cc CO CO CO LO CD t>^ I>^I> coco t>i t^CD in ^ C^ CDC^ i>ic^coco CO c^ t>^ c^ CO LO^ CO CM t>^ t>^t> cot>coi>tr^ CD l> CO cot>t> I>^ CDt> t^I>i •>* CO -* -* CM PC CM CO ^cc co^co c/) CM -^ CO-* CO CO-*'* CO CO COTf -*-*co COCO ^ "*-# ^^co ■^ "^ "^ ^ ^ CO -t-*^w^-:-' ^ -* coco 00 00 t> 00 1—1 Oj 00 ci O3t>00 t> CAJ t> 00 ai [>t>cy5ooaJ I> 00 00 00 o^ 00 00 00 00 00 00 00 00 00 CTi C5 t> Ol 05 o oo go oo o o Q ooo (-, oo oo t^ 00 c:^ CM CO -* LO CO t^oocr- ^ — (C^1 co^ JK CO coco '^t "^ ^-* xt< "* -* -*-*-* LO LOLO LOLO c5 "^ a 2 1 JL, 1-H f-H .-H >-H .-H J-l J^ fH r^ i-i .-H t-H rH r-H f-H •— 1 <-H '~D C^OO a-, o — iCM CO ^ LO cot- 00 o-- O— 1 c^ico CO coco co-^t ^-* ^ -* ^ -*-*-* ^ LOLO LOLO (U > 3CM ^ LO " CM == ^ - 00 " CM s s S' == = ^ r—{ 3 CMS — ) MENTAL LIFE OF MONKEYS AND APES 77 CO CD (M C^ CO Ci o CM CO G-. t> ^ 00 00 t^ :=; o CI ^ :=; r> CO 00 00 t>LOLnt> c^ ^T C^0C^3t>t> COC--00 CD CD LO ^ ^ CO ^ CVI C^l coimnco co cc 00 00 CO CO t^COCM -^ ^ LO CD CD 00 t- 00 00 00 00 c^ X t^ t>X t>00 t- CDI> LC '^ C^ COXCD LOX'* X ^COCO o t3< 00 CD CD t>- CO'XjCD'^ XC^t>- LO'^'^C^ CO -^-^LOCDCD t> od^LOLO CDXOC CDLOLOOO ^ TfLOLOCDLO CO ^CO CO '^ CO CO ^ ^ o-i CO ^^CO'^CO-^'* CO C^l ^ CO ^ ^ '*'stX ^Ci I>-CDX^CDCDCO t-^'^X^XCO ^CD LD^ • • CD • . . . (^ . . . CTi LOCr, X Cd'^ ^ LO^C-j^C^OLO t^LO^^LOCTl^XLC^XLO'^XX CD cor- CD X '^(y> CD cDLOt-cricDCDcdx c^cb'-icbLnLn'-HC--t>^cDT^cDCDCD c. t-oix t- c-xL-ot- t>! t> CO ci t> c>^ t>^ ^-^ LD X r>^ i> c^ ^- cvi CD CM "* t^ LO ID LO t> CD ^^ LTD COM< LO LO "^ ^ CM cvi ^ LO LO Lfi LO COLO LO ^ LO LOLO^CO^'* ^ ^LOCO^-^-Nf LO LOLOLOLOLOLn LO ^ ■^•^LCCOLOLOCOLO LO CO CM CO CO C^l CO CM CO CO CO CO CO 00 C0COC^]C0C0COCO C^ICM' CM CMCM CvlCMCM CM CO COCOCOCO0OCO:^:iCMC^]CO co CO CMCMCOCO C^1COCO CO CM CO CM CM _ . _ _ . _ t— X t— Loxt>. CD Tj LO 00 C- D^ CO C- CD 00 LO 00 LO LO t^ CD tj- C- xcbooxoc cDLOLOxcb 'e' co CDOOLOOOCO CD 00 00X00 c^cbcbcb xlo c- c-c-t^co t-c-t- t^ -r ^'X> cdcdcd CO LC LO CD LO LO CD CD CO CD LO LO CO LO CO LO LO LO CO CD Cd' LO CO CO CO' -^ LO LO CD CD UO LO LO LO LO LO CD CO -^ t> CO -^ CO CD C^ CO CD CO CO CD CD CD C-~ LO CD LO I> CO CO CO CO CO ^CO"— ICO ^ CO COCOCOCO ■^CO'^ X C-. tr-XO ir-c-oix X X c^ c^ cr. c^ X cr.oo XX cr. X X cj; c^ cr- c\ cr. X X C7-. cr. xt^c-. Oi XX c^c^o-. C^X (T.CC XX X t^ ^ — ^ i6 LO 1 LO c X LO oooo a> o ^ cvT LO CD CD CO o CO CD c CO OCOO LO CD t>- X CO CO CO CO OOO OO'-i CDC-C- o o o s o CO LO 5 LO LO LO XO-jO— 1 LO LO CO CD CM CO CD •^ LO CO r- CD CD CD CO XCiO COCDC^ c^ C^l c-^ ^ c^ LO r- 78 ROBERT M. YERKES < Di o ^ CM 00 CO 00 00 CO LO CO r^ CO g o o "o2 o '"' CM CM CM LO Oi Di '-' ^ -^ ^ ^ -< -I ^ O) CO -* CM -H CM CM LO CM 00 « t> CO 00 (js LO CM ^ LO CO c^ CO 00 00 00 CD 00 t^ CO C- t>.cyi OiCTiOO Pi LO ,^ CO ^ CM CM CM Tj< CM CO ■* ^ 00^ r-I^CM 00 LOLO t^ LO o t^ oo LO 00 t> c^ cot> CO cnio t^ 00 C^ t> t^C^tOCD LO t^'^CO (/) -*COLO CDC-- t> t> 00 00 LO 00 00 -* t> t> CDC- C^ CD LOLO CO LOIOLO LOCO CD LO CD CD CD CD CD ^ '^ LO Tl ";*< 00 D t>^ CO CD -* C75 CD 00 COT}-' 00 CD -^OOOO 00 00 00 CO LO t>^ 00 t^ OOO 00 LO'* 00 00 CD 00 CO CM CO cb^ "*t>cDcocoa^ LO f^ CD CO CD I> CO -^ CO LO CD Tf '^ LO '^ LO ^3" ^ t^ LO Tj- LO LO ■>* LO ^LO ^^ CO '^ '^ c\i -f ^ '^ LO LO LO LO LO ^ LO lOrj LO (/) lO CD t> '^C^ 00 t:^ 00 00t>. CO CDt>- oor^cD -^ CD LO '^LO ^ LO t^ CD C- CD LO t> -^ ^in coLOrr ^ CD LOLO LO in CO LO -^LO •<*co cb^ COLO T^J-CM cbcD CDLO^ '^ CD LO LO ^ ^ ,_^ CD CD CO CD CO CM ^ LO LO CDLO 1— J LO ^ ^r- ( ^CM LO CO-^LO-NfCOLO^COOOTa- ^^LO t^ vO CO CO -* LO ■<* CD CD CO CO C-CD CD ^ U) CO CD LOCO C^ C^ CDt>- C^ LO t> t> CO CM CD LO CO -^C^] CD LO UO CD f^ LO CO CD LO COLO LOLO COLO C^l '^ CO CO ^ ■^rH ^co^ 00 1/5 f«5 00 -^-^OOCMrf!^ COCOCO ■<* CM ■rf ^ Tt* -O -^'*'^ CM CO CO '^COCO CO CO CO CO C^-*C^ J^CM CM CM CO CM CM CM CM CM CM CM CM c/5 00 00 00 OO 00 t> 00 X oi 00 a; 00 X en 00 00' 00 00 00 00 t^ t^ c^ t^ O- OO C- 00 t> t- I>t> t>-t>c^ d 2 en o o 00 o 00 o o 00 0 c c o o o o 01 CO ^ LO CD t^ 00 00 00 00 00 00 00 00 -^ C^l CO 'Tf LO CO t^ 00 oo 00 00 00 00 00 oo Si ooo -HCMCO a; C7i a-. 4J fLO CD - [> ^ — - - ro - - C^] 2 3 CM - oo CM- 01 MENTAL LIFE OF MONKEYS AND APES 79 g CM CM to s o in CM O O to CM LO O t^ in CO CM CM '-^ o ^ Tl --1 -H — 1 O O o o ^ ^ ^ ^ -^ ^ ^ ^ ^ ^ - -' s ^ CM to CM ^ CM O CO t- t> ^ o C5 00 ^ 00 CD 00 2 ^ 2 15 2 tnt^c^ •* -^ COLO to to to ^OOO^OOCMLOLOCOCO ^ CM CO CM CM CM LO traooco to O t^lO ^ '^ ^ tOCOCM'.OCMOOLOlOt>C- to 00 1> 00 00 00 in t> c^ l> c^ I>- I>t> :> (O to 00t>00 00 00 lO to t> in 00 t>t> C£> t> to to to to t> t> I>I>tOt>tOt>t^t^C>tO to 00 O C^ t> t- l> CO CO ■^ CO CM "* CO CO CO CO CO CO CM 00 CO CO 00^ CO ^ CO '^^^'^^COCOCOCOCO CO CO CO CO CO CO CO ooo 00 [>t>C-C^ 00 t^t^ 00 00 00 00 t> to to to 00 C^ 00 tOLOOOOO 00 00 t>^5n 00 00 to to 00 [>itO C3 C^ to t^ to to 00 00 to 00 00 00 tr^ 00 !> to t> O t^OOO <£>LOtJ< t^t>ot>-too ooot>t>t>t>toi>-t>LOCitoooooooaiC^to S '^ a> -* CO r3 ■* LO LO a; ^ '^ ■^* ^* ^* ^* ^* T^CO-* -^co^ LO Tl< '^^ in '^ LO Tj 00 § t> t- C^ t> C^ LOtO c^ IOI>LO cpo t> c-to bfl l> to 00 [:^t^tOt>tO!>toc^t>toiot>c-t>t^c^t>to c •^'^ X2 LOLO tCXOLO >> in tncT) CM^'-( J2 to in LO LO LO LO LO m'^'# LO ^^DtoinLO to LO to intoLOLOtoLOLotoiOLo Tf LO to in in in in '^'^O ^ .2 in Tj-t> en to to toto to to LOLOCMLnCO to to to t- t^ to toc^c^tototototot-to to to to to to to t> c o •a -t-> coco CO C>0 CO CO 'sf'^COCOCO ^co CO CO CM CO CO coco ^ CO Tf -^^^'^^CO'^-^COCO CO CO CO CO CO CO CO 00 00 CO 00 OJ t^ c^ o- c^ 8 00 00 C5 00 00 Ci ■ Oi t> t> t> c~-t> 5 00 t- 00 ooi:^t>ooooootr-oot>oo I>- C- 00 00 00 00 c^ ooo O 00 05 oo o o o o CM § OOOOOOOOOO -TT LO to tr- 00 Ol O — 1 M CO .-h^^^^h^h^hCMC^JCMCM ooooooo '-?• LO 'O tr- X c. O CM C-) M r] 03 M CO t— 1 .— I ■— ( CO^IO Ci C5 Qi s M C/3 o 2 CM CO^LOtOt-OOO-. O— 'CM ^^^^^^,-iCMC^JCM CO -* in to t^ 00 cr- CM OJ CV] CVl CJ C^] CM g5- ^ :- to C^J ^ « G o CO CO - §jCMs C05^3LO33tO 3 < :i t-5 =i Cn=: 2 80 ROBERT M. YERKES Ratio of RtoW 1: 1.14 1: 1.00 1: 0.67 1: 0.67 ^ S 2 CO 00 Di -rj< O CM CM ^ t>'*CD-*-^^'*^ « CO CD Ti< CD CO CO CD (D S. 10 3.4.5.6.7.8 i>o6 t> odcdt^ t^ot^oo cd-^LnC^COCDCDCD S.9 1.2.3.4 CO COCOCMCOCOCOCOCO S.8 1.2.3.4.5 6.7.8.9 00 001>0000 0000 00t>COl>t>00t>C7) CDLO-^cbcOC>CDt>^ C^ £ .2 c S.7 2.3.4.5 S. 6 1.2.3 CM CM CMCMCOCMCMC^CMOO C3 o 05 a; S.5 4.5.6.7.8 C^ !> t> t> t> t> t> CDLTJCOCDCDCOCOC^ S.4 1.2.3.4.5.6 LO LO •"^LnLn-^LOLOi-OLO S.3 2.3.4.5.6.7 CO LO CD CD CO CD CD CD CD S. 2 1.2.3.4 CO ^COCOCOCOCOCOCO S. 1 7.8.9 0000000000000000 t> c>i i>^ t> ai c~- 1> oi No. of trials oooooooo ^CVlCO'^LOCOt^OO cocococococococo o ^ c^i CO -* LO CD r^ cocococococococo Q c/) ?.0- ^^ CM^ CTl^ Ofl,— 1 - ,—1 ' .—1 - ,— 1 - 3 < < MENTAL LIFE OF MONKEYS AND APES 81 wholly different method. Even over night a new method might develop. In the monkeys, although there was occasionally something comparable with this, it was by no means so evident. After two hundred and fifty trials on problem 2 had been given Julius, it seemed desirable to introduce a radical change in method in order to stimulate him to maximal effort. It was therefore decided to force him to make a round trip through the apparatus in connection with each choice, and to let this forced labor serve, in the place of confinement, as punishment for mistakes. This new method yielded peculiar and characteristic results. They differ from those previously obtained largely because of the orang utan's remarkably strong tendency to reenter the box through which he had just passed. This occurred so persist- ently, as may be seen in table 9 (June 17, second series, June 18, etc.), that a further modification of method was introduced in that after the same wrong box had been entered five times in succession, the experimenter on the next choice of the box confined the animal for a stated interval, say sixty seconds, in it, and then allowed it to escape by way of the exit door and choose repeatedly until it finally located the right box. Were it not for this particular feature of the method, the number of choices recorded after June 17 would unquestionably be very much greater than the table indicates. The new method proved a severe test of the orang utan's patience and perseverance, for he had to work much harder than formerly for his reward, and often became much fatigued before completing the regular series of ten trials. Early in the use of this method, he developed the habit of rolling around from exit door to starting point by a series of somersaults. When especially discouraged he would often bump his head against the floor so hard that I could hear the dull thud. As has been noted, I found it desirable to vary the procedure repeatedly. It proved especially interesting to give one series per day with the round trip as punishment and another series with confine- ment as punishment. Day after day, as the experiment progressed, slight or great fluctuations of the ratios of right to wrong choices appeared, but without consistent improvement. There was, to be sure, as the last column of table 9 shows, a radical improvement during the first six hundred and fifty trials, for the number of 82 ROBERT M. YERKES right choices per series increased from 0 to 8. But, as the obser- vations were continued from day to day, it became more and more evident that the animal was merely passing from tendency to tendency — method to method — mixing tendencies, and occa- sionally developing new ones, without approach to the solution of the problem. This fact would have led me to discontinue the work much earlier than I actually did had it not been for the peculiarity of the results obtained with problem 1. It seemed not improbable that at any time Julius might succeed in perfectly solving this problem over night precisely as he had solved the first problem. A curiously interesting bit of behavior appeared for the first time on June 29. Julius had gone to the first box at the right end of the group, and instead of entering, he had wheeled around toward his right, and turning a complete circle, faced the right box, which he promptly entered. Subsequently, the tendency developed and the method was used with increasing frequency. On June 30, it appeared in the first series, four times, in the second series, six times; on July 1, in the first series, three times, and in the second series, four times; on July 2, in the first series, five times, and in the second series, nine times. It was indeed only by accident that the animal failed to fulfill the technical requirement for perfect solution of the problem in this series. Yet, had he done so, his subsequent trials would doubtless have revealed the lack of any other idea than that of turning com- pletely around before entering a box. This odd bit of behavior proved peculiarly interesting and significant in that the tendency to turn became dissociated from the position (in front of the first box at the right end of the group) in connection with which it originally developed. After a few days, Julius would enter the reaction-chamber and instead of proceeding directly to the right end of the group, would stop suddenly wherever he happened to be, turn toward his right in a complete circle, and hasten into the box nearest to him which, as often as not, proved to be the wrong one. Thus the idea of turning completely about, which had it continued its associa- tion with the idea of facing the first box at the right, would have yielded success, instead became useless because of its dissociation. That the orang utan is capable of using free ideas seems clear enough in the light of this behavior. That he proved incapable MENTAL LIFE OF MONKEYS AND APES 83 of getting the idea of second from the right end is as clearly shown by the detailed results of table 9, — the fruits of weeks of experimenting. Certain other interesting tricks developed in Julius's behavior. Thus, on July 5, there appeared the tendency to move as though about to enter the right box (feint), then to stop suddenly and promptly enter another box, which was, of course, a wrong one. The reason for the development of this tendency could not be discovered, but in connection with it, there appeared another tendency which possibly can be explained. Julius took to back- ing into the chosen box so that he could face the experimenter. He would then, after a period of hesitation, come out and promptly enter one of the other boxes. This tendency was apparently due to the fact that during one or two series the experimenter growled at the orang utan every time he made a mistake. The growl startled him and caused him to look around. He evidently felt the need of keeping his eyes on the experi- menter,— hence the backing into the open box. The tendency disappeared shortly after the experimenter gave up the use of the growl as a method of punishing the animal for what were suspected to be careless choices. Curiously enough, it was not until July 10 that direct choice of the right box was made at all frequently. Previously, selec- tion of it had been made almost invariably after approach to other boxes. But in the second series for July 10 there was an extraordinary improvement in method. This developed in the presence of two visitors, and it is therefore all the more sur- prising. The choices were made not only directly, but with decision and evident certainty that was quite at variance with the previous behavior of the animal. All the while through variation of methods, I was seeking to discover the best means of holding the orang utan to his maxi- mum effort and care in attempting to select the right box. One day it would seem as though forcing him to make round trips with rewards only for correct first choices proved most satis- factory, and the next it might seem equally clear that punish- ment by confinement for thirty seconds or sixty seconds, with reward for correct choice in every trial, yielded better results. •In the end I had to admit that no best method had been demon- strated and that I had failed to develop conditions which served 84 ROBERT M. YERKES to compel the animal's attention to the problem and to lead him to work without discouragement. There were, it is true, days on which it seemed practically certain that the problem would be solved, but as it turned out, Julius never succeeded in choos- ing correctly throughout a series of ten trials. As a last resort, in order to make perfectly sure that the orang utan was doing his best, I decided to introduce corporal punish- ment in a mild form. For this purpose, I placed my assistant in charge of the apparatus and the series of trials, and stationed myself in one corner of the reaction-chamber with a whip in my hand. Whenever Julius entered a wrong box, I approached him with the whip and struck at him, being careful not to injure him and rarely striking him at all, for the threat was more effec- tive than a blow. He was extremely afraid of the whip and would begin to w^hine and attempt to get out of the way as soon as he saw it. This method was introduced on August 10, but no improve- ment resulted, and in the end there was no reason to consider it more satisfactory than the other procedures. I am now wholly convinced that Julius did his best to choose correctly in the majority of the numerous series which were given him in connection with problem, 2. From trials 1001 to 1100, a radical departure from the pre- vious methods was introduced in that the right box was indi- cated to the animal by the slight and momentary raising of its exit door. Of course no records of the choices for this group of one hundred trials appear in table 9, for the simple reason that the animal inevitably and immediately entered the right box. It was thought that this method might serve to break up the previously developed tendencies toward inadequate forms of response and so encourage the animal that he would later solve the problem when given opportunity to select the right box without aid from the experimenter. But as a matter of fact, while the ratio of right to wrong first choices was 1 to .67 in the series preceding this change of method, it was 1 to 1.50 in the first series following its use. There is no satisfactory^ evidence that Julius profited by this experience, though as a matter of fact he did succeed in making his best daily record, eight right to two wrong choices, on August 4, after 1190 trials.' The curve of learning for this problem has been plotted and MENTAL LIFE OF MONKEYS AND APES 85 is presented in figure 19. It is of course incomplete and it is offered only to indicate the extreme irregularity in performance. Problem la. First at the Right End It was decided on August 19 that the further continuation of the work of Julius on problem 2 was not worth while. He had become much discouraged, and although willing to work for food, gave no indications whatever of improvement and seemed to have exhausted his methods. It seemed wise instead of giving up work with him in the multiple-choice method to return to a form of problem 1. We may designate it as problem la. The right box is definable as the first at the right end of the series instead of the first at the left end as in the original problem 1. It was thought possible that Julius might quickly solve this problem by a process similar to that used for problem 1. Work was begun on problem la, August 20, and for six suc- cessive days two series of trials per day were given, the settings for which as well as the resulting choices are given in table 10. Most notable in these results is the large number of cases in which Julius chose first the second box from the right end of the series, or in other words that box which had been the right one in prob- lem 2. Contrary to expectation, he showed no inclination to abandon this tendency to choose the second from the right end, and the ratio of right to wrong choices changed in the direction opposite from expectation, beginning with 1 to 4 and ending on the sixth day with 0 to 20. It was obviously useless to continue the experiment further since Julius had given up his attempts to locate the right box in the first choice and was apparently satisfied to discover it by a process of trial and error. He had, it would seem, satisfied himself that the problem was insoluble. These results obtained in problem la constitute a most interesting comment on the effects of problem 2 on the orang utan. Behavior similar to that which he developed well might have been obtained from a child of three to four years placed in a like situation and forced to strive, day after day, to solve a problem beyond its ideational capacity. In many respects the most interesting and to the experimenter the most surprising result of this long series of observations with Julius was the lack of consistent improvement. It seemed 86 ROBERT M. YERKES X o ^ o o 8 8 8 8 8 ^'oS ■^ -* cr. a~. Ci s ^ Di - ^ ^ ^ ^ o ^ ^ to 00 cri cTi g oi -* ^ CM --H — 1 o ^ C^CD Oi t>C7iC7)2'3^«^'22 o X CO— 1 - CO— I-HO--I— lOO o S. 10 1.2.3.4 5.6.7 t>t> C^ c- to t^ to t> t> t> LO I>- c^ LO «o^o in tOLOtOtOtJ^^tOtO "* N© CO c/) o:; LOCO 'X> to to to to to to to to (M UOLO LO tOLOLOlOLOLOLOLO LO 00 p^ CO -^ C^l CM CM CM CM (MC^J CM CMCMCM'-HCM'-H^'-i ^ x> 00 c-oc to 0000 0000 00 j^ C/) ^ OCI>00 t:^t>-cot-o-xc^to 00 t>c^to c^ tOtOCO^LOt^tOl-O t> 1 CA) ^ ro CO CO CO CO CO CO CO CO CO roc^] CM COCMCMCMCMCMCMCM CM i> in l^-t^-t>. tr^ c/) ^■ t>t^ t> t^c^t>t>t>tototo t£> 00 CO^ ^ tOtOtOLOLOLOtOLO LO Ift -* ^ LO LO LOLO LO LOLOLOLOLO ^LO '* Tt^ ^ ^■^^^rT'^LOCOCO CO C7i 0^ a-. ooo: 00 oi ex en 00 ex 00 t> S. 3 6.7.8. ooi:^ooi>.a:a^CT)a-. ocoocooc^oooooi ooot-^ooc^ooooooot^c^c^t^t^c^t^c^ coc^a 3t-0 oot^c^tototototototototo ^ c/) ^ ^rf 'vf T^ Tf ■^ ^ ^ ^ "^ ^ CO coco CO cocococococococo CM '"' sO c/i ^ to to tOtOODtOtOtOtOtO to tOLO LO lOLOLOLOlOlOLOLO LO No. of trials oo ^CM o CO CM c;oooooQO -sfLOtot^ooaiO-H co^LOtot>oocr-o § Q SO- < CM - CO.. ^- LO- to -CM-CM-CM-CM 3 MENTAL LIFE OF MONKEYS AND APES 87 almost incredible that he should continue, day after day, to make incorrect choices in a particular setting while choosing correctly in some other setting which from the standpoint of the experimenter was not more difficult. The evidence suggests that in this young orang utan idea- tional learning tended to replace the simpler mode of problem solution by trial and error. Seemingly incapable of solving his problems by the lower grade process, he strove persistently, and often vainly, to gain insight. He used ideasin effectively. Animals far lower in intelligence (e. g., the pig), surpass him in ability to solve these relational problems because they use the method of elimination by trial consistently and effectively. Julius, in these experiments, made a poor showing because his substitute for trial and error is only slightly developed. Would he have succeeded better with the same problems if mentally mature? There are many important features of the results which, for lack of space, have not been indicated or discussed. They can be developed from later comparative studies of the data, for in the tables appear all of the essential facts of response apart from those mentioned in the text. IV RESULTS OF SUPPLEMENTARY TESTS OF IDEATIONAL BEHAVIOR 1. Julius, Pongo pygmcBUS Box Stacking Experiment In addition to the multiple-choice experiments which have been described in detail in the previous section, it was possible to conduct certain less systematic tests of ideational behavior in the monkeys and the orang utan. From, the technical stand- point these tests were relatively unsatisfactory because onl}^ inexactly describable. But their results are in many respects more interesting, if not also more important, in the light which they throw on ideation than are those previously presented. First, in order of time, comes a test which may be desig- nated as the box stacking experiment. The method will now be described in connection with an account of the behavior of Julius as contrasted with that of a child of three years and four months of age. In the large central cage labelled Z, figure 12, which was twenty -four feet long, ten feet wide, and ten to twelve feet deep, the following situation was arranged. From the center of the wire covering of the cage, a banana was suspended on a string so that it was approximately six feet from the floor, five feet from either side of the cage, and twelve feet from either end. From all approaches it was far beyond the reach of Julius, since it was impossible for him to climb along the wire roof and thus reach the string. Two boxes were placed on the floor of the cage several feet from the point directly under the banana. The one of these boxes was heavy and irregular in shape, as is shown in figures 21, 23 and 24 of plate V. Its greatest height was twenty-one inches; its least height, eighteen inches; its other dimensions, twelve and sixteen inches respec- tively. The smaller and lighter box measured twenty-two by twelve by ten inches. According to the experimenter's calcu- MENTAL LIFE OF MONKEYS AND APES 89 lations, the only way in which JuHus could obtain the banana was by placing the smaller box upon the larger and then climb- ing upon them. At 10 a. m. on March 5, Julius was admitted to the large cage, and the banana was pointed out to him by the experi- menter. He immediately set about tr\^ing to get it, and worked diligently during the whole of the period of observation, which, because of the unfinished condition of some of the cages, was limited to slightly over ten minutes. Within this period he made upward of a dozen fairly well directed attempts to obtain the food. Chief among them were three attempts to reach the banana from different positions on the left wall of the cage (as the experimenter faced the laboratory) ; two attempts to reach it from dift'erent positions on the right wall; two from the large box in positions nearly under the banana; two from the large box with the aid of the experimenter's hand; and one from the distant end of the cagef?). There occurred, also, less definite and easily describable efforts to get at the reward. On account of the unfinished condition of the cages, the experimenter had to remain in the large cage with Julius during the test. This interfered with the experiment because the animal tended both to try to escape and to get the experimenter to help him with his task. Particularh^ interesting is the latter sort of behavior. After the orang utan had made two or three futile attempts to obtain the food he came to the experimenter, who was standing in one corner of the cage, took him by the hand, and led him to a point directly under the banana. He then looked up toward the banana, grasped the experimenter's arm, raised it, and then tried to pull himself up. He was not allowed to get the food by climbing up on the experimenter. A few minutes later, he again led the experimenter toward the banana, but receiving discouragement in this activity', he pro- ceeded to devote himself to other methods. Apart from the distractions which have been mentioned above, Julius's attention to the food was surprisingly constant. Whatever his position with respect to it, he seemed not for an instant to lose his motive, and to whatever part of the cage he W'Cnt and whatever he did during the interval of observation was evidently guided by the strong desire to obtain the banana. Frequently he would look directly at it for a few seconds and 90 ROBERT M. YERKES Explanation of Plate V Orang utan, Julius, obtaining banana by piling boxes or by using pole Figure 21. — Julius in act of setting larger box on end. Figure 22. — Placing smaller box on larger. Figure 23. — Balancing on larger box preparatory to reaching for banana. Figure 24. — Balancing and reaching to the utmost. Figure 25. — Standing on three boxes (after stacking them) and reaching for reward. Figure 26. — Lifting smaller box up toward banana. Figure 27. — The act of stacking the boxes. Figure 28. — Sequel to figure 27. Figure 29. — Box and pole experiment. Pushing the second pole into the box. Figure 30. — Pushing pole into box. Figure 31. — Enjoying the reward of success. PLATE V MENTAL LIFE OF MONKEYS AND APES 91 then try some new method of reaching it. His gaze was delib- erate and in the handHng of the boxes he accurately gauged distances. Several times he succeeded in placing the larger box almost directly under the banana, and repeatedly he located that portion of the side wall from which he could most nearly reach the coveted prize. From my notes I quote the following comment on the results of the initial experiment: " Despite all that has been written concerning the intelligent behavior of the orang utan, I was amazed by Julius's behavior this morning, for it was far more deliberate and apparently reflective as well as more persistently directed toward the goal than I had anticipated. I had looked for sporadic attempts to obtain the banana, with speedy dis- couragement and such fluctuations of attention as would be exhibited by a child of two to four years. But in less than ten minutes Julius made at least ten obvious and well directed attempts to reach the food. There were also wanderings, efforts to obtain aid from the experimenter, and varied attempts to escape from the cage." Before proceeding further with the description of the behavior of Julius in the box stacking test, I shall describe for contrast the behavior of a boy three years four months of age when confronted with a situation practically identical with that which the ape was given an opportunity to meet. For the child, the banana was suspended, as previously described, from the roof of the cage. The same two boxes were placed on the floor at considerable distances from the banana, and in addition, a light stick, about six feet long, and a piece of board, the latter by accident, were on the fl-oor. The child was asked to get the banana for Julius, and he eagerly and confidently volunteered to do so. His behavior may best be described by enumeration of the b eral attempts made. They include (1) placing the larger boA nearly under the banana and reaching from it. (2) Stand- ing of the larger box on end with resulting failure because the child could not stand on the sloping edges of the top of the box. (3) The larger box was turned on its side and the lighter box drawn up opposite it and stood on end. The child then mounted the larger box and from it stepped to the top of the smaller. But the boxes had not been placed beneath the banana, 92 ROBERT M. YERKES and when the child reached for it, he found himself several feet away from his prize. (4) The boxes were moved to a position nearly under the banana and another futile attempt was made to reach it without placing the smaller box on top of the larger one, the only position from which the child could readily obtain it. (5) The piece of board was placed on top of the larger box and from this height the child again reached upward. (6) The six-foot stick was taken up and an attempt was made to strike the banana and thus dislodge it, but it was too securely fastened to be obtained thus. (7) Attention shifted to other things, and the child played for a time with the board. Reminded of the banana by the experimenter, he again tried method (3). (8) He again used the stick on the banana. (9) The effort to knock the prize to the floor having failed, he became discouraged and said that he must go home. (10) When told that Julius was very hungry and wanted the banana, he repeated efforts similar to those described in (3) and (6). Up to this time the observations had covered a period of twenty minutes. The child was now taken from the cage and allowed to play about for fifteen minutes. Asked then whether he would go back and try to get the banana, he replied, " No, 'cause I don't want to get it," thus indicating his discourage- ment with the situation. When taken into the cage, he, never- theless, made the additional attempts indicated below : (11) Use of one of the boxes. (12) He remarked, " Now I know, I'll get it," and after so saying, repeated (3). (13) Failing, he turned to me and said, " I could get it if I was on your head," but he did not, as Julius had done, lead me to the proper place and try to reach the banana by climbing up or by urging me to lift him. (14) Later, he played in the boxes, apparently forget- ful of his task. Finally he remarked: " I'll get the banana," but he made no attempt to do so, and instead, watched the monkeys intently. Thereafter, he showed no further interest in the solution of the problem, and the experiment, after a total period of fifty-five minutes, was discontinued. Comiparison of the behavior of the ape with that of the child indicates a greater variety of ideas for the latter. Julius gauged his distances much more accurately than the child, attended more steadily, and worked more persistentl}^ to obtain the reward, but he did not so nearly approach the idea of stacking MENTAL. LIFE OF MONKEYS AND APES 93 the boxes as did the child, for the latter, in placing the board on one of the boxes, exhibited in ineffective form the idea which should have yielded the solution of the problem. The child was given no further opportunity to work at the problem, whereas Julius, as I shall now describe, continued his efforts on subsequent days under somewhat different conditions. On Wednesday, March 10, the banana was suspended as formerly, and three boxes, all of them small and light enough to be readily handled by the ape, were placed in distant parts of the cage. The six-foot stick which had been present in the test with the child, but not in the first test with Julius, was also placed in the cage. Julius was allowed to work for about an hour. As formerly, he was sufficiently hungry to be eager to get the food and evi- dently tried all of the possible ways which occurred to him. Chief among these were (1) the use of the various boxes sepa- rately or in pairs in very varied positions but never with one upon another, — the only way in which the banana could be reached; (2) climbing to various points on the sides of the cage, with infrequent attempts to reach the banana. Usually his eyes saved him the vain effort. Unlike the child, Julius paid little attention to the six-foot stick. Two or three times he took it up and seemingly reached for the banana, but in no case did he try persistently to strike it and knock it from the string. It is but fair, however, to remark that such an act is very difficult for the young orang utan, as compared with the child, because of the weakness of the legs and the awkwardness of striking from a sitting posture. As previously, the steadiness of attention and the persistence of effort toward the end in view were most surprising. At one time Julius walked to the end of the cage and there happened to see one of the monkeys eating. He watched intently a few seconds and then hastened back to the banana as if his task had been suggested to him by the sight of the feeding animal. Most interesting and significant in this behavior was the suddenness with which he would turn to a new method. It often looked precisely as though a new idea had come to him, and he was all eagerness to try it out. On March 11, Julius was given another opportunity to obtain the banana by the use of the three boxes. Although he used 94 ROBERT M. YERKES them together he made no effort to place one upon another. Certain of his methods are shown in plate V, especially by figures 21, 23 and 24. This experiment was continued on April 2 under yet different conditions, for this time only two boxes were placed in the cage, the one of them the heavy, irregularly shaped box and the other the smaller, lighter one originally used. On the end of the heavier box had been nailed a two by two inch wooden block in order to increase the difficulty in using this box alone. As previously, Julius made varied attempts to obtain the banana, but on the whole his interest and attention seemed somewhat weaker than previously and there were indications of discourage- ment because of repeated failures. He handled the boxes conspicuously well, and it seemed at times that he would certainly succeed in placing the one upon the other and in reaching the food. After one series of attempts from the sides of the cage and frorn the large box, he deliberately turned away from the box and neatly executed a somersault on the floor of the cage, as much as to say, " I am disgusted with the whole situation." Again, later on the same day, after falling from the top of the larger box, which tilted over very easily, he rolled himself into a ball, and childlike, played with his feet. An additional evi- dence of his changed affective attitude toward his task, especially in connection with definite failures, appeared in his rough hand- ling and biting of the boxes. When most impatient, he worked very roughly. Julius was allowed to work for the reward from thirty to ninety minutes, or, as a rule, until he had become completely discouraged on April 3, 5, 6, 7, 8, 9 and 13. His behavior was interesting and significant, but nothing new appeared except that his willingness to work gradually disappeared, and on April 13, although previously hungry, he made only a single attempt to obtain the banana and then paid no further atten- tion to it. The prolonged and varied efforts to obtain the banana were due in a measure at least to three accidental successes. Thus on April 2, 3 and again on the 5th, by fortunate combinations of circumstances, he succeeded in getting the banana, contrary to the intention of the experimenter. MENTAL LIFE OF MONKEYS AND APES 95 Although active at first on April 6, he soon wearied of his task and quit work. The same was true on April 7, and again on the Sth and 9th. On these days, although hungry, he did not care to enter the large cage and worked only a few minutes each day, seldom making more than two or three half-hearted attempts to obtain the banana. His attitude toward the task had changed completely, in that hopelessness had taken the place of eager expectancy. By the 13th of April he had so nearly given up voluntary efforts to solve the problem that it seemed worth while to test his ability to get the idea by watch- ing the experimenter. For this purpose the following test of imitation was made. On the morning of April 14, having placed a banana in the usual position, I took Julius into the large cage, dragged the two boxes to the proper position beneath the banana, placed the smaller one upon the larger one and then climbed up on them to show the ape that I could reach the banana. I then stepped down and gave him a chance to climb on the boxes. He did so immediately and obtained the food. Another piece of banana was supplied, the boxes w^re placed in distant corners of the cage, and fifteen minutes were allowed Julius so to place them that he could obtain his reward. He gave no indications of having profited by my demonstration, but worked with the boxes singly, usually with the larger one. On April 16, with the banana in position and the two boxes also in the cage, Julius was admitted and allowed to work for five minutes, but again without success. I then placed the boxes properly for him and he immediately climbed up and got the banana. While he was eating, the boxes were carried to distant corners of the cage and another banana placed in posi- tion. Now thirty minutes were allowed him for unaided work on the problem. As formerly, the larger box was used repeatedly and attempts to reach from the side of the cage appeared, but there was no tendency to try to use the two boxes together. He worked fairly persistently, however, and showed clearly the stimulating and encouraging effect of aid from the experimenter. Once more, on April 17, Julius was taken into the cage and allowed to watch me place the boxes in proper position. He then climbed up and obtained the desired food. After the bait had been renewed and the boxes displaced, he immediately tried 96 ROBERT M. YERKES to use the larger one, then he reached for the small one as though to use both together. But the impulse died out and he turned again to the larger box as usual, standing it on end, and per- sistently trying to balance himself on it. Nothing else of special interest happened during the interval of unaided effort. Similarly, I placed the boxes for the ape on April 19, allowed him to get the banana and then gave him opportunity to try for himself after the boxes had been displaced. This time he immediately reached for the smaller box and moved it about a little, thus indicating a new association. He next turned to the larger box and worked with it persistently. Later, he once more worked with the smaller box in an unusual manner. He repeatedly stood on it, but made no attempt to lift it or to place it on the larger box. Clearly the usually neglected smaller box had become associated with the satisfaction of obtaining the banana. The same method was carried out on April 20. As I placed the boxes in position beneath the banana, Julius watched with unusual intentness, and when it came his turn to try to obtain the food by the use of the boxes, he began at once to work with the smaller box, but as on April 19, he soon aban- doned it and turned to the other. While I was making note of this particular feature of his behavior, he suddenly seized the smaller box by two corners with his hands and by one edge with his teeth, and after a few attempts placed it on top of the larger box, climbed up, and obtained the banana. Because of bad weather on April 21, the next test was made on April 22, with everything as usual. Unaided, the ape was given an opportunity to obtain the coveted reward, while I stood ready to obtain records of his behavior with my camera. He wasted no time, but piled the smaller box on top of the larger one immediately, and obtained his reward. As soon as opportunity was offered, he repeated the performance. The same thing happened on April 23 and several succeeding dates. Julius had got the idea, and the only further improvement possible was in skill in manipulating the boxes. One of the curious performances which appeared during the imitative period is pictured in figure 26, plate V, where the ape is seen lifting the smaller box into the air. This he did three or four times one day, raising it toward the banana each time as though he expected thus to obtain the reward. As he did MENTAL LIFE OF MONKEYS AND APES 97 not go up with the box (according to his expectation?), he aban- doned this method, and looking about, discovered the larger box in a distant corner. Thereupon, he promptly pulled the boxes to their proper position beneath the banana, stacked them, and obtained his food. After considerable skill had been acquired in the placing of the boxes, the one upon the other, the height of the banana above the floor was increased so that three boxes were neces- sary. Figure 25 of plate V shows him standing on three boxes and reaching upward, and figures 22, 27 and 28 show various modes of handling the boxes and of reaching from them. He was not at all particular as to the stability of his perch, and often mounted the boxes when it seemed to the experimenter inevitable that they should topple over and precipitate him to the floor. Only once, however, during the several days of experimentation did he thus fall. Obviously important is the evident change in the animal's attention on April 20. He watched with a keenness of interest which betokened a dawning idea. Before he had succeeded in stacking the boxes, I had written in my note-book, " He seemed much interested today, in my placing of the boxes." Interest- ing, and important also, is the ease and efficiency with which he met the situation time after time, after this first success. " Trial and error " had no obvious part in the development of the really essential features of the behavior. The ape had the idea and upon it depended for guidance. Except for the fact that Julius was immature, probably under five years of age, it is likely that he would have stacked the boxes spontaneously instead of by suggestion from the experi- menter or imitatively. No unprejudiced psychologist would be likely to interpret the activities of the orang utan in the box-stacking experiment as other than imaginal or ideational. He went directly, and in the most business-like way from point to point, from method to method, trying in turn and more or less persistently or repeat- edly, almost all of the possible ways of obtaining the coveted food. The fact that he did not happen upon the only certain road to success is surprising indeed in view of the many ineffec- tive methods which he used. It seemed almost as though he avoided the easv method. 98 ROBERT M. YERKES It is especially important, in connection with these results, to point out the risk of misinterpretation of observations on the anthropoid apes. If they can imitate human activities as readil}^ and effectively as Julius did in this particular experiment, we can never be sure of the spontaneity of their ideational behavior unless we definitely know that they have had no opportunity to see human beings perform similar acts. Of all the methods of eliciting ideational or allied forms of behavior used in my study of the monkeys and ape, none 3aelded such illuminating results as the box stacking test, and although from the technical standpoint, it has many shortcomings, as a means to qualitative results it has proved invaluable. Other Methods of Obtaining the Reward Some weeks later, I tried to discover how Julius would obtain the much desired banana when the boxes were absent. I placed in the large cage a stick about six feet long and an old broom. When admitted, he looked about for the boxes, but not seeing them, picked up the broom and placing it with the splints down, beneath the banana, he tried to climb it, but as it fell over with him, he abandoned this after a few trials, went to his cage, and picking up some old bags which he used at night as covers, he dragged them out and placed them on the floor beneath the banana. He next put the broom upon them and tried to climb up. This general type of behavior persisted for several min- utes, everything within reach being used as were the bags, as a means of raising him in the desired direction. Finally, he placed his feet on the broom where the handle joins the splints, seized the handle near the top with his hands, drew himself up as far as possible, and then launched himself in the air and tried to seize the banana. On the third attempt he succeeded. Later, he was given a plain stick about five feet long. Figure 32 of plate VI shows him using this to obtain the banana in the manner described above. He would grasp it with one* or both feet, usually one, ten to fifteen inches from the floor of the cage, meanwhile holding with his hands near the top of the stick. He would then, with all his strength, draw himself up suddenly and jump toward the banana. Often he came down rather hard on the cement floor, much to his disgust. Yet another method of obtaining the reward developed a day MENTAL LIFE OF MONKEYS AND APES 99 or two later. A light red- wood stick about five feet long and an inch in its other dimensions was the only object in the cage which could possibly be of use in obtaining the banana. The aim of the experimenter was to discover whether Julius would use this as a club. Previously, in connection with the use of the boxes, he had taken up the same stick two or three times and reached for the banana with it, but in no case had he struck at it or clearly tried to knock it from the string, as did the child most readily and naturally. When provided with this same stick, and it alone, as a means of obtaining the food, he hit upon the fol- lowing interesting method. Placing one end of the stick be- tween a wooden brace and the wire side of the cage, he climbed up to a level with the banana as is shown in figure 33 of plate VI. Then holding with one hand and one foot to a timber of the cage and to the stick with his other foot, he swung outward as far as possible and reached the banana with his free hand. Having once succeeded by the method, he used it whenever given an opportunity. It was impossible for him to make the reach without the use of the small stick, while with it he suc- ceeded fairly easily and regularly. Box and Pole Experiment Following the box stacking test, Julius was given an oppor- tunity to exhibit ideation in another type of experiment. This may be designated the box and pole test. The conditions are describable thus. A strong wooden box eighty-four inches long, by four inches wide, by four inches deep, with open ends, was built with one side hinged. Hasps and padlocks enabled the experimenter to lock this "lid" after food had been placed in the center of the box. This box could be placed in the center of the large cage and there fastened by means of cross bars. It is well shown in position in figure 29, plate V. Two poles each eight feet long and approximately one and a half inches in their other dimensions were the only additional materials in the experiment. On May 1, Julius was allowed to see the experimenter place a half banana in this box, close the lid, lock it in position, and securely fasten the box by means of the cross bars. He was then given opportunity to try to get the banana. The two 100 ROBERT M. YERKES poles lay on opposite sides of the box and near the edges of the cage. Doctor Hamilton and the writer were in the cage watch- ing. Julius looked into the box through one end, and seeing the banana, reached for it. He could not obtain it in this way, so he began to bite at the box and to pull at it with all his strength. During the fifteen minutes allowed him, he worked at the box in a great variety of ways, fooling with the locks which had been attached to the hasps as well as with the cross bars and con- tinually reaching in at the one or the other end. He was some- what distracted by the presence of the two observers and attended rather unsatisfactorily to the task in hand. Not once did he touch the poles, and it is doubtful whether he even noticed them. He was not very hungry at this time, and after a few minutes active work he virtually gave up trying to get the food. Two days later, on May 3, the box was once more placed in position, this time with a half banana in the middle and a small piece of banana near each open end. The two poles lay on the floor of the cage, each several feet distant from the box. Julius was eager for food. When released he went immediately to the box, reached in and obtained a piece of banana from the end nearer the laboratory. He then looked in and saw the piece near the middle of the box. His next move was to pick up the eight foot pole and push it into the box, but before pushing it all the w^ay through, he stopped and began to pull at the box in various ways. Shortly he returned to the pole and twice thrust it in as far as he could reach. The first time, after thrust- ing it all the way through, he pulled it out and examined the end as though expecting the banana to come out with it. After a third attempt he looked into the box, presumably seeing the banana, then turned a backward somersault, came to the end of the cage, and looked at me. Had it been at all possible, he would have taken me by the hand and led me to the box as a helper. After a few seconds, he returned to the pole, pried the lid of the box with it, then gnawed at the pole. For about five minutes he worked fairly rapidly and steadily, using the poles, pulling, gnawing, and walking about. His next move was to go to the opposite end of the box, look in, take the piece of banana which was near the opening, then pick up the second pole, which had not previously been noticed, and after a number of attempts, push it into and through PLATE VI Figure 32. — Julius obtaining banana by using pole to climb up on and spring from. Figure 33. — Using pole to swing out on so that banana could be grasped. Figure 34. — Using stick to draw carrot within reach. MENTAL LIFE OF MONKEYS AND APES 101 the box, looking after it and then pulling it out and looking into the box. Having done this he again came to my end of the cage, and from there returned to tr}^ once more with the pole which he had first used. He pushed this pole all the way through, then walked to the other end of the box, looked in and reaching in, obtained the banana which had been pushed far enough along to be within his grasp. Figures 29, 30 and 31 of plate V show stages of this process. Julius had worked twenty-four minutes with relatively little lost time before succeeding. He had shown almost from the start the idea of using the pole as an instrument, and his sole difficulty was in making the pole serve the desired purpose. The experiment was rendered still more crucial on May 5 by the placing of the two poles upright in opposite corners of the large cage. For a few minutes after he entered the cage, Julius did not see them, and his time was spent pulling and gnawing at the box. Then he discovered one of the poles, seized it, and pushed it into the box. He tried four times, then went and got the other pole and pushed it into the opposite end of the box. Twice he did this, then he returned to the original pole, bringing the second one with him. He pushed it in beside the first, and as it happened, shoved the banana out of the opposite end of the box. But he did not see this, and only after several seconds when he happened to walk to that end of the box did he discover the banana. The total time until success was fifteen minutes. Subsequently the ape became very expert in using the pole to obtain the banana, and often only a minute or two sufficed for success. It was not possible for him to direct the stick very accurately, for when he was in such a position that he could look through the box, he could not work the stick itself. It was, therefore, always a matter of chance whether he obtained the banana immediately or only after a number of trials. Although it is possible that the use of the poles in this experi- ment was due to observation of human activities, it seems prob- able in the light of what we know of the natural behavior of the anthropoid apes that Julius would have solved this problem independently of human influence. It was the expectation of the experimenter that the pole would be used to push the banana through the box, but as a matter of fact the ape used it, first 102 ROBERT M. YERKES of all, to pull the food toward him, thus indicating a natural tendency which is important in connection with the statements just made. Subsequently he learned that the banana must be pushed through and obtained at the farther end of the box. I am not prepared to accept the solution of this problem as satisfactory evidence of ideation, but I do know that few ob- servers could have watched the behavior of the orang utan with- out being convinced that he was acting ideationally. Draiv-in Experiment An interesting contrast with the box and pole test is fur- nisl\ed by what may be called the draw-in experiment. This was planned as a simple test of Julius's ability to use a stick to draw things into his cage from beyond the wire side. A board was placed, as is shown in figure 34 of plate VI, with sides to hold a banana, carrot, or some other bit of food, in position. In the actual test either a carrot or a banana was placed about two feet from the wire netting and a stick two feet long was then put into the cage with the ape. When this situation was first presented to Julius, he looked at the banana, reached for it, and failing, picked up a bag from the floor of the cage and tried to push it through the wire mesh toward the banana. He also used a bit of wire in the same way, but was unable thus to get the food. As soon as a stick was placed in his cage, he grasped it and used it in a very defi- nite, although unskillful, way to pull the banana toward him. He was extremely eager and impatient, but nevertheless per- sistent in his efforts, and within five minutes from the beginning of the first trial, he had succeeded in getting two pieces of banana, using always his left hand to manipulate the stick. This test was repeated a number of times with similar results. He had from the first the ability to use a stick in this way, and the only difficulty with the test as a means of obtaining evidence of ideational behavior is that the possibility of imitation of man cannot be certainly excluded. Lock and Key Test By my assistant it was reported on May 5 that the orang utan had been seen to place a splinter of wood in a padlock which was used on the cages and to work with it persistently. MENTAL LIFE OF MONKEYS AND APES 103 It looked very much like imitation of the human act of using the key, and I therefore planned a test to ascertain whether Julius could readily and skillfully use a key or could learn quickly to do so by watching me. The first test was made on May 15 with a heavy box whose hinged lid was held securely in position by means of a hasp and a padlock. The key, which was not more than an inch in length, was fastened to a six inch piece of wire so that Julius could not readily lose it. With the animal opposite me, I placed a piece of banana in the box, then closed the lid and snapped the padlock. I next handed Julius the key. He im- mediately laid it on the floor opposite him and began biting the box, rolling it around, and occasionally biting also at the lock and pulling at it. During these activities he had pulled the box toward his cage. Now he suddenly looked up to the posi- tion where the banana had been suspended in the box experi- ment. Evidently the box had suggested to him the banana. For thirty minutes he struggled with the box almost contin- uously, chewing persistently at the hinges, the hasp, or the lock. Then he took the key in his teeth and tried to push it into one of the hinges, then into the crack beneath the lid of the box. Subsequentl}^ I allowed him to see me use the key repeatedly, and as a result, he came to use it himself now and then on the edge of the box, but he never succeeded in placing it in the lock, and the outcome of the experiment was total failure on the part of the animal to unfasten the lock of his own initiative or to learn to use the key by watching me do so. I did not make any special attempt to teach him to use the key, but merely gave him opportunity to imitate, and it is by no means impossible that he would have succeeded had the key been larger and had the situation required less accurately coordinated movements. However, it is fair to say that the evidence of the idea of using the key in the lock was unconvincing. My assistant's observation was, perhaps, misleading in so far as it suggested that idea. It may and probably was purely by ac- cident that the animal used the splinter on the padlock. 104 ROBERT M. YERKES 2. Skirrl, Pithecus irus Box Stacking Experiment The monkey Skirrl was tested by means of the box stacking experiment much as JuHus had been. On August 23, with a carrot suspended six feet from the floor of the large cage and three boxes in distant corners, the animal was admitted and his behavior noted. The boxes, which were made of light, thin material, ranged in size from one six inches in its several dimensions to one twenty inches long, thirteen inches wide, and eleven inches deep. Only by using at least two of these boxes was it pos- sible for the animal to reach the carrot. Immediate^ on admission to the cage, Skirrl began to gnaw at the boxes, trying with all his might to tear them to pieces. After some thirty minutes of such effort, interrupted by wander- ings about the cage and attempts to get at the other monkeys, he suddenly went to the largest box of all, set it up on end almost directly under the carrot, mounted it, and looked up at the food. It was still beyond his reach and he made no effort to get it, but instead, he reached from his perch on the big box for the next smaller box, which was approximately sixteen inches, by fourteen, by twelve. This he succeeded in pulling toward him, at the same time raising it slightly from the floor, but his efforts caused the large box to topple over and he quit work. The experiment was discontinued after a few minutes, the total period of observation having been thirty- five minutes. Skirrl handled the boxes with ease and with evident pleasure and interest. He also noticed the carrot at various times during the interval, but his attention was fixed on it only for short periods. The test was continued on August 24 when, instead of a carrot, a half banana was used as bait. It was placed only five feet from the floor, and three boxes were as formerly placed in distant corners of the cage. When admitted, Skirrl looked at the banana, then pulled one of the boxes toward it, but in- stead of mounting, he went to the smallest box and began to gnaw it. Shortly, he mounted the middle sized box and looked up toward the banana, but the box was not directly under the MENTAL LIFE OF MONKEYS AND APES 105 bait, and in any event, it would have been impossible for him to reach it. He next went to the largest box, gnawed it vigor- ously, turned it over several times, and then abandoned it for the middle sized box, from which by skillful use of his teeth and hands, he quickly tore off one side. By this time, apparently without ver^' definitely directed effort on the part of the monkey, all three of the boxes were in the center of the cage and almost directly beneath the banana. Skirrl climbed up on the largest box and made efforts to pull the middle sized one up on to it, the while looking at the banana every few seconds. He did not succeed in getting the boxes properly placed, and after a time began moving them about restlessly. His behavior plainly indicated that hunger was not his chief motive. He was more interested in playing with things or in working with them than in eating, and the satisfaction of tear- ing a box to pieces seemed even greater than that of food. It is especially noteworthy that when Skirrl attempts to dismember a box, instead of starting at random, he searches carefully for a favorable starting point, a place where a board is slightly loosened or where a slight crack or hole enables him to insert his hand or use his teeth effectively. Many times during this experiment he was observed to examine the boxes on all sides in search of some weak point. If no such weak point were found, he shortly left the box; but if he did find a favorable spot, he usually succeeded, before he gave up the attempt, in doing considerable damage to the box. Following the behavior described above, Skirrl returned to the middle sized box, placed it on end under the banana, mounted, and looked upward at the bait, but as it was a few inches beyond his reach, he made no attempt to get it, but instead, after a few seconds, went to the smallest box, and finding a weak point, began to tear it to pieces. Later he rolled what was left of the smallest box close to the other two boxes, nearly under the banana, and the remainder of his time was spent gnawing at the boxes and playing with pieces which he had succeeded in tearing from them. During the remainder of the thirty minute interval of observation, no further attention was given the bait. Again, on August 25, the test was tried, but this time with 106 ROBERT M. YERKES boxes whose edges had been bound with tin so that it was impossible for the monkey to destroy them. He spent several minutes searching for a starting point on the middle sized box, but finding none, he dragged it under the banana, looked up, mounted the box, but, as previously, did not reach for the bait because it was beyond his reach. He then played with the boxes for several minutes. Finally he worked the two smaller boxes to a position directly under the banana, put the middle sized one on end, mounted it, and looked at the bait, but again abandoned the attempt without reaching. During the thirty minutes of observation he made no definite effort to place one box upon another. Three times he mounted one or another of the boxes when it was under the banana or nearly so, but in no case was it possible for him to reach the bait. From the above description of this monkey's behavior, it seems fairly certain that with sufficient opportunity, under strong hunger, he would ultimately succeed in obtaining the bait by the use of two or more boxes. For his somewhat abor- tive and never long continued efforts to drag two boxes together or to place the one upon the other clearly enough indicate a tendency which would ultimately yield success. The possibility of imitation is not excluded, for Skirrl had opportunities to see Julius and the experimenter handle the boxes. Because of the other work which seemed more important at the time, this experiment was not continued further. The re- sults obtained suggest the desirability of testing thoroughly the ability of monkeys to use objects as only the anthropoid apes and man have heretofore been thought capable of using them. Box and Pole Experiment Skirrl was first tested with the box and pole experiment on August 12. As in the case of JuHus, a half banana was placed in the middle of the long box and the attention of the monkey was attracted to the bait by small pieces of carrot placed near each open end. Two poles were placed near the box on the floor of the cage. When admitted to the cage Skirrl went almost directly to the ends of the box, took the pieces of carrot which were in sight, but apparently failed to perceive the bait in the middle of the box. For a while he played with the locks on the box, shoved it about, and amused himself with it, showing MENTAL LIFE OF MONKEYS AND APES 107 no interest in obtaining the food. Later he looked through the box and saw the banana. He then dragged the box about, apparently trying to get it into his cage, but he gave no atten- tion to the poles nor did he make any evident effort to obtain the banana which w^as easily visible in the center of the box. The period of observation was only twelve minutes. On August 24 this experiment was repeated with an import- ant modification of the apparatus in that the wooden lid of the long box had been replaced by a wire cover through which the animal could see the bait. Two poles were as formerly on the floor of the cage, not far from the box. Skirrl almost immediately noticed the banana and tried to get it by gnawing at the box. He did not once reach in at the ends of the box, but he did handle the poles, throwing them about and pound- ing with them. There was not the slightest attempt to use them in obtaining the bait. This experiment was later repeated three times at intervals of a number of days, but in no case did Skirrl show any ten- dency to use the poles as means of obtaining the food. Draw-in Experiment This also was arranged in the same manner as for Julius, and on each of five days Skirrl was allowed at least thirty minutes to work for the bait. Either a banana or a carrot was each day placed on the board well beyond his reach, and one or two, usually two, small sticks were put into his cage. Not once during the several periods of observation did Skirrl make any attempt to use a stick or any other object as a means of drawing the food to him. Instead, he reached persistently with his arm, pulled and gnawed at the wires which were in his way, and occasionally picked up and gnawed or pounded with the sticks in the cage. His attention every now and then would come back to the food, but it tended to fluctuate rather rapidly, and in the regular period of observation, thirty minutes, it is unlikely that he attended to the bait itself for as much as five minutes. In this respect as well as many others, Skirrl's be- havior contrasts sharply with that of the orang utan. The results of this experiment indicate the lack in the monkey of any tendency or ability, apart from training, to use objects as means of obtaining food. Ways of using objects as tools 108 ROBERT M. YERKES which apparently are perfectly natural to the anthropoid apes and to man are rarely employed by the lower primates. Hammer and Nail Test One day I happened to observe Skirrl playing with a staple in his cage. He had found it on the floor where it had fallen and was intently prodding himself with the sharp points, ap- parently enjoying the unusual sensations which he got from sticking the staple into the skin in various portions of his body, and especially into the prepuce. A few days later I saw him playing in similar fashion with a nail which he had found, and still later he was seen to be using a stick to pound the nail with. This suggested to me the hammer and nail test. A heavy spike was driven into an old hammer to serve as an indestructible handle. This hammer, along with a number of large wire nails and a piece of redwood board, was then placed in the monkey's cage. Skirrl immediately took up the hammer, grasping the middle of the handle with his left hand, and with his right hand taking up a nail. He then sat down on the board, examined the nail, placed the pointed end on the board, and with well directed strokes by the use of the head of the hammer drove the nail into the board for the distance of at least an inch. He then tried to pull it out, but was forced to knock it several times with the hammer before he could do so. This performance, during the next few minutes, was repeated several times with variations. Often the side of the hammer was used instead of the head, and occasionally, as is shown in figure 8 of plate H, he seized the hammer well up toward the juncture of the same with the spike. This figure does justice to the performance. At the moment the picture was taken, Skirrl's attention had been attracted by a monkey in an adjoining cage, and he had momentarily looked up from his task, the while holding nail and hammer perfectly still. This test was repeated on various days, and almost uniformly Skirrl showed intense interest in hammer and nails and used them more or less persistently in the manner described. Occasionally, apparently for the sake of variety, he would put the blunt end of the nail on the board and hammer on the point. Again, he would try persistently to drive the nail into the cement floor, MENTAL LIFE OF MONKEYS AND APES 109 and once by accident, when hammer and nails were left in his cage over night, he succeeded in making several holes in the bottom of his sheet iron water pan. There was no doubting the keen satisfaction which the animal took in this form of activity. It is impossible to say that the behavior was not imitative of man, for Skirrl, along with all of the other monkeys, had had abundant opportunity to see carpenters working. But this much can be said against the idea of imitation, — no one of the other animals, not excepting the orang utan, showed any interest what- ever in hammer and nails. Occasionally they would be played with momentarily or pushed about, but Sobke, Jimmie, Gertie, Julius, although given several opportunities to exhibit any ability which they might have to drive nails, made not the least attempt to do so. Evidently we must either conclude that Skirrl had a peculiarly strong imitative tendency in this direction, or in- stead, a pronounced disposition or instinct for the use of objects as tools. It would seem fair to speak of it as an instinct for mechanical activity. Under this same heading may be described Skirrl's reactions to such objects as a handsaw, a padlock, and a water faucet. The saw was given to him in order to test his ability to use it in human fashion, for if he could so expertly imitate the carpenter driving nails, it seems likely that he might also imitate the use of the saw. As a matter of fact, he showed no tendency to use the saw as we do. Instead, he persistently played with it in various ways, at first using it as a sort of plane to scrape with, later often rubbing the teeth over a board so that they cut fairly well, but never as effectively as in the hands of a man. After two or three days' practice with the saw, Skirrl hit upon a method which is, as I understand, used by man in certain countries, namely, that of placing the saw with the teeth up, holding it rigid, and then rubbing the object which is to be sawed over it. This Skirrl succeeded in doing very skillfully, for he would sit down on the floor of the cage, grip with both feet the handle of the saw, with the teeth directed upward, then holding either end in his hands, he would repeatedly rub a stick over the teeth. In this way, of course, he could make the saw cut fairly well. But still more to his liking was the use of a spike instead 110 ROBERT M. YERKES of a stick as an object to rub over the teeth, for with this he was able to make a noise that would have satisfied even a small boy. Further light is shed on the force of the tendency to imitate man by the saw test. After Skirrl had been given an oppor- tunity to show what he could do with the tool spontaneously, I demonstrated to him the approved human way of sawing. Often he would watch my performance intently as though fas- cinated by the sound and motion, but when given the tool he invariably followed his own methods. Although I repeated this test of imitation several times on three different days, the re- sults were wholly negative. Other Activities One day as Skirrl was being returned to his own cage by way of the larger cage, he picked up an unfastened padlock and carried it into the cage with him. For more than an hour he amused himself almost without interruption by playing with this lock. The things which he did with it during that time would require pages to describe. Flis interest in it was very similar to that which he had exhibited in hammer and nails, saw, and indeed any objects which he could play with. The lock was pounded in various ways, bitten, poked with nails, hooked into the wires of the cage, used to pull on, pounded with a stick, used to hammer on the floor of the cage with, and in fine, manipulated in quite as great a variety of ways as a human being could have discovered. Finally it was hooked to the side of the cage and snapped shut, and as Skirrl was unable to dislodge it from this position, he shortly gave up playing with it. At the end of the large cage and just outside the wire netting was a faucet to which a hose was usually attached. The valve could be opened by turning a wheel-shaped hand piece. Both Skirrl and Julius learned to turn this wheel in order to get water to play with, but usually the former's strength was not sufficient to turn on the water. The latter could do it readily. The indications are that both animals profited by seeing human beings turn on the water. This unquestionably attracted their attention to the faucet, and probably by playing with it they accidentally happened upon the proper movement. At any MENTAL LIFE OF MONKEYS AND APES 111 rate, Skirrl's behavior was significant in this connection, for he would pick up the hose to see if water were flowing, and if it were not, he would throw it down, go directly to the faucet, and try to turn the wheel. The association of the wheel with the desired flow of water was therefore definitely established. Shall we describe the act as ideational? It seems the natural thing to do. 3. Sobke, Pithecus rhesus Box Stacking Experiment For this test, in the case of Sobke, three light boxes made of redwood about one-third of an inch thick were used. The smallest, box 1, was six inches in each direction, the next larger, box 2, was twelve inches, and the third, box 3, eighteen inches. As in the case of the other animals, bait, either banana or carrot, was suspended from the middle of the roof of the large cage at such distance from the floor as to be reached by the animal only by the use of the boxes. The first observations on Sobke were made on June 14. The three boxes had been placed in the form of a pyramid directly under the banana, which hung about eighteen inches above the uppermost box. Sobke' s attention while in his cage had been attracted to the bait by seeing me fastening it in position, but when admitted to the large cage, he simply glanced at it and then wandered about the cage, picking up bits of food and struggling to get at the other monkeys. This he did for about five minutes. He then went to the boxes, placed his hands on top of the bottom one, but did not climb up on it. A few minutes later he returned to the box again, climbed up, and readily reached the food, which he ate while resting on boxes 1 and 2. I now replaced the bait and gave the monkey a second chance to obtain it. Almost immediately he climbed up as far as the second box, but although he could reach the banana only from the uppermost box, he deliberately shoved it off to the ground and sat down upon box 2. As he was unable to obtain the banana from this, he soon began to gnaw and pull at it, and as he was succeeding all too well in his efforts to tear the box to pieces, he had to be returned to his cage. The most important features of his behavior were, first, his 112 ROBERT M. YERKES stealthy and indirect manner, and second, his failure to use other means of obtaining the bait than that supplied by the observer. Instead of looking straight at the experimenter, or at the object which he wished to obtain, he apparently looked and attended elsewhere. For this reason it was often difficult to decide whether or not he had noticed the bait or the boxes. Finally I was led to conclude that he usually knew exactly what was going on and had in his furtive way noted ail of the essential features of the situation, and that his manner was extremely indicative of his mental attitude of limited trust. Both Julius and Skirrl went to the opposite extreme in the matter of direct- ness, or as we should say in human relations, frankness. They would look the experimenter directly in the eye, and they usu- ally gazed intently at anything, such for example as the bait, that interested them. Sobke, even when very hungry, instead of going directly toward the bait, and trying to obtain it, usually did various other things as though pretending that he had no interest in food. On the following day, June 15, the three boxes were again placed nearly under the banana, but this time the two smaller boxes, numbers 1 and 2, were pushed to the extreme end of the lower box and so far from the bait that it could not be reached from box 1. It was necessary then for the animal to push boxes 1 and 2 along on box 3 until they were nearer the bait. Sobke, when admitted to the cage, evidently noticed the banana, but as formerly, he made no immediate effort to obtain it. After w^andering in search of food and quarreling with the other monkeys for several minutes, he went to the boxes, pushed the topmost one, number 1, off on to the floor, and then carried it into his cage where he quickly tore one side off. He next returned to the large cage, climbed up on box 2, and he was able, by jumping, to reach and obtain the banana. As Sobke was very good at jumping, his new method ren- dered the box stacking experiment of uncertain value, since it was next to impossible so to arrange the spatial relations of bait and boxes that he should be neither discouraged by too great a distance nor encouraged to jump by too small a distance. Evidently it would be more satisfactory to simplify the con- ditions by trying to discover, first of all, whether he would use a single box as a means of reaching the reward. MENTAL LIFE OF MONKEYS AND APES 113 In pursuance of this idea, I suspended a piece of bread five feet from the floor of the cage, and a few feet to one side of it, I placed a box from which it could be reached, or at least easily seized by jum.ping. Sobke shortly walked to a point beneath the bait and leaping into the air, seized it. I then replaced the bait, raising it to a height of five feet ten inches from the floor of the cage. When I had retired, Sobke placed himself in the proper position beneath, looked up at it, but went away without jumping for it. During the remaining ten minutes of observation, he paid no further atten- tion to the bait, having satisfied himself evidently that it was beyond his reach. ]My use of this test was concluded on June 16 when once more I suspended a piece of bread six feet from the floor and placed a few feet to one side the eighteen inch box, number 3, from which had the monkey pushed it to a point directly under the bread, he could have obtained the food easily. Sobke noticed the food promptly, and from time to time as he wandered about, he glanced at it out of the corner of his eye, but not once did he sit down and look at it steadily and directly as Julius and Skirrl might have done. In the first twenty minutes of observation the monkey made no attempt either to use the box or to reach the food by jump- ing. I then placed the box directly under the bait, and scarcely had I withdrawn from the cage before Sobke climbed up on it and looked toward the food. He could not reach it without jumping, and he made no effort to get it. I had left a second box in the cage, — one which I had been using as a seat. Sobke now went to this box, placed his hands on it, looked toward the bait, and then went to a distant part of the cage. No further indications were obtained during the remainder of the period of observation of interest in the boxes as possible means of obtaining the desired food. It is of course obvious that this experiment was not long enough continued to justify the conclusion that either Sobke or Skirrl could not use the boxes or even learn to place one box upon another in order to obtain the bait. The experiment, like several others which are being described briefly, was used to supplement the multiple-choice experiment, and the experi- menter's chief interest was to discover the number and variety 114 ROBERT M. YERKES of methods which would be used by the animal in the first few presentations of a situation. It is practically certain that both of these monkeys would have succeeded ultimately in solving the problem of obtaining the food had they been left in the cage with a number of boxes, for Skirrl very early indicated interest in moving the boxes about, and Sobke showed a ten- dency in that direction which perhaps was inhibited partially by his distrust of the experimenter. Draw-in Experiment For Sobke, as for Julius and Skirrl, the draw-in test was made by putting food on a shelf outside the cage, beyond the reach of the animal, and placing in the cage with the animal one or two sticks long enough to be used for drawing in the bait. Sobke was first given this test on July 24. He tried per- sistently to reach the banana with his hand, seized the box which supported the bait, shook it, picked up one or other of the sticks, and chewed at it repeatedly, but not once did he make any move to use a stick to draw the food toward him. This experiment was repeated on July 27, 29, 30 and 31, a period of thirty minutes being allowed on each day for obser- vation. At no tim.e did Sobke show any inclination to use either a stick or any other object as a means of reaching the bait. Instead, he confined himself strictly to the use of hands and teeth. This test makes it fairly certain that Sobke had no natural tendency to use objects as tools. In so far as he attended to things about the cage or laboratory, it seemed to be rather to play with them in a general way than to use them ideationally or otherwise for definite purposes. The definitely negative result of the draw-in experiment ren- dered needless prolonged observation with the box and pole test, whose results are now to be presented. Box and Pole Experiment The eighty-four inch box, previously used for a similar test with Julius, was presented to Sobke on August 24, the wooden cover having been replaced by a wire one so that the monkey could readily see the bait in the middle of the box. Sobke, MENTAL LIFE OF MONKEYS AND APES 115 when admitted to the large cage, went directly to the box and at once discovered the banana which was midway between the ends. He evidently desired it. Shortly, he went to one end of the box and looked in. This he repeated later. He also shook the box and tried to pull it about and tear it with his teeth, but to the tw^o poles lying nearby on the floor of the cage he gave not the slightest attention during a thirty minute period of observation. The experiment was not repeated because of more important work. Other Activities In more respects than I have taken time to enumerate in the above descriptions of behavior, the relations of Sobke to objects differed from those of Skirrl, and still more from those of Julius. Hammer, nails, saw, stones, sticks, locks, and various other objects received relatively little attention from Sobke unless they happened to come in his way ; then they were usually pushed aside with but scant notice. Rarely he would carry something to the shelf of his cage with him, but as a rule only to lay it down and attend to something else. Skirrl, on the contrary, attended persistently to anything new in the shape of a movable object. He was extremely partial to objects which could be manipulated by him in various ways, and especially to any- thing with which he could make a noise. His interest in hammer and nails, saw, locks, etc., seemed never to wane. I have seen him play for an hour almost uninterruptedly with a hammer and a nail, or even with a big spike which he could use to pry about his cage. In the absence of anything more interesting, even a staple or a small nail might receive his undivided atten- tion for minutes at a time. How important is the species difference in this connection, I have no means to judge, but if we may not consider these different modes of behavior character- istic of P. rhesus as contrasted with P. irus, we must conclude that remarkable individual differences exist among monkeys, for whereas Skirrl is by nature a mechanical genius, Sobke has apparently no such disposition. I can imagine no more fasci- nating task than the careful analytical study of the tempera- ments of these two animals. Skirrl's behavior has importantly modified my conception of genius. MISCELLANEOUS OBSERVATIONS 1. Right- and lejt-handedness Several years ago Doctor Hamilton reported to me observa- tions which he had made on preference for the right or left paw in dogs. He has not, I believe, published an account of his work. Subsequently, Franz observed a similar preference in monkeys which, according to his report, exhibit marked tendency to be right-handed, left-handed, or ambidextrous. My own observations, although they are wholly incidental to my other work, seem worthy of description at this point. I noted, first of all, that the orang utan Julius tended to use his left hand. He by no means limited himself to this, but in difficult situations he almost invariably reached for food or manipulated objects in connection with food getting with the left hand. Figures 23 and 24 of plate V, show him reaching for a banana with the left hand. Likewise, figure 34 exhibits the use of the left hand in the draw-in experiment. So marked was Julius's preference for his left hand that I became interested in observing similar phenomena in the monkeys. Skirrl, when driving nails, held the hammer with his left hand and the nail with his right hand. The fact that he never was observed to reverse the use of the hands is surprising, for other observations indicate that he preferred the right hand for certain acts. Stimulated by the obvious left-handedness, in certain con- nections, of Julius and Skirrl, I tested the preference of several of the monkeys in the following simple way. Standing outside the cage I would hold out a peanut to a hungry animal, keeping it so far from the cage that the monkey could barely reach it with its fingers. I noted the hand which was used to grasp the food. Next I varied the procedure by placing the peanut on a board in order to make sure that I was not definitely directing the animal's attention. MENTAL LIFE OF MONKEYS AND APES 117 With Sobke the following results were obtained. In forty trials given on two different days, he reached for and obtained the food each time with his left hand. Only by holding the bait well toward the right side of his body was it possible to induce him to use the right hand. So far as ma^^ be judged from these observations and from others in connection with the experiments, this animal is definitely left-handed. With Skirrl the results are strikingly different. As stated above, he used the hammer consistently with his left hand, but in twenty attempts to obtain food by reaching, he used his right hand seventeen times and his left only three times. It was quite as difficult to induce him to use his left hand for this purpose as it was to induce Sobke to use his right. We must therefore conclude that Skirrl is right-handed in connection with certain movements and left-handed in others. The monkey named Gertie in the reaching experiment con- sistently used her left hand, never once using the right. Jimmie, so far as it was possible to make tests with him, also used his left hand, but it should be said that the results are unsatisfactory because he was at the time extremely pug- nacious and paid attention to the experimenter rather than to the food. Scotty, in the first series of ten trials, used his right hand eight times, his left twice. In the second series, given the following day, he used the right hand three times and the left seven times. From this we should have to infer that he is ambidextrous. A female rhesus monkey which had been brought to the labo- ratory only a few days previously showed a preference for the right hand by the use of it fourteen times to six. In connection with these data which are, I should repeat, too scanty to be of any considerable value, I wish to describe my own experience. Although naturally left-handed, I am by training right-handed to the extent of having been able to use my hands in writing and in various other activities equally well at the age of twelve. I am at present ambidextrous in that there are many things which I do with equal readiness and skill with either hand. Delicate, exact, and finely coordinated movements, such as those of writing and using surgical instru- ments, I perform always with my left hand, while grosser move- 118 ROBERT M. YERKES merits involving the whole hand or arm, I am rather likely to perform with my right hand. It seems not improbable in the light of my own experience that we shall find some specialization among the lower animals with respect to preference for right and left hand or arm. I should not be at all surprised to discover that it is the rule for animals to possess or to develop readily definite preference for one hand in connection with a given act of skill and to have quite as definite a preference for the other hand in connection with a radically different kind of act. 2. Instinct and emotion Of the rnany presumably instinctive modes of behavior which were observed, only those which have to do with social relations seem especially worth reporting. From among them I shall select for description a few which have already been referred to in connection with the experimental observations. Maternal Instinct Aspects of the maternal instinct I had opportunity to observe in Gertie, who on February 27 gave birth to a male infant, I present below the substance of a previously published note on her behavior (Yerkes, 1915). "On February 27 one of the monkeys of our collection gave birth, in the cages at Montecito, to a male infant. The mother is a Macacus cynomolgtis rhesus [P. irits rhesus) who has been described by Hamilton (1914, p. 298) as 'Monkey 9, Gertie, M. cynomolgus rhesus {P. iriis rhesus). Age, 3 years 2 months. (She is now. May 1, 1915, 4 years and 6 months.) Daughter of monkeys 3 and 10. First pregnancy began September, 1913.' The result of this pregnancy was, I am informed, a still-birth. " The second pregnancy, which shall now especially concern us, resulted likewise in a still-birth. Parturition occurred Satur- day night, and the writer first observed the behavior of the mother the following Monday morning. In the meantime the laboratory attendant had obtained the data upon which I base the above statements. "At the time of parturition Gertie was in a 6 by 6 by 12 foot out-door cage containing a small shelter box, with an excep- MENTAL LIFE OF MONKEYS AND APES 119 tionally quiet and gentle male (not the father of the infant) who is designated in Hamilton's paper as Monkey 28, Scotty. " My notes record the following exceptionally interesting and genetically important behavior. On March 1 , when I approached her cage, Gertie was sitting on the floor with the infant held in one hand while she fingered its eyelids and eyes with the other. Scotty sat close beside her watching intently. When disturbed by me the mother carried her infant to a shelf at the top of the cage. Repeatedly attempts were made to remove the dead baby, but they were futile because Gertie either held it in her hands or sat close beside it ready to seize it at the slightest disturbance. " Especially noteworthy on this, the second day after the birth of the infant, are the male's, as well as the female's, keen interest in the body and their frequent examinations of the e^^es, as if in attempts to open them. Often, also, the mother searched the body for fleas. " Observations were made from day to day, and each day opportunity was sought to remove the body without seriously frightening or exciting the female. No such opportunity came, and during the second week the corpse so far decomposed that, with constant handling and licking by the adults, it rapidly wore away. By the third week there remained only the shriv- eled skin covering a few fragments of bone, and the open skull from the cavity of which the brain had been removed. This the mother never lost sight of: even when eating she either held it in one hand or foot, or laid it beside her within easy reach. " Gradually this remnant became still further reduced until on March 31 there existed only a strip of dr^^ skin about four inches long with a tail-like appendage of nearly the same length. " The male, Scotty, on this date was removed to another cage. Gertie made a great fuss, jumping about excitedly and uttering plaintive cries when she discovered that her mate was gone. Whenever I approached her cage she scurried into the shelter box and stayed there while I was near. This behavior I never before had observed. It continued for two da^^s. On i\.pril 2, it was noted that she had lost her recently acquired shyness and she no longer made any attempts to avoid me. As usual, on this date, she was carrying the remnant about with her. " The following day, April 3, Gertie was lured from her cage to a large adjoining compartment for certain experimental obser- 120 ROBERT M. YERKES vations. After she had been returned to her own cage the remnant was noticed on the floor of the large cage. I picked it up. Gertie evidently noticed my act, for although at a distance of at least ten feet from, me, she made a sharp outcry and sprang to the side of the cage nearest me. I held the piece of skin (it looked more like a bit of rat skin than the remains of a monkey) out to her and she immediately seized it and rushed with it to the shelf at the top of the cage. " Two days later the remnant was missing, and careful search failed to discover it in the cage. It is probable that Gertie had carelessly left it lying on the floor whence it was washed out when the cages were cleaned. On this date Gertie seemed quieter than for weeks previously. " Thus it appears that during a period of five weeks the in- stinct to protect her offspring impelled this monkey to carry its gradually vanishing remains about with her and to watch over them so assiduously that it was utterly impossible to take them from her except by force. "After reading this note in manuscript. Doctor Hamilton in- formed me that Gertie had behaved toward her first still-birth as toward her second. And, further, that Grace, a baboon, also carried a still-birth about for weeks. " I am now heartily glad that my early efforts to remove the corpse were futile, for this record of the persistence of ma- ternal behavior seems to me of very unusual interest to the genetic psychologist." Fear In connection with the multiple-choice experiments Skirrl exhibited what seemed to be instinctive fear as a result of his unfortunate experience with nails in the floor of box 1. He seemingly referred his misadventure to some unseen enemy under the floor, and this in spite of the fact that he was given abundant opportunity to examine the floor of the box, but not until after the dangerous nails had been clinched. His long continued avoidance of the experiment boxes and his still more persis- tent hesitanc}^ in entering them, coupled with his almost ludi- crous efforts to see beneath the floor through the holes cut for the staples on the doors, gave me the impression of supersti- tious fear of the unseen. As I watched and recorded his beha- vior day after day during the period of most pronounced fear, MENTAL LIFE OF MONKEYS AND APES 121 I could not avoid the thought that the instinctive fear of snakes had something to do with his pecuhar actions, and although I have never studied either the natural or the acquired re- sponses of monkeys to snakes, I suspect that lacking such in- stinctive equipment, Skirrl would have behaved differently as a result of the pricks which he received from the nails. It is needless to redescribe his acquired fear of whiteness as it manifested itself in the freshly painted apparatus. Accompanying these instructive modes of response and their emotions are suggestions of peculiarly interesting problems as well as of modes of attacking them. As a matter of fact, Skirrl's fear-reactions did much to alter my conception of the constitu- tion of his mind. I should not have been surprised by the features of behavior exhibited, but I was by no means prepared for their persistence, and for the highly emotional attitude toward the particular situation. Only an organism of complexly constituted nervous system and fairly highly developed affec- tive life could be expected to respond as did this monkey. As has been suggested above, I find the appeal to instinct, modi- fied by experience, a natural mode of accounting for the unex- pected features of Skirrl's behavior. Sympathy The instinctive playfulness of the young monkey Tiny con- trasted most strikingly with the more serious, if not more sedate, modes of behavior of the older individuals. During the greater part of my period of observation Tiny was cage-mate of Scotty, the most calm and apparently lazy of all the monkeys. Tin^^ delighted in teasing Scotty, and her varied modes of mildly tormenting him and of stirring him to pursuit or to retaliation were as interesting as they were amusing. Her most common trick was to steal up behind him and pull the hair of his back, or seize his tail with her hands or teeth. Often when he was asleep she would suddenh^ run to him, give a sudden jerk at a handful of hairs, and leap away. He was surprisingly patient, and I never saw him treat her roughly in retaliation. Another of Tiny's favorite forms of amusement was that of trying to stir up the other monkeys to attacks on one another. She very cleverly did this by pretending that she herself w^as 122 ROBERT M. YERKES being attacked. The instant the older animals began to show hostility toward one another she would leap out of the way and watch the disturbance with evident satisfaction. It was this mode of behavior in the little animal which ultimately pro- vided opportunity for the observations which I wish now to report as indicative of sympathetic, possibly I may say al- truistic, emotions. Tiny was confined with Scotty in a cage adjoining the one in which Jimmie and Gertie were being kept. The cages were separated by wire netting of half -inch miesh. One morning as I was w^atching the behavior of the animals in the several cages, I noticed Tiny dressing with her teeth a wounded finger. It had evidently been bitten by one of the other animals, in all probability either by Jimmie or Gertie. Tiny was trimming away the loose bits of skin very neatly and cleansing the wound. After w^orking at this task for a few minutes, she quickly climbed up to the shelf near the top of her cage, and by rushing to the partition wire between the two cages, she lured Gertie to an attempted attack on her. Gertie sprang up to the partition, placed her hands on it, with the fingers projecting through the meshes, and attempted to seize Tiny's fingers with her teeth. But the latter was too quick for her, and withdrawing her hands, like a flash seized in her teeth the middle finger of Gertie's left hand. She then bit it severely and with all her might, at the same time pulling and twisting violently, often placing the entire weight of her body on the finger. Her sharp teeth cut to the bone, and it was impossible for the larger and stronger monkey to tear away. For several seconds this continued, then Gertie succeeded in escaping, whereupon she at once retreated to the opposite end of her shelf and proceeded to attend to her injured finger. She cried, wrung her hands, and from time to time placed the finger in her mouth as though in an effort to relieve the pain. By this time Jimmie's attention had been attracted by the dis- turbance and he rushed up to the shelf, and facing Gertie, watched her intently for a few seconds. The look of puzzled concern on his face was most amusing. Apparently he felt dimly that something in which he should have intelligent interest was going on, but was unable wholly to understand the situation. After watching Gertie for a time and tr3'ing to discover what MENTAL LIFE OF MONKEYS AND APES 123 she was doing, which was rendered difficult by her tendency to turn away from him, in order to shield her injured finger, he rushed over to the wire partition and made strenuous efforts to seize Tiny with his hands and teeth. But although she continued close to the partition and often crowded against it with face and hands flattened on the wires, he was not able to get hold of her, and after a few vain attempts he returned to his mate, and again with evident solicitousness and the most troubled expression, watched her wringing her hands and chew- ing or sucking at her injured finger. Shortly he again returned to the partition and renewed his attempts to seize the young monkey. Thus he went back and forth from one place of in- terest to the other several times, but being unable to achieve anything at either point, he finally gave up and returned to his breakfast on the floor of the cage. I report this incident fully because the behavior of Jimmie was in marked contrast with the usual behavior of the mon- keys. Selfishness seemed ever^n^^here dominant, while clear in- dications of sympathetic emotions were rare indeed. The above is undoubtedly the best evidence of anything altruistic that I obtained. It is possible that Tiny's action was retaliatory, but although it is practical^ certain that either Gertie or Jimmie inflicted the wound on her finger, I of course cannot be sure that the spirit of revenge stirred her to punish Gertie so severely. Jim- mie's part in the whole affair is, however, perfectly intelligible from our human point of view, and there seems no reason to doubt that he did experience something like a feeling of sym- pathy with his mate, coupled with a feeling of resentment or anger against Tin}^ VI HISTORICAL AND CRITICAL DISCUSSION OF IDEATIONAL BEHAVIOR IN MONKEYS AND APES It is my purpose in this section to indicate the relations of my work on monkeys and apes to that of other investigators. Although throughout the report I have used freely the psy- chological terms idea and ideation, it has been my aim to describe the behavior of my animals rather than to interpret it or speculate concerning its accompaniments. Certain acts I have designated as ideational simply because they seemed to exhibit the essential features of what we call ideational behavior in man. Further study may, and probably will, modify my opinion concerning this matter. It is of prime importance to analyze ideational behavior so that it may be accuratel}^ described and satisfactorily defined in terms of its distinguishing charac- teristics. I had hoped to be able to present a tentative analysis in this report, but the results of my efforts are so unsatisfactory that I do not feel justified in publishing them. The terms idea and ideation have been used to designate contents of consciousness which are primarily representative. Nowhere have I attempted to indicate different types or grades of ideational behavior and nowhere have I found it necessary to emphasize differences between image and idea. In general, the acts which I have called ideational have been highly adap- tive, and the learning processes in connection with which they have appeared have differed strikingly from those of the selec- tive sort in their abruptness of appearance. Extremely interesting and valuable definitions of ideation and discussions of the characteristics of different sorts of ideas in the light of original observations on monkeys have been pre- sented bv Thorndike (1901, pp. 1, 2; 1911, p. 174); Kinnaman (1902, p.' 200); and Hobhouse (1915, p. 270). As these authors have contributed importantly to our knowledge of the behavior of monkeys, their discussions of the meaning of terms are especi- ally valuable. Serviceable definitions are to be found, also, in MENTAL LIFE OF MONKEYS AND APES 125 Romanes (1900), Morgan (1906), Washburn (1908), and Holmes (1911). Evidences of Ideation in Monkeys Aside from anecdotal and traveller's notes on the behavior of monkeys and apes we have only a scanty literature. In fact, the really excellent articles on the behavior and mental life of these animals may be counted on one's fingers; and not more than half of these are experimental studies. I shall, in this brief historical sketch, neglect entirely the anecdotal litera- ture, since my own work is primarily experimental, and since its results should naturally be compared with those of other experimenters. Thorndike (1901), the American pioneer in the application of the experimental method to the study of mind in animals, pub- lished the first notable paper on the psychology of monkeys. His results force the conclusion that "free ideas" seldom appear in the monkey mind and have a relatively small part in beha- vior. That the species of Cebus which he observed exhibits various form.s of ideation he is willing to admit. But he insists that it is of surprisingly little importance in comparison with what the general behavior of monkeys as known in captivity and as described by the anecdotal writers have led us to expect. It is important to note, however, that Thorndike' s observations were limited to Cebus monkeys which, as contrasted with various old world types, are now considered of relatively low intelligence. In many respects the most thoroughgoing and workmanlike experimental study of monkeys is that of Kinnaman (1902), who has reported on the study of various forms of response in P. rhesus. He presents valuable data concerning the learn- ing processes, sensory discrimination, reaction to number, and to tests of imitation. His results indicate a higher level of intelligence than that discovered by Thorndike, but this is almost certainty due to difference in the species observed. Kin- naman goes so far as to say "We have found evidence, also, of general notions and reasoning, both of low order " (p. 211). The contribution of Hobhouse (1915) to our knowledge of the mental life of monkeys, although in a measure experimental, is based upon relatively few and unsystematic observations as contrasted with those of Thorndike and Kinnaman. It ap- pears, however, that Hobhouse's experiments were admirably 126 ROBERT M. YERKES planned to test the ideational capacit}^ of his subjects, and one can not find a more stimulating discussion of ideation than that contained in his " Mind in Evolution." The results of his tests made with a P. rhesus monkey are similar to those of Kinnaman, for almost all of them indicate the presence and importance of ideas. Watson (1908) in tests of the imitative abiHty of P. rhesus saw relatively little evidence of other than extremely simple forms of ideation. But in contrast with his results, those ob- tained by Haggerty (1909), in a much more extended investiga- tion in which several species of monkey were used, obtained more numerous and convincing evidences of ideation in imita- tive behavior. Although this author wholly avoids the use of psychological terms, seeking to limit himself to a strictly objec- tive presentation of results, it is clear from, an unpublished manuscript (thesis for the Doctorate of Philosophy, deposited in the Library of Harvard University) that he would attribute to monkeys simple forms of ideational experience. Witmer (1910) reports, in confirmation of Haggerty's results, intelligently imitative behavior in P. iriis. The work of Shepherd (1910) agrees closely, so far as evi- dences of ideation are concerned, with that of Thorndike. He obviously strives for conservatism in his statements concerning the adaptive intelligence of his monkeys, all of which belonged to the species P. rhesus. At one point he definitely states that they exhibit ideas of a low order, or something which corres- ponds to them. Satisfactory evidences of reasoning he failed to obtain. Franz's (1907, 1911) studies of monkeys, unHke those men- tioned above, have for their chief motive not the accurate de- scription of various features of behavior but instead knowledge of the functions of various portions of the brain. His results, therefore, although extremely interesting and of obvious value to the comparative psychologist, throw no special Hght upon the problem of ideation. The investigation by Hamilton (1911) of reactive tendencies in P. rhesus and irus yielded preeminently important data con- cerning complex behavior. For the ingenious quadruple-choice method devised by this observer showed that mature monkeys exhibit fairly adequate types of response. As Hamilton's in- MENTAL LIFE OF MONKEYS AND APES 127 terest centered in behavior, he did not discuss ideation, but this does not prevent the comparison of his data with those of the present report, and the agreement of his findings with my own is obvious. My work contrasts sharply with that briefly mentioned above in that I apphed systematically and over a period of several months an experimental method suited to reveal problem solv- ing ability. Previously, the so-called problem or puzzle-box method had been used as a means of testing for the presence of ideas. For this I substituted the multiple-choice method. One of the chief advantages of this new method is the possi- bility of obtaining curves of learning for the solution or attempted solution of relational problems of varying difficultness. I am confident that these curves of learning will prove far more valuable than such data as are yielded by the puzzle-box method. The Pithecus monkeys, which I studied intensively, yielded relatively abundant evidences of ideation, but with Thorndike I must agree that of "free ideas" there is scanty evidence; or rather, I should prefer to say, that although ideas seem to be in play frequently, the}^ are rather concrete and definitely at- tached than " free." Neither in my sustained multiple-choice experiments nor from my supplementary tests did I obtain convincing indications of reasoning. What Hobhouse has called articulate ideas, I believe to appear infrequently in these ani- mals. But on the whole, I believe that the general conclusions of previous experimental observers have done no injustice to the ideational ability of monkeys. It is clearly important, how- ever, that we always should take into account the species of animal observed, for unquestionably there are extreme differ- ences in mental development among the monkeys. As I view my results in the light of their relations to earlier work, I am strongly impressed with the importance of the use of improved methods for the study of complex behavior. The delayed reaction method of Hunter, the quadruple-choice method of Hamilton, and my multiple-choice method offer new and promising approaches to forms of activity which thus far have been only superficially observed. The ability exhibited by Skirrl to try a method out and then to abandon it suddenly is characteristic of animals high in intelHgence. Most of the problems which I presented to my 128 ROBERT M. YERKES animals would be rated as difficult by psychologists, for as a rule the}^ involved definite relations and demanded on the part of the subject both perception of a particular relation and the ability to remember or re-present it on occasion. I was greatly surprised by the slow progress of the monkeys toward the solution of these problems. It had been my suppo- sition that they would solve them more quickly than any lower type of mammal, but as a matter of fact they succeeded less well than did pigs. Their behavior throughout the work proved that of far greater significance for the experimenter than the solution of a problem is definite knowledge of the modes of behavior exhibited from moment to moment, or day to day. This is true especially of those incidental or accidental modes of response which so frequently appeared in connection with my work that I came to look upon them, the surprises of each day, as my chief means of insight. Evidences of Ideation in Apes Reliable literature of any sort concerning the behavior and mental life of the anthropoid apes is difficult to find, and still more rare are reports concerning experimental studies of these .animals. There are, it is true, a few articles descriptive of tests of mental ability, but even these are scarcely deserving of being classed as satisfactory experimental studies of the psychology of the ape. I have the satisfaction of being able to present in the present report the first systematic experimental study of any feature of the behavior of an anthropoid ape. Among the most interesting and valuable of the descriptions ■ which may be classed among accounts of tests of mental ability is Hobhouse's (1915) study of the chimpanzee. The subject was an untrained animal, so far as stated, of somewhat unsatisfac- tory condition because of timidity. Nevertheless, Hobhouse was able to obtain from him numerous and interesting responses to novel situations, some of which may be safely accepted as evi- dences of ideation of a fairly high order. Similar in method and result to the work of Hobhouse is that of Haggerty (unpublished thesis for the Doctorate of Philos- ophy, deposited in the Library of Harvard University). Hag- gerty's tests of the ability of young orang utans and chim- MENTAL LIFE OF MONKEYS AND APES 129 panzees to solve simple problems and to use tools in various ways yielded results which contrast most strikingly with those obtained in his experimental study of the imitative tendency in monkeys. His observations, had he committed himself to anything approaching interpretation, doubtless would have led him to conclusions concerning the ideational life of these animals very similar to those of Hobhouse. Koehler, working in the Canary Islands, has, according to information which I have received from him by letter, made certain experiments with orang utans and chimpanzees similar to those of Hobhouse and Haggerty. His results I am. unable to report as I have scanty information concerning them. They are, presumably, as yet unpublished. In his laboratory at Montecito, California, Hamilton has from time to time kept anthropoid apes, but without special effort to investigate their ideational behavior. He has most inter- esting and valuable data concerning certain habits and instincts, all as yet unpublished. To a congress of psychologists Pfungst (1912) briefly reported on work with anthropoid apes in certain of the German zoolo- gical gardens. His preliminary paper does not enable one to make definite statements concerning either his methods or such results as he may have obtained concerning ideational behavior. So far as I know, he has not as yet published further concerning his investigation. Mobius (1867) has described interesting observations concern- ing the mental life of the chimpanzee. But this, like all of the work previously mentioned, is rather in the nature of casual testing than thoroughgoing, systematic, and analytic study. In addition to the above reports, there are a few concerning the behavior of apes which have been especially trained for purposes of exhibition. Most interesting of these is that of Witmer (1909), who studied in exhibitions and in his own labo- ratory the behavior of the chimpanzee Peter. The varied forms of intelligently adaptive behavior exhibited by this ape con- vinced Witmer of ideational experience and even of an approach to reasoning. In his brief report he expresses especial interest in the possibility of educating this "genius among apes" to the use of language. A chimpanzee named Consul was observed several years ago 130 ROBERT M. YERKES by Hirschlaff (1905), and his tricks were interestingly described from the pedagogical standpoint. Similar in character is Shepherd's (1915) brief description of the stage behavior of Peter and Consul, both chimpanzees. It is impossible to determine from the account whether these animals are the same as were observed by both Witmer and Hirschlaff. As no reference is made in Shepherd's paper to other descriptions of the behavior of these animals and as he adds nothing to w^hat had already been presented, the reader obtains no additional light on ideation. I have mentioned only samples of the articles on trained anthropoids. All are necessarily descriptions of the behavior of individuals who had been trained not for psychological pur- poses but for the vaudeville stage, and although such observa- tions unc}uestionably have certain value for comparative psy- chology, it is w^ell known that unless an observer knows the history of an act, he is not able to evaluate it in terms of intel- ligence and is especially prone to overestimate its value as evidence of ideation. There remain studies of the apes, dealing primarily with behavior and mental characteristics, which are slightly if at all experimental and deserve to be ranked as naturalistic accounts. Such is, for example, the book of Sokolowski (1908), in which attention is given to the characteristics of young as well as fairly mature specimens of the gorilla, chimpanzee and orang utan. The various publications of Garner (1892, 1896, 1900) deal especially with the language habits of monkcA^s and apes, but observations bearing on ideation are reported. Wallace (1869) describes certain features of the behavior of an infant orang utan whose mother he shot in Borneo. He also reports observations concerning the behavior of adult orang utans, many specimens of which were shot by him during his travels. Early in the last century, Cuvier (1810) interested himself in studies of the intellectual characteristics of the orang utan, and his data, taken with those of Wallace, Sokolowski, and others similarly interested in the natural history of mind, give one a valuable glimpse of the life of the anthropoid ape. Finally, the data brought together by Brehm (1864, 1875, 1888) in his famous Tierleben; by Darwin (1859, 1871) in " The MENTAL LIFE OF MONKEYS AND APES 131 Origin of Species," and other works, by Romanes (1900), especi- ally in his books on mental evolution, by C. Lloyd Morgan (1906) in his several works on comparative psychology, and by Holmes (1911) in his discussion of the evolution of intelligence, contribute not unimportantly to our all too meagre knowledge of the mental life of the anthropoid apes. My own results, viewed in the light of what one may learn from the literature, stand out as unique because of the method of research. Never before, so far as I have been able to learn, has any ape been subjected to observation under systematically controlled conditions for so long a period as six months. More- over, my multiple-choice method has the merit of having yielded the first curve of learning for an anthropoid ape. This fact is especially interesting when one considers the nature of the particular curve. For so far as one may say by comparing it with the curves for various learning processes exhibited by other mammals, it' is indicative of ideation of a high order, and possibly of reasoning. I do not wish to exaggerate the importance of my results, for as contrasted with what might be obtained by further study, and with what must be obtained if we are ade- quately to describe the mind of the orang utan, they are meager indeed. Especially noteworthy, as evidences of ideation, in the results yielded by the multiple-choice method are (1) the use by the orang utan of several different methods in connection with each problem; (2) the suddenness of transition from method to method; (3) the final and perfect solution of problem 1 without diminu- tion of the initial errors; (4) the dissociation of the act of turn- ing in a circle from that of standing in front of a particular box. To these features of behavior others of minor importance might be added. But as they have been sufficiently empha- sized in the foregoing detailed descriptions, I need only repeat my conclusion, from the summation of evidence, that this young orang utan exhibited numerous free ideas and simple thought processes in connection with the multiple-choice experiment. His ultimate failure to solve the second problem is peculiarly interesting, although in the light of other features of his beha- vior by no means indicative of inferior intelligence. The various supplementary experimental tests which I em- ployed are in no wise importantly distinguished from those 132 ROBERT M. YERKES used by other observers. The box stacking experiment has, according to my private information, been used by Koehler. It is obviously important that such tests be applied in the same manner to individuals not only of the different genera of anthro- poid apes, but of different ages, sex, and condition of training. The box stacking experiment, although it yielded complete success only as a result of suggestion on my part, proved far more interesting during its progress than any other portion of my work. In connection with it, the orang utan exhibited surprisingly diverse and numerous efforts to meet the demands of the situation. It is fair to characterize him as inventive, for of the several possible ways of obtaining the banana which were evident to the experimenter, the ape voluntarily used all but two or three, and one of these he subsequently used on the basis of imitation. Had Julius been physically and mentally mature, my results would undoubtedly have been much more impressively indica- tive of ideas, but even as matters stand, the survey of my ex- perimental records and supplementary notes force me to con- clude that as contrasted with the monkeys and other mammals, the orang utan is capable of expressing free ideas in considerable number and of using them in ways highly indicative of thought processes, possibly even of the rational order. But contrasted with that of man the ideational life of the orang utan seems poverty stricken. Certainly in this respect Julius was not above the level of the normal three-year-old child. In common with other observers, I have had the experience of being profoundly impressed by the versatility of the ape, and however much I might desire to disprove the presence of free ideas and simple reasoning processes in the orang utan, I should feel bound to accept many of the results of my tests as evidences of such experience. I have attempted to indicate briefly the historical setting of my investigation. I propose, now, in the concluding section, to look forward from this initial research and to indicate as well as I may in a few words the possibilities of results im- portant for mankind from the thorough study of the monkeys and anthropoid apes. VI [ PROVISION FOR THE STUDY OF THE PRIMATES, AND ESPECIALLY THE MONKEYS AND ANTHROPOID APES' I should neglect an important duty as well as waste an oppor- tunity if in this report I did not call attention to the status cf our knowledge concerning the monkeys and apes and present the urgent need of adequate provision for the comparative study of all of the primates. Although for centuries students of nature have been keenly interested in the various primates, the information which has been accumulated is fragmentary and wholly inadequate for generally recognized scientific and practical needs. There is a voluminous literature on many aspects of the organization and lives of the monkeys and apes, but when one searches in it for reasonably connected and complete descriptions of the organisms from any biological angle, one is certain to meet disappointment' Concerning their external characteristics we know much; and our classifications, if not satisfactory- to all, are at least emi- nently useful. But when one turns to the morphological sciences of anatomy, histolog>% embryology, and pathology, one dis- covers great gaps, where knowledge might reasonably be ex- pected. Even gross anatomy has much to gain from the care- ful, systematic examination of these organisms. With still greater force this statement applies to the studies of finer struc- tural relations. Little is known concerning the embr^^ological development and life history of certain of the primates, and almost nothing concerning their pathological anatomy. Clearly less satisfactory- than our knowledge of structure is the status of information concerning those functional processes which are the special concern of physiolog}^ and pathology. Certain important experimental studies have been made on the nervous system, but rarely indeed have physiologists dealt sys- tematically with the functions of other systems of organs. There * Much of the material of this section was published originally in Sciettce (Yerkes, 1916). 134 ROBERT M. YERKES Sub-divisions of the Order Primates Order Sub-orders Families a. Prosimii (Lemurs and Aye-Ayes) i. Hapalidae (Marmosets) Primates. . . { ii. Cebidae (Howling Monkeys, Tee Tees, Squirrel Monkeys, Spider Monkeys, and Capuchin Monkeys) b. Anthropoidea . iii. Cercopithecidae (Baboons and Macaques) iv. Simiidae (Gibbons, Orangs, Chimpanzees, and Gorillas) V. Hominidae (Man) are almost no satisfactory physiological descriptions of the monkeys, anthropoid apes, or lower primates. When we turn to the science of genetics we meet a similar condition, for the literature reveals only scattered bits of infor- mation concerning heredity in the primates. No important experimental studies along genetic lines have been made with them, and such general observations from nature as are on record are of extremely uncertain value. Were one to insist that we know nothing certainly concerning the relation of her- edity in other primates than man, the statement could not well be disputed. Occasionally in recent years students of human diseases have employed monke^^s or apes for experimental tests, but aside from the isolated results thus obtained, extremely little is known concerning the diseases peculiar to the various types of infra- human primates or the significant relations of their diseases to those of man. Next in order of extent to our morphological knowledge of these organisms is that of their behavior, mental life, and social relations. But certainly no one who is conversant with the behavioristic, psychological and sociological literature could do otherwise than emphasize its incompleteness and inadequacy. For our knowledge of behavior has come mostly from natural- istic observation, scarcely at all from experimentation; our knowledge of social relations is obviously meager and of un- certain value; and finally, our knowledge of mind is barely more than a collection of carelesslv drawn inferences. MENTAL LIFE OF MONKEYS AND APES 135 This picture of the status of scientific work on the primates, although not overdrawn, will doubtless surprise many readers, and even the biologist may find himself wondering why we are so ignorant concerning the lives of the organisms most nearly akin to us, and naturally of deepest interest to us. The reasons are not far to seek. Most scientific investigators are forced b}' circumstances to work with organisms which are readily ob- tained and easily kept. The primates have neither of these advantages, for many, if not most of them, are expensive to get and either difficult or expensive to keep in good condition. Clearly, then, our ignorance is due not to lack of appreciation of the scientific value of primate research but instead to its difficultness and costliness. Strangely enough, the practical importance of knowledge of the primates has seldom been dwelt upon even by those biolo- gists who are especially interested in it. It is, therefore, appro- priate to emphasize the strictly human value of the work for which I am seeking provision. During the past few years it has been abundantly and con- vincingly demonstrated that knowledge of other organisms may aid directly in the solution of many of the problems of experi- mental medicine, of physiology, genetics, psychology, sociology, and economics. In the light of these results, it is obviously desirable that all studies of infrahuman organisms, but especi- ally those of the various primates, should be made to contri- bute to the solution of our human problems. To me it seems that thoroughgoing knowledge of the lives of the infrahuman primates would inevitably make for human betterment. Through the science of genetics, as advanced by experimental studies of the monkeys and anthropoid apes, prac- tical eugenic procedures should be more safely based and our ability to predict organic phenomena greatly increased. Sim- ilarly, intensive knowledge of the diseases of the other primates in their relations to human diseases should contribute import- antly to human welfare. And finally, our careful studies of the fundamental instincts, forms of habit formation, and social relations in the monke^'-s and apes should lead to radical im- provements in our educational methods as well as in other forms of social service. Along theoretical lines, no less than practical, systematic 136 ROBERT M. YERKES research with the primates should rapidly justify itself, for upon its results must rest the most significant historical or genetic biological descriptions. It is beyond doubt that genetic psy- chology can best be advanced to-day by such work, and what is obviously true of this science is not less true of all the biolo- gical sciences which take account of the developmental or genetic relations of their events. In view^ of the probable values of increasingly complete ac- counts of primate Hfe, it seems far from extravagant to insist that the securing of adequate provision for systematic and long continued research is the most important task for our genera- tion of biologists and the one which we shall be least excusable for neglecting. Indeed, when one stops to reflect concerning the situation, it seems almost incredible that the task has not been accomplished. Some ten years ago Professor John B. Watson (1906) entered a plea for the founding of a station for the experimental study of behavior. He made no special mention of work w4th the monkeys and apes, but it is clear from the problems which he enumerates that he would consider them most important subjects for observation. Professor Watson's plea has appar- ently been forgotten by American biologists, and it seems not inappropriate to revive it at this time. For surely we have advanced sufficiently along material and scientific Hnes during the last ten years to render possible the realization of his hope. To my knowledge, only one definite attempt has thus far been made to gain special provision for the study of the pri- mates. Somew^here about the year 1912 there was established on Tenerife, one of the Canary Islands, a modest station for the study of the anthropoid apes. I have already referred to it briefly on page 1. The plan and purpose of this station, which is of German origin, have been presented briefly by Roth- mann (1912). From personal communications I know that a single investigator has been in residence at the station since its founding and that psychological and physiological results of value have been obtained, but no published reports have come to my attention. When I first heard of the existence of the German anthropoid station I naturally thought of the possibility of cooperative work, but the events of the past two years have rendered the MENTAL LIFE OF MONKEYS AND APES 137 chances of cooperation so remote that it now seems wholly desirable and indeed imperative to seek the establishment of an American station, which, unlike the German station, shall provide adequately not only for the study of the anthropoid apes but for that of all of the lower primates. It should be the function of such a station or research institute (1) to bring together and correlate all the information at present available; (2) to fill in existing gaps observationally and thus complete and perfect our knowledge of these organisms; (3) to seek to bring all available information to bear upon the problems of human life. Hitherto the unsatisfactoriness of progress has been due to the lack of a definite plan and program. Every investigator has gone his own way, doing what little his personal means and opportunity rendered possible. The time has at last come when concerted action seems feasible as well as eminently desirable. I am therefore offering a plan and program which, if wisely developed, should lead ultimately to fairly complete and prac- tically invaluable knowledge of the lives of all of the primates. There should be provided in a suitable locality a station or research institute which should offer adequate facilities (1) for the maintenance of various types of primate in normal, healthy condition; (2) for the successful breeding and rearing of the animals, generation after generation; (3) for systematic and continuous observation under reasonably natural conditions; (4) for experimental investigations from ever^^ significant bio- logical point of view"; (5) for profitable cooperation with existing biological institutes or departments of research throughout the world. The station should be located in a region whose climate is highly favorable to the life of many of the lower primates as well as to that of man. Such a location is by no means easy to find. Because of my intense interest in the subject, I have, during the past five years, prospected in various parts of the world for a satisfactory^ site. I shall now attempt to indicate the chief requirements and also the foremost advantages and disadvantages of several regions which have been considered. It is first of all requisite that the climate be such as to agree with the organisms to be studied and such, also, as to render their breeding normal and dependable. Second in importance 138 ROBERT M. YERKES is its satisfactoriness for the life and scientific productiveness of the observer. While certain tropical localities would meet the first requirement perfectly, they would prove extremely un- satisfactory for research activity. It therefore seems essential to find a region whose climate shall reasonably meet the needs of the experimenter while adequately meeting those of the animals to be studied. A further factor which has important bearing upon the pro- ductiveness of the observer is the degree of isolation from civili- zation and from other scientific work. No scientist can long work effectively, even in a reasonably healthy and stimulating climate, if entirely cut oft" from similar interests and activities. It is therefore desirable, if at all possible, to discover a location in the midst of civilization and with reasonably good oppor- tunities for scientific associations. With these several desiderata before us, I shall call attention to a number of possible sites for a station, several of which I have visited. Southern California, and especially the portion of the State between Santa Barbara and San Diego, promises fairly well. It is definitely known that certain, if not all, spe- cies of monkey will breed there fairly satisfactorily, and al- though it has not yet been demonstrated, there is no reason to suppose that in certain regions the anthropoid apes might not also be kept in perfect health and successfully bred. The main advantages of this general region are (a) a climate which promises to be reasonably satisfactory for many if not all of the primates; (b) admirable climatic conditions for investiga- tors; (c) wholly satisfactory scientific and cultural environment for the staff of a station. The most significant disadvantages are (a) a temperature, which is at times a trifle too low for the comfort of certain of the monkeys and apes. It is by no means certain, however, that they would not usually adapt themselves to it. (b) The necessity of importing all of the animals and of having to rely upon successful acclimatization. Of course it is to be assumed that importation would be necessary only at the outset of such work, since the animals later should replenish themselves within the confines of the station. Florida offers possibilities somewhat similar to those of south- ern California, but as I have not had opportunity to examine the conditions myself, I can say only that in view of such infor- MENTAL LIFE OF MONKEYS AND APES 139 mation as is available the advantage seems to be greatly in favor of the latter. Cuba, Jamaica, Porto Rico, and for that matter, several of the West Indies, offer possible sites for a successful station. I have reasonably intimate personal knowledge only of the con- ditions in Jamaica. The major advantages in the West Indies are (a) suitable climatic conditions and food supply for the animals; and (b) reasonably satisfactory climatic conditions for the staff. These are, however, more than counterbalanced in my opinion by the following serious disadvantages: (a) the relative isolation of the investigators from their fellow scien- tists; (b) the necessity of importing all of the animals originally used; (c) the risk of destruction of the station by storms. It is definitely known that anthropoid apes as well as mon- keys can be successfully kept, bred, and reared in the West Indies. During the past year, on the estate of Dona Rosalia Abreu, near Havana, Cuba, a chimpanzee was born in cap- tivity. A valuable account of this important event and of the young ape has been published by Doctor Louis Montane (1915). It therefore seems practically certain that regions could be found readily on Jamaica, Porto Rico, or smaller islands, which would be eminently satisfactory for the breeding of apes. There are obvious reasons why an American station for the study of the primates should be located on territory controlled by the United States Government, and if a tropical location proves necessary, it would probably be difficult to find more satisfactory regions, aside from the inconveniences and risk of importation and the relative isolation of the investigators, than are available on Porto Rico. I have not seriously considered the possibility of locating an American station on the continent of Africa, for although two- of the most interesting and importaat of the anthropoid apes, the gorilla and the chimpanzee, are African forms, while many species of monkey are either found there or could readily be imported, it has seemed to me that the islands of the West and East Indies and the portions of the United States referred to above are much to be preferred over anything available in Africa. In the East, Borneo, the Philippine Islands, and Hawaii are well worth considering. Borneo is the home of the gibbon and 140 ROBERT M. YERKES of at least one species of orang utan, and in addition to these important assets, it presents the advantages of (a) a wholly suitable climate and food supply for monkeys and apes; and (b) climatic conditions for investigators which, I am informed by scientific friends, are nearly ideal. For investigators the most serious disadvantage here, as in all other parts of the East, would be the isolation from other scientific work and workers. The possibilities of Central America I considered several 3^ears ago when it seemicd to me possible that work might profitably be done with mionkeys and apes on the Canal Zone. The advan- tages are (a) a climate which promises fairly well for the animals; and (b) reasonable accessibility from the United States. The disadvantages are (a) a far from ideal climiate for long con- tinued scientific work; and (b) an environment which from the cultural and scientific point of view leaves much to be desired. Were a permanent psycho-biological station for the study of the primates to be established in southern California, it would, even though wholly satisfactory conditions for the breeding, rearing, and studying of the animals were maintained, fu-nish more or less inadequate opportunity for the observation of the animals under free, natural conditions. It would therefore be necessary to supplement the work of such a station by field work in Borneo, Sumatra, Africa, India, South America, and such other regions as the species of organism under considera- tion happen to inhabit. Considering equally the needs of the experimenter and the demands of the animals, it seems to me reasonable to conclude that southern California should be definitely proved unsuitable before a more distant site were selected. For the information which I have been able to accumulate convinces me that .it would in all probability be possible successfully to breed and keep the primates there, and it is perfectly clear that in such event the output of a station would be enormously greater because of the more favorable conditions for research than in any tropical region or in a more isolated location. Assuming that satisfactory provision in the shape of a scien- tific establishment for the study of the primates in their rela- tions to man were available, the following program might be followed: (1) Systematic and continuous studies of important forms of individual behavior, of social relations, and of mind; MENTAL LIFE OF MONKEYS AND APES 141 (2) experimental studies of physiological processes, normal and pathological, and especially of the diseases of the lower pri- mates, in their relations to those of man; (3) studies of heredity, embryology, and life history; (4) research in comparative anat- omy, including gross anatomy, histology, neurolog\', and path- ological anatomy. Each of these several kinds of research should be in progress almost continuously in order that no materials or opportuni- ties for observation be needlessly wasted. Because of the nature of the work, it would be necessars^ to provide, first of all, for those functional studies which demand healthy and normally active organisms, whose life history is intimately and com- pletely known. This is true of all studies in behavior, whether physiological, psychological, or sociological. Simultaneously with behavioristic observations and often upon the same individuals, genetic experiments might be conducted. This would be ex- tremely desirable because of the relativeh^ long periods between generations. After the usefulness of an animal in behavioristic or genetic inquiries had been exhausted, it might be made to render still further service to science in various experimental physiological, or medical inquiries. And finally, the same indi- vidual might ultimately be used for various forms of anatomical research. Thus, it is clear that the scientific usefulness of a lemur, a monkey, or an ape might be maintained at a high level throughout and even beyond the period of its life history. The program thus briefly sketched would provide either directly or indirectl}' for work on every aspect of primate life. Especially important would be the intimacy of interest and cooperation among investigators, for the comparative method should be • applied consistently and to the limit of its value. The results of various kinds of observation should be correlated so that there should ultimately emerge a unitary and prac- tically valuable account of primate life, to replace the patch- work of information which we now possess. Because of the costliness of maintaining and breeding the monkeys and apes, it is especially desirable that the several kinds of research mentioned above should be conducted. In- deed, it would seem inexcusably wasteful to attempt to main- tain a primate or anthropoid station for psychological observa- tions alone, or for any other narrowly limited biological inquir^^ 142 ROBERT M. YERKES Furthermore, the station should be permanent, since for many kinds of work it would be essential to have intimate knowledge of the life history and descent of an individual. With the lower primates, a generation might be obtained in from two to five years; w^ith the higher, not more frequently, probably, than from ten to fifteen years. It therefore seems not improbable that the value of the work done in such a station would con- tinue to increase for many years and would not reach its max- imum short of fifty or even one hundred years. A staff of several highly trained and experienced biologists would be needed. The following organization is suggested as desirable, although, as indicated below, not necessarily essential in the beginning: (1) An expert especially interested in the problems of behavior, psychology, and sociology, with keen ap- preciation of practical as well as of theoretical problems; (2) an assistant trained especially in comparative physiology; (3) an expert in genetics and experimental zoology; (4) an assistant with training and interests in com.parative anatomy, histology, and embryology; (5) an expert in experimerital medicine, who could conduct and direct studies of the diseases of man as well as of the lower primates and of measures for their control ; (6) an assistant trained especially in pathology and neurolog;^>^ To this scientific staff of six highly trained individuals there should be added a business manager, a clerical force of three individuals, a skilled mechanician, a carpenter, and at least four laborers. The annual expenditures of an institute with such a working staff, would in southern California, approximate fifty thousand dollars. It would therefore be necessary that it have an en- dowment of approximately one million dollars. In the absence of this foundation it would, of course, be pos- sible to make a reasonably satisfactory beginning on the work which has been outlined in the following less expensive manner. A working plant might be established, on ground rented or purchased at a low figure, for about ten thousand dollars; the salary of a director, assistants, a clerical helper, and combined mechanic and laborer might be estimated at the same figure; the cost of animals and of maintenance of the plant would approximate five thousand dollars. Thus, we should obtain as an estimate of the expenditures for the first year twenty-five MENTAL LIFE OF MONKEYS AND APES 143 thousand dollars. Without expansion, the work might be con- ducted during the second year for fifteen thousand dollars, and subsequently it might be curtailed or expanded, resources permitting, according as results achieved and in prospect iusti- fied. An institute established on such a modest basis as this still might render largely important scientific service through its own research and through organized cooperation with other existing research establishments. Thus, for example, suppos- ing that behavioristic, psychological, sociological, and genetic inquiries were conducted in the institute itself, animals might be supplied on a mutually satisfactory basis to institutes for experimental medicine, for physiological research, and for anat- omical studies. Under such conditions, it is conceivable that extremely economical and good use might be made of all the available primate materials. But it is not improbable that even cooperative research would prove on the whole more profitable, except possibly in the case of morphological work, if investiga- tors could conduct their studies in the institute itself rather than in distant laboratories. In any event, the idea of coopera- tion should be prominent in connection with the organization of a research station for the study of the primates. For thus, evidently, scientific achievement in connection with these im- portant types of animal might be vastly increased over what would be possible in a single relatively small institution with a limited and necessarily specialized staff of workers. Despite the fact that biologists generally recognize the import- ance of the work under consideration and are eager to have it done, it is perfectly certain that we shall accomplish nothing un- less we devote ourselves confidentl3\ determinedly and unitedly, with faith, vision, and enthusiasm, to the realization of a definite plan. Our vision is clear, — if we are to gather and place at the service of mankind adequate comparative knowledge of the life of the primates and if we are to make this possible harvest of scientific results count for human betterment, we must bend all our efforts to the establishment of a station or institute for research. VIII BIBLIOGRAPHY Brehm, a. 1888. COBURN, C. 1915. Tierleben. A. and Yerkes, R. M. A study of the behavior of the crow, Corvus Americaniis Aud. by the multiple-choice method. Joiirnal oj Animal Behavior, 5, 75-114. CuviER, Frederic. Description d'un orang-outang, et observations sur ses faculties 1810. intellectuelles. Annales dii Museum d'Histoire nalureUe, 16, 46-65. Darwin, C. Origin of species. 1859. The descent of man, and selection in relation to sex. G. A review of the primates. New York. 1871. Elliot, D 1913. Franz, S. 1907. 1911. Psychological 1913. 1911. 1914. Journal of Journal of [. On the functions of the cerebrum: the frontal lobes. Archives of Psychology, no. 2, 1-64. On the functions of the cerebrum: the occipital lobes. Monographs, 13„ 1-118. Observations on the preferential use of the right and left hands by monkeys. Journal of Animal Behavior, 3, 140-144. Garner, R. L. The speech of monkeys. London. 1892. 1896. Gorillas and chimpanzees. London. 1900. Apes and monkeys; their life and language. Boston and London. Haggerty, M. E. Imitation in monkeys. Journal of Comparative Neurology and 1909. Psychology, 19, 337-455. Hamilton, G. V. A study of trial and error reactions in mammals. Animal Behavior, 1, 33-66. A study of sexual tendencies in monkeys and baboons. Animal Behavior, 4, 295-318. Hirschlaff, L. Der Schimpanzee Konsul. Zeit. f. pad. Psy., 7, 1. 1905. HoBHOUSE, L. T. Mind in evolution. London. 1915. Holmes, S. J. Evolution of intelligence. New York. 1911. Hrdlicka, a. Anatomical observations on a collection of orang skulls from west- 1907. em Borneo; with a bibliography. Proceedings of the United States National Museum, 31, 539-568. Kinnaman, a. J. Mental life of two Macacus rhesus monkeys in captivity. Amer- 1902. ican Journal of Psychology, 13, 98-148, 173-218. Lashley, K. S. and Watson, J. B. Notes on the development of a young monkey. 1913. Journal of Animal Behavior, 3, 114-139. MoBlus, K. Zur psychologic des Schimpanse. Zool. Garten., 8, 279-280. 1867. MoNTANii, L. Un chimpance Cubano. Havana, El Siglo XX. 1915. Morgan, C. L. Introduction to comparative psychology. London. 1906. Pfungst, O. Zur Psychologic der AfTen. Bcr. iiber d. V. Kongress f. exp. Psycliol.. 1912. 200-205. MENTAL LIFE OF MONKEYS AND APES 145 Romanes, G. J. Mental evolution in animals. New York. 1900. ROTHMANN, M. Ueber die errichtung einer Station zur psychologischen und hirn- 1912. physiologischen Erforschung der Menschenaffen. Berliner klin. Wochenschr., Nr. 42. Shepherd, W. T. Some mental processes of the rhesus monkey. Psychological 1910. Monographs, 12, 1-61. 1915. Some observations on the intelligence of the chimpanzee. Journal of Animal Behavior, 5, 391-396. SoKOLOWSKY, A. Beobachtungen iiber die Psyche der Menschenaffen. Frankfurt 1908. a. M. Thorndike, E. L. The mental life of the monkeys. Psychological Monograph, 1901. 3, 1-57. 1911. Animal intelligence. New York. Wallace, A. R. The Malav Archipelago. London. 1869. Washburn, M. F. The animal mind. New York. 1908. Watson, J. B. The need of an experimental station for the study of certain prob- 1906. lems in animal behavior. Psychological Bulletin, 3, 149-156. 1908. Imitation in monkeys. Psychological Bulletin, 5, 169-178. 1909. Some experiments bearing upon color vision in monkeys. Journal of Comparative Neurology and Psychology, 19, 1-28. Witmer, L. a monkey with a mind. Psychological Clinic, 3, 179-205. 1909. 1910. Intelligent imitation and curiosity in a monkey. Psychological Clinic, 3, 225-227. Yerkes, R. M. The study of human behavior. Science, 39, 625-633. 1914. 1915. Maternal instinct in a monkey. Journal of Animal Behavior, 5, 403- 405. 1916. Provision for the study of the monkeys and apes. Science, 43, 231-234. Yerkes, R. M. and Coburn, C. A. A study of the behavior of the pig Sus scrofa 1915. by the multiple-choice method Journal oj Animal Behavior, 5, 185-225. Behavior Monographs VOLUME 1 No. 1 The development of certain instincts and habits in chicks. By Frederick S. Breed. Pp. iv + 78, $1.00, postpaid. No. 2 Methods of studying vision in animals. By Robert M. Yerkes and John B. Watson. Pp. iv + 90, $1.25, postpaid. No. 3 An experimental study on the death-feigning of Belostoma (-Zaitha Aucct.) flumineum Say and Nepa apiculata Uhler. By Henry H. P. Severin and Harry C. Severin. Pp. iii + 47, $.65, postpaid. No. 4 The biology of Physa. By Jean Dawson. Pp. iii + 120, $1.50, postpaid. No. 5 The function of the vibrissae in the behavior of the white rat. By Stella Burnhara Vincent. Pp. iv + 81, $1.15, postpaid. VOLUME 2 No. 6 The delayed reaction in animals and children. By Walter S. Hunter. Pp. v + 86, $1.15. No. 7 The Canada porcupine : a study of the learning process. By Leroy Walter Sackett. Pp. iii + 84, $1.15. No. 8 Audition and habit formation in the dog. By Harry Miles John- son. Pp. iv + 78, $1.00. No. 9 Habit formation in a strain of albino rats of less than normal brain weight. By Gardner Cheney Basset. Pp. iv + 46, $.65. No. 10 Distribution of effort in learning in the white rat. By John Linck Ulrich. Pp. iii + 51, $.65. No. 11 The effect of age on habit formation in the albino rat. By Helen B. Hubbert. Pp. v + 55, $.65. VOLUME 3 No. 12 The mental life of monkeys and apes: a study of ideational behavior. By Robert M. Yerkes. Pp. iv + 145, $1.50. Directory of American Psychological Periodicals American Journal of Psychology — Worcester, Mass.: Florence Chandler. Subscription $5. 600 pages annually. Edited by G. Stanley Hall. Quarterly. General and experimental psychology. Founded 1887. Pedagogical Seminary — Worcester, Mass. : Florence Chandler. Subscription $.5. 575 pages annually. Edited by G. Stanley Hall. Quarterly. Pedagogy and educational psychology. Founded 1891. Psychological Review — Princeton, N. J.: Psychological Review Company. Subscription (with Psychological Bulletin) $5. 480 pages annually. Bi-monthly. General. Founded 1894. Edited by Howard C. Warren. Psychological Bulletin — Princeton, N. J. : Psychological Review Company. Subscription $2.75. 480 pages annually. Psychological literature. Monthly. Founded 1904. Edited by Shepherd I. Franz. Psychological Monographs — Princeton, N. J.: Psychological Review Company. Subscription $4. 500 pages per vol. Founded 1895. Ed. by James R. Angell. Published without fixed dates, each issue one or more researches. Psychological Index — Princeton, N. J.: Psychological Review Company. Subscription $1. 200 pages. Founded 1895. Edited by Madison Bentley. An annual bibliography of psychological literature. Journal of Philosophy, Psychology and Scientific Methods — ^New York: Science Press. Bi-weekly. 728 pages per volume. Founded 1904. Subscription $3. Edited by F. J. E. Woodbridge and Wendell T. Bush. Archives of Psychology — Substation 84, New York: Archives of Psychology. Subscription $5. 600 pp. ann. Founded 1906. Edited by R. S. Woodworth. Published without fixed dates, each number a single experimental study. Journal of Abnormal Psychology — Boston: Richard G. Badger. Subsa-iption $4. 480 pages annually. Edited by Morton Prince. Bi-monthly. Founded 1905. Entire field of abnormal psychology. Psychological Clinic — Philadelphia: Psychological Clinic Press. Subscription $1.50. 280 pages annually. Edited by Lightner Witmer. Monthly (9 numbers). Orthogenics, psychology, hygiene. Founded 1907. Training School Bulletin— Vineland, N. J. : The Training School. Subscription $1. 160 pages annually. Ed. by E. R. Johnstone. Founded 1904. Monthly (10 numbers). Psychology and training of defectives. Journal of Religious Psychology — Worcester, Mass. : Louis N. Wilson. Subscription $3. 480 pages annually. Founded 1904. Edited by G. Stanley Hall. Published without fixed dates. 4 numbers constitute a volume. Journal of Race Development — Worcester, Mass. : Louis N. Wilson. Subscription $2. 460 pages annually. Founded 1910. Quarterly. Edited by George H. Blakeslee and G. Stanley Hall. Journal of Educational Psychology — Baltimore: Warwick & York. Subscription $2.50. 600 pages annually. Founded 1910. Monthly (10 numbers). Managing Editor, J. Carleton Bell. (Educational Psychology Monographs. Edited by Guy M. Whipple. Published separately at varying prices. Same publishers.) Journal of Animal Behavior — Cambridge, Mass. : Emerson Hall. Subscription $3, Foreign $3.50. 450 pages annually. Founded 1911. Bi-monthly. Robert M. Yerkes, Managing Editor. The Behavior Monographs — Cambridge, Mass.: Emerson Hall. Subscription $3. 450 pages per volume. Edited by John B. Watson. Published without fixed dates, each number a single research. Psychoanalytic Review — New York: 64 West 56th Street. Subscription $5. 500 pages annually. Psychoanalysis. Quarterly. Founded 1913. Ed. by W. A. White and S. E. Jelliffe. Journal of Experimental Psychology — Princeton, N. J.: Psychological Review Company. 480 pages annually. Experimental. Founded 1916. Subscription $3. Bi-monthly. Edited by John B. Watson. The Brandow Printing Comfany Albany, N. Y Behavior Monographs Volume 3, Number 2. 1916 Serial Number 13 Edited by JOHN B. WATSON The Johns Hopkins University A Study of Perseverance Reactions in Primates and Rodents G. V. HAMILTON Published at Cambridge. Boston, Mass. HENRY HOLT & COMPANY 34 West 33d Street, New York G. E. STECHBRT & CO., London, Pari* and Leipzie, Foreign Aflent* The Journal of Animal Behavior An organ for the publication of studies concerning the instincts, habits and intelligence of organisms The Journal contains a Department of Notes in which appear brief accounts of especially interesting and valuable observations of behavior. Published bi-monthly at Cambridge, Boston, Mass., by Henry Holt and Company, New York. Each volume contains at least 450 pages with plates and text-figures. The subscription price is $5.00 per volume (foreign subscription $5.50) postpaid. Manuscripts, subscriptions, and all business correspondence should be addressed to THE JOUENAL OF ANIMAL BEHAVIOR Emerson Hall, Cambridge, Massachusetts Behavior Monographs For the publication of studies in behavior and intelligence which are too lengthy or too expensive for acceptance by The Journal of Animal Behavior Published at irregular intervals at Cambridge, Boston, Mass., in connection with the Journal of Animal Behavior, by Henry Holt and Company, New York. Each volume contains approximately 450 pages with plates and text-figures. The subscription price is $3.00 per volume (foreign subscription $3.50) postpaid. Monographs may be purchased separately at prices varying with the cost of manufacture. Manuscripts and inquiries concerning terms of publication should be addressed to the Editor of the Behavior Monographs, JOHN B. WATSON, The Johns Hopkins University, Baltimore, Md. Subscription to The Journal of Animal Behavior and the Behavior Monographs should be sent to Emerson Hall, Cambridge, Massa- ohusetts. jnii f~i\j iv/i# Behavior Monographs Volume 3, Number 2, 1916 Serial Number 13 Edited by JOHN B. WATSON The Johns Hopkins University A Study of Perseverance Reactions in Primates and Rodents G. V. HAMILTON Montecito, California TWO FIGURES Published at Cambridge, Boston, Mass. HENRY HOLT & COMPANY 34 West 33d Street, New York G. E STECHERT & CO., London, Paris and Leipzig, Foreign Agents CONTENTS Page Introduction 1 Description of Apparatus 4 Method of Conducting Experiments 7 List of Subjects 9 Discussion of Subjects 10 I. Presentation of Results 11 II. Reaction-type Tables (Table 2) 14 III. Discussion of Table 2 20 IV. Determinants of Initial Direction of Activity 32 V. Studies of Habit Formation by the Multiple Choice Method 38 VI. Conclusions 43 Appendix : 1. Individual Records of Twenty Girls 46 2. Individual Record of White Rat 5 57 3. Analysis of White Rat 5's Record (Calculations for Table 8) 58 A STUDY OF PERSEVERANCE REACTIONS IN PRIMATES AND RODENTS G. V. H-\MILTON Montecito, California INTRODUCTION The present study is a continuation of a previously published one (1), which was also devoted to an investigation of per- severance (trial and error) reactions. The method that was employed in both studies requires the subject to find his way out of an enclosure from which there are four apparently pos- sible exit places. During a series of 100 trials each of the four exit places affords escape 25 times, and during any given trial only one of these places can be used for this purpose. The right exit place to try for escape from the enclosure varies from trial to trial, i.e., it is useless, during any given trial, to seek exit from the enclosure by way of the place that afforded escape during the immediately preceding trial. While the experimen- ter's order of varying the right place to try for exit contains this one negatively directive feature, it contains no specifically directive positive feature that has thus far been discovered by any of the subjects. The appearance of the apparatus, as it is seen from within by the subject, affords no clue whatsoever as to whether a given exit place is a right or a wrong one to try until actual trial is made. It follows, therefore, that this method operates against the formation of a specifically adap- tive habit, i.e., a habit which would involve no unproductive quests of an avenue of escape from the apparatus. The relation of this method to Yerkes' (3) multiple choice method requires more explicit definition than it has yet been given. Yerkes' method, which is of later development than mine, embodies some of its secondary features but has a quite different intention. Both methods exclude all specifically direc- tive sensory stimuli, and require the subject to vary the place through which he escapes from the apparatus from trial to trial. 2 G. V. HAMILTON But Yerkes' method, unlike mine, supplies a specifically direc- tive feature which is deliberately introduced with the intention of enabling the subject to form a definite, specifically adaptive habit. The construction of his apparatus is such that now one, now another group of apparently possible exit places (or their equivalents) can be exposed to the subject, with trial to trial variations of the position and constitution of such groups. For example, in one series of trials by the Yerkes' method any three, five, seven, nine or eleven of twelve exit places may be exposed to the subject, who will invariably find that the middle exit place of any three, five, seven, nine or eleven that may be ex- posed is the right one to try. In other series it is the place at the extreme left or at the extreme right of whatever group the experimenter may expose that should be sought for escape. Watson's comparison of the two methods (5) implies a sim- ilarity of intention which does not exist. While it is true that Yerkes' method, like most methods that are employed for the study of mammalian behavior, affords information as to various types of searching reaction, it does so only incidentally, and since it encourages habit formation rather than prolonged varia- bility of response its intention is practically opposite to that of my method. Watson's (5) reference to my method as a multiple stimuli one, is in my opinion, somewhat misleading. One would not refer to ordinary maze experiments as conforming to the multiple stimuli type, since they employ confinement as the primary and unvarying stimulus to reaction. Even where this stimulus is reinforced by the introduction of features which render con- finement particularly apt to lead to prompt reaction (e.g., elec- trical stimulation within the enclosure) it remains, in a sense, a single, general stimulus. The descriptions of my method and apparatus that are given in subsequent pages will show that unless the term " multiple stimuli method " may properly be used to designate maze experiments, my method cannot be regarded as a multiple stimuli one. The relation of my work to a general behavioristic program cannot be satisfactorily defined without reference to the stand- point from which I have proceeded to a study of behavior. A mammalian's initial adjustments to a situation which is not characteristically encountered by the members of its species PERSEVERANCE REACTIONS IN PRIMATES AND RODENTS 3 are apt to consist of, (a) various non-adaptive activities, and (b) a more or less accidental activity which terminates reaction by withdrawing the subject from the stimuli supplied by the situation. If the situation be frequently encountered the sub- ject tends more or less gradually to abandon the non-adaptive activities and to develop a habit of manifesting only the adaptive one whenever the situation is encountered anew. It is in this familiar circumstance that the behaviorist finds many of his problems. Unfortunately, the non-adaptive activities, which are accounted " errors," are usually given no more attention than is involved in recording their number, duration, distribution and mode (gradualness or abruptness) of disappearance. The ex- perimenter's interest is apt to center in what is significant for estimating the reactive value of the stimuli involved in terms of the subject's sensory equipment or of its capacity to learn. That behaviorists have displayed so little interest in the quali- tative aspect of " errors " (non-adaptive activities) is in part due, I believe, to the influence of pedagogical traditions which direct attention toward the positive determinants of learning. My interest in the qualitative aspect of non-adaptive activi- ties is largely due to the fact that I have approached the study of behavior from the standpoint of the clinical psychopatholo- gist. In spite of the fact that I have never been able to assim- ilate the doctrines of the Vienna and Zurich schools of psycho- analysis to my own experience and conceptions without many heretical reservations and qualifications, I am indebted to these doctrines for- certain behavioristic concepts which, it seems to me, underlie them. If I do not misconceive Freud's and Jung's writings they assume that even,' reaction, no matter how in- appropriate it may be as an attempt at adjustment to the par- ticular situation that elicits it, is the expression of an innate tendency which enters as a functional unit into the composition of the organism's total reactive equipment. Although a given reaction may prove to be inimical to the welfare of both the individual and his race, the tendency of which it is an expression possesses in itself conservative value either for the individual or for his race, or for both. The innate tendencies of an organism are disclosed in the first instance, of course, by his reactions and are identified through studies of them. But reactions do not alwavs stand 4 G. V. HAMILTON out as clean-cut entities : response often consists of manifold activities which lack a common objective reference. The task of identifying relatively separate types of reaction is, therefore, a difficult one, and is not apt to appeal to the investigator who is impatient to have done with roughly exploratory methods and to apply only highly exact ones. At this point it is to be remembered that in a technological field of comparative psychology — psychoanalytic psychopathology — a rough and ad- mittedly often unnecessarily crude exploration is currently revealing material for behavioristic investigation, the existence of which was barely if at all suspected by certain psychopatholo- gists who were and still are committed to the exclusive use of highly exact methods. I can think of no better justification of a method that is exploratory in intention and that seeks to prepare a way for the identification of reactive tendencies than is contained in Holt's (2) statement that Freud has given us " the first key which psychology has ever had which fitted, and moreover I believe that it is the only one that psychology will ever need." Apparatus for Rodents. — Figure 1 gives a view of the appa- ratus as a whole. Its essential feature is a semicircular enclo- sure, which is built on a table top. The walls of this enclosure are interrupted by five apertures. The one that divides the straight part of the wall into two equal parts is shown as blocked by a box in fig. 1, and can best be seen in fig. 2. This aperture is for entrance. The curved part of the wall is broken by four apertures, each of which opens into its own exit way or alley. Figure 2 gives the plan of the entire table top, and is there- fore a floor plan of the semicircular enclosure, the entrance and exit alleys, the bases of the wall and the parts of the table top that lie outside the wall. The shaded parts of the figure marked with W's represent the bases of the enclosure walls, which are also the walls of the alleys. The bases or bottoms of these box- like walls are of 1-inch redwood, as are also their tops, which are of the same dimensions. They afford attachment for vertical sides of sheet tin. The horizontal dimensions of the walls are given in fig. 2. The sides have a uniform elevation of 15 cm. The curved wall forms a half circle, and has a 36 cm. radius. Reference to fig. 2 will show that the exit alleys are equidistant PERSEVERANCE REACTIONS IN PRIMATES AND RODENTS ^ from the midpoint of the inner boundary Hne of the exit alley, hence the subject, on first entering the apparatus, is subjected to none of the influences that might be contained in inequalities of distance between his starting point and the four possible places for exit. The floor of the semicircular space and that of the entrance and exit alleys consist of alternate 3 mm. strips of iron and 6 mm. strips of parafline. The first, third, fifth, etc., iron strips are connected for continuous electrical conduction by a copper wire, which is soldered to one end of each strip. The opposite ends of the second, fourth, sixth, etc., iron strips are similarly con- nected by a second copper wire. The paraffine strips insulate adjacent iron strips from one another. The two copper wires are carried to the poles of an induction coil, the position of which on the table is shown in fig. 1. This coil is in turn con- nected with the dry cells shown in fig. 1, and is equipped with a rheostat. One of the connecting wires between the coil and the dry cells is permanently attached, whilst the other is unattached at its battery end, thus obviating the need of a switch. I find that it is convenient and practically noiseless to close the circuit between batteries and coil by merely touching the loose end of the wire to the appropriate pole. It is obvious that when a subject is within the enclosure or its tributary alleys he will receive a shock whenever the experimenter closes the battery- coil circuit. Figure 1 shows fii'e transfer boxes, one of which, it will be seen, is drawn in sufficient detail to render separate representa- tion unnecessary. These boxes are made of 1-inch Oregon pine, and are all alike. They have the following inner dimensions: top, 5.5 cm. X 18 cm.; bottom, 5.5 cm. x 20.5 cm.; sides, 9.2 cm. X 20.5 cm. Near the inner edges of the sides are vertical grooves for the reception of tin slides, which are used for opening and closing the box at either of its two ends. The box top is shorter than the bottom, hence it does not interfere with the vertical slides as they are moved up and down within their containing grooves. The slides are 5.1 cm. wide by 14.4 cm. long, and each is perforated near its upper edge for the reception of a lifting hook, to be described below. The five transfer boxes serve to block the outer apertures of the entrance and exit alleys; their vertically sliding, removable 6 G. V. HAMILTON doors (tin slides) are the only ones with which the apparatus is equipped. When a subject is about to be given the first of his daily ten trials he is carried from his living quarters to the en- trance end of the apparatus in one of these transfer boxes. When the box is placed in position to block the outer end of the entrance alley the slide at its inner end is elevated by means of a cord and hook device (fig. 1). The cord is permanently at- tached to the table top near the edge at which the experimenter is stationed, and from there it passes to the middle perforation in the horizontal arm of the standard shown in fig 1. After it passes through this perforation it hangs down and carries a hook at its end. The subject usually refuses to leave the transfer box for the less desirable entrance ailey and enclosure beyond, and his failure to do so requires the removal of the slide in the outer end of the box for the admission of a wooden plunger (fig. 1), which is used to force the animal out of the box. Four exit alley cords, which have permanent attachments at the positions shown in figs. 1 and 2, pass from there through perforations in the horizontal arm of the standard at the en- trance end of the apparatus, and from these perforations to eyes in the upright standards at the exit end of the apparatus (fig. 1). Each standard carries an eye through which passes a cord with a hook at its free end. When, now, a subject enters an exit alley from which he is to be allowed to escape into the transfer box at the exit alley mouth, the slide in the inner end of the box is elevated by means of the cord and hook device. Since the five cords have their places of permanent attachment near the end of the tabl.e at which the experimenter sits, any of the five slides that block the entrance and exit alleys can be raised or lowered from that end of the table. When a subject completes a trial, i.e., when he has been forced from the box at the entrance end of the apparatus into the en- trance alley and into the semicircular enclosure, and has passed from there through the right exit alley into the transfer box at its end, the box that has just been vacated at the entrance end and the one at the exit-alley end that contains the subject are made to exchange places. This brings the subject again to the entrance end, ready to be forced into the enclosure for another trial. sc sc sc sc sc PERSEVERANCE REACTIONS IN PRIMATES AND RODENTS 7 It is undesirable that a subject, while within the semicircular enclosure, shall be able to see whether the slides in the boxes at the outer mouths of the exit alleys are in an open or a closed position. On this account the apertures that allow passage from the semicircular enclosure into the exit alleys extend only half way from the top of the apparatus. The subject must, there- fore, climb over a 7.5 cm. partition in order to get into any exit alley, and he cannot see the lower half of the slides at the exit alleys mouths from any part of the semicircular enclosure. His tendency to cling to the top of the partition in an effort to keep his feet off the apparatus floor is discouraged by means of a copper wire which runs parallel to and just in front of the upper edge of the metal partition. This wire is connected with one pole of the induction coil and the partition is connected with the other pole. The walls of the apparatus and the surfaces of the slides that are visible from within the apparatus are battleship gray. The interior of the transfer boxes and the inner surfaces of the slides are black. The top of the apparatus is glass. Apparatus for Primates. — This type of apparatus is described in a previous communication (1), hence it is necessary to give here merely its essential points of difference from the rodent apparatus just described. The primate subject's willingness to enter the apparatus for the sake of reward for escape there- from renders transfer boxes and an equipment for electrical stimulation unnecessary. The mouths of the alleys are blocked by means of doors, which are hinged to the apparatus. The main enclosure is fan-shaped rather than semicircular. The only important difference between its reactive value and that of the rodent apparatus is contained in its lack of all means for rein- forcing the confinement-stimulus with the more definitely dis- agreeable electrical one. Method of Conducting Experiments First Preliminary Trial. — The rodent subject is forced into the entrance alley and is given an electrical shock. The strength of this stimulus is determined for each individual by increasing it until a definite escape-response is obtained. After the initial shock the rodent is allowed to explore the apparatus, and the stimulus is not repeated until visible escape-responses cease. 8 G. V. HAMILTON The first of any of the exit alleys entered by the subject is allowed to give escape into the transfer box at its end. The monkey subject is tolled into the enclosure and confined therein by the closure of the entrance alley door. He is "allowed to use the first exit alley entered for escape into a court, where he is at once given food. The human subject (only children were used in these studies) is told that escape from the apparatus will be rewarded by a toy credit, i.e., each escape counts one toward the number of escapes necessary to secure a reward. Four toy credits secure a reward on the day that is devoted to preliminary trials; after that ten credits secure a reward. The human subject, like the rodent and the monkey, is allowed to escape by the first exit alley entered. Second Preliminary Trial. — The exit alley used during the first preliminary trial will not afford escape, but any one of the other three exit alleys may be used for this purpose. Third Preliminary Trial. — Either of the two exit alleys that have not yet been used for escape will afford escape. Fourth Preliminary Trial. — -Only the one exit alley that has not yet been tried will afford escape. These four preliminary trials, which are given to the sub- ject in as rapid succession as is possible, are followed the next day by the first ten of a series of 100 formal trials. Ten formal trials are given daily for ten successive days. In describing the method of conducting these trials and in subsequent discussions of results I will use the following terms: Alley No. i — the exit alley at the subject's extreme left as he faces the row of exit alleys from within the inclosure. Alley No. 2 — the second exit alley from the left. Alley No. 3 — the third exit alley from the left. Alley No. 4 — the fourth exit alley from the left. The Right Alley of a given trial is the only one that will afford escape. The Impossible Alley of a given trial is the one that afforded escape during the immediately preceding trial. There is always one inferentially impossible alley for the subject who learns that it is useless, during a given trial, to try the right alley of the immediately preceding trial. PERSEVERANCE REACTIONS IN PRIMATES AND RODENTS 9 The Three Possible Alleys are the three exit alleys that would not afford escape during the immediately preceding trial. Although only one of these alleys will actually afford escape during a given trial, all three are inferentially possible ave- nues of escape from the adult human subject's standpoint. The following order is observed by the experimenter in choos- ing right alleys for each of the 100 formal trials: 1st trial — alley No. 1 will afford escape 2nd » " " 2 " " 3rd « « a 3 « « 4th " u u 4 u « 5th « u « 2 (( « 6th « (1 u 3 u u 7th (( u » 4 u u 8th " u u 1 u a 9th " u u 3 K » 10th " u » 4 u u 11th " (( u 1 u a 12th « u u 2 U li 13th (1 u « 4 » u 14th " u u 1 a u 15th " u u 2 u a 16th li ti a 3 u u This order is repeated so that, e.g., the right alley of the first trial is also the right one for the seventeenth trial, the right alley for the second trial is also the right one for the eigh- teenth trial, etc. This order is such that each of the four exit alleys is the right alley for 25 of the 100 formal trials. Subject Girl 1 Girl 2 Girl 3 . . . A 3^ 3" 4 4 5 5 5 5 Lis /ears . u u 11 a u u it u a u u it TABLE 1 ;t of Subjects Race or Species Mexican Irish Girl 4 . . . " — sister of Girl 7 Girl 5 Girl 6 Girl 7 Girl 8 . Mexican . American . Irish^feeble minded, sister of Girl 4 . " —sister of Girt 12 Girl 9 Girl 10 Girl 11 5 6 7 7 7 7 7 9 9 . Italian .Mexican — feeble minded Girl 12 Girl 13 Girll4 Girl 15 Girl 16 . Irish — sister of Girl 8 . German . Italian . Mexican Girl 17 « 10 G. V. HAMILTON TABLE I — Continued Subject Age Girl 18 9 years. Girl 19 9 " . Girl 20 12 " . Baboon 2.^ Adult. Monkey o Monkev 16 Monkev 18 Monkey 26 Mouse 1 Gray Rat 1 . Gray Rat 2. Gray Rat 3. Gray Rat 4 . Grav Rat 5. Black Rat I . Black Rat 2. Black Rat 3. Black Rat 5. White Rat White Rat W'hite Rat White Rat White Rat White Rat White Rat 8 White Rat 10 Gopher 1 . Gopher 2. Gopher 3 . Gopher 4 . Gopher 5 . Gopher 6 . Race or Species , Mexican . Italian .Mexican — feeble minded . Species unknown . Pithecus rhesus .Mus musculus . Mus rattus alexandrinus . Mus rattus . Mus norvegicus albinus — ^strain A litter 3-4 strain B . Geomvs bursarius All but four of the subjects given in table 1 had at least 100 formal trials each: I was unable to complete the series with Gopher 6 and with Giris 2, 6 and 19. Baboon 2 had 500 trials and Monkey 26 had 200 trials. The giris are inmates of an orphanage, a circumstance on which largely depends the dissimilarity of their records to those of the human subjects of my earlier studies (1). It is significant that three of the orphans are distinctly feeble minded. Although some children of good stock are found in our local orphanage, many of them are descended from parents who are mentally inferior, socially delinquent or shiftless and incompetent. The most inferior strains of the native Mexican population are also represented in our institution. PERSEVERANCE REACTIONS IX PRIMATES AND RODENTS 11 The marked individual differences presented by the five infra- human primate subjects reflect a poHcy of selecting subjects in whom oddities of general reactive equipment had been observed. Monkey 26, for example, could be successfully bullied by even the smallest and youngest of my monkeys in spite of the fact that he was the largest but one of a band of sixteen monkeys and baboons that lived together for several months in a single band. Monkey 16, whose record is in striking contrast to that of Monkey 26, is the least distractible monkey that I have ever owned. Monkey 5, on the other hand, is not only distractible to a high degree, but unusually stupid in her social relations. The gophers, roof rats, black rats and mouse were all wild subjects. None was used until it had been at least two weeks in captivity. The white rats belong to two different strains. White Rats 5, 6 and 8 were born in the laboratory, their parents being White Rats v-? and 4 of the list. PRESENTATION OF RESULTS I Six general types of searching response were manifested by the various subjects: 1. Response with reference to the experimenter s rule that no alley is the right alley for two successive trials. It is obvious that if a subject respond consistently with refer- ence to this rule (but without inferred or other knowledge of the experimenter's order) he will have an average chance of being required to try but one of the three inferentially possible alleys during each of one-third of his trials. If he learns the always-one-impossible-alley rule during his preliminary trials, and never departs from it during his 100 formal trials he will have an average chance of effecting his escape from the appa- ratus 100 times by trying the various alleys 200 times. In constructing tables of results I have employed the desig- nation, " Type A" to indicate reactions of the kind just de- scribed. The cases in which the subject tries only one or two of the three inferentially possible alleys cannot properly be classified as Type A reactions, since one cannot be sure that an inferentially possible alley would have been chosen had a third 12 G. V. HAMILTON choice been necessary. For example, alley No. 3 was the right alley for the sixteenth trial of any subject, alley No. 2 was the inferentially impossible alley and alleys No. 1, 4 and right alley No. 3 were inferentially possible ones. If a subject tried alley No. 4 first during this trial, then alley No. 1, then (successfully) alley No. 3, a Type A reaction was recorded. But if he tried only alley No. 3 or only alleys No. 4 and 3 his reaction did not enter into any of the tables that deal with the six general reac- tion-types now under discussion. The statement that a subject responds with reference to a given rule and the use of such terms as " inferentially possible " and ' ' inferentially impossible ' ' are meant to imply neither aware- ness of rules nor capacity for inference on the part of the sub- jects. In the case of a given subject a few Type A reactions may be manifested during his series of 100 trials, even where an examination of all his reactions may disclose a tendency to favor rather than to avoid the trial-to-trial varying impos- sible alley. 2. Response to the rule that it is useless to try any alley more than once during a given trial; all four alleys tried, and in an irregular order. This is the Type B reaction of the tables. An example of this type would be afforded by a subject if, during his sixteenth trial, he were to try the alleys in the following order: 4, 1, 2, 3; or, 1, 4, 2, 3; or, 2, 1, 4, 3, etc. If this tendency alone were operative during the formal series the subject would have an average chance of effecting his 100 escapes by trying the various alleys 250 times. Since many of the subjects, including the children, were manifestly but little influenced by the one- impossible-alley rule the reaction-type tables omit all reactions in which no alley was entered more than once during a given trial and one or more of the three inferentially possible alleys was not tried, even though the impossible alley was tried. For example, during the sixteenth trial alley No. 3 was the right one and alley No. 2 the impossible one. If the subject tried only alleys No. 1, 2 and 3 or only alleys No. 2 and 3 his reac- tion was not recorded in the tables. These incomplete reac- tions enter, however, into another reckoning of results, as will be seen in subsequent pages. PERSEVERANCE REACTIONS IN PRIMATES AND RODENTS 13 3. When No. 4 is the right alley an effort to escape by trying alleys No. 1, 2, 3 and 4 once each, in the order given; or the reverse of this when No. 1 is the right alley. This reaction-type, which is recorded as Type C, is an ex- pression of an interesting tendency to stereotype a systematic mode of searching for a varying place of escape. Some subjects have a marked tendency to try first the alley at the extreme right or extreme left of the row, and to follow this by trying the other alleys in the order of their occurrence along the row until the right one is found. Individual monkeys and a type of feeble-minded child will be found to manifest this type of reaction with especial frequency. It will be discussed at greater length in subsequent pages. It is obvious that if a subject were invariably to initiate his reaction by trying alley No. 1 and to follow this by tr>dng the other alleys in order until the right one were found he would manifest 25 Type C reactions per 100 trials, and would try the various alleys 250 times. The same holds true in cases where alley No. 4 is the first choice for each trial and the reverse order is followed. 4. More than one separate effort to escape by a given alley during the same trial, but with an interruption of such efforts by an inter- val of effort to escape by one or more of the other alleys. This is the Type D reaction of the tables. In most cases it probably involves the operation of two different general ten- dencies, viz., (a) a tendency to vary searching activities until adjustment is effected and (b) a tendency to repeat an activity that has once been manifested. Examples of this reaction during the sixteenth trial are as follows: 4, 1, 2, 4, 3. 2,1,2,3. 4, 1, 4, 3. 5. During a given trial the subject enters an alley which does not afford escape, leaves it and reenters it one or more times without having tried any other alley; or, having tried a group of two or three alleys in a certain order he reenters all alleys of the group in the same order one or more times. Examples of the Type E reactions during the sixteenth trial are as follows: 4, 4, 4, 3. 1, 2, 4, 1, 1, 1, 1, 1, 1, 1, 3. 2, 1, 4, 2, 1, 4, 2, 1, 4, 3. 1, 2, 1, 2, 1, 2, 1, 2, 1, 2, 1, 2, 1, 2, 4, 1, 2, 3. 6. During a given trial the subject persistently avoids the right alley until he has tried the other alleys at least six times. 14 G. V. HAMILTON The limitations of my method render it impossible to dis- tinguish clearly in all cases between this " persistent avoidance " reaction-type and the reaction described above as Type E. Since the two reaction-types have a single general characteristic I have classified them together in the tables as Type E reac- tions. At times a subject will start to enter an alley, obviously inhibit the impulse to do so, and pass on to other alleys, which are entered without hesitation. If the " inhibited " alley hap- pens to be the right one for that trial persistence of this inhibi- tion may lead to the entrance into the other alleys many times, and with constant variation of choice among them. I have usually found that such episodes are characteristically manifested under conditions which contain adventitious stimuli that are conducive to either excitement or distraction, or both. In most cases where a subject tries other alleys than the right one during a given trial many times one cannot be sure whether the subject is manifesting (1) a persistent inhibition of an originally positive impulse to enter a given alley or (2) merely a persistent positive impulse to enter a given alley or alleys, so that one alley appears to be avoided when it is not actually an " inhibited " alley. II Reaction-type Tables. — In table 2, which directly follows, the number of each type of reaction that was manifested by the individual subject is indicated both by a horizontal line of dashes and a numeral at the end of the line. Thus the letters " A," " B," " C," " D " and " E " appear after Girl I's place in the table, and in Hne with "A" are 17 dashes, followed by the numeral 17. This is meant to indicate that this subject mani- fested 17 Type A reactions during her 100 formal trials. The reactions obtained during the preliminary trials do not appear in this table. TABLE 2 Subject Age Number of each Reaction-Type Girl 1 3 years... A. 17 B. 3 C. 4 D. 11 E. 6 PERSEVERANCE IN REACTIONS PRIMATES AND RODENTS 15 TABLE 2— Continued Subject Age Number of each Reaction-Typs Girl 2 3 years. .A. 3. This subject had only 75 trials. B. 2 C. 13 D. 3 E. 4 Girl 3 4 vears. . . A. 12 B. 0 C. D. 0 [ 20 E. 0 Girl 4 4 vears. . .A. 0. Sister of Girl 7 B. 0 C. D. 0 [— 22 E. —1 Girl 5 5 vears. ..A. • 17 B. -1 C. D. 6 [ 22 E. 2 Girl 6 5 vears. . .A. 5. This subject had only 90 trials. B. 4 C. 6 D. 9 E. 0 Girl 7 A. — 1. Sister of Girl 4 — feeble-minded. B. —1 C. D.O [ 19 E. 0 Girl 8 5 years. . .A. • 14. Sister of B. 0 [Girl 12 C. 2 D. 3 E. 11 Girl 9 5 years ... A. — 1 B. 4 C. 4 D. 5 E. 12 Girl 10 6 years. . .A. 0. Feeble-minded. B. 0 C. D. 4 [ 21 E. 0 Girl 11 7 vears... A. 14 B. 4 C. 3 D. 13 E. 0 16 G. V. HAMILTON TABLE 2 — Continued Subject Age Number of each Reaction-Type Girl 12 7 years. ..A. 13. Sister of B. 0 Girl 8. C. 3 D. E. — 1 [ 19 Girl 13 7 years. . . A. 7 B. 0 C. 7 D. 4 E. 0 Girl 14 7 years. ..A. 12 B. —1 C. 15 D. 8 E. 0 Girl 15 7 years. ..A. -9 B. 0 C. 11 D. 10 E. 0 Girl 16 9 vears...A. 15 B. 0 C. 13 D. 4 E. 0 Girl 17 9 years. ..A. 12 B. —1 C. 5 D. 14 E. —1 Girl 18 9 years... A. —1 B. 0 C. 8 D. 11 E. 0 Qirl 19 9 years A. • —6. This subject had only 50 trials. B. 0 C. 5 D. 2 E. 0 Girl 20 . . . .12 years..A. H- Feeble-minded. B. 6 C. 4 D. 4 E. 0 Baboon 2 Adult ...A. 8. Trials 1-100. . B. 8 C. 6 D. 8 E. 7 PERSEVERANCE REACTIONS IN PRIMATES AND RODENTS 17 TABLE 2— Continued Subject Age Number of each Reaction-Type Baboon 2 A. 14. Trials B. 3 [101-200 C. 7 D. 12 E. —1 Baboon 2 A. 11. Trials 201-300. B. 14 C. 15 D. 4 E. — 1 Baboon 2 A. 7. Trials 301-400. B. 9. C. 7 D. 7 E. 2 Baboon 2 A. 12. Trials 401-500. B. 4 C. 10 D. 8 E. — 1 Monkey 5. . . . Adult .... A. 5 B. 9 C. 2 D. 12 E. 18 Monkey 16. . .Adult. . . .A. ■ 10 B. 9 C. 11 D. 2 E. 3 Monkey 18. . .Adult ... .A. 7 B. 3 C. 5 D. 13 E. 8 Monkey 26. . .Adult. . . .A. 2. Trials 1-100. B. — 1 CO D. 9 E. 30 Monkey 26 A. 8 Tnals 101-200. B. 5 C— 1 D. 15 E. 7 18 G. V. HAMILTON Subject Age Mouse 1 Adult . Grav Rat 1 TABLE 2— Continued Number of each Reaction-Type A. 4 B. 7 C. 7 D. 5 E. Adult. .. A. B. CO D.- E. - -24 -33 Gray Rat 2, . Adult. A. B. C. 2 D. E. -14 Grav Rat 3.. Adult. Gray Rat 4. . Adult Grav Rats.. Adult. A. - B. ~ CO D.- E. - A. B. C D. E. A. B. C D. E. -27 -22 -33 Black Rat 1 ..Adult. A. B. C D. E. 3 -10 -16 Black Rat 2 ..Adult. Black Rats ..Adult. B. 3 C 2 D. E. A. B. 4 CO D. E. -10 -23 -20 PERSEVERANCE REACTIONS IN PRIMATES AND RODENTS 19 TABLE 2— Continued Subject Age Number of each Reaction-Type Black Rat 5 . .Adult \. 4 B. —1 C. 2 D. 7 E. White Rat 1 . .Adult. ...A. 9 B. 5 C. 14 D. 3 E. White Rat 5. .Adult A. B. —1 D. — - E. — - -30 -20 White Rat 2. .Adult. .A. 5 B. —1 C. 2 D. 12 E. 17 White Rat 3. .Adult. . . A. 9 B. 5 C. 3 D. 12 E. 12 WhiteRat 4..Adult. .. .A. 6 B. 0 C 3 D. 3 E. -40 -23 White Rat 6. Adult. . ..A. 12 B. 4 C. 4 D. 12 E. 18 White Rat 8. .Adult. .. .A. 5 B. —1 C. 5 D. 7 E. 18 White Rat 10. Adult. . . .A. 2 B. —1 CO D. 8 E. -26 20 G. V. HAMILTON TABLE 2 — Continued Subject Age Number of each Reaction-Type Gopher 1 Adult ... .A. 2 B. 0 C. 2 D.O E. 53 Gopher 2 Adult. . . .A. — 3 B. — 1 C. — 1 D. 6 E. 45 Gopher 3 Adult. . . .A. 4 B. 2 C. 3 D. 10 E. 34 Gopher 4 Adult A. 4 B. — 1 C.— 1 D. 2 E. 51 Gopher 5 Adult. . . .A. 6 B. 2 CO D. 4 E. 46 Gopher 6 Adult A. 2. This subject had onlv 80 trials. B. — 1 CO D. 2 E. 37 III Discussion of Tabic 2. — This table discloses the fact that marked individual differences are elicited by the quadruple choice method. The behaviorist who is accustomed to methods which favor habit formation is apt to encounter such differences only during the first few trials with animals for whom the ex- PERSEVERANCE REACTIONS IN PRIMATES AND RODENTS 21 perimental situations are unfamiliar and complex. If, e.g., a difficult maze or puzzle box situation be presented to a half dozen mammalians of the same age and species their first-trial random movements will be found to vary markedly from indi- vidual to individual. Unfortunately, only such differences as can be tabulated in terms of reaction-time and numbers of errors are apt to be recorded in experimental studies of habit forma- tion. One of the most important generalizations, perhaps, that can be based upon Table 2 is this: When a mammalian is confronted by a series of situations for which he is unable to discover and stereotype a specifically adequate and invariably successful mode of response he tends to vary his responses in a manner which is less a species than an individual characteristic. Mammalian neural organization is so complex and its func- tions are so plastic that there is excellent ground to assume, a priori, that slight individual differences of reactive equipment will be reflected by marked differences, as to detail, of behavior which conforms to the random movement type. Further ex- emplification of this point can be found in comparisons of the specimen individual records given in the appendix. Although a Type A reaction presents the objective character- istics of a response which has reference to an inferred principle or experiment rule, its occurrence can be regarded as such with certainty only when, in the course of a large number of trials, it clearly predominates over all other types It was the most frequently manifested reaction of three-year old Girl 1, five-year old Girl 8, seven-year old Girl 11, nine-year old Girl 16, nine- year old Girl 19 and feeble-minded, twelve-year old Girl 20. No animal subject presented an excess of Type A reactions over any other type of reaction during the first series of 100 trials. Baboon 2 had an equal number of Type A and Type B reac- tions during his first 100 trials, and during his second and fifth series he had a plurality of Type A reactions. These reactions are wholly absent from the records of but two subjects, viz., four- year old Girl 4 and feeble-minded, six-year old Girl 10. These two girls stereotyped a method of response which, as can be seen from a glance at their records in the appendix, precluded the possibility of " by-product " or accidental Type A reactions. 22 (1. V. HAMILTON With such marked individual differences to take into account we cannot attach much importance to group averages. Never- theless, the averages given in table 3 are of some interest, as showing that only 2 (Monkey 16 and White Rat 6) of the 28 animal subjects from whom complete series were obtained equalled or exceeded the girl-group average of 9.18 Type A reactions per 100 trials. Table 3 should be directly compared with table 4 since it will suggest that Type B reactions (which involve trying the one impossible alley as well as the three possible ones) may be manifested as an equivalent of Type A reactions in subjects who are capable of learning that it is useless to enter any one alley twice during the same trial. It is not to be forgotten that even a few Type A reactions may express a feebly potent tendency to avoid the varying impossible alley. On the other hand a reaction of this type may be accidental in the sense that failure to try the impossible alley may be due to the operation of other tendencies than that which would lead to an avoidance of it. E.g., if a subject stereotypes a method of initiating all trials by first entering alley No. 1 and of following this activity by entering, wherever it is necessary, the other alleys in the order of their occurrence from left to right until the right one is found he will necessarily manifest an equal number of Type A and Type C reactions. The same will obtain if he adopts the reverse method, i.e., if he initiates each trial by entering alley No. 4 and working from right to left in search of the right alley. TABLE 3 TABLE 4 Average number Average number A-reactions B-reactions Group per 100 trials per 100 trials Girls (incomplete records excluded) 9. 18 1 . 24 Girls (feeble-minded cases excluded) .... 10. 29 1 . 00 Baboon and Monkeys 6. 40 6. 00 (atypical Monkeys 5 and 26 excluded) . . 8.33 6.67 Mouse 4.00 7.00 Gray Rats 2.80 5.20 Black Rats ^ 6.2.5 2.50 WhiteRats . 6.38 2.25 Gophers 3.80 1.20 PERSEVERANCE REACTIONS IN PRIMATES AND RODENTS 23 Type P Reactions. — Table 2 shows that 10 of the 20 girls manifested no Type B reactions, whilst only 2 of the 29 animal subjects — a white rat and a gopher — failed to manifest any of these reactions. The maximum number of Type B reactions for the first 100 trials of individual subjects is found in the records of Monkeys 5 and 16 and Gray Rat 2, each of whom had 9 of these reactions. Baboon 2 had five series of 100 trials each. He had 8 B-reactions in his first series, 3 in his second series, 14 in his third series, 9 in his fourth series and only 4 in his fifth series. The girls, black rats, white rats and gophers — 3)d< subjects — afford but one case of an individual whose B- reactions are equal to or exceed the average number of B-reac- tions for either the baboon-monkey or gray rat group. This subject. Girl 20, is feeble minded. Type C Reactions. — The reactions that are classified under this heading have the essential objective characteristics of a quite definite reaction-type, hence their distribution in the records of individual subjects (table 2) and in the group aver- ages (table 5) is of considerable interest. The findings recorded in table 2 suggest that the tendency toward manifesting C- reactions is an individual rather than a species characteristic, but this is an acceptable interpretation in only a very much qualified sense. It is to be remembered that when an animal seeks to escape from conjinement his reaction to baffled effort is apt to enter as a determinant of his behavior, and that such reac- tion is influenced by various individual traits, particularly by timidity, distractibility and excitability. Whenever any of these traits enters into the situation as a determinant of reaction the subject is apt to make repeated efforts to escape by way of a particular possible outlet or, at least, to return to a previously tried possible outlet before all have been tried. Individual mem- bers of each of the various mammalian species with which I have laboratory familiarity (with the possible exception of gophers) have displayed a clearly instinctive tendency toward C-reactions, hence I believe that this tendency is widely dis- tributed among mammalian species. In my earlier studies (1) one of my dogs had 15 C-reactions in 100 trials. It was mani- fested by one of five cats that were used as subjects — Cat 1 of those studies had 5 C-reactions in 100 trials. With the exception of the gopher group, each group of sub- 24 G. V. HAMILTON jects presented one or more individuals whose records display an abrupt appearance of C-reactions, their persistence for a number of trials and their abrupt disappearance. It would seem that the mammalian, both primate and infra-primate, has a repertoire of more or less distinct reactive tendencies for meeting situations which call for efforts to escape, and that where a specifically adaptive adjustment is not learned, now one, now another of these tendencies will come to expression. From the standpoint of almost any of the behavior technologies it is highly important to explore for the conditions that char- acteristically set each of the various known tendencies in opera- tion. This problem must remain an unsolved one so far as the present studies are concerned, but I believe that it will be found that C-reactions are not apt to occur under conditions which are conducive to intensity of affective response. Some of the girls, particularly Girls 4, 7 and 10, tended to alternate a first choice of alley No. 1 with a first choice of alley No. 4 from trial to trial in starting the systematic (Type C) search for the right alley. This largely accounts for the fact that Girls 4 and 10 did not manifest a single A-reaction, although a habit of displaying C-reactions is apt to yield as many " by product " or incidental A-reactions as C-reactions. Another factor which accounts for these two girls' failure to give any A-reactions was this: whenever a given end alley proved to be the right one it was tried first at the next trial, regardless of the demands of the self -established alternating rule. Girl 7's one A-reaction occurred during her sixtieth trial, when this factor of recency failed to distract her from her habit of alternat- ing her first choice of end alley with which to begin C-reactions. None of the animals displayed this trial-to-trial alternation of first choice of end alley. Occasionally, however, a subject who seemed to have fixed a habit of always trying first the alley at a given end of the row would reverse his choice. E.g., all of Monkey 16's C-reactions from his sixteenth to his fifty-sixth trial were initiated by a first choice of alley No. 1, but on enter- ing the apparatus for his fifty-seventh trial he tried No. 4 first, and since No. 1 was the right alley for that trial he followed this by entering alleys No. 3, 2 and 1 in the order given. During the remainder of his series all of his-C reactions conformed to the 4-3-2-1 order. PERSEVERANCE REACTIONS IN PRIMATES AND RODENTS 25 TABLE 5 Average number of C-reactions per 100 trials Group Girls (incomplete records excluded) 10. 77 Girls (feeble minded cases excluded) 9 . 93 Baboon and Monkeys 4 . 80 (atypical Monkeys 5 and 26 excluded) 7. 33 Mouse 7.00 Gray Rats 1 . 00 Black Rats 3. 50 White Rats 4.88 Gophers 1 . 40 Type D Reactions. — This is a form of discontinuous repetition of a non-adaptive direction of activity which enters into the typical perseverance reactions of animals and even of human adults in the presence of difficult and unusual or highly exciting situations. At this point it will be found convenient to examine the individual records of D-reactions in table 2 and the group averages in table 6: TABLE 6 Average number of D-reactions per 100 trials Group Girls (incomplete records excluded) 6. 82 Girls (feeble-minded cases excluded) ' 7. 71 Baboon and Monkeys 8 . 80 (atypical Monkeys 5 and 26 excluded) 7. 67 Mouse 5. 00 Gray Rats 6. 20 Black Rats 6.50 White Rats 8. 13 Gophers 4.40 Since the records of D-reactions are not easily interpreted without constant reference to those of the E-reactions, both types will be discussed together, and table 7 will follow directly : 2 2 03 ,43 13 6 20 .00 24. 00 26 00 22. 25 21. 75 45. 80 '26 Ci. V. HAxMILTOX TABLE 7 Average number of E-reactions per 100 trials Group Girls (incomplete records excluded ) Girls (feeble-minded cases excluded) Baboon and Monkeys (atypical Monkeys 5 and 26 excluded) Mouse Gray Rats Black Rats White Rats Gophers Although the individual records of D-reactions (table 2) at first sight seem to be quite perplexing on account of the indi- vidual distribution of these reactions, these records and those of E-reactions serve to throw some light on the general problem that concerns the determinants of individual differences of reaction. Table 2 shows that seven-year old Girl 12 had 19 D-reactions, whilst four-year old Girls 3 and 4 and five-year old Girl 7 had none. Monkey 16 had only 2 D-reactions whilst Monkey 18 had 13; the maximum nimiber for any individual of the black rat group is 10 and the minimum 3; for white rats the maximum and mininium numbers are respectively 12 and 3, and for gophers they are 10 and 0. Girl 12, who gave 19 D-reactions, was bashful, easily distracted and much less attentive to the apparatus situation than to the experimenter. vShe and her sister, Girl S, were two pathetic little Irish girls in a group of swarthy, self-assertive Mexican children, a circumstance which led me to establish bonds of friendship with them that had better have been delayed until the experiments were over. They were easily rattled by their own blunders, and never performed well when they were not allowed to a]ipear in their Sunday clothes. Under these condi- tions Girl 12 fell into the error of reentering an already tried alley after an interval of effort to escape by an as yet (for the trial) untried alley. Her younger sister. Girl 8, under similar conditions was more apt to display the more i^rimitive E-reac- tions, of which she had 11 in 100 trials. PERSEVERANCE REACTIONS IN PRIMATES AND RODENTS 27 Nine-year old Girl 17, with a record of 14 D-reactions, belongs to a highly distractible, prematurely erotic type of native (Mex- ican) California children who are the despair of local school teachers. Distractibility in school during the period of awak- ening erotic interests, unwillingness or inability to attend to things toward which there is no instinctive inclination and easy excitahil'ty are, of course, more or less characteristic of most children. But these traits are present in an extreme degree in children of the type to which I refer. Girl 18, with 11 D-reac- tions, belongs to this type. Girl 11, with \3 D-reactions, was in more or less constant terror of an embarrassing enuresis which was especially apt to assert itself during the experiments. She is a naturally intel- ligent, attentive child, and would probably have given me records similar to those of the children of my earlier studies had she not thus been handicapped. Girl 15, with 10 D-reactions, is unusually excitable. During her trials she would rush impulsively from alley to alley, up- braiding herself all the while for her stupidity. Girl 6, with 9 of these reactions in 90 trials, was a homesick, listless, dejected newcomer to the orphanage. She wandered through the appa- ratus in apparent indifTerence to the problem before her. Monkey 16's variation from the average from his group is sufficiently marked to require explanation: he had only 2 D- reactions, whilst Baboon 2, the individual of the baboon-monkey group whose record of these reactions most nearly approximates his own, had 8 such reactions during his first 100 trials. Mon- key 16, who was " Sobke " of Yerkes' subsequent experiments (4), is less easily distracted or excited than is any other monkey of which I have knowledge. Type E Reactions are wholly absent from the records of 12 of the 20 girls, and only four of these subjects have more of these reactions than the low average for the group. As might have been expected, three-year old Girls 1 and 2 displayed more of the primitive Type E reactions than did any of the older girls with the exception of the two special cases to be cited. Girl 8, sister of Girl 12, has already been discussed in the para- graph devoted to an explanation of the occurrence of D-reac- tions in the girls' records. This five-year old girl, like her sister, was bashful, easilv distracted and more attentive to the 28 . G. V. HAMILTON experimenter than to the apparatus. The detailed record of her reactions, which is given in the appendix, shows lapses from periods during which she quite obviously reacted with refer- ence to the one-impossible-alley principle into periods of grossly inadequate behavior. During one of her lapses she was almost exclusively attentive to her clothing, and after her day's trials were over she wept, declaring that she felt humiliated to have appeared before me so unattractively clad. Girl 9, with 12 E-reactions, is a five-year old Italian. She is excitable, inor- dinately distractible and a favorite among the other children and the teachers on account of her merry ways. It was im- possible to make her take the experiments seriously. Monkey 5, with 18-E reactions, and Monkey 26, with 30 E- reactions, have already been discussed as atypical animals. In my earlier studies five monkeys gave an average of 4.8 E-reac- tions per 100 trials, hence the records of Baboon 2 and Monkeys 16 and 18 of these studies (an average of 6 E-reactions per first 100 trials) may be taken as fairly typical in this respect. It is important for subsequent discussions of the determinants of D and E-reactions that only three of the 25 human and infra- human primate subjects of these studies (Girl 9 and atypical Monkeys 5 and 26) have individual records of E-reactions which equal or exceed that of the rodent subject which has the mini- mum number of such reactions for the entire group of 24 rodents. It is clear that primitiveness of reactive equipment favors the manifestation of E-reactions. Although the various rodent species have records which dis- close marked individual differences, both the table of averages of E-reactions (table 7) and the table of individual reactions (table 2) support the view that in the presence of a situation for which they have no specifically adaptive instinct these sub- jects have, without exception, a marked tendency toward the repetition of unsuccessful as well as of successful attempts at adjustment. The high average of E-reactions for gophers (45.80 E-reactions per 100 trials) is consistent with my observations of the field and cage behavior of these animals. I have a small experimental garden, in one part of which a gopher persistently attacked the roots of a loganberry vine. One of the numerous mouths of his network of burrows was close to a hydrant, so that I was able to turn water into it without much trouble to PERSEVERANCE REACTIONS IN PRIMATES AND RODENTS 29 myself. For several weeks it was my daily practice to remove the little mound of earth with which the animal occluded this burrow mouth and to flood the burrow with water. My per- sistence was equalled by the gopher, who invariably returned to the muddy burrow during the night and concealed its opening with earth. This involved much labor on the gopher's part without advantage, since these animals will not occupy a muddy burrow and will not feed on roots that are growing in well satu- rated ground if roots can be found in moist rather than muddy ground. In this case the gopher's burrows extended to logan- berry roots in desirable feeding ground. A much clearer example of gopher persistence in the face of disadvantage was obtained when I put 12 gophers into a large cage in which there were 12 small nests. Each gopher was driven into a nest and I hoped to avoid the slaughter that usually follows any effort to keep a number of these animals in a common cage. It was soon observed, however, that if a gopher, having left his own nest to secure food, invaded another nest after his cheek pockets were filled with grain and bits of carrot, a battle would ensue which would terminate only when one of the combatants was either dead or a helpless cripple. No matter how large and powerful the occupant of the invaded nest might be, nor how small the disputant of his occupancy, the invader did not seem to be deterred by the disadvantageous consequences of his persistence. He would dart into the nest, only to reappear in a moment, thrust out by its rightful occupant, re-enter and suffer a second expulsion, and so on, until exhaustion or death terminated his stupidly persistent repetition of the nest-seeking reaction. A little variation of reaction would have led him to his own nest or, at least, to one less well defended. Within a fortnight all but two of the 12 gophers were slain in this manner. The survivors were two large males who occupied nests at oppo- site ends of the cage. During the experiments it was more or less characteristic of all rodent subjects, and particularly so of gophers, that a dis- advantageous experience in a given alley did not seem to deter the animal from reentering it. After a subject had entered a given alley a number of times during the same trial he was apt to become inactive within that alley, so that electrical stimula- tion was necessary to bring about a resumption of activity. 30 G. V. HAMILTON This penalty for failure to continue would cause the animal to run out of the alley into the semicircular enclosure, following which he would be as apt to reenter the alley from which he was thus driven as any of the untried alleys. It might be expected that the disadvantage growing out of lack of success in attempt- ing to escape through a particular alley and electrical stimula- tion for " loafing " in it would quickly lead to avoidance of that alley. On the contrary, my records contain many instances of persistent reentrance of an alley during a given trial when as many as 10 successive punishments therein failed to direct the subject's activities toward the untried alleys. I had feared that the introduction of disadvantageous stimulation (electric shock) would lead to avoidances of particular alleys, but this expectation was set at naught by actual experiment. This fact suggests that disadvantage that is not of a kind which is avoided by reason of inherent, specifically appropriate features of reactive equipment is a considerably less important determinant of reac- tion than the traditional assumptions of comparative psychology would lead us to suspect. I am not prepared, however, fully to subscribe to Watson's (6) view that advantage and disadvan- tage as such have nothing to do with habit formation. This point will be discussed in another part of these studies. The automatic repetitions of unproductive activities and of inappropriate inhibitions that one observes in the schizophrenic psychoses are ascribed to purely psychical determinants by the psychoanalysts, but there is much of value both to psycho- pathology and to comparative psychology in the fact that such behavior is apt to assert itself in the mammalian under conditions that are unfavorable to the manifestation of more adequate reac- tive tendencies and under certain natural conditions which are characteristically encountered by infra-primate species. The ob- servations that are recorded in these studies, supplemented by various field and clinical observations, lead me to believe that any of the following conditions are apt to precipitate reaction of the D type and E type described in preceding pages : 1. Inherent primitiveness of reactive equipment, such as is pos- sessed by rodents and the young of more highly developed species. — In my earlier studies (1) a twenty-six-months old infant gave 13 E-reactions and 10 D-reactions, whilst none of 10 human subjects, whose ages ranged from eight years to maturity gave PERSEVERANCE REACTIONS IN PRIMATES AND RODENTS 31 a single reaction of this type with the exception of a fifteen-year old boy who, in a moment of embarrassment reentered an alley that he had just tried unsuccessfully. In the present studies only the two three-year old girls and two temperamental five- year old girls of a total of 20 human subjects exceeded the group average of two E-reactions per 100 trials. The individual records of the 24 rodents, on the other hand, disclose a minimum of 12 such reactions per 100 trials for any individual and a maximum of 53. 2. Excitability. — When a subject is directing his activities toward adjustment to a situation which has initiated reaction the repetitive tendency is favored by the concurrence of adjus- tive reactions and strong emotional reactions. Even the average human adult, if trapped and badly frightened in a burning hotel, will rush madly again and again to a part of the building which obviously will not afford escape from the building, thereby divert- ing time and effort from as yet untried possibilities of escape which would readily occur to him if he were not excited. An angry man will stupidly repeat an empty phrase instead of varying his efforts to effect a verbal adjustment to the situation. If now, any subject, either human or infra-human, is easily excited by a situation which elicits reaction by failure to effect a ready adjustment to it or by intercurrent or adventitious stimuli, he is more apt to manifest reactions of the primitive D and E types than is the subject whose affective responses to such con- ditions are of less intensity. 3. Distractibility. — Among primates and, to a certain extent, among infra-primate mammalians one is apt to encounter more or less marked individual differences of capacity for sustained exclusive responsiveness to an unexciting situation which precipi- tates adjustive reaction thereto. Distractibility consists in a tendency toward quick shifts of responsiveness from possible stimulus to possible stimulus, with consequent absence of sus- tained responsiveness to a particular stimulus. When a stimu- lus or stimulus-complex S-1 leads to adjustive activities, and intercurrent stimuli S-2, S-3, etc., cause an incomplete shifting of responsiveness from S-1 to S-2, S-3, etc., so that activities directed toward adjustment to S-1 do not wholly cease, we may expect the subject to manifest reactions of the D or E type. 32 G. V. HAMILTON 4. Feeble responsiveness. — A given set of stimuli may elicit activities which are directed toward adjustment to them in a manner to suggest that the threshold of responsiveness has barely been crossed. An example has but this moment come within my own experience: I was interrupted by a nurse who wished to consult me about an unimportant matter which con- cerns a patient for whose care I am responsible. I responded to his question by seeking, in a rather indifferent frame of mind, an answer that would satisfy, not so much the patient's needs as the professional amenities of the situation. I found myself saying, inadequately, " The woolen cap — the woolen cap — the woolen cap." I was barely responsive to the situation, and the result was a reaction which consisted in the repetition of an unsuccessful attempt at adjustment. Had I been more respon- sive, attempts at verbal adjustment would have varied without repetition until a successful one had been found. Where habit serves as a guide for the direction of activity such repetition is less apt to occur, since a successful adjustment is at once effected, thereby removing the subject from the influence of the situation. In the schizophrenic psychoses the subject characteristically withdraws his interest from reality, and is therefore apt to be but feebly responsive to the outside world. There is also apt to be a deterioration of the habits that were accjuired before the advent of the psychosis, hence this disorder presents conditions which we have already found to be favorable to the manifes- tation of D and E reactions (the cases of Girls 6, 8 and 12, alreadv cited). IV Determinants of the Initial Direction of Activity. — When a subject first enters the enclosure he may first try any one of the four alleys for escape. The fact that this first choice, as it will hereafter be designated, was varied from trial to trial or at least every few trials by many of the subjects calls for an examina- tion of the records for the purpose of discovering, if possible, the determinants of first choices. With this in view the record of each subject has been given a detailed analysis, the results of which are presented in table cS. This table requires the following explanations : 1. In the column that contains the list of subjects an as- terisk (*) is placed opposite the name of any subject who had less than 100 trials. Girl 2 had only 75 trials. Girl 6, 90 trials, PERSEVERANCE REACTIONS IN PRIMATES AND RODENTS 33 Girl 19, 50 trials, and Gopher 6, 80 trials. All other subjects had 100 trials each, and only the first 100 trials of subjects who had more than a single series are given in table 8. 2. The heading, First Choice, is the one under which is given 1234 the number of times that each subject entered first each of the four alleys at the beginnings of trials. E.g., the first horizontal group of figures under this heading is meant to show that Girl 1 initiated reaction in 17 of her 100 trials by entering alley No. 1; that in 28 of her trials she made first choice of alley No. 2, etc. 3. Under the heading. Alley Entrances, is shown the number I 2 J 4 Total of times that each of the various subjects entered each of the four alleys during his 100 trials. It also shows under the sub- heading, Total, the total number of separate alley-entrances that were made by each of the subjects. E.g., Girl 1, in the course of her 100 trials, entered alley No. 1 49 times, alley No. 2 83 times, etc. In her efforts to escape from the apparatus 100 times she entered the various alleys a total number of 272 times. Possible Determinants 4. Under the heading, of First Choice is shown the R. F. R.-F. number of times that the individual subject's first choices sus- tained the relations of (a) recency, (b) frequency and (c) both recency and frequency to previous trials. If a subject began a trial by first entering the alley that was last entered during the immediately preceding trial one recency reaction was counted for that trial. If the alley first chosen was the one that had been most frequently entered during all of that subject's preceding trials one frequency reaction was recorded. If the alley first chosen sustained both the relation of recency and that of fre- quency to preceding trials one recency-frequency reaction was recorded, and neither a recency nor a frequency reaction was recorded in a case of this kind. The records showed that there were a few cases where two or even three of. the four alleys had been entered an equal number of times at the end of a given number of trials, and that each of these two or three alleys might be said to sustain the relation of frequency to previous trials. In such cases first choice of any of these two or three alleys was recorded as a frequency reac- tion for that trial. 34 G. V. HAMILTON TABLE 8 First choice Subject L 2. 3. 4. Girl 1 17 28 21 34 Girl 2* 46 14 7 8 Girl 3 54 3 1 42 Girl 4 32 5 18 45 Girl 5 16 11 8 65 Girl 6* 6 35 27 22 Girl 7 58 3 1 38 Girl 8 10 19 45 26 Girl 9 26 65 6 3 Girl 10 79 21 0 0 Girl 11 20 33 32 15 Girl 12 15 32 46 7 Girl 13 22 28 30 20 Girl 14 31 16 12 41 Girl 15 25 19 25 31 Girl 16 28 21 26 25 Girl 17 18 25 42 15 Girl 18 24 27 34 15 Girl 19* 8 10 17 15 Girl 20 16 33 37 14 Baboon 2 46 25 12 17 Monkev 5... 16 30 31 23 Monkey 16... 64 1 1 34 Monkev 18... 68 18 2 12 Monkey 26... 13 24 11 52 Mouse 1 11 24 23 42 Grav Rat 1 . Gray Rat 2 . Gray Rat 3. . Gray Rat 4.. Gray Rat 5.. Black Rat 1 . Black Rat 2. Black Rat 3 . Black Rat o . 11 26 29 34 26 20 13 41 16 17 26 41 6 8 28 58 5 14 27 54 2 18 21 59 24 43 15 18 3 68 19 10 43 29 17 11 White Rat 1. 11 11 White Rat 2. 9 24 White Rat 3. 79 9 White Rat 4. 18 31 White Rat 5. 39 27 White Rat 6. 12 20 White Rat 8. 12 16 White Rat 10. 17 43 Gopher 1 11 20 Gopher 2 20 25 Gopher 3 7 23 Gopher 4 42 25 Gopher 5 8 13 Gopher 6*.... 9 26 25 53 33 34 11 1 35 16 22 12 43 25 47 25 30 10 19 50 16 39 32 38 16 17 33 46 21 24 Allev entrances Possible determinants of first choice 4. Total R. F. R.-F. 49 58 61 55 48 29 70 37 54 49 39 50 48 49 52 41 21 46 131 75 109 78 114 83 59 59 62 77 68 62 77 49 129 82 82 81 89 59 67 69 73 83 62 31 63 51 59 60 82 66 50 95 114 58 55 67 69 68 79 69 29 62 60 272 33 201 56 235 52 229 80 287 44 207 49 231 50 259 59 351 29 251 41 244 40 266 33 186 56 240 47 233 40 230 39 253 40 212 20 101 40 211 81 83 79 106 77 56 110 78 101 98 54 49 267 93 106 384 56 67 256 52 71 311 55 239 493 70 96 99 96 361 70 71 97 78 58 71 73 93 71 112 154 110 76 92 72 124 111 95 94 139 108 157 65 99 90 85 90 105 97 115 129 123 77 113 157 109 108 181 144 132 86 115 420 61 116 322 73 124 377 88 116 379 88 105 385 76 91 296 77 78 321 72 102 357 86 64 414 104 105 377 124 74 394 53 328 67 432 51 444 86 369 76 349 89 384 69 132 128 119 98 100 124 189 525 82 153 487 132 117 439 73 141 480 173 113 575 124 128 528 8 21 7 8 20 8 9 42 9 44 1 1 10 37 11 12 26 7 11 34 23 4 42 1 8 46 17 5 63 4 31 10 34 10 17 11 14 12 1 10 16 14 18 13 17 6 24 7 31 14 7 9 16 29 10 24 20 7 8 48 15 9 54 13 10 38 12 15 27 7 21 18 3 20 29 12 17 36 10 10 47 9 9 37 13 12 42 17 12 24 7 8 49 19 17 36 6 22 19 6 11 21 7 8 50 19 23 20 7 17 20 11 14 31 13 25 15 7 20 16 3 11 39 10 21 23 14 12 34 12 20 21 10 22 11 5 12 12 3 PERSEVERANCE REACTIONS IN PRIMATES AND RODENTS 35 Discussion of Table 8. — The figures given in this table were obtained from an analysis of 4,795 trials that were given to 49 subjects, and since the relation of each trial to the trials that preceded it was obtained by a necessarily somewhat compli- cated method of calculation, a presentation of illustrative speci- mens of this work will be reserved for the appendix. Table 8 presents in detail most of the data on which the following discus- sion is based, but occasional reference to what will hereafter be designated " calculations for table 8 " will be necessary. Alley position was evidently in itself a partial determinant of some of the subject's first choices. Feeble-minded Girl 10, for example, tried alley No. 1 first during each of 79 trials, and it is probable that the mere position of this alley as the first one to the left determined her preference for it as the first one to try. Of course a marked tendency toward Type C reactions (trying alleys in the order 1-2-J-4 or 4-J-2-1) would account for a preference of alley No. 1 or alley No. 4 because of their ter- minal positions with reference to the row of alleys. Neverthe- less, table 8 shows that of the 4,795 first choices made by the subjects, 1199 were for alley No. 1, 1,126 for alley No. 2, 1,093 for alley No. 3 and 1,377 for alley No. 4 — a finding w^hich suggests that there was no marked disproportion between the average reactive values of the terminal and the middle alleys. Twelve of the subjects have the maximum number of their first choices in the alley No. 1 column, 7 in the alley No. 2 col- umn, 11 in alley No. 3 column and 19 in alley No. 4 column. Feeble-minded Girl 10 made no first choices whatsoever of either alley No. 3 or alley No. 4. Recency as a factor in determining first choices is shown in table 8 to have played a less important role than the factor of frequency. It is to be remembered that my method is such that it is to the subject's disadvantage to initiate a trial by choosing first the alley that was last entered during the immediately pre- ceding trial. The general tendency of mammalians to repeat the most recently manifested activity on reencountering a given situation might be expected to result in more than 25 of these " recency first choices " as we may conveniently designate them, per 100 trials in subjects who are incapable of learning to react with reference to the disadvantage of such first choices. On the other hand, a subject who has more or less capacity for learning 36 G. V. HAMILTON this would be expected to manifest less than 25% of recency- first choices. Table 8 shows that only 1 of the 49 subjects (four-year old Girl 4) had more than 25% of recency first choices if we exclude the cases where first choices sustained the relations of both recency and frequency to previous trials. If we do not exclude such cases, 17 subjects — 2 girls, 1 baboon, 1 monkey, 2 gray rats, 2 black rats, 6 white rats and 3 gophers — had each more than 25% of recency first choices. Frequency was apparently an important determinant of first choices. In making the calculations for table 8 a frequency reaction was recorded whenever the subject made first choice of the alley that had been most frequently entered during all preceding trials. The table shows that 27 subjects — 11 girls, 1 baboon, 3 monkeys, 1 mouse, 4 gray rats, 3 black rats, 2 white rats and 2 gophers — had each more than 25% of frequency first choices if we exclude the cases where first choices sustained the relation of recency as well as of frequency to previous trials. If we do not exclude such cases, 40 of the subjects had each more than 25% of frequency first choices. The averages for the primate and rodent groups given in table 9 are of some interest. Subjects which had less than 100 trials each are excluded from this table: TABLE 9 Averages of Recency, Frequency and Recency-Frequency First Choices Group Recency Frequency Recency-Frequency Primates (22) 11.55% 30.68% 9.09% Rodents(23) 15.96% 28.91% 9.87% « This table suggests various interesting problems. It would be of considerable importance to have sufficient experimental data on which to base a conclusion as to whether with descent in the phyletic scale the relative importance of recency and frequency changes in favor of recency. Table 9 suggests that with ascent of the phyletic scale the role of recency as a deter- minant of habit formation decreases whilst that of frequency increases, but the individual variations from the averages are too marked to render this a legitimate interpretation. The fact that the rodents, as a group, were not more influ- enced by the factor of recency under conditions which rendered the operation of this factor disadvantageous did not attract my PERSEVERANCE IN REACTIONS PRIMATES AND RODENTS 37 attention until I read Watson's (7) contention that advantage and disadvantage as such have nothing to do with habit forma- tion. It was then too late to use most of my subjects for an experimental investigation of the problem that is dealt with in subsequent pages. The repetitive tendency that finds expression in reactions that I have classified as belonging to types " D " and " E " may play an important role in habit determination. If one may safely generalize to the effect that what an organism has done most frequently in past encounters with a situation it is most apt to do, all things else being equal, in reencounters with that situation it necessarily follows that whatever leads to the repe- tition of a given activity before a definite habit is formed must be regarded as a possible determinant of the ultimately estab- lished habit. We have found that various factors which are subject to but limited and imperfect control may operate at any time to precipitate persistent repetition of an unsuccess- fully directed activity, even though no past experience and no constant reactive value of the situation could be regarded as directly accounting for such repetition. Inwardly arising adven- titious stimuli or unpreventable causes of excitement or dis- traction that operate from without may thus account for the relative frequency of a given activity, and this frequency may in turn act as a factor in determining the composition of the ulti- mately established habit. Reference to the calculations for White Rat 5 (appendix) affords the following example: At the beginning of White Rat 5's twentieth trial she had entered alley No. 1 10 times, alley No. 2 16 times, alley No. 3 22 times and alley No. 4 18 times. During the twentieth trial a quite incalculable factor precipitated a Type E reaction, so that she entered alleys No. 1 and No. 2 7 times each. At the end of her twenty-fifth trial she had entered alley No. 2 ?>2 times and alley No. 3 only 31 times. At the completion of her series of 100 trials alley No. 2 had been entered 161 times, whilst alley No. 3 had been entered only 133 times and alley No. 4 only 56 times. In other words, during four preliminary trials (these necessarily enter into the calculations for table 8) and 19 formal trials this subject had shown a preference for alley No. 3 over all other alleys, and a preference for alley No. 4 over alley No. 2; but an episode of repetitive behavior which 38 G. V. HAMILTON could not be ascribed to any change in the intrinsic reactive values of the various alleys and which suddenly increased the frequency with which alley No. 2 had been tried established a preference for this alley which was apparent at the end of the twenty-fifth trial and was maintained throughout the remainder of the series. V Studies of Habit Formation by the Multiple Choice Method. — ■ These studies were limited to an investigation of the value of recency and frequency of advantageous reaction, and are presented here to supplement the findings of the foregoing pages, which include some information as to the value of these factors when no element of advantage is associated with their operation. Black Rat 5 and White Rats 2, 5, 6, 8 and 10 were used as subjects for these experiments. On the day following a sub- ject's conclusion of his 100 formal trials he was given a series of habit forming trials which differed from the formal series in the following features: 1. The right alley, i.e., the only one that would afford escape from the enclosure, remained the same for a given subject through- out this series, so that the subject had only to learn always to seek this one alley for exit in order to acquire a specifically adaptive habit. 2. In most cases the alley selected by the experimenter as the right alley for the series was the one that had been entered by the subject the least number of times during his 100 formal trials. If, however, it appeared toward the end of the formal series that an episode of repetitive (E type) reaction had inclined the subject toward the alley of least frequent entrance another alley was selected by the experimenter as the right one. 3. The trials were repeated during a single session until the subject failed to try any alley but the right one during six suc- cessive trials. The records for each of the six subjects will be presented separately : Black Rat 5 During this subject's 100 formal trials she entered alley No. 1 154 times, alley No. 2 110 times, alley No. 3 89 times and alley No. 4 64 times. Allev No. 1 was her first choice in 43 trials, PERSEVERANCE REACTIONS IN PRIMATES AND RODENTS 39 alley No. 2 in 29 trials, alley No. 3 in 17 trials and alley No. 4 in 11 trials. Alley No. 3 rather than alley No. 4 was chosen as the one through which she would be allowed to escape during her habit forming trials since she had entered the former alley but 14 times during her last 20 formal trials, whilst she had entered alley No. 4 22 times, and had entered it 7 times during her ninety-sixth trial. Table 10 gives her reactions during the 20 trials that she had before she ceased to try other alleys than the right alley, No. 3. TABLE 10 Alleys Entered by Black Rat 5 During Each of 20 Habit-Forming Trials 1st trial— 3 2nd " 2-1-4-1-4-2-3 3rd " 2-4-2-3 4th " 1-4-2-4-2-4-2-4-1-4-1-4-1-3 5th " 2-4-2-4-1-3 6th u 3 7th " 2-4-2-4-1-3 8th " 1-4-1-4-1-4-1-4-1-4-2-4-3 9th " 3 10th " 2-4-1-4-1-4-1-4-2-3 11th " 2-4-1-4-4-2-4-1-4-2-4-1-3 12th « 2-3 13th " 2-3 14th " 4-2-3 15th « 2-3 16th " 2-3 17th " 2-3 18th " 2-3 19th " 3 20th " 2-3 The table shows that this subject, while forming the habit of trying first and only alley No. 3, entered alley No. 1 19 times, alley No. 2 25 times, alley No 3 20 times and alley No. 4 30 times. The repetitive episode in the ninety-sixth trial, already referred to, was probably a factor in determining the relatively frequent attempts to escape by alley No. 4. The significance of this table is contained in the following facts: 1. A habit of entering alley No. 3 first was formed under con- ditions which rendered such first choice invariably advantageous. 2. During this subject's 100 formal trials, when the opera- tion of the factor of recency was invariably disadvantageous, she manifested only 17 pure recency first choices and 6 recency- frequency ones. During the 20 habit forming trials under dis- 40 G. V. HAMILTON cussion she manifested only 3 recency first choices, but after these trials, during which the operation of the factor of recency was invariably advantageous she manifested 100% of recency first choices. 3. During the 100 formal trials this subject entered alley No. 3 less frequently than she entered either alley No. 1 Or No. 2. During her 20 habit forming trials both alley No. 2 and alley No. 4 were entered more frequently than was alley No. 3, hence the factor of frequency, as such, cannot be held to account for the formation of the habit. These findings suggest that frequency of an advantageous re- sponse is a more important determinant of habit formation than is frequency of a response which carries with it no invariable advantage. This point will be given further discussion in the presentation of the records that follow: White Rat 2 This subject displayed 11 pure recency and 7 recency-fre- quency reactions during his 100 formal trials. The record of his first choices is as follows: alley No. 1 was the first choice in 9 trials, alley No. 2 in 24 trials, alley No. 3 in 33 trials, alley No. 4 in 34 trials. During his 100 formal trials he entered alley No. 1 72 times, alley No. 2 124 times, alley No. 3 124 times and alley No. 4 74 times. Table 11 gives the record of the 14 trials that it was necessary to give him before he ceased to try other alleys than right alley No. 1 : Alleys Entered by White Rat 2 During Each of 14 Habit-Forming Trials TABLE 11 T 2 During Eaci 1st trial- -3-2-1 2nd " 2-3-1 3rd " 3-2-1 4th " 3-2-1 5th " 2-1 6th " 3-2-1 7th " 2-1 8th " 3-2-1 9th " 1 10th " 1 11th " 1 12th " 2-1 13th " 1 14th " 2-1 PERSEVERANCE REACTIONS IN PRIMATES AND RODENTS 41 The effects of previous frequency of efforts to escape by alleys No. 2 and 3 are reflected in this table; these alleys were evidently preferred during the first 8 habit forming trials, after which there were no further efforts to escape by alley No. 3 and. only 2 efforts to escape by alley No. 2. Since alley No. 1 had been entered no more frequently than had alley No. 2 at the end of the eighth habit forming trial it is likely that fre- quency of advantageous response was the essential determinant of the habit. One must not forget that in this case the subject displayed only 18% of combined pure recency and recency- frequency first choices during the 100 trials in which the factor of recency always operated disadvantageously, whilst he dis- played 100% recency first choices after he had been given 14 trials during which this factor invariably operated advantageously. White Rat 5 TABLE 12 Alleys Entered by White Rat 5 During 5 Habit-Forming Trials 1st trial— 3-3-4 2nd " 3-3-4 3rd " 3-4 4th " 3-4 5th " 3-4 This subject. had 17 recency and 11 recency-frequency first choices, which were distributed as follows: alley No. 1, 39 trials; alley No. 2, 27 trials; alley No. 3, 22 trials; alley No. 4, 12 trials. She entered alley No. 1 108 times, alley No. 2 157 times, alley No. 3 128 times and alley No. 4 51 times. It is quite clear that alley No. 4 was for her the one of least preference, both for first choice and for entrances subsequent to first choice. The rapidity with which White Rat 5 formed the habit of trying alley No. 4 first and only is striking, and affords an ex- ample in which the factor of frequency merely as frequency and not as frequency of advantageous response could have played no important part in determining the habit. During the five habit forming trials alley No. 3 was entered 7 times, whilst alley No. 4 was entered but 5 times. White Rat 6 This subject had 14 pure recency and 13 recency-frequency first choices, which were distributed as follows: alley No. 1, 12 42 G. V. HAMILTON trials; alley No. 2, 20 trials; alley No. 3, 43 trials; alley No. 4, 25 trials. During his 100 formal trials he entered alley No. 1 65 times, alley No. 2 99 times, alley No. 3 119 times and alley No. 4 86 times. TABLE 13 Alleys Entered by White Rat 6 During 14 Habit-Forming Trials 1st trial— 3-4-3-4-3-2-3-4-1 2nd 2-3-2-1 3rd 3-2-1 4th ' 3-2-1 5th ' 3-1 6th ' 3-2-1 7th ' 3-2-1 8th ' 3-2-1 9th ' 2-1 10th ' 2-1 11th ' 4-3-2-1 12th ' 3-1 13th 3-1 14th 3-1 In this table, as in tables 10, 11 and 12, we have an example of habit formation which evidently depended not only upon the operation of the factors of recency and frequency as such, but upon recency and frequency of advantage. White Rat 8 This subject had 25 pure recency and 7 recency-frequency first choices, which were distributed as follows: alley No. 1, 12 trials; alley No. 2, 16 trials; alley No. 3, 47 trials; alley No. 4, 25 trials. It is significant that this subject displayed a larger number of recency first choices and a smaller number of fre- quency first choices than did any subject of the six that were used for the studies of habit formation. He entered alley No. 1 90 times in 100 formal trials, alley No. 2 85 times, alley No. 3 98 times and alley No. 4 76 times. TABLE 14 Alleys Entered by White Rat 8 During 8 Habit-Forming Trials 1st trial- -2-1-3-2-4 2nd " 3-2-3-2-4 3rd " 4 4th " 4 5th " 3-1-3-4 6th " 4 7th " 3-4 8th " 3-4 PERSEVERANCE REACTIONS IN PRIMATES AND RODENTS 43 In this case the factor of frequency as such, regardless of the advantage or disadvantage resulting from its operation, could have played only a secondary role since alley No. 4 was entered but once more frequently during the habit forming trials than was alley No. 3 and since, as is shown in table 8, frequency as such was a possible determinant of his first choices in only 22% (15% pure frequency and 7% recency-frequency) of his 100 formal trials. White Rat 10 This subject was the only representative of his strain. He had 20 recency and 3 recency-frequency first choices. His first choices were distributed as follows: alley No. 1, 17 trials; alley No. 2, 43 trials; alley No. 3, 30 trials; alley No. 4, 10 trials. He entered alley No. 1 90 times, alley No. 2 105 times, alley No. 3 100 times and alley No. 4 89 times. Although this subject had 15 habit forming trials he did not once enter any other alley than right alley No. 4. This alley is the right one for the one-hundredth formal trial, hence the factor of recency (it is evident that the factor of frequency is of secondary importance in this case) may have accounted for his first choice of alley No. 4 when he was given his first habit forming trial. During the trials that followed both recency and increasing frequency of advantage seem to have determined the habit of trying only alley No. 4. VI CONCLUSIONS 1. When a mammalian is confronted by a series of situations for which he is unable to discover and stereotype a specifically adequate and invariably successful mode of response he tends to vary his response in a manner which is less a species than an individual characteristic. 2. The multiple choice method that was employed in these studies operates against habit formation and favors a varying of modes of searching for a place of exit throughout an entire series of 100 trials. Such variation of response was manifested by a baboon throughout 500 trials. The following types of response were elicited from children, infra-human primates and five different rodent species: 44 G. V. HAMILTON Type A. — Response to confinement by trying three trial-to- trial varying inferentially possible exit alleys, the one trial-to- trial varying inferentially impossible exit alley not being tried by the subject. Type B. — Response by trying all four exit alleys, but once each and in an irregular order. Type C. — Response by trying all four alleys but once each and in regular order from left to right or from right to left. Type D. — Response by entering a given alley more than once during a given trial, with an interval of effort to escape by another alley between entrance and reentrance of the same alley. Type E. — (a) Response includes two or more separate efforts to escape by the same alley or group of alleys during a given trial without intervals of effort to escape by other alleys, (b) Response includes persistent avoidance of a given alley dur- ing a given trial whilst six or more separate efforts to escape by the other alleys are manifested. 3. None of the subjects of these studies manifested a con- sistent appreciation of the inferentially impossible, trial- to-trial varying one impossible exit alley as the one alley that would not afford escape from confinement, hence Type A reactions were not exclusively manifested by any subject. The use of an inferior class of human subjects must be held to account for the fact that the maximum number of Type A reactions was manifested by a three-year old and a five-year old girl in a group of 20 children whose ages ranged from three to twelve years. Nevertheless, the children manifested a higher average number of these reactions than did any of the animal groups. The infra-human primates average of A-reactions was higher than that of any of the rodent groups. 4. If we exclude the case of the single mouse that was used in these studies, B -reactions were of more frequent occurrence among infra-human primates than among any of the other groups. The only human subject who manifested a consider- able number of B -reactions was feeble-minded twelve-year old Girl 20. 5. The girls manifested the highest average number of C- reactions. Their average in this respect is slightly lowered if PERSEVERANCE REACTIONS IN PRIMATES AND RODENTS 45 the feeble-minded cases are excluded. If the two atypical mon- keys are excluded the infra-primate group has the next highest average number of C-reactions. Among rodents (the mouse being excluded) the white-rat group has the highest average of these reactions. 6. The group averages of D-reactions (table 6) do not suf- ficiently differ from one another to be of much significance, since these averages were much affected by extreme individual cases. 7. In spite of the influence of atypical cases on the primate averages for E-reactions, table 7 shows that the tendency toward the manifestation of these reactions is much more marked among rodents than among primates. The infra-human primate aver- age number of E-reactions is conspicuously higher than the girl's average number, but conspicuously lower than any rodent group's average. 8. The following factors favor the manifestation of Type D and, particularly, of Type E reactions: (a) Inherent primitiveness of reactive equipment, such as is possessed by rodent species and by the young or mentally defec- tive of more highly developed species. (b) Excitability. (c) Distractibility. (d) Feeble responsiveness to a situation which elicits reaction. 9. The first direction that is given to a subject's activities under conditions which elicit escape responses is apt to be de- termined by such factors as the spacial relations of the various apparently possible avenues of escape, the recency with which each of these various avenues have been previously tried for escape and the frequency with which each has been previously tried for escape. 10. Recency and frequency, in the sense in which they are used by Watson (7) are apt to be in themselves factors in deter- mining the direction of a subject's activity, but when either of these factors repeatedly operates advantageously it is apt to act more strongly as a behavior determinant than where either no constant advantage or actual disadvantage attend its operation. 11. A comparison of the studies made under conditions which operate against habit formation with those which favor habit formation suggests that a response which has been frequently 46 G. V. HAMILTON manifested with invariable advantage is more apt to recur in the behavior of the individual than is a still more frequently manifested response which has not brought invariable advantage. 12. These studies suggest a possibility, which they by no means prove, that with descent of the phyletic scale the factor of recency increases in importance as a determinant of habit formation whilst that of frequency relatively decreases. APPENDIX I. Individual Records of Tw^enty Girls. II. Individual Record of White Rat 5. III. Analysis of White Rat 5's Record (calculations for table 8). PERSEVERANCE REACTIONS IN PRIMATES AND RODENTS 47 Girll Girl 2 Girl 3 Girl 4 Trial (3 years) (3 years) (4 years) (4 years) number Alleys entered Alleys entered Alleys entered Alleys entered 1 1 4-3-2-1 1 1 2 4-3-2 4-3-2 1-2 2 3 3 4-3 4-3 3 4 4 2-1-4 3-4 4 5 3-2 2 1-2 1-2 6 1-2-3 3 2-3 2-3 7 4 4 4 4 8 3-2-1 2-3-4-1 1 4-3-2-1 9 4-3 2-3 1-2-3 1-2-3 10 3-2-3-4 4 4 2-3-4 11 1 1 1 1 12 4-3-2 2 4-3-2 2 13 4 3-4 1-2-3-4 3-4 14 3-2-1 1 1 4-3-2-1 15 4-3-2 3.4.1.4.3.4.3-2 4-3-2 2 16 4-3 1-2-3 1-2-3 3 17 4-3-2-1 1 1 4-3-2-1 18 4-3-2 2 4-3-2 1-2 19 4-3 3 1-2-3 3 20 3-2-1-2-3-4 4 4 4 21 2 1-2 1-2 1-2 22 4-3 3 4-3 3 23 4 2-3-4 1-2-3-4 4 24 3-2-1 1 1 4-3-2-1 25 4-3 1-2-3 4-3 1-2-3 26 4 1-2-3-4 1-2-3-4 4 27 4-3-2-1 1 1 4-3-2-1 28 4-3-2 0 4-3-2 1-2 29 4 T-2-3-4 4 3-4 30 3-2-1 2-3-4-3-4-3-2-3-1 1 4-3-2-1 31 2 1-2 1-2 1-2 32 4-3 2-1-2-3 4-3 3 33 2-1 1 1 4-3-2-1 34 2 1-2 4-3-2 1-2 35 2-1-2-3 1-2-3 1-2-3 3 36 4 1-2-4 4 4 37 2 1-3-4-1-2 1-2 4-3-2 38 3 2-3 4-3 1-2-3 39 2-1-2-3-4 1-2-3-4 1-2-3-4 4 40 1 1 1 4-3-2-1 41 3 1-2-3 1-2-3 1-2-3 42 4 1-2-3-4 4 4 43 2-3-4-3-2-1 1 1 4-3-2-1 44 2 1-2 4-3-2 1-2 45 4 2-3-4 1-2-3-4 3-4 46 3-2-1 1 1 4-3-2-1 47 2 1-1-2 4-3-2 1-2 48 1-2-3 1-2-3 1-2-3 3 49 2-1 1 4-3-2-1 4-3-2-1 50 1-2 1-2 1-2 1-2 48 G. V. HAMILTON Girl 1 Girl 2 Girl 3 Girl 4 Trial (3 years) (3 years) (4 years) (4 years) number Alleys entered Alleys entered Alleys entered Alleys entered 51 2-3 1-2-3 1-2-3 1-2-3 52 4 1-2-3-4 1-2-3-4 4 53 3-2 1-2 1-2 4-3-2 54 4-3 1-2-3 4-3 1-2-3 55 1-2-3-4 1-2-3-4 1-2-3-4 4 56 1 4-3-2-1 4-3-2-1 4-3-2-1 57 1-2-1-2-3 1-2-3 1-2-3 1-2-3 58 2-3-4 1-2-3-4 4 4 59 2-3-4-3-2-1 1 1 4-3-2-1 60 1-2 2 4-3-2 1-2 61 2-1-2-1-2-3-4 1-2-3-4 1-2-3-4 1-2-3-3-4 62 3-4-3-2-3-4-3-2-1 1 4-3-2-1 4-3-2-1 63 3-2 7 4-3-2 1-2 64 2-3 i-2-3 1-2-3 3 65 2-3-4-3-2-1 3-4-3-2-1 4-3-2-1 4-3-2-1 66 4-3-2 1-2 1-2 1-2 67 2-1-2-3 1-2-3 4-3 3 68 2-1-2-3-4 1-2-3-4 1-2-3-4 4 69 3-2 1-2 4-3-2 4-3-2 70 4-3 1-2-3 1-2-3 1-2-3 71 1-2-3-4 1-2-3-4 4 4 72 2-3-2-3-4-3-4-3-2- 1 4-3-2-1 1 4-3-2-1 73 4-1-2-3 1-2-3 4-3 1-2-3 74 3-2-1-4 1-2-3-4 1-2-3-4 4 75 1 3-2-3-4-3-2-1 4-3-2-1 4-3-2-1 76 3-2 1-2 1-2 77 3-2-3-4 4 3-4 78 1 1 4-3-2-1 79 2 2 1-2 80 4-3 4-3 3 81 2-1 1 4-3-2-1 82 4-3-2 4-3-2 1-2 83 4-3 2-3 3 84 2-1-4 4 4 85 3-2 1-2 4-3-2 86 2-1-4-1-4-3 4-3 1-2-3 87 2-4 1-2-3-4 4 88 1 4-3-2-1 4-3-2-1 89 1-2-3 1-2-3 1-2-3 90 2-3-4 4 4 91 2-3-4-3-2-1 1 4-3-2-1 92 4-3-2 4-3-2 1-2 93 4 1-2-3-4 3-4 94 3-2-1 4-3-2-1 4-3-2-1 95 4-3-4-3-2 1-2 1-2 96 1-2-4-3 4-3 3 97 2-1-4 1-2-3-4 4 98 3-2-1 1 4-3-2-1 99 1-2 4-3-2 1-2 100 4-3 1-2-3 3 PERSEVERANCE REACTIONS IN PRIMATES AND RODENTS 49 Girl 5 Giri6 Girl 7 Girl 8 Trial (5 years) (5 years) (5 years) (o years) No. Alleys entered Alleys entered Alleys entered Alleys entered 1 1 1 1 1 2 2 2 3-4-1-2 2 3 4-3 3 1-2-3 3 4 1-2-3-4 4 2-3-4 4 5 1-2 2 1-2 3-2 6 1-2-3 2-3 1-2-3 3 7 1-2-3-4 4 1-2-3-4 4 8 1 2-3-4-2-1 1 3-2-1 9 1-2-3 1-2-3 1-2-3 2-3 10 1-2-3-4 4 2-3-4 4 11 1 4-3-2-1 1 1 12 2 2 1-2 2 13 4 3-2-4 1-2-3-4 3-4 14 2-4-3-2-1 3-2-1 1 3-2-3-4-3-2-3-4-3-2 15 3-4-3-2 2 1-2 2 [3-2-3-4-3-2-1 16 2-1-3 3 1-2-3 3 17 3-4-2-1 3-4-2-1 1 4-3-2-1 18 1-2 2 1-2 3-2 19 4-3 3 1-2-3 1-2-1-2-3 20 1-2-3-4 4 4 4 21 4-3-2 4-3-2 1-2 1-2 22 4-3 2-3 1-2-3 4-3 23 4 3-4 4 3-2-1-2-3-4 24 4-3-2-1 2-3-4-3-2-1 1 3-2-1 25 4-3 2-3 1-2-3 3 26 4 3-4 4 4 27 4-3-2-1 3-4-3-2-1 1 3-2-3-4-3-2-1 28 3-2 2 1-2 2 29 4 2-3-4 1-2-3-4 3-4 30 2-1 2-1 1 3-2-1 31 4-3-2 4-3-2 1-2 1-2 32 4-3 2-3 4-3 3 33 4-3-2-1 4-3-2-1 1 4-3-2-3-4-3-2-1 34 4-3-2 2 1-2 4-3-2 35 4-3 3 1-2-3 4-3 36 4 4 1-2-3-4 4 37 4-3-2 2 1-2 3-2 38 4-3 2-3 4-3 4-3 39 4 2-3-4 1-2-3-4 2-3-4 40 4-3-2-1 3-4-2-1 1 3-2-1 41 4-3 4-3 1-2-3 1-2-3 42 3-2-1-2-3-4 2-1-2-3-4 4 4 43 4-3-2-1^ 2-1 1 3-2-3-4-3-2-1 44 4-3-2 2 1-2 2 45 4 3-4 4 3-2-3-4 46 4-3-2-1 3-4-3-2-1 1 3-2-3-2-1 47 4-3-2 2 1-2 2 48 4-3 3 2-3 3 49 4-3-2-1 4-3-2-1 4-3-2-1 4-3-2-3-1-3-2-1 50 4-3-2 3-2 1-2 2 50 G. V. HAMILTON Girl 5 Girl 6 Girl 7 Girl 8 Trial (5 years) ( 5 years) (5 years) (5 years) No. Alleys entered Alleys entered Alleys entered Alleys entered 51 4-3 4-3 1-2-3 1-2-3 52 1-2-3-4 2-1-2-3-4 4 4 53 4-1-2 4-3-2 1-2 3-2 54 4-3 2-3 4-3 3 55 4 2-3-4 1-2-3-4 4 56 3-2-1 2-3-4-2-1 4-3-2-1 3-2-3-4-3-4-3-2-3 57 4-3 3 1-2-3 3 [4-3-2-1 58 2-1-2-3-4 4 4 4 59 4-3-2-1 2-3-2-1 1 3-2-1 60 4-3-2 3-2 4-3-2 3.4.3.4.3-2 61 4 4 1-2-3-4 1-2-3-4 .62 3-2-1 3-2-1 4-3-2-1 3-2-1 63 4-3-2 3-4-2 1-2 2 64 4-3 3 4-3 3 65 2-1 4-3-2-1 4-3-2-1 4-3-2-3-4-3-2-1 66 4-3-2 3-2 1-2 2 67 4-3 4-2-1-3 4-3 3 68 4 2-1-3-4 4 4 69 3-2 2 4-3-2 3-2 70 2-1-2-3 3 1-2-3 3 71 1-2-3-4 4 4 4 72 4-3-2-1 3-2-1 4-3-2-1 3-2-1 73 4-3 2-3 1-2-3 2-3 74 2-1-2-3-2-3-4 3-4 4 4 75 4-3-2-1 2-3-4-3-2-1 1 3-2-1 76 4-3-2 1-2 1-2 1-2 77 4 3-4 4 3-4 78 4-3-2-1 4-3-2-1 4-3-2-1 3-2-1 79 4-3-2 1-2 1-2 2 80 4-3 3 4-3 3 81 4-3-2-1 4-3-2-1 4-3-2-1 4-3-2-3-4-3-2-3-4 82 4-3-2 2 1-2 2 [3-2-3-4-1 83 4-3 3 1-2-3 4-3 84 2-1-2-3-2-1-2-3-2 4 4 2-1-2-3-4 85 4-3-2 [1-2-3-4 1-2 4-3-2 2 86 1-2-3 2-3 4-3 3 87 4 2-3-4 1-2-3-4 4 88 4-3-2-1 1 4-3-2-1 3-2-1 89 4-3 2-3 1-2-3 1-2-3 90 2-1-2-3-4 4 4 4 91 4-3-2-1 4-3-2-1 3-2-1 92 4-3-2 1-2 2 93 1-2-3-4 4 3-4 94 4-3-2-1 4-3-2-1 3-2-1 95 4-3-2 ■ 1-2 2 96 1-2-3 4-3 3 97 4 4 4 98 3-2-1 4-3-2-1 3-2-1 99 4-3-2 1-2 2 100 4-3 4-3 3 PERSEVERANCE REACTIONS IN PRIMATES AND RODENTS 51 Girl 9 Girl 9 Trial (5 years) Trial (5 years) No. Alleys entered No. Alleys entered 1 1 51 1-2-3 2 2 52 2-3-4 3 2-3 53 2 4 2-3-4 54 2-3 5 2 55 2-3-4 6 2-3 56 2-3-4-3-4-3-2-1 7 2-3-4 57 2-3 8 2-3-4-2-3-4-2-3-4-2-3-4-1 58 2-3-4 9 1-2-3 59 1 10 1-2-3-2-1-2-3-1-2-3-2-3-2-3- -2-3-1-2 60 2 [3-1-3-2-3-2-3-2-3-2-3-2-3-4 11 1 61 1-2-3-4 12 2 62 2-3-4-1 13 1-2-3-4 63 9 14 1 64 1-2-3 15 2 65 2-3-4-2-3-4-2-3-4-2-3-4-3-2- 1 16 2-3 66 1-2 17 2-3-4-1 67 1-2-3 18 2 68 1-2-3-4 19 2-3 69 1-2 20 2-3-4 70 2-3 21 1-2 71 2-3-4 22 2-3 72 2-3-4-3-2-3-4-2-3-4-3-2-4-3-2-3-4-1 23 4 73 2-3 24 2-3-4-3-2-3-4-1 74 1-2-3-4 25 2-3 75 1 26 4 76 1-2 27 3-2-3-4-3-4-1 77 2-3-4 28 4-3-2 78 2-3-4-3-4-3-2-1 29 3-2-1-2-3-4 79 1-2 30 2-3-4-3-2-3-2-3-4-3-2-3-4-3- -2-3-4-3-2- •1 80 2-3 31 1-2 81 3-2-1 . 32 3 82 2 33 2-3-4-1 83 2-3 34 2 84 2-3-4 35 3 85 2 36 2-3-4 86 2-3 37 1-2 87 3-2-4 38 2-3 88 1 39 2-3-4 89 2-1-2-3 40 2-3-4-1 90 2-3-4 41 1-2-3 91 1 42 2-3-2-3-4 92 2 43 1 93 2-1-2-3-4 44 2 94 2-3-4-3-2-1 45 2-3-4 95 2 46 2-3-4-3-2-1 96 2-3 47 2 97 2-3-2-1-2-3-4 48 2-3 P 98 1 49 2-3-4-3-2-3-4-3-2-1 99 2 50 2 100 2-3 52 G. V. HAMILTON Girl 10 Girl 11 Girl 12 Girl 13 Trial (6 years) (7 years) (7 years) (7 years) number Alleys entered Alleys entered Alleys entered Alleys entered 1 1 1 1 1 2 1-2 1-2 3-4-2 2 3 1-2-3 1-2-3 2-3 3 4 1-2-3-4 1-2-3-4 3-4 4 5 1-2 1-2 3-4-3-2 1-2 6 1-2-3 4-3 1-2-3 3 7 1-2-3-4 1-2-3-4 3-4 4 8 1 2-3-4-2-1 3-4-3-2-1 1 9 2-3 3 2-3 3 10 1-2-3-4 2-3-4 2-3-4 4 11 1 1 1 1 12 1-2 1-2 3-4-3-2 2 13 1-2-3-4 3-4 3-4 3-4 14 1 2-3-4-2-1 2-1 1 15 1-2 3-4-3-2 2 2 16 1-2-3 2-1-3 2-3 3 17 1 2-3-4-2-1 2-3-4-3-2-1 4-3-2-1 18 1-2 3-4-2 2 1-2 19 1-2-3 2-1-2-3 3 2-3 20 1-2-3-4 4 3-2-1-2-3-4 3-4 21 1-2 1-2 1-2 1-2 22 1-2-3 4-3 3 2-3 23 1-2-3-4 2-1-3-4 3-4 3-4 24 1 3-4-2-1 3-4-3-2-1 1 25 1-2-3 3 3 2-3 26 1-2-3-4 2-1-2-3-4 2-1-2-3-4 4 27 1 2-1 2-1 1 28 1-2 3-4-2 2 2 29 1-2-3-4 3-4 3-2-1-2-3-4 4 30 1 3-2-1 1 1 31 1-2 1-2 1-2 1-2 32 1-2-3 3 3 2-3 33 1 2-1 3-4-3-2-1 1 34 1-2 2 3-2 2 35 1-2-3 1-2-3 3 3 36 1-2-3-4 4 3-4 4 37 1-2 2 3-2 1-2 38 1-2-3 1-2-3 3 2-3 39 1-2-3-4 2-3-4 2-1-2-3-4 4 40 1 2-3-4-2-1 2-1 1 41 1-2-3 1-2-3 2-3 1-2-3 42 1-2-3-4 4 3-4 2-3-4 43 1 2-1 3-2-1 4-1 44 1-2 2 3-2 2 45 1-2-3-4 4 2-1-2-3-4 2-3-4 46 1 2-1 2-3-4-3-2-1 4-3-2-1 47 1-2 4-3-2 3-4-3-2 2 48 2-3 3 3 2-3 49 1 2-1 2-1 3-4-3-2-1 50 1-2 3-4-2 1-2 2 PERSEVERANCE REACTIONS IN PRIMATES AND RODENTS 53 Girl 10 Girl 11 Giri 12 Girl 13 Trial (6 years) (7 years) (7 years) (7 years) lumber Alleys entered Alleys entered Alleys entered Alleys entered 51 1-2-3 1-2-3 3 1-2-3 52 1-2-3-4 2-1-3-4 3-2-1-2-3-4 2-3-4 53 1-2 3-2 3-2 1-2 54 1-2-3 1-3 2-1-2-3 2-3 oo 1-2-3-4 2-1-4 2-1-2-3-4 4 56 1 1 2-1 4-3-2-1 o7 1-2-3 2-1-2-3 3 4-3 58 1-2-3-4 3-2-1-4 2-1-2-3-4 3-2-1-2-3-4 59 1 3-2-1 3-2-1 4-3-2-1 60 1-2 2 2 1-2 61 1-2-3-4 1-2-3-4 1-2-3-4 1-2-3-4 62 1 2-1 3-2-1 3-2-1 63 2 3-2 2 2 64 1-2-3 2-1-3 1-2-3 3 65 1 1 2-1 4-3-2-1 66 o 2 3-2 2 67 1-2-3 2-1-2-3 1-2-3 3 68 1-2-3-4 4 3-4 4 69 1-2 2 3-2 3-2 70 1-2-3 2-1-3 3 2-1-2-3 71 1-2-3-4 4 1-2-3-4 1-2-3-4 72 1 3-2-1 3-2-1 3-2-1 73 2-3 4-3 2-3 2-3 74 1-2-3-4 2-1-2-3-4 1-2-3-4 3-4 7o 1 2-1 3-2-1 3-2-1 76 1-2 3-2 3-2 9 77 1-2-3-4 3-2-1-2-3-4 2-1-2-3-4 3-4 78 2-3-4-3-2-1 3-2-1 3-2-1 3-2-1 79 9 4-3-2 4-3-2 3-2 80 2-3 4-3 2-1-2-1-2-3 3 81 1 1 1 1 82 2 2 2 2 83 T-2-3 3 3 3 84 2-3-4 4 2-1-2-3-4 4 85 1-2 3-2 4-3-2 3-2 86 2-3 3 2-1-2-3 3 87 2-3-4 2-1-2-3-4 4 4 88 2-3-4-3-2-1 3-2-1 3-2-1 3-2-1 89 1-2-3 4-3 3 3 90 2-3-4 3-2-1-2-3-4 2-1-2-3-4 2-1-2-3-4 91 1 1 1 1 92 2 2 2 2 93 1-2-3-4 3-4 3-4 4 94 2-3-4-3-2-1 3-2-1 3-2-1 3-2-1 95 2 3-2 4-3-2 2 96 2-3 3 1-2-3 3 97 2-3-4 4 4 4 98 3-4-3-2-1 3-2-1 3-2-1 3-2-1 99 2 3-2 4-3-2 2 100 2-3 3 4-3 3 54 G. V. HAMILTON Girl 14 Girl 15 Girl 16 Girl 17 Trial (7 years) (7 years) (9 years) (9 years) number Alleys entered Alleys entered Alleys entered Alleys entered 1 2-3-4-3-1 1 1 1 2 2 2 1-2 2 3 3 3 2-3 3 4 3-4 4 1-2-3-4 4 5 1-2 1-2 1-2 2 6 3 3 2-3 4-3 7 4 4 1-2-3-4 2-1-2-3-4 8 1 1 1 2-1 9 4-3 2-3 1-2-3 2-3 10 4 4 , 1-2-3-4 1-2-3-4 11 2-3-4-1 1 1 1 12 3-2 2 2 2 13 3-4 3-4 3-4 3-4 14 1 1 1 2-1 15 3-2 2 2 3-4-2 16 3 3 3 1-2-3 17 4-3-2-1 4-3-2-1 4-3-2-1 3-4-3-2-1 18 4-3-2 4-3-2 1-2 9 19 1-2-3 1-2-3 3 2-3 20 4 1-2-3-4 4 2-1-2-3-4 21 2 1-2 1-2 1-2 22 3 3 2-3 23 4 4 4 4 24 1 4-3-2-1 1 2-1 25 2-3 4-3 2-3 3 26 4 2-1-2-3-4 4 4 27 1 2-1 4-3-2-1 1 28 3-2 4-3-2 1-2 2 29 4 1-2-3-4 2-3-4 3-4 30 4-3-2-1 2-3-4-3-2-1 4-3-2-1 3-2-1 3] 1-2 1-2 1-2 1-2 32 2-3 3 2-3 4-3 33 4-3-2-1 4-3-2-1 4-3-2-1 2-1 34 4-3-2 4-3-2 4-3-2 4-3-2 35 1-2-3 4-3 3 4-3 36 4 1-2-3-4 2-1-2-3-4 4 37 4-3-2 4-3-2 4-3-2 1-2 38 1-2-3 1-2-3 1-2-3 2-1-3 39 4 4 2-1-2-3-4 2-1-2-3-4 40 4-3-2-1 3-2-1 3-2-1 2-3-4-2-3-1 41 4-3 1-2-3 1-2-3 1-2-3 42 1-2-3-4 3-2-1-2-3-4 4 3-4 43 1 4-3-2-1 1 3-2-1 44 2 4-3-2 2 2 45 4 1-2-3-4 3-4 2-1-2-3-4 46 ] 4-3-2-1 3-2-1 3-2-1 47 1-2 3-2 3-2 1-2 48 4-3 3 2-3 4-3 49 1 2-3-2-1 4-3-2-1 4-3-2-1 50 4-3-2 3-2 4-3-2 3-2 PERSEVERANCE REACTIONS IN PRIMATES AND RODENTS 55 Girl 14 Gid 15 Girl 16 Girl 17 Trial (7 years) (7 years) (9 years) (9 years) number ■ Alleys entered Alleys entered Alleys entered Alleys entered 51 1-2-3 2-3 1-2-3 1-2-3 52 4 3-2-1-2-3-4 3-4 4 53 2 3-2 4-3-2 3-2 54 1-2-3 4-3 2-1-2-3 1-2-3 55 4 2-1-2-3-4 3-4 3-2-1-2-3-4 56 1 2-3-4-3-2-1 4-3-2-1 1 57 2-3 1-2-3 4-3 3 58 1-2-3-4 3-4 3-2-1-2-3-4 3-2-1-2-3-4 59 2-3-4-3-2-1 2-3-4-3-2-1 3-2-1 3-2-1 60 3-2 3-2-1 2 3-2 61 1-2-3-4 4 1-2-3-4 1-2-3-4 62 3-2-1 3-2-1 4-3-2-1 2-1 63 1-2 4-3-2 2 1-2 64 4-3 1-2-3 3 2-3 65 1 4-3-2-1 4-3-2-1 3-2-1 66 4-3-2 4-3-2 3-2 2 67 4-3 2-1-2-3 1-2-3 4-3 68 2-1-2-3-4 4 4 4 69 1-2 3-2 3-2 3-2 70 4-3 1-2-3 1-2-3 3 71 1-2-3-4 4 1-2-3-4 1-2-3-4 72 4-3-2-1 3-2-1 3-2-1 3-2-1 73 4-3 4-3 2-3 3 74 1-2-3-4 2-1-2-3-4 4 3-2-1-2-3-4 75 4-3-2-1 4-3-2-1 3-2-1 4-3-2-1 76 4-3-2 1-2 4-3-2 3-2 77 1-2-3-4 3-4 4 3-2-1-2-3-4 78 4-3-2-1 1 3-2-1 3-2-1 79 4-3-2 2 4-3-2 3-2 80 1-2-3 3 1-2-3 3 81 1 4-1 1 1 82 4-3-2 2 2 3-2 83 1-2-3 1-2-3 3 3 84 4 4 4 2-1-2-1-2-3-4 85 1-2 1-2 3-2 3-4-2 86 4-3 3 1-2-3 3 87 2-1-2-3-4 4 4 2-1-3-4 88 4-3-2-1 1 3-2-1 3-2-1 89 4-3 3 2-3 3 90 2-1-2-3-4 4 3-4 3-2-3-4 91 1 1 1 1 92 4-3-2 2 2 3-4-2 93 2-1-2-3-4 3-4 1-2-3-4 2-1-3-2-4 94 3-2-1 1 3-2-1 3-2-1 95 4-3-2 2 2 3-2 96 2-1-2-3 3 3 2-1-2-3 97 4 4 4 3-2-1-2-3-4 98 1 1 3-2-1 3-2-1 99 4-3-2 2 2 3-2 100 2-1-2-3 3 3 4-3 56 G. V. HAMILTON Girl 18 Girl 19 Girl 20 Trial (9 years) ( 9 years) (12 years) number Alleys entered Alleys entered Alleys entered 1 1 1 1 2 9 2 9 3 3 3 3 4 4 4 4 5 1-2 4-2 2 6 3 2-1-2-3 3 7 4 4 3-4 8 1 3-2-1 2-3-4-2-1 9 2-3 4-3 3 10 4 2-1-2-3-4 3-4 11 1 1 1 12 2 2 2 13 3-4 3-4 3-4 14 4-3-2-1 4-3-2-1 2-3-4-2-1 15 3-4-3-2 1-2 3-4-2 16 3 3 4-2-1-3 17 4-3-2-1 4-3-2-1 4-2-1 18 3-2 2 2 19 '3 4-3 4-2-1-3 20 3-2-1-2-3-4 4 3-4 21 1-2 1-2 1-2 22 3 3 3 23 4 4 4 24 2-3-4-3-2-1 4-3-2-1 3-1 25 2-3 2-3 3 26 1-2-3-4 4 4 27 2-3-4-3-2-1 3-2-1 2-1 28 2 2 2 29 2-3-4 3-4 3-4 30 1 3-2-1 2-1 31 1-2 1-2 1-2 32 3 3 3 33 4-3-2-1 4-3-2-1 4-3-2-1 34 2 1-2 3-4-2 35 3 3 3 36 3-2-1-2-3-4 4 4 37 2 3-2 2 38 1-2-3 1-2-3 1-2-3 39 4 4 4 40 2-1 3-2-1 2-1 41 1-2-3 1-2-3 1-2-3 42 3-4 3-4 3-4 43 1 3-2-1 3-2-1 44 2 2 2 45 3-4 3-4 3-4 46 1 3-2-1 1 47 3-2 2 3-4-2 48 4-3 3 2-3 49 2-1 4-3-2-1 4-3-2-1 50 2 2 3-4-2 PERSEVERANCE REACTIONS IN PRIMATES AND RODENTS 57 Girl 18 Girl 20 (9 years) (12 years) Trial number Alleys entered Alleys entered 51 1-2-3 1-2-3 52 3-4 3-4 53 1-2 2 54 3 1-2-4-3 55 4 2-1-3-4 56 1 2-1 57 3 3 58 1-2-3-4 2-1-3-4 59 2-3-4-3-2-1 2-1 60 3-2 2 61 1-2-3-4 1-2-3-4 62 1 3-2-1 63 3-4-3-2 2 64 3 3 65 1 1 66 2 2 67 3 3 68 4 2-4 69 1-2 4-2 70 3 1-3 71 1-2-3-4 1-2-3-4 72 2-3-4-3-2-1 3-1 73 3 2-1-3 74 1-2-3-4 4 75 2-3-4-3-2-1 2-1 76 4-3-2 3-4-2 77 2-1-2-3-4 2-1-3-4 78 2-1 3-2-1 79 2 2 80 3 3 81 1 1 82 2 2 83 3 3 84 3-4 4 85 4-3-2 2 86 3 2-1-3 87 3-4 3-1-3-4 88 2-1 3-2-1 89 3 2-1-3 90 2-3-4 2-4 91 1 1 92 2 3-2 93 3-4 2-3-4 94 3-4-3-2-1 2-1 95 3-2 3-2 96 3 1-3 97 4 3-2-1-3-4 98 2-1 3-2-1 99 2 3-1-2 100 4-3 4-3 58 G. V. HAMILTON White Rat 5 White Rat 5 Trial (Adult) Trial (Adult) No. Alleys entered No. Alleys entered 1 4-4-3-4-3-2-1 51 1-2-3 2 4-3-2 52 1-2-3-4 3 1-1-2-4-3 53 1-2 4 2-1-1-3-4 54 1-2-1-2-1-2-3 5 4-3-2 55 1-2-3-2-3-2-1-2-3-4 6 3 56 2-3-4-4-3-2-1 7 4 57 1-2-3 8 4-3-2-1 58 3-4 9 3 59 1 10 4 60 1-2 11 3-2-1 61 3-2-3-1-2-1-2-3-1-2-3-4 12 3-2 62 2-3-2-3-4-3-2-3-4-3-2-1 13 4 63 1-2 14 3-2-1 64 2-1-2-3 15 3-2 65 3-2-1 16 4-3 66 1-2 17 3-2-1 67 3 18 3-2 68 1-2-3-2-1-2-3-2-1-2-3-4 19 4-3 69 1-2 20 3-2-1-2-1-2-1-3-2-1-3-2 1-2-1-2-1-3-4 70 1-2-3 21 4-3-2 71 2-3-2-3-2-3-2-1-2-3-2-3-2-3-2-3-2-3-2-1-3 1-2-1-2-3-4 22 2-1-3 72 2-3-4-3-2-3-2-3-4-3-4-3-2-3-2-3-1 23 1-2-3-4 . 73 3 24 2-3-2-1 74 1-2-3-4 25 1-2-1-2-1-2-1-2-3 75 2-3-4-3-4-3-2-3-2-1 26 3-2-1-2-1-2-3-4 76 2 27 1 77 2-1-2-3-2-3-4 28 4-2 78 3-2-4-3-2-1 29 1-3-2-1-4 79 2 30 2-1 80 2-1-2-1-2-1-2-3 31 1-2 81 2-3-2-3-2-3-2-3-4-3-2-3-4-3-4-3-2-1 32 2-3 82 3-4-3-2 33 2-1 83 1-2-3 34 2 84 1-2-3-4 35 1-2-3 85 4-3-1-2 36 1-2-3-4 86 1-2-3 37 1-2 87 3-1-2-3-1-4 .38 2-3 88 2-3-2-1 39 1-2-3-2-1-2-3-4 89 3 40 1 90 2-2-1-2-3-4 41 1-2-3 91 1 42 1-2-3-4 92 2 43 1 93 2-3-2-3-2-1-2-1-2-3-1-2-3-1-2-2-3-2-1-3-4 44 1-2 94 3-2-2-3-1 45 1-2-3-4 95 2 46 1 96 2-3 47 1-2 97 2-3-2-1 48 1-2-3 98 3-1-2 49 1 99 3 50 1-2 100 2-1-3-1-2-1-3-2-1-2-1-2-1-3-4 PERSEVERANCE REACTIONS IN PRIMATES AND RODENTS 59 Analysis of White Rat 5's Record Trial number 1 Alleys 2 3 4 Alley first entered Right alley Relation of first choice to previous trials 1 1 1 4 1 5 2 5 3 4 1 Frequency 2 2 0 5 1 7 1 8 1 4 2 Frequency 3 2 2 6 1 8 1 9 1 1 3 4 4 2 7 1 9 1 10 1 2 4 5 6 0 8 1 10 1 11 1 4 2 Recency-Frequency 6 6 0 9 0 11 1 12 • 0 3 3 7 6 0 9 0 12 0 12 1 4 4 Frequency 8 6 1 9 1 12 1 13 1 4 1 Recency-Frequency 9 7 0 10 0 13 1 14 0 3 3 10 7 0 10 0 14 0 14 1 4 4 Frequency 11 7 1 10 1 14 1 15 0 3. 1 12 8 0 11 1 15 1 15 0 3 2 Frequency 13 8 0 12 0 16 0 15 1 4 4 14 8 1 12 1 16 1 16 0 3 1 Frequency 15 9 0 13 1 17 1 16 0 3 2 Frequency 16 9 0 14 0 18 1 16 1 4 3 17 9 1 14 1 19 1 17 0 3 1 Recency-Frequency 18 10 0 15 1 20 1 17 0 3 2 Frequency 19 10 0 16 0 21 1 17 1 4 3 60 G. V. HAMILTON Analysis . OF W HiTE Rat 5's Record- -Continued Trial number 20 1 10 7 Alleys 2 3 16 22 7 4 4 18 1 Alley first entered 3 Right alley 4 Relation of first choice to previous trials Recency-Frequency 21 17 0 23 1 26 1 19 1 4 2 Recency 22 17 1 24 1 27 1 20 0 2 3 Recency 23 18 1 25 1 28 1 20 1 1 4 24 19 1 26 2 29 1 21 0 2 1 25 20 4 28 4 30 1 21 0 1 3 Recency 26 24 2 32 3 31 2 21 1 3 4 Recency 27 26 1 35 0 33 0 22 "o 1 1 28 27 "o 35 1 33 0 22 4 2 29 27 2 36 1 33 1 23 1 1 4 30 29 1 37 1 34 0 24 0 2 1 Frequency 31 30 1 38 1 34 0 24 0 1 2 Recency 32 31 0 39 1 34 1 24 0 2 3 Recency-Frequency 33 31 1 40 1 35 0 24 0 2 1 Frequency 34 32 0 41 1 35 0 24 0 2 2 Frequency 35 32 1 42 1 35 1 24 0 1 3 36 33 1 43 1 36 1 24 1 1 4 37 34 1 44 1 37 0 25 0 1 2 38 35 0 45 1 37 1 25 0 2 3 Recency-Frequency PERSEVERANCE REACTIONS IN PRIMATES AND RODENTS 61 Analysis of White Rat 5's Record — Continued Trial number 1 Alleys 2 3 4 Alley first entered Right alley Relation of first choice to previous trials 39 35 2 46 3 38 2 25 1 1 4 40 37 1 49 0 40 0 26 0 1 1 41 38 1 49 1 40 1 26 0 1 3 Recency 42 39 1 50 1 41 1 26 1 1 4 43 40 1 51 0 42 0 27 0 1 1 44 41 1 51 1 42 0 27 0 1 2 Recency 45 42 1 52 1 42 1 27 1 1 4 46 43 1 53 0 43 •0 28 0 1 1 47 44 1 53 1 43 0 28 0 1 2 Recency 48 45 1 54 1 43 1 28 0 1 3 49 46 1 55 0 44 0 28 0 1 1 50 47 1 55 1 44 0 28 0 1 2 Recency 51 48 1 56 1 44 1 28 0 1 3 52 49 1 57 1 45 1 28 1 1 4 53 50 1 58 1 46 0 29 0 1 2 54 51 3 59 3 46 1 29 0 1 3 55 54 2 62 4 47 3 29 1 1 4 56 56 1 66 2 50 2 30 2 2 1 Frequency 57 57 1 68 1 52 1 32 0 1 3 Recency 62 G. V. HAMILTON Analysis OF White Rat 5's Record- -Continued Trial lumber 58 1 58 0 Alleys 2 3 69 53 0 1 4 32 1 Alley first entered 3 Right alley 4 Relation of first choice to previous trials Recency 59 58 1 69 0 54 0 33 0 1 1 60 59 1 69 1 54 0 33 0 1 2 Recency 61 60 3 70 4 54 4 33 1 3 4 62 63 1 74 4 58 5 34 2 2 1 Frequency 63 64 1 78 1 63 0 36 0 1 2 Recency 64 65 1 79 2 63 1 36 0 2 3 Recency-Frequenc 65 66 1 81 1 64 1 36 0 3 1 Recency 66 67 1 82 1 65 0 36 0 1 2 Recency 67 68 0 83 0 65 ' 1 36 0 3 3 68 68 3 83 5 66 3 36 1 1 4 69 71 1 88 1 69 0 37 0 1 2 70 72 1 89 1 69 1 37 0 1 3 71 73 4 90 12 70 10 37 1 2 4 Frequency 72 77 1 102 5 80 8 38 3 2 1 Frequency 73 78 0 107 0 88 1 41 0 3 3 74 78 1 107 1 89 1 41 1 1 4 75 79 1 108 3 90 4 42 2 2 1 Frequency 76 80 0 111 1 94 0 44 0 2 2 Frequency PERSEVERANCE REACTIONS IN PRIMATES AND RODENTS 63 Analysis of White Rat 5's Record — Continued Trial number 1 Alleys 2 3 4 Alley first entered Right alley Relation of first choice to previous trials 77 80 1 112 3 94 2 44 1 2 4 Recency-Frequency 78 81 1 115 2 96 2 45 1 3 1 79 82 0 117 1 98 0 46 0 2 2 Frequency 80 82 3 118 4 98. 1 46 0 2 3 Recency- Frequency 81 85 1 122 6 99 8 46 3 2 1 Frequency 82 86 0 128 1 107 2 49 1 3 2 83 86 1 129 1 109 1 50 0 1 3 84 87 1 130 1 110 1 50 1 1 4 ' 85 88 1 131 1 111 1 51 1 4 2 Recency 86 89 1 132 1 112 1 52 0 1 3 87 90 2 133 1 113 2 52 1 3 4 Recency 88 92 1 134 2 115 1 53 0 2 1 Frequency 89 93 0 136 0 116 1 53 0 3 3 90 93 1 136 3 117 1 53 1 2 4 Frequency 91 94 1 139 0 118 0 54 0 1 1 92 95 0 139 1 118 0 54 0 2 2 Frequency 93 95 5 140 9 118 6 54 1 2 4 Recency-Frequency 94 100 1 149 2 124 2 55 0 3 1 95 101 0 151 1 126 0 55 0 2 2 Frequency 34 G. V. HAMILTON Analysis OF White Rat 5's Record- -Continued- Trial lumber 96 1 101 0 Alleys 2 3 152 126 1 1 4 55 0 Alley first entered 9 Right alley 3 Relation of first choice to previous trials Recency-Frequency 97 101 1 153 2 127 1 55 0 2 1 Frequency 98 102 1 155 1 128 1 55 0 3 9 99 103 0 156 0 129 1 55 0 3 3 100 103 6 156 5 130 3 55 1 9 4 Frequency Total 109 161 133 56 The above analysis requires the following explanation: The upper one of the two horizontal rows of figures that are opposite each trial number indicates the number of times that each of the four alleys was entered dttring all preceding trials. For example, opposite trial No. 1 are the figures 1, 4, 5 and 5. They are meant to show that White Rat 5 had entered alley No. 1 once, alley No. 2 four times, alley No. 3 five times and alley No. 4 five times when she began her first formal trial, i.e., she had made these alley entrances during her preliminary trials. During her first formal trial she entered alley No. 1 once, alley No. 2 once, alley No. 3 twice and alley No. 4 three times. These figures appear in the second horizontal row opposite trial No. 1, and are appropriately distributed under the figures 1, 2, 3 and 4, which have the legend, "Alleys" for their heading. The two sets of figures opposite trial No. 1 and under the head- ing "Alleys" are added to give the uppermost horizontal row of figures opposite trial No. 2, and these sums show the number of times that each of the four alleys had been entered in all trials previous to formal trial No. 2. The figures under the heading, "Alley first entered" indicate the subject's first choice of alley to enter during each of the 100 formal trials. The figures under the heading, " Right Alley " indicate the alley that actually afforded escape during each trial. The heading, " Relation of first choice to previous PERSEVERANCE REACTIONS IN PRIMATES AND RODENTS 65 trials " is self-explanatory- to readers who have followed fore- going discussions of the factors of recency and frequency in these studies. REFERENCES 1. Hamilton, G. \'. A study of trial and error reactions in mammals. Journal of 1911. Animal Behavior, 1, 33-66. 2. Holt, E. B. The Freudian Wish. vii. New York. 1915. 3. Yerkes, R. M. The study of human behavior. Contributions from the (Bos- 1913. ton) Psychopathic Hospital, 9-10. 4. Yerkes, R. M. The mental life of monkeys and apes, etc. Behavior Mono- 1916. graphs, 3, 1. 5. Watson, J. B. Behavior. 62. New York. 1914. 6. Watson, J. B. Same, 257. 7. Watson, J. B. Same, 262. The Behavior Monographs Edited by JOHN B. WATSON The Johns Hopkins University, Baltimore, Md. VOLUME 1 No. 1 The development of certain instincts and habits in chicks. By Fi'ederick S. Breed. Pp. iv 4- 78, .$1.00, postpaid. No. 2 Methods of studying vision in animals. By Robert M. Yerkes and John B. Watson. Pp. iv + 90, |l.25, postpaid. No. 3 An experimental study on the death-feigning of Belostoma (-Zaitha Aucct.) flumineum Say and Nepa apiculata Uhler. By Henry H. P. Severin and Harry C. Severin. Pp. iii + 47, $.65, postpaid. No. 4 The biology of Physa. By Jean Dawson. Pp. iii + 120, $1.50, postpaid. No. 5 The function of the vibrissae in the behavior of the white rat. By Stella Burnham Vincent. Pp. iv + 81, Si. 15, postpaid. VOLUME 2 No. 6 The delayed reaction in animals and children. By Walter S. Hunter. Pp. v + 86, $1.15, postpaid. No. 7 The Canada porcupine : a study of the learning process. By Leroy Walter Sackett. Pp. iii + 84, $1.15, postpaid. No. 8 Audition and habit formation in the dog. By Harry Miles John- son. Pp. iv + 78, Si. 00, postpaid. No. 9 Habit formation in a sti'ain of albino rats of less than normal brain weight. By Gardner Cheney Basset. Pp. iv + 46, $.65, postpaid. No. 10 Distribution of effort in learning in the white rat. By John Linck Ulrich. Pp. iii + 51, $.65, postpaid. No. 1 1 The effect of age on habit fonnation in the albino rat. By Helen B. Hubbert. Pp. v + 55, $.65, postpaid. VOLUME 3 No. 12 The mental life of monkeys and apes: a study of ideational be- havior. By Robert M. Yerkes. Pp. iv + 145, $1.50, postpaid. No. 13 A study of perseverance reaction in primates and rodents. By G. V. Hamilton, iv + 65. $.75. Regular subscription $3 per volume, postpaid. THE BEHAVIOR MONOGRAPHS Emerson Hall, Cambridge, Mass Behavior Monographs Volume 3. Number 3. 1917 Serial Number 14 Edited by JOHN B. WATSON The Johns Hopkins University An Analysis of the Learning Process in the Snail, Physa gyrina Say BY ELIZABETH LOCKWOOD THOMPSON Published at Cambridge, Boston, Mass. HENRY HOLT & COMPANY 34 West 33d Street, New York G. E. STECHERT & CO., London, Paris and Leipzig, Foreign Agents The Journal of Animal Behavior An organ for the publication of studies concerning the instincts, habits and intelligence of organisms The Journal contains a Department of Notes in which appear brief accounts of especially interesting and valuable observations of behavior. Published bi-monthly at Cambridge, Boston, Mass., by Henry Holt and Company, New York. Each volume contains at least 450 pages with plates and text-figures. The subscription price is $5.00 per volume (foreign subscription $5.50) postpaid. Manuscripts, subscriptions, and all business correspondence should be addressed to THE JOURNAL OF ANIMAL BEHAVIOR Emerson Hall, Cambridge, Massachusetts Behavior Monographs For the publication of studies in behavior and intelligence which are too lengthy or too expensive for acceptance by The Journal of Animal Behavior Published at irregular intervals at Cambridge, Boston, Mass., in connection with the Journal of Animal Behavior, by Henry Holt and Company, New York. Each volume contains approximately 450 pages with plates and text-figures. The subscription price is $3.00 per volume (foreign subscription $3.50) postpaid. Monographs may be purchased separately at prices varying with the cost of manufacture. Manuscripts and inquiries concerning terms of publication should be addi-essed to the Editor of the Behavior Monographs, JOHN B. WATSON, The Johns Hopkins University, Baltimore, Md. Subscription to The Journal of Animal Behavior and the Behavior Monographs should be sent to Emerson Hall, Cambridge, Massa- chusetts. Behavior Monographs Volume 3, Number 3. 1917 Serial Number 14 Edited by JOHN B. WATSON TTie Johns Hopkins University An Analysis of the Learning Process in the Snail, Physa gyrina Say ELIZABETH LOCKWOOD THOMPSON Contribution from the Zoological Laboratory of the University of Michigan No. 147 Ann Arbor, Mich. 28 FIGURES Published at Cambridge. Boston, Mass. HENRY HOLT & COMPANY 34 West 33d Street, New York G. E. STECHERT & CO.. London. Paris and Leipzig. Foreign Agents TABLE OF CONTENTS Page Introduction 1 I. Modifiability studied by a method analogous to the Pawlow salivary reflex method — learning through conditioned stimuli 1 1. Outline of problem 1 2. Material and methods in general 3 3. Response to the food stimulus 5 A. The response a reflex 5 B. The number and duration of reactions in normal food stimulus response 7 4. Response to the conditioned food stimulus, simultaneous food and pressure stimuli H A. Special apparatus and methods 12 B. Experiments which show lack of response of untrained snails to pressure alone 12 C. Experiments to determine response to the conditioned food stimulus, first food-pressure series 13 D. Experiments with pressure alone on trained snails; asso- ciation i 24 E. Experiments to determine the effect of training on the food-pressure response, second food-pressure series 27 5. Summary of observations 31 6. Discussion of modifiabihty through the method of simultaneous stimuli 3o II. Modifiability studied by the labyrinth method 39 1. Introduction 39 2. Apparatus and material 40 3. Experiments on preference 42 A. Experiments to test right-left preference 43 B. Experiments to test preference for an ascending or des- cending path 46 4. Experiments on learning the U-shaped labyrinth: choice of right and left paths; punishment; failure to get air 47 5. Experiments with Y-shaped labyrinth; roughness as a warning signal ; electric-shock punishment 60 6. Experiments with Y-shaped labyrinth; mechanical stimulus as warning signal; electric-shock punishment 72 7. Summary of observations 81 8. Discussion of learning by the labyrinth method 84 III. Conclusions °' IV. Literature cited 88 List of illustrations 89 AN ANALYSIS OF THE LEARNING PROCESS IN THE SNAIL, PHYSA GYRINA SAY INTRODUCTION The experiments included in this paper were begun with the purpose of studying the abiUt\^ of snails to discriminate between stimuli. To this end a method was adopted analogous to the salivary reflex method of Pawlow. Its use showed, among other things, a form of modifiability of behavior that could be inter- preted only as learning — the capacity to form associations. The original plan to use this capacity as a basis for studies in discrimination was then changed. The discrimination experi- ments were deferred and the work was continued in order to determine if the snail, which can form associations, can also solve a simple labyrinth. The paper thus falls into two parts. Part one deals with modifiabitity as disclosed by the method of conditioned stimuli — with the power to form simple associa- tions. Part two takes up the further question of the ability of the snail to solve a labyrinth and discusses the relation of this to the formation of simple associations. It is the purpose of the writer to return to the study of discrimination. To Professor Jacob Reighard the author is indebted for sug- gesting the problem and for advice during the progress of the work: to Professor John F. Shepard she is indebted for numer- ous criticisms and suggestions. I. MODIFIABILITY AS STUDIED BY A METHOD ANALOGOUS TO THE PAWLOW SALIVARY REFLEX METHOD 1. Outline of the Problem The snail, Physa gyrina Say, like many other species, has the habit of crawling suspended from the surface film of the water. Its shell is then below and the ventral surfaces of the muscular foot and the head are exposed to view from above. The mouth may then often be seen to open and close many times in suc- cession. Dawson (1911) found that when the region within a 1 2 ELIZABETH LOCKWOOD THOMPSON millimeter or two of the mouth was touched with a bit of food or a clean glass rod, a chewing motion of the mouth parts began which " would continue for perhaps an hour or so, even if the animal were ])laced in fresh clean water." She found that the same stimuli applied to other parts of the animal produced no movements of the mouth parts. Familiarity with the mouth movements of this snail suggested that it might be a reflex of the same definite character as the salivary reflex made use of by Pawlow and his pupils in their studies of association and dis- crimination in the dog. Pawlow (1904) opened the duct of the parotid gland of a dog to the outside by means of a fistula. He then measured the secretion and determined its quality (viscosity), when induced, first by the odor or sight of food unaccompanied by other experi- mental stimuli. Pawlow called this food stimulus an " uncon- ditioned " stimulus. He then measured the secretion induced by food stimulation in the presence of an auxiliary (or secondary) stimulus such as sound or color. This he called a " conciitioned " stimulus. After using the conditioned stimuhis for a time, he omitted the primary stimulus (food) and found that the secretion was induced by the auxiliary stimulus, the tone or color, alone. He believed that he had shown conclusive!}^ that the animal had formed an association betw^een the two stimuli since at first it had not reacted to the secondary stimulus, but after this stimulus had been used for a time in connection with the pri- mary or food stimulus it had reacted to the secondary stimulus alone. Then by changing slightly the tone, color or other sec- ondary stimulus used, he determined whether the animal re- acted to this altered stimulus and to what extent. A change in the reaction was taken as evidence that the animal discrim- inated between the original and the altered stimulus. Thus power to discriminate between stimuli was tested. Pawlow and his followers believed that the method could be used on only a limited number of mammals. " Its obvious limi- tation appears in the number of organisms with which it may be employed. Evidently it cannot be used for the study of animals which lack salivary glands, and even among those animals which do possess these glands there are many which surely would not lend themselves satisfactorily to the method. It seems therefore as if Pawlow's method were especially important in animal psy- AN ANALYSIS OF THE LEARNING PROCESS IN THE SNAIL 3 cholog^^ as a means to the intensive study of the mental Hfe of a hmited number of mammals." (Yerkes and Morgulis, 1909.) Theoretically, however, this method is applicable to any organism that responds to a stimulation by a specific, measurable reflex of muscles or glands. Such reflexes occur in many of the lower forms. The mouth movement of Physa is an example. It was decided then to study; (1.) The character of the response following a single appli- cation of a food stimulus to the mouth parts of the snail. Such a response consists of one or more reactions. A single complete movement — one reaction — may be compared to the secretion of a drop of saliva in the Pawlow experiments. A determination of the number of reactions per response was deemed comparable to Pawlow's measurements of the salivary secretion induced by food alone or resulting from an " unconditioned " stimulus. (2.) Following this the application of two stimuli together, that is, food to the mouth parts and pressure at a fixed .distance from the mouth, would, it was believed, correspond to Pawlow's " conditioned " stimulus. The number of reactions constitut- ing each response would correspond to his measurements of saliva under such conditions. (3.) It was then planned to apply the associated or auxiliary stimulus alone in the absence of food, in order to determine from the presence or absence of reactions, whether or not an association had been formed between the two stimuli. 2. Material and Methods The snails* used in the experiments were collected in an oxbow pond that had been completely cut oft' from a small creek. Several hundred were brought into the laboratory and placed in large glass aquaria containing water plants such as Elodea and Spirog\Ta. An effort was made to have the environ- ment in these aquaria as nearly like the natural habitat as pos- sible. Besides the water plants upon which snails usually feed, fresh lettuce was placed in the aquaria. The\^ ate this in com- paratively large quantities, and seemed to prefer it to other plant food. * The author is indebted to H. B. Baker and Harold Cummins for the identifica- tion of the snails used in the experiments. 4 ELIZABETH LOCKWOOD THOMPSON When the tank containing recently captured snails was jarred or disturbed in any way, they would instantly expel the air from the lung, retract into their shells, and drop from the surface film upon which they were crawling, to the bottom of the tank. Many of them frequently remained motionless on the bottom for an hour or more before they again sought the surface and refilled their lungs. It became necessary therefore, to " tame " the snails which were to be used in the experiments, to so accustom them to handling that they would remain extended and retain the air in the lung while they were being worked upon. About twenty specimens of approximately the same size were selected for the tests. These were placed in two bacteria dishes. In each dish was about a liter of water with water plants and as much lettuce as the snails would eat. The snails were taken in the hand at intervals and moved about under water. Each was held in the hand beneath the water until it emerged from the shell and suspended itself from the surface film. They gradually became accustomed to this handling to such an extent that they could be moved from dish to dish, at the will of the operator, without retracting or ex]3elling the air from the lung. These " tamed " snails were then divided into two groups. Each individual was numbered by means of white water-proof paint applied to the shell after it had been thoroughly dried. These groups were worked on alternate days. Each group was starved for a period of twenty-four hours before tests were made. The snails belonging to the group were thus believed to be in approximately the same state of hunger at the time of experi- mentation. After a group had been worked on it was allowed to feed for twenty-four hours and was again starved, before it was used. In this way an attempt was made to obtain physio- logical uniformity among the snails directly under observation. The physical conditions were also kept as uniform as possible. The whole series of daily experiments was carried on at approxi- mately the same hour. Each snail in turn was placed in a clean dish of fresh, filtered tap water. The water was of the same depth and at the same (room) temperature as that in which the snails had been living and the lighting was at all times uniform with that to which they were accustomed. Each individual was put through a series of ten tests per day with one exception (check series, Table I, p. 6) and a separate record made of each. AN ANALYSIS OF THE LEARNING PROCESS IN THE SNAIL 5 3. The Response to the Food Stimulus In order to utilize the snails it was necessary to know: (1) Whether the mouth reactions occur only as the result of external stimulation, and (2) Their constancy as to the nimiber of reactions and the duration of the response following a single stimulus. Should they occur regularly as the result of external stimulation and only then, they must be regarded as " involuntary " responses or reflexes available for the purpose of the experiments. Upon the constancy and duration of the response depends its value for quantitative uses. A. 1 he response to the food stiniulus a reflex. — It was found by observation not only that the mouth movements of Physa were induced by food and mechanical stimulation of the mouth region, but that they also sometimes occurred when no such stimuli could be detected. Such apparently voluntary move- ments might be due to the stimuli from microscopic particles in the water. A check series of experiments in filtered *water was deemed necessary in order to determine whether or not the mouth reactions might normally occur without external stimu- lation and if so under what conditions and with what frequency. In these tests each snail, under the controlled conditions just mentioned, was in turn held in the hand of the operator until it suspended itself from the surface film. It was then permitted to move about, and was touched only when necessary to keep it from crawling down the sides of the dish. At such times it was pushed gently toward the center of the tank with a sterile glass rod. Each animal was thus kept crawling on the surface film for an average period of forty-five minutes under constant observation and a record was made of each mouth reaction. It was found to be impossible to keep the snail absolutely free from all chance stimulation of the mouth. It has been shown that the animal uses the same method of locomotion in crawling on the surface film that it does in moving over any substratum, that is, it crawls along a mucous path which it secretes as it moves and which remains behind it on the surface of the water. (Parker, 1912), (Dawson, 1911). The mucus thus left on the surface mav reach the mouth of the snail as it moves about. In *A Berkefeldt filter was used. 6 ELIZABETH LOCKWOOD THOMPSON several instances this mucus was observed to touch the mouth parts and such contact was followed b}^ reactions. These are noted in the following table. Stimulation of this sort is the probable cause of all other reactions that occur in the test series. TABLE I Showing the number of reactions of the mouth of six individuals of Physa gyrina Say, kept under observation in filtered tap water for 32.75 hours, when not stim- ulated by the application of food to the mouth region. The snails had been with- out food for 24 hours preceding the tests. They were divided into two groups of three each and tested on alternating days. Period of No. of mouth No. of mouth observation reactions, reactions, No. of days in minutes no visible mucous Snail No. tested each day 60 stimulus stimulation 18 1 0 0 2 45 0 . 0 3 60 0 0 4 60 0 0 5 60 0 0 6 60 0 0 19 1 60 0 0 2 45 2 14 3 60 0 5 4 60 1 0 5 60 0 0 6 60 0 0 20 1 60 0 0 2 45 0 0 *3 60 0 0 4 60 0 0 5 60 0 0 21 1 30 0 0 2 60 0 . 0 3 60 0 0 4 60 0 0 5 60 0 0 6 60 0 0 22 1 30 0 0 2 60 0 0 3 60 0 0 4 60 0 0 5 60 0 0 6 60 0 0 23 1 30 7 0 2 60 4 0 3 60 0 2 4 60 3 0 « 5 60 0 0 6 60 0 0 * Indicates that an egg mass had been deposited within the 24 hours preceding the test. AX ANALYSIS OF THE LEARNING PROCESS IN THE SNAIL 7 TABLE I — {Continued) Total time of observation • 32 . 75 hrs. Total number of reactions 38 Total number of reactions, no visible cause 17 (44 . 73%) Total number of reactions, mucous stimulation 21 (55 . 26%) Average number of reactions per hour 1 . 16 Average number of reactions, no visible cause, per hour 0.51 This table shows that when six snails were kept under obser- vation for nearly thirty-three hours (about five and a half hours each) in filtered tap water onty 38 mouth reactions were observed, an average of 1.16 per hour. If we allow for each reaction 1.25 seconds (Table 11, p. 8) the duration of the i'& reactions is 47.50 seconds. Of the 38 reactions 21 or 55.26% were observed to be due to mucous stimulation and the remainder were probably due to the same cause. Owing to the fact that the mucus is colorless, and as it occurs on the surface of the water, trans- parent, it is exceedingly difficult to determine its presence at all times. Any attempt to bring it into view by the addition of pigment such as powdered carmine would only serve to make it more eft"ective as a stimulus. The reactions which did occtir can then be accoiuited for. The mouth reactions of Physa may then be regarded as a true reflex, an " luilearned response " (Watson, 1914), which occtirs normally as the result of external stimulation of the mouth or the region immediately surrounding it. For experimental purposes we may say that the reaction occurs only as the result of external stimulation. If now, the snail were placed in water freed from microscopic particles by filtration and the mouth reactions were practically alwa^^s in- duced by bringing food directly into contact with the mouth, the reactions must be regarded as reflexes, that is, involuntary. Such a series of reactions then might be interpreted as an " un- conditioned reflex " (Pawlow, 1904). The ntimber of reactions and the duration of the series should then prove an available substitute for the measurements of saliva in inquiries into the behavior of a group of animals widely separated from mammals. B. Number and duration of reactions in the normal jood- stimulus response. — The snails that were used in all of the re- maining tests had been fed upon lettuce while in the laboratory; for this reason lettuce was used as a food stimulus in all the experiments. A small piece taken from a fresh leaf was applied to the normally inactive mouth of the snail by means of forceps 8 ELIZABETH LOCKWOOD THOMPSON and immediately withdrawn. This stimulus served to start the movements of the mouth parts, and an exact record of the num- ber of reactions thus induced was kept for each individual. Each complete opening and closing movement is spoken of as a reaction. The number of reactions (whether one or several) induced by a single stimulation of the mouth parts is termed a response. During work a record was also made of the length of time that the reactions persisted; that is, the duration of the response (Table II). TABLE II Showing the number of reactions of the mouth of two individuals of Physa gyrina Say, resulting from a single application of food to the mouth. The tests were made in filtered water on alternating days. The snails had been starved for 24 hours preceding the tests. Snail No. 4 No. for No. of No. of No. of "food trial reactions Duration days stimulus" for per of response tested series day response in sec. 1 1 1 1 1 2 2 0 0 3 3 1 1 4 4 6 10 5-10 5-10 0 0 Totals 10 8 Av. per response 2 2.66 11 1 1 1 12 2 6 10 . 13 3 38 50 14 4 * 0 15 5 1 1 16 6 1 1 17 7 3 5 18 8 1 1 19-20 9-10 0 0 Totals 10 51 Av. per response 3 7.28 21 1 0 0 22 2 1 23 3 1 24 4 1 25 5 ' 1 26 6 1 27 7 2 3 28 8 0 0 29 9 1 1 30 10 0 0 Totals 10 8 Av. per response 1.14 * Cases in which lack of response was due to known mechanical disturbance. AN ANALYSIS OF THE LEARNING PROCESS IN THE SNAIL 9 TABLE II — {Continued) Snail No. 4 — (Continued No. for No. of No. of No. of "food trial reactions Duration days stimulus" for per of response tested series day response in sec. 4 31 1 3 4 32 2 2 2 33 3 1 1 34 4 1 1 35 5 1 1 36 6 2 2 37 7 1 1 38 8 2 2 39-40 9-10 0 0 Totals 10 13 Av. per response 5 1.62 41 1 11 15 42 2 6 8 43 3 0 0 44 4 3 3 45 5 1 1 46 6 0 0 47-50 7-10 * 0 Totals 10 21 Av. per response 5.25 Total number of trials 50 Total number of stimuli followed by response 29 Per cent of stimuli followed by response 58% Total number of reactions 101 Average number of reactions per trial 2.02 Average number of reactions in only those trials in which response occurred 3 . 48 Total time of reactions 2 . 18 min. Average time p^r reaction 1 . 29 sec. Average number of reactions per minute 46 . 51 * Cases in which lack of response was due to known mechanical disturbance. 10 ELIZABETH LOCKWOOD THOMPSON TABLE U— (Continued) Snail No. 5 No. for No. of No. of No. of "food trial reactions Duration days ! stimulus" for per of response tested series day response in sec. 1 1 1 24 30 2-4 2-4 0 0 5-10 5-10 * 0 Totals 10 24 Av. per response . 2 24 11 1 5 10 12 2 1 1 13 3 * 0 14 4 0 0 15 5 1 1 16 6 1 1 17 7 2 3 18 8 1 1 19-20 9-10 * 0 Totals 10 11 Av. per response 3 1.83 21 •1 1 1 22 2 1 1 23 3 1 1 24 4 2 3 25 5 2 3 26 6 1 1 27-30 7-10 * 0 Totals 10 8 Av. per response 4 1.33 31 1 3 3 32 2 4 5 33 3 4 5 34 4 2 2 35 5 1 1 36 6 1 1 37 7 1 1 38 8 1 1 39 9 4 4 40 10 1 1 Totals 10 22 Av. per response 5 2.2 41 1 19 22 42 2 •20 22 43 3 5 6 44 4 10 12 45 5 4 8 46 6 6 5 47 7 1 1 48 8 0 0 49 9 3 4 50 10 6 8 Totals 10 74 Av. per response 8 '^'^ ' Cases in which lack of response was due to known mechanical disturbance. AN ANALYSIS OF THE LEARNING PROCESS IN THE SNAIL 11 TABLE U— {Continued) Total number of trials 50 Total number of stimuli followed by response 32 Per cent of stimuli followed by response 64% Total number of reactions 139 Average number of reactions per trial 2 . 78 Average number of reactions in only those trials in which response occurred 4 . 35 Total time of reactions 2.81 min. Average time per reaction 1.21 sec. Average number of reactions per minute 45 . 9 Summary, Snails 4 and 5: Total number of stimuli (trials) 100 Total number of stimuli followed by response 61 Per cent of stimuli followed by response 61% Table II shows that when a food stimulus is applied to the mouth region of Physa gyrina Say, response occurs in 61% of 100 trials with two snails. The individual records show that the number of reactions making up the response varies from 1 to 38. The first 30 trials of the series of 50 trials for Snail 4 shows 67 reactions, while the following 20 trials give 3-4 reactions with an average of 2.83 reactions per response. In the case of Snail 5 the corresponding numbers are 43 and 96 reactions with an average per response of i.i and 5.05. This indicates in the case of Snail 5 a tendency to recover during the second part of the series from the inhibiting effect of the manipulation (adaptation). In Snail 4 the recovery may be delayed beyond the limits of the series. The series is in both cases too short to afford anything more than a suggestion as to the occurrence of adaptation. The absence of reaction in certain trials and the wide variation in their number in those trials in which they occur and in the differences between individual snails, are brought out in the summary (p. 31) and are discussed later. The average number of reactions per response furnishes a means of comparison of the results here obtained with those recorded in the part of this paper which deals with the simultaneous use of two stimuli. 4. Response to the Conditioned Food Stimulus: Simultaneous Appli- cation OF Food and Pressure There are then determined the number of reactions of the mouth parts induced by a single definite stimulation and constituting a single response and the duration of the response and of the single reaction. The stiniulus according to Pawlow's terminology would be " unconditioned " or primary. An attempt vras next 12 ELIZABETH LOCKWOOD THOMPSON made to use a " conditioned " or double stimulus by the simul- taneous application of pressure and food. A. Special apparatus and methods. — This required the con- struction of a special apparatus which would make possible the application of a uniform pressure at a fixed distance from the food stimulus when the latter was applied to the mouth of the snail. Out of the several appliances tried, the one here pictured (fig. 1, Plate I) gave the best results. (A) was a wire nail which served as a handle. A small strip of sheet metal (B) 5 mm. in width, and bent in the form of a s: uare cornered U, was soldered to the end of A. B contained two smooth-edged holes, one directly above the other and both as large as would permit the free movement through them of a smooth steel rod (C) which was made from a No. 5 sewing needle. C was kept from dropping through the holes, by a fine wire (D) which was passed through the eye of the needle. To the lower end of the rod C was soldered a U-shaped piece made of No. 30 platinum wire. One limb of the U (E) 16 mm. in length, ended in a small hook. The other limb (F) 18mm. in length, ended in a three-pronged, dull-pointed platinum fork 3 mm. wide and turned at right angles to the plane of the U-shaped piece. A was held in the hand of the operator and the fork on the lower end of F was allowed to rest on the exposed ventral surface of the snail. At the same time food (a small piece of lettuce rolled between the fingers to form an inverted cone) suspended from the hook at the lower end of E was applied to the mouth of the snail. The free movement of the lower part of the appara- tus permitted by the smooth-edged holes in B, made it possible for the operator to manipulate the apparatus in such a way that it remained practically stationary upon the surface of the snail during the period of stimulation in spite of the constant motion of the animal. B. Ex perhncuts wJiicJi show lack of response of untrained snails to pressure alone. — Before using the apparatus in experiments with the conditioned stimulus, a series of twenty trials was made on each of six snails in order to learn whether the application to the foot of the pressure fork alone induced response. The snails were starved for twenty-four hours previous to the experi- ments and each was tested in filtered tap water. The above AN ANALYSIS OF THE LEARNING PROCESS IN THE SNAIL 13 described apparatus was applied without food to the upturned ventral surface of the foot. It was placed upon the animal in such a way that the food hook was just above the mouth. Pressure was thus applied to the same part of the foot that would receive it if the double stimuli, food and pressure, were being applied together. In the total of 120 trials 4 responses occurred (3.33%) and these were probably due to mucous stimulation. Pressure alone, as applied with the apparatus used is followed by response in so small a percentage of the trials that its occur- rence does not vitiate the experiments which foll9w. C. Experiments to determine response to the conditioned food stimidus: first food and pressure series. — By means of the food- pressure apparatus food and uniform pressure were next applied simultaneously in a long series of trials. Both stimuli were immediately withdrawn and the number of reactions of the mouth counted and recorded as before. Table III shows the record obtained in this attempt to form an association between two dissimilar stimuh applied simultaneously. TABLE III Showing the number of reactions of the mouthsof two individuals of Physa gyrina Say, resulting from the simultaneous application of two dissimilar stimuli, food and pressure. The tests were made in filtered tap water on alternating days. The snails had been starved for 24 hours preceding the tests. Sn.\il No. 4 No. for "food and No. of No. of No. of pressure" trial reactions Duration days stimulus for per of response tested series day response in sec. 6 1-10 1-10 0 0 7 11-20 1-10 0 0 8 21-30 1-10 0 0 9 31-40 1-10 0 0 10 41-50 1-10 0 0 11 51-60 1-10 0 •0 Totals 60 0 12 61 1 4 5 62 2 1 1 63 3 2 2 Totals 64-70 4-10 10 0 7 2.33 0 Av. per response . 14 ELIZABETH LOCKWOOD THOMPSON TABLE III— {Continued) Snail No, 4 — (Continued) No. for "food and No. of No. of No. of pressure" trial reactions Duration days stimulus for per of response tested series day response in sec. 13 71-80 1-10 0 0 14 81 1 0 0 82 2 1 1 83 3 8 10 84-90 4-10 0 0 Totals 10 9 Av. per response . 15 4.5 91 1 1 1 92 2 1 1 93 3 0 0 94 4 2 2 95-100 5-10 0 0 Totals 10 4 Av. per response. 16 1.33 101-110 1-10 0 0 17 111 1 16 18 112 2 6 8 113 3 0 0 114 4 1 1 115 5 1 1 116 6 0 0 117 7 1 1 118 8 1 1 119 9 2 2 120 10 0 0 Totals 10 28 Av. per response. 4.0 18 121 1 1 1 122 2 0 0 123 3 2 2 124 4 2 2 125 5 5 6 126 6 2 2 127 7 1 1 128 8 1 1 129-130 9-10 0 0 Totals. 10 14 Av. perlresponse . 19 2.0 131 1 25 27 132 2 6 8 133 3 30 33 134 4 4 5 135 5 7 8 136-140 6-10 0 0 Totals 10 72 A v. per response. 14.4 AN ANALYSIS OF THE LEARNING PROCESS IN THE SNAIL 15 TABLE III— {Continued) Snail No. ^—{Continued) No. for "food and No. of No. of No. of pressure" trial reactions Duration days stimulus for per of response tested series day response in sec. 20 141 1 19 22 142 2 30 35 143 3 8 9 144 4 10 12 145 5 3 3 146 6 0 0 147 7 1 1 148 8 * 0 149-150 9-10 0 0 Totals 10 71 Av. per response . 11.83 21 151 1 8 9 152 2 8 10 153 3 2 2 154 4 3 4 155 5 * 0 156 6 2 2 157 7 0 0 158 8 9 10 159-160 9-10 0 0 Totals 10 34 Av. per response . 5.66 22 161 1 1 1 162 2 0 0 163 3 1 1 164 4 1 1 165 ' 5 1 1 166 6 1 1 167-170 7-10 0 0 Totals 10 5 Av. per response . 1.0 23 171 1 10 13 172 2 12 16 173 3 2 2 174 4 1 1 175 5 1 1 176 6 4 5 177 7 1 1 178 8 9 12 179 9 3 4 180 10 1 1 Totals 10 44 Av. per response . 4.4 * Cases in which lack of response was due to known mechanical disturbance. 16 ELIZABETH LOCKWOOD THOMPSON TABLE lU— {Continued) Snail No. -i— (Continued) No. for "food and No. of No. of No. of pressure" trial reactions Duration days stimulus for per of response tested series day response in sec. 24 181-190 1-10 0 0 25 191 1 4 5 192 2 1 1 193 3 4 5 194 4 1 1 195 5 1 1 196 6 4 5 197 7 1 1 198-200 8-10 0 0 Totals 10 16 Av. per response. 2.28 26 201 1 2 2 202 2 3 4 203 3 6 8 204 4 1 1 205 5 1 1 206 6 2 2 207 7 2 2 208 8 3 4 209 9 2 2 210 10 2 2 Totals 10 24 Av. per response. 2.4 27 211 1 3 4 212 2 4 5 213 3 5 6 214 4 0 0 215 5 1 1 216-220 6-10 0 0 Totals 10 13 Av. per response . 3.25 28 221 1 4 5 222 2 1 1 223 3 2 2 224 4 1 1 225 5 1 1 226 6 4 5 227 7 1 1 228 8 1 1 229 9 2 2 230 10 1 1 Totals 10 18 Av. per response . 1.8 AN ANALYSIS OF THE LEARNING PROCESS IN THE SNAIL 17 TABLE III— (Continued) Snail No. 4 — {Continued) No. for "food and No. of No. of No. for pressure trials reactions Duration days stimulus" for per of response tested series day response in sec. 29 231 1 1 232 2 1 233 3 1 234 4 1 235 5 1 236-240 6-10 0 0 Totals 10 5 Av. per response. 1.0 30 241 1 10 12 242 2 1 243 3 1 244 4 1 245 5 1 246 6 1 247-250 7-10 0 0 Totals 10 15 Av. per response 2.5 31 251 1 4 5 252 2 5 6 253 3 5 6 Totals 3 14 Av. per response . 4.66 Total number of trials 253 Total number of stimuli followed by response 99 Per cent of stimuli followed by response 35% Total number of reactions 393 Average number of reactions per trial 1 . 55 Average number of reactions in only those trials in which response occurred 3 . 96 Total time of reactions 7 . 58 min. Average time per reaction 1 . 15 sec. Average number of reactions per minute 51 . 84 18 ELIZABETH LOCKWOOD THOMPSON TABLE in— (Continued) Snail No. 5 No. of days tested No. for "food and pressure" stimulus series No. of trials for day No. of reactions per response Duration of response in sec. Totals . . . 6 7 8 9 10 11 12 13 1-10 11-20 21-30 31-40 41-50 51-60 61-70 71-80 1-10 1-10 1-10 1-10 1-10 1-10 1-10 1-10 80 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Totals Av. per response . . 14 81-90 10 14 14.0 16 Totals 15 91-100 10 0 0 Totals 16 101-110 10 0 0 Totals Av. per response . . 17 111-120 10 24 4.8 28 Totals Av. per response . . 18 121-130 10 11 1.83 12 Totals Av. per response . . 19 131-140 10 74 9.25 85 Totals Av. per response.'. 20 141-150 10 23 2.87 27 Totals 21 151-160 10 0 0 Totals 22 161-170. 10 0 0 Totals Av. per response . . 23 171-180 10 13 3.25 16- Totals Av. per response . . 24 181-190 10 3 1.0 3 Totals Av. per response . . 25 191-200 10 13 3.25 15 Totals Av. per response. . 26 201-210 10 17 1.7 19 Totals 27 211-220 10 0 0 Totals 28 221-230 10 0 0 Totals Av. per response . . 29 231-240 10 5 1.25 5 Totals Av. per response . . 30 241-250 10 28 3.11 34 AN ANALYSIS OF THE LEARNING PROCESS IN THE SNAIL 19 TABLE lll^iContinued) Snail No. '3^ (Continued) Total number of trials 250 Total number of stimuli followed by response 62 Per cent of stimuli followed by response 24% Total number of reactions 225 Average number of reactions per trial 0.9 Average number of reactions in only those trials in which response occurred 3 . 62 Total time of reactions 4 . 33 min Average time per reaction 1.11 sec. Average number of reactions per minute 51 . 92 Summary Snails 4 and 5: Total number of stimuli (trials) 503 Total number of stimuli followed by response 161 Per cent of stimuli followed by response 32% TABLE III-A Showing the distribution of responses of each 10 trials of the 250 trials shown in Table III. Snail No. 4 Total duration Average Total in sec. of duration No. of no. of responses per of Trials no. responses reactions 10 trials reactions 1- 60 0 0 0 0 61- 70 3 7 8 1.14 71- 80 0 0 0 0 81- 90 2 9 11 1.22 91-100 3 4 4 1.0 101-110 0 0 0 0 111-120 7 28 32 1.14 121-130 7 14 15 1.07 131-140 5 72 81 1.12 141-150 6 71 82 1.15 151-160 6 34 39 1.14 161-170 5 5 5 1.0 171-180 10 44 56 1.27 181-190 0 0 0 0 191-200 7 16 19 1.18 201-210 10 24 28 1.16 211-220 4 13 16 1.23 221-230 10 18 20 1.11 231-240 5 5 5 1.0 241-250 6 15 17 1.13 251-253 3 14 Snail No. 5 17 1.21 1- 80 0 0 0 0 81- 90 1 14 16 1.14 91-100 0 0 0 0 101-110 0 0 0 0 111-120 5 24 28 1.16 121-130 6 11 12 1.09 131-140 8 74 85 1.14 20 ELIZABETH LOCKWOOD THOMPSON TABLE Ul-A—{Contmiied) Snail No. 5— {Continued) Total duration Average Total in sec. of duration No. of no. of responses per of Trials no. responses reactions 10 trials reactions 141-150 8 23 27 1.13 151-160 0 0 0 0 161-170 0 0 0 0 171-180 4 13 16 1.23 181-190 3 3 3 1.0 191-200 4 13 15 1.15 201-210 10 17 19 1.11 211-220 0 0 0 0 221-230 0 0 0 0 231-240 4 5 5 1.0 241-250 9 28 Snail No. 6 34 1.21 1-110 0 0 0 0 111-120 1 6 7 1.16 121-130 0 0 0 0 131-140 2 9 11 1.22 141-150 2 2 2 1.0 151-160 0 0 0 0 161-170 0 0 0 0 171-180 0 0 0 0 181-190 3 3 3 1.0 191-200 2 2 2 1.0 201-210 5 6 6 1.0 211-220 0 0 0 0 221-230 5 5 5 1.0 231-240 4 7 7 1.0 241-250 9 25 Snail No. 7 29 1.16 1- 50 0 0 0 0 '51- 60 1 1 1 1.0 61- 70 1 11 13 1.18 71- 80 0 0 0 0 81- 90 2 2 2 1.0 91-100 3 10 13 1.3 101-110 0 0 0 0 111-120 4 27 31 1.11 121-130 3 6 7 1.16 131-140 5 32 41 1.21 141-150 3 38 44 1.15 151-160 3 4 4 1.0 161-170 4 11 13 1.18 171-180 4 46 55 1.19 181-190 4 4 4 1.0 191-200 10 35 44 1.25 201-210 8 11 12 1.09 211-220 10 13 13 1.0 221-230 4 7 8 1.14 231-240 5 29 33 1.13 241-250 8 50 56 1.12 AN ANALYSIS OF THE LEARNING PROCESS IN THE SNAIL 21 TABLE lU-A—iConlinued) Snail No. 8 Total duration Average Total in sec. of duration No. of no. of responses per of Trials no. responses reactions 10 trials reactions 1- 80 0 0 0 0 81- 90 1 1 1 1.0 91-100 0 0 0 0 101-110 0 0 0 0 111-120 1 . 2 3 1.5 121-130 2 2 2 1.0 131-140 1 2 2 1.0 141-150 2 3 3 1.0 151-160 5 9 9 1.0 161-170 0 0 0 0 171-180 3 4 4 1.0 181-190 0 0 0 0 191-200 4 4 4 1.0 201-210 4 4 4 1.0 211-220 0 0 0 0 221-230 9 10 10 1.0 231-240 10 21 24 1.14 241-250 8 9 9 1,0 251-253 3 13 Snail No. 14 15 1.15 1- 60 0 0 0 0 61- 70 3 7 8 1.14 71- 80 0 0 0 0 81- 90 1 4 5 1.25 91-100 2 3 3 1.0 101-110 2 3 4 1.33 111-120 0 0 0 0 121-130 3 13 16 1.23 131-140 5 29 35 1.2 141-150 3 4 4 1.0 151-160 2 5 6 1.2 161-170 3 8 9 1.12 171-180 0 0 0 0 181-190 1 1 1 1.0 191-200 0 0 0 0 201-210 7 25 31 1.24 211-220 6 32 38 1.18 221-230 0 0 0 0 231-240 4 31 37 1.12 241-250 5 17 19 1.11 22 ELIZABETH LOCKWOOD THOMPSON TABLE III-B Showing the distribution of responses of each 50 trials of the 250 trials of the food-pressure " stimulus series. Snail No. 4 Total duration Average Total in sec. of duration No. of no. of responses per of Trials no. responses reactions 50 trials reactions 1- 50 0 0 0 0 51-100 8 20 23 1.13 101-150 25 185 210 1.13 151-200 28 99 119 1.2 201-250 35 75 Snail No. 5 86 1.14 1- 50 0 0 0 0 51-100 1 14 16 1.14 101-150 27 132 152 1.15 151-200 11 29 34 1.17 201-250 23 50 Snail No. 6 58 1.16 1- 50 0 0 0 0 51-100 0 0 0 0 101-150 5 17 20 1.17 151-200 5 5 5 1.0 201-250 23 43 Snail No. 7 47 1.09 1- 50 0 0 0 0 51-100 7 24 29 1.2 101-150 15 103 123 1.19 151-200 25 100 120 1.2 201-250 35 110 Snail No. 8 122 1.1 1- 50 0 0 0 0 51-100 1 1 1 1.0 101-150 6 9 10 1.11 151-200 12 17 17 1.0 201-250 31 44 Snail No. 14 47 1.06 1- 50 0 0 0 0 51-100 6 14 16 1.14 101-150 13 49 59 1.2 151-200 6 14 16 1.14 201-250 22 105 125 1.19 From Table III it appears that when two stimuU were apphed simultaneously, food to the mouth and pressure to the foot, there was at first an inhibitory effect upon the reflex. This is shown in the first 60 to 80 trials in which no reaction followed the application of the double stimulus. When this table is AX ANALYSIS OF THE LEARNING PROCESS IN THE SNAIL 23 compared with Table II it is seen that a sHght inhibition con- tinued throughout this series of tests as indicated in the average number of responses per 50 trials (Table III-B) and the per cent of stimuli followed by response (Tables III and VI). The aver- age number of reactions per response was also less than in the normal food series. For the food-pressure response the average number of reactions in the numerical order of the six snails was 3.96, 3.62, 1.96, 4.1, 1.58, 3.78, or a mean for all six snails of 3.18. The corresponding values for the normal food response (Table II) are 3.48, 4.35, 6.5, 6.5, 3.4, 10.93, with a mean value of 5.85 for the six snails. Only in the case of Snail No. 4 is there a greater average number of reactions per response to the " con- ditioned " stimulus than to the " unconditioned." Table III shows further that the average duration of the reactions in the food pressure response is less than that of the normal food re- sponse. Expressed in the number of reactions per minute the values are 51.84, 51.92, 65.45, 51.72, 56.7, 50.84, with a mean of 54.74. The corresponding values for the normal food response are 46.51, 45.9, 46.15, 46.51, 44.44, 48.00 with a mean of 46.25. Under food-pressure stimulation the reactions thus occur at the rate of 8.49 more per minute than under the normal food stimu- lus. Along with the reduction in the number of reactions there is in the food-pressure response an acceleration of the individual reaction as compared with that of the normal food response. The distribution of the reactions over the series of 250 trials is shown for the six snails in Tables III-A and III-B and is represented in graphs (figs. 2-13, Plates II-III). Table III-A and figs. 2-7 (Plate II) show the great fluctuation in the number of reactions per response. This fluctuation causes the reaction curve to vary from the response curv^e although the two are in general similar in direction. The response curve shows a some- what more uniform rise than the reaction curve. A general fact to be noted is the relative uniformity in the average duration of the individual reactions throughout the series. The method of timing the response duration by means of the stop watch is necessarily inaccurate, especially with values which, as in this case, are frequently in the neighborhood of a second. If a more accurate method were available greater uniformity might be expected. With the method employed it is noteworthy that the agreement is so close. 24 ELIZABETH LOCKWOOD THOMPSON The food-pressure series thus shows the following character- istics: (1) An absence of reaction at the beginning, (2) a simi- larity in the reaction and response curves, (3) in general, a gradual rise of the response curve, (4) a less average number of responses per 50 trials and a less average number of reactions per response as compared with the normal food series, with the exception of Snail No. 4, (5) an increase in the number of reactions per minute, i. e. an acceleration of the individual response, (6) a fairly constant reaction duration. This is characteristic of all the series. D. Experiments ivith pressure alone on trained snails: asso- ciation. — The third part of the problem called for the applica- tion of the secondary stimulus alone. Any mouth reactions which might follow (beyond those attributed to chance stimu- lation of the mouth), would according to the Pawlow theory be evidence that an association had been formed between the essen- tial, primary or " unconditioned " stimulus and the secondary — or in this case, between food and pressure. The series of tests was carried on in exactly the same manner as the experiments just described. The same apparatus was used but no food was placed upon the food hook. vSince the food hook was shorter than the pressure fork (fig. 1, Plate I) it did not touch the mouth, which therefore received neither food nor mechanical stimulation. The pressure fork was applied to the usual place on the ventral surface of the snail. The following table (IV) shows the results obtained. AN ANALYSIS OF THE LEARNING PROCESS IN THE SNAIL 25 TABLE IV Showing the number of reactions of the mouth of two individuals of Physa gyrina Say, resulting from the application of pressure alone as a stimulus, after an attempt had been made to establish an association between the two stimuli, food and pressure (Table III). The tests were made in filtered tap water on alternating days. The snails had been starved for 24 hours preceding the tests. No. of days tested No. for "pressure" stimulus series Snail No. 4 No. of trial for day No. of reactions per response Duration of response in sec. 31 1 *4 3 4 2 5 11 3 6 11 4 7 11 5 8 11 6 9 11 7 10 0 0 Totals 7 8 Av. per response 1 . 33 32 8 1 1 1 9 2 11 10-17 3-10 0 0 Totals 10 2 Av. per response 1.0 33 18-27 1-10 0 0 34 28-37 1-10 0 0 Total number of trials (last 20 trials omitted) 17 Total number of stimuli followed by response 8 Per cent of stimuli followed by response 47% Total number of reactions 10 Average number of reactions per trial 0 . 58 Average number of reactions in only those trials in which response occurred. 1 . 25 Total time of reactions in seconds 11 Average time per reaction in seconds 1.1 Average number of reactions per minute (estimated) 54.54 Snail No. 5 No. for No. of No. of No. of "pressure" trial reactions Duration days stimulus for per of response tested series day response in sec. 31 1 1 11 14 2 2 11 3 3 11 4 4 12 14 5 5 11 6 6 5 6 7 7 3 4 8-10 8-10 0 0 Totals 10 34 Av. per response • 4 . 85 *For trials 1-3 on the 31st day see Table III (31st day). 26 ELIZABETH LOCKWOOD THOMPSON TABLE l\— {Continued) Snail No. 5 — {Continued) No. for No. of No. of Duration No. of "pressure" trials reactions of days stimulus for per response tested series day response in sec. 32 11 1 1 1 12 2 1 1 13-20 3-10 0 0 Totals 10 ■ 2 Av. per response 1.0 33 21-30 1-10 0 0 34 31-40 1-10 0 0 Total number of trials (last 20 trials omitted) 20 Total number of stimuli followed by response 9 Per cent of stimuli followed by response 45% Total number of reactions 36 Average number of reactions per trial 1.8 Average number of reactions in only those trials in which response occurred 4 Total time of reactions in seconds 43 Average time per reaction in seconds 1 . 19 Average number of reactions per minute (estimated) 50.42 Summary Snails 4 and 5. Total number of stimuli (trials) 37 Total number of stimuli followed by response 17 Per cent of stimuli followed by response 45 . 9% Table IV shows that when the secondary stimulus, pressure, alone was applied to the inputh of Physa, the mouth reacted in 17 out of v37 trials with two individuals, or in 45.9% of the trials. No mucous stimulation was observed. The average number of reactions per response was for Snail 4, 1.25 and for Snail 5, 4.0 (Table IV). Snails Nos. 6, 7, 8, 14 were given a similar series of tests which are not here tabulated in detail but are summarized in Table V-C. These showed the follow^ing average number of reactions per response 4.92, 6.09, 8.0, 3.8. The mean for the six snails is 4.67. The mean for the normal food series is for six snails 5.86. For the food-pressure series it is 3.18. The average duration of reactions for the pressure response of trained snails is less than for the normal food response. The average rate per minute is 53.9, 7.65 reactions per minute in excess of that of the normal food response. These trials followed 48 hours after the completion of the AN ANALYSIS OF THE LEARNING PROCESS IN THE SNAIL 27 training in the food-pressure series. In the case of Snail 4, response followed the first nine trials. The last two of these are separated by an interval of 48 hours from the first seven. The effect of training thus persisted for 96 hours. Similar results were obtained from Snails 5, 6, 7, 8 and 14. Immediately following the mouth response to pressure of trained Snail No. 4 (Table IV) seven successive applications of pressure gave no response. Similar results were obtained with the other five snails. Snails which before training gave no response to pressure, gave, following training, a response consisting of a series of mouth reactions. This response is given on the first trial and continues to be given in trials extending over a period of 96 hours. It then suddenly ceases. There is no response of the mouth to subsequent stimulation of the foot by pressure. E. Experiments to determine the effect of training on food- pressure response: second food-pressure series. — It was believed that another series of tests on the same snails, reintroducing the same " conditioned " stimulus, would be of interest in show- ing the ease or difficulty in re-establishing the reflex. As noted in Table IV, the snails were tested on two experimental days after pressure alone had failed to be effective as a stimulus. Allowing for the alternation of experimental days, made neces- sary^ by the twenty-four hours of starving previous to experi- mentation, the second series of tests using food and pressure as a stimulus was started approximately 120 hours after the dis- appearance of reaction upon the application of pressure as a stimulus. Fiftv tests were given and the results appear in Tables V and VI-D. 28 ELIZABETH LOCKWOOD THOMPSON TABLE V Showing the number of reactions of the mouths of two individuals of Physa gyrina Say, resulting from the simultaneous application of food and pressure after the complete disappearance (Table IV) of the effect of the association established between these two stimuli (Table III). The tests were made in filtered water on alternating days. The snails had been starved for 24 hours preceding the tests. Snail No. 4 No. for "food and pressure" No. of No. of Duration No. of stimulus, trial reactions of days second for per response tested series day response in sec. 35 1 1 11 13 2 2 10 12 3 3 2 3 4-10 4-10 0 0 Totals 10 23 Av. per response . 7.66 36 11 1 3 5 12 2 3 5 13 3 7 10 14 4 6 8 15 5 3 5 16 6 2 3 17 7 1 1 18 8 2 3 19 9 2 3 20 10 2 3 Totals 10 31 Av. per response. 3.1 37 21 1 8 11 22 2 6 8 23 3 3 5 24 4 8 10 25 5 2 4 26 6 5 8 27 7 6 9 28 8 7 10 29 9 5 7 30 10 3 5 Totals 10 53 Av. per response . 5.3 38 31 1 29 33 32 2 9 12 33 3 9 12 34 4 4 6 35 5 4 6 36 6 3 5 37 7 6 9 38 8 4 6 39 9 3 5 40 10 9 11 Totals 10 80 Av. per response. 8 AN ANALYSIS OF THE LEARNING PROCESS IN THE SNAIL 29 TABLE V— {Continued Snail No. 4 — {Continued) No. for "food and pressure" No. of No. of Duration No. of stimulus, trial reactions of davs second for per response tested senes day response in sec. 39 41 1 3 5 42 2 3 5 43 3 2 3 44 4 2 3 45 5 3 5 46 6 2 3 47 7 3 5 48 8 2 3 49 9 3 5 50 10 1 1 Totals 10 24 Av. per response . 2.4 Total number of trials 50 Total number of stimuli followed by response 43 Per cent of stimuli followed by response 86% Total number of reactions 211 Average number of reactions per trial 4 . 22 Average number of reactions in only those trials in which response occurred. 4 . 9 Total time of reactions in minutes 4.9 Average time per reaction in seconds 1 . 39 Average number of reactions per minute 43 . 16 Sn.\il No. 5 No. for "food and pressure" No. of No. of Duration No. of stimulus, trial reactions of davs second for per response tested series day response in sec. Totals 35 1-10 10 21 26 Av. per response . 4.2 Totals 36 11-20 10 37 52 Av. per response . 4.62 Totals 37 21-30 10 64 91 Av. per response. 6.4 Totals 38 31-40 10 35 73 Av. per response . 3.5 Total number of trials 40 Total number of stimuli followed by response 33 Per cent of stimuli followed by response 82 . 5% Total number of reactions 157 Average number of reactions per trial 3 . 92 Average number of reactions in only those trials in which response occurred. 4 . 75 Total time of reactions in minutes 4 . 33 30 ELIZABETH LOCKW'OOD THOMPSON TABLE V— {Continued) SXAIL No. 5 — {Continued) Average time per reaction in seconds 1 . 54 Average number of reactions per minute 38 . 96 Summary Snails 4 and 5: Total number of stimuli (trials) 90 Total number of stimuli followed by response 76 Per cent of stimuli followed by response 84 . 4% Table V shows that movements of the mouth occurred in 76 otit of 90 trials, with two individuals, or in 84.4% of the trials. The first series of tests in which the " conditioned " stimulus was used showed reactions occurring in 61 out of 100 trials or 61%, when food alone was used as a stimulus (Table II). These differences show the wearing away of the inhibition caused by the secondary stimulus (Table III) and by the experimental manipttlation. The mean number of reactions per response for snails 4 and 5 was 4.84, 0.91 less than for the normal food response and 1.01 more thati in the food-pressure series with trained snails. The inhibition of the response to food by the pressure stimultis has nearly disappeared (^adaptation). TABLE VI Showing a summary of the reactions of the mouths of four individuals of Pliysa gyrina Say, resulting from (A) food stimulus alone, cf. Table II; (B) food and pressure used simultaneously, cf. Table III; (C) pressure alone, cf. Table IV; (D) food and pressure used simultaneously, 2d series, cf. Table V. The tests were made in filtered tap water on alternating days. The snails had been starved for 24 hours preceding the tests. g^^g-j gnail Snail Snail 6 7 8 14 A. Stimulus Food — • Total number of trials 42 50 60 60 Total number of stimuli followed by response . . 8 38 47 32 Per cent of stimuli followed by reaction 19% 76% 78.3% 53.3% Total number of reactions 52 247 143 350 Av. number of reactions per trial 1 . 23 5.14 2 . 38 5 . 83 Av. number of reactions in only those trials in which response occurred 6.5 6.5 3.4 10 . 93 Total time of reactions in minutes 1.13 5.33 3.23 7.35 Av. time per reaction in seconds 1.3 1 . 29 1 . 39 1 . 25 Av. number of reactions per minute 46 . 15 46 . 51 44 . 44 48 B. Stimulus Food and Pressure Total number of trials 250 250 253 250 Total number of stimuli followed bv response .33 82 53 47 Per cent of stimuli followed by reaction 13 . 2%, 32 . 8%: 20 .9% 18 .8% Total number of reactions 65 337 84 182 A V. number of reactions per trial 0 . 26 1 . 34 0 . 03 0 . 72 Av. number of reactions in only those trials in which response occurred 1 . 96 4.1 1 . 58 3 . 87 Total time of reactions in minutes 1.2 6. 56 1.5 3.6 Av. time per reaction in seconds 1.1 1.16 1 . 07 1.18 Av. number of reactions per minute 65.45 51.72 56.7 50.84 AN ANALYSIS OF THE LEARNING PROCESS IN THE SNAIL 31 TABLE Yl—iContinued) Snail Snail Snail Snail 6 7 8 14 C. Stimulus Pressure -— — Total number of trials 30 20 17 20 Total number of stimuli followed bv response . . 13 11 3 5 Per cent of stimuli followed by reaction 43.3% 55% 15.5% 25% Total number of reactions 64 67 24 19 Av. number of reactions per trial 2 . 13 3 . 35 1.41 0 . 95 Av. number of reactions in only those trials in which response occurred 4.92 6 . 09 8 3.8 Total time of reactions in minutes 1 . 28 1 . 35 0.5 0 . 38 Av. time per reaction in seconds 1.2 1.2 1.25 1.21 Av. number of reactions per minute 64 . 16 64 . 16 48 49 . 54 D. Stimulus Food and Pressure (Second series) Total number of trials 50 50 50 50 Total number of stimuli followed bv response . . 16 36 44 25 Per cent of stimuli followed by reaction 32% 72% 88% 50% Total number of reactions 123 173 158 120 Av. number of reactions per trial 2 . 46 3 . 46 3.16 2.4 Av. number of reactions in only those trials in which response occurred 7 . 68 4.8 3 . 59 4.8 Total time of reactions in minutes 2 . 3l 3 . 78 3 . 28 2 . 53 Av. time per reaction in seconds 1.21 1.31 1 . 24 1 . 26 Av. number of reactions per minute 49 . 54 45 . 8 48 . 38 47 . 61 5. Summary of Observations 1. All observations and experiments were made on Physa gyrina Say while suspended from the surface film in a dish of filtered tap-water. In this position the lower surfaces of the foot and head are exposed and the movements of the mouth are visible. 2. When food (lettuce) or an inert substance (glass rod) is quickly applied to such a snail on or near the mouth and im- mediately withdrawn there results a response which consists of one or more opening and closing movements of the mouth. A single opening and closing of the mouth is referred to as a reaction. A response consists of one or many reactions in succession and takes place after the withdrawal of the stimulating substance. 3. Before the snails were used they were " tamed " b}^ handling until they no longer visibW reacted to the mechanical disturb- ances incidental to the experiments. 4. All experiments were condticted under approximately uni- form physical conditions, that is in filtered tap water at room temperature and with constant light intensity. 5. In all tests except those of the preliminary series each snail was used at intervals of fortv-eight hours. It was allowed 32 • ELIZABETH LOCKWOOD THOMPSON to feed freely during the first half of this period but was given no food during the second half of the period. By thus controlling the food an attempt was made to have all snails in the same physiological state at the time of using them. 6. As a me0,ns of detecting variations in the physiological state a record was kept of the egg masses deposited by each snail, but no relation was found between the egg laying activity and any results of the experiments. 7. Each snail was distinguished by a number painted on the shell and a separate record was kept for each. 8. After a preliminary series of experiments involving six snails (Nos. 18-23), six other snails were used in the remaining experiments. For two of these (4, 5) complete records are presented in Tables II, III, IV, V, pp. 8, 13, 25 and 28. For the remaining four (6, 7, 8, 14) a summar}^ of results is given in Table VI, p. 30. 9. Reactions of the mouth do not occur in the absence of ex- ternal stimulation (preliminary test series). Six snails (Nos. 18-23, Table I, p. 6) were kept under observation for a total period of 32.75 hours (an average of 5.47 hours each), without stimulation. During this time but eight responses occurred with a total of 38 reactions. Three of the responses were seen to follow contact of the mouth of the snail with parts of the mucous track left by the animal in its normal movements on the surface film. The remaining five responses probably fol- lowed similar contact with this nearly invisible mucus. For experimental purposes the mouth response may be said to occur only as the result of the external stimulation. It is " involun- tary." The error introduced into later experiments as a result of unobserved mucous stimulation is not enough to vitiate the experiments. 10. The normal food-rcspousc. — When food was applied to the mouth and at once withdrawn response followed in 61% of the tests (average for Snails 4 and 5). The average response consisted of 3.93 reactions. The average time of reaction was 1.25 sec. or 46.2 reactions per minute. Table II, p. 8. 11. The average number of reactions constituting the normal food-response varied with the individual snail as follows; No. 4, 3.48; No. 5, 4.35; Nos. 6 and 7, 6.5; No. 8, 3.4; No. 14. 10.93. The mean for all six snails is 5.86. AN ANALYSIS OF THE LEARNING PROCESS IN THE SNAIL 33 12. The average duration of a single reaction in the normal food-response is much more nearly the same in different indi- viduals than is the number of reactions constituting a response. This average in seconds for the six* snails in the order of paragraph 11 is, 1.29, 1.21, 1.3, 1.29, 1.39, 1.25. The values are obtained by dividing the duration of the response in seconds (as obtained with a stop watch) by the number of its component reactions. At the above rate the number of reactions per minute for the six snails would be 46.51, 45.9, 46.15, 46.51, 44.44, 48.0, or a mean of 46.25 reactions per minute. 13. An apparatus (food-pressure apparatus, fig. 1, Plate I) was devised for the simultaneous application of two unlike stimuli, food to the mouth and pressure to the foot at a fixed distance from the mouth. 14. Pressure response of untrained snails. — By means of the food-pressure apparatus pressure was applied to the foot at a fixed distance from the mouth. No stimulus was applied to any other part of the animal. In one hundred twenty applica- tions of pressure to six snails (20 each) but four responses were obtained. Response thus occurred to C).^>^yc of the applications of pressure. The responses were probably due to chance stimu- lation by mucus. Response to pressure on the foot with the apparatus used occurs in so small a per cent of cases that it does not affect the experiments summarized below. 15. Food-pressure response. — When food was applied to the mouth and pressure to the foot at the same instant no response followed in the first 60 to 110 trials, in six series of 250 trials each (Snails 4, 5, 6, 7, 8 and 14). Response followed in the remaining trials of each series. In a total of 376 responses for the six snails, the average number of reactions per response was 3.42 as compared with 5.86 with the same snails when the food stimulus alone was used (Tables III and VI-B, pp. 13 and 30). The snails are referred to as " trained." 16. Pressure response of trained snails. — Forty-eight hours after the completion of their training, pressure alone was applied to the foot of the six trained snails (paragraph 15). In the case of Snail No. 4 response followed the first nine trials. The last two of these were separated by an interval of 48 hours from the first seven. The effect of training thus persisted for 96 34 ELIZABETH LOCKWOOD THOMPSON hours. Similar results were obtained from vSnails 5, 6, 7, 8 and 14 (Tables IV and VI-C, pp. 25 and 31). 17. Cessation of pressure response in trained snails. — Imme- diately following the ninth response to pressure of trained Snail No. 4 (paragraph 16) seven successive applications of pressure gave no response ; 48 hours later ten applications gave no response ; 96 hours after the last response ten applications failed to give any response. Similar results were obtained with the other five snails. Cessation of respotise to pressure after training is sudden and final (Tables IV and VI-C, pp. 25 and 31. 18. One hundred and twenty hours after the last response to pressure of Snail No. 4 (paragraph 16) a scries of 50 trials was run on this snail by the simultaneous application of food and pressure. Response followed the first three trials and occurred in 86% of the subsequent trials (cf. paragraph 15), with an average of 4.9 reactions. Similar results were obtained w^ith Snail No. 5. The average number of reactions per response was 4.84 for the two snails, 0.91 less than for the normal food re- sponse (paragraph 10) and 1.01 more than in the first food- pressure series (paragraph 15). The interference of the pres- sure stimulus with the response to food has nearly disappeared (Tables IV and VI-B, pp. 25 and 30). 19. The average number of reactions in the food-pressure response is less than the number in the normal food response. The numbers for the six snails in their numerical order is 3.96, 3.62, 1.96 (Table III, p. 13), 4.1, 1.58 and 3.87 (Table VI-B, p. 30). The average for the six snails is 3.18. Application of pressure to the foot at the same time with food to the mouth shortens the response. 20. The average duration of the reaction is less in the food- pressure responses than in the normal food response. The num- ber per minute for the six snails in numerical order is 51.84, 51.92, 65.45, 51.72, 56.7, 50.84 (Tables III and VI-B, pp. 13 and 30). This is a mean distribution of 54.74, or 8.49 reactions per minute in excess of the normal food response (paragraph 12). 21. In the food-pressure series, the average number of reactions per ten trials increases irregularly toward the end of the series, except in the case of Snail No. 4. The average number of responses per ten trials shows a more uniform rise (Table III-A, p. 19, figs. 2-7, Plate II). AN ANALYSIS OF THE LEARNING PROCESS IN THE SNAIL 35 22. The average number of reactions per response resulting from the appHcation of pressure alone to the foot of trained snails (paragraph 16) is greater than the average number result- ing from the application of either food or food-pressure stimula- tion. The numbers for the six snails in their numerical order is 1.25, 4.0, 4.92, 6.09, 8.0, 3.8, or a mean of 4.67 reactions per response. These averages differ more than those of either food or food-pressure responses (Tables II, III and VI-A and VI-B, pp. 8, 13 and 30). 23. The average duration of the reactions resulting from the pressure stimulus alone on the foot of the trained snails is less than that of the normal food response. This is shown by the average number of reactions per minute for the six snails, 53.9, which is 7.65 more than with the normal food response (Tables IV and VI-C, pp. 25 and 31). In all trials the -reaction duration is notably constant. 6. Discussion A. Method. — The experimental method used in this research will be called the method of training by the use of simultaneous stimuli. The term may be applied to any procedure in which two unlike stimuli are applied together or in immediate succes- sion, to one of which the animal has previously responded by a definite reflex, but to the other of which it has failed to respond b}^ the same reflex. The purpose is to establish response to the second stimulus alone. The stimuli may be referred to as prim_ary and secondary and the primary stimulus may be called " conditioned " when used with the secondary (Pawlow's terms). The term, as thus defined, includes the salivary reflex method of Pawlow. Comparison of the method with the salivary reflex method. The Pawlow^ method has been described (p. 2). It requires that the responses to the primary stimulus be a reflex, a so-called involun- tary response, and that it be measurable. In order to meet the first requirement the response must, (a) occur only in response to external stimulation, (b) " always " occur in response to exter- nal stimulation; that is, it should occur with sufficient regu- larit}^ to be available for experimental uses. That the mouth response of Physa occurs only in response to stimulation of the mouth region has been shown. In six animals carefully guarded from such stimulation for a total 36 • ELIZABETH LOCKWOOD THOMPSON period of nearly thirty-three hours (about five and a half hours to each animal) the response occurred but eight times. Three of these responses were due to observed accidental stimulation and the rest were doubtless due to the same cause. On the other hand, as the tables show, stimulation of the mouth region induces the response in a sufficient percentage of trials to make it avail- able for experimental purposes. The response is therefore a reflex available for these purposes. The response of Physa to stimulation of the mouth region unaccompanied by other stimulation will be referred to as the normal food response. It consists of one or more opening and closing movements of the mouth. Each complete opening and closing movement is spoken of as a reaction. The response is measurable in terms of the number of its reactions as well as in duration. An inspection of the tables shows ver>'^ great fluctua- tion in the numerical value of the response. B. Probable causes of fluctuation in response are not difficult to find. They arise in part from the accidents of manipulation and are in part due to the physiological state of the snails. It is not an easy matter to apply a stimulus to the mouth of a moving snail, delicately suspended from the surface film. The difficulty is increased when two stimuli must be accurately placed at the same time. If the pressure is applied with the pressure fork strictly vertical, it is only that of the weight of the movable part of the apparatus and should be constant. wSlight deviation from the vertical is inevitable and must result in occasional variation in pressure and in the manner in which the food reaches the mouth. These accidents should shorten response. That unskilled manipulation sometimes inhibits response entirely is clear from those cases marked x in the tables. It is fair to assume that they partially inhibit other responses. They depend in part on the phy.siological state of the operator. Added to these personal factors the tendencies to produce irregularity in response are vibrations of an unstable building due to slamming of doors, passing of streets cars and movements of people. Other accidents prolong response. Probably at times portions of food are retained in the mouth of the snail and the response may be prolonged. At times accidental stimulation from unobserved miucus secreted by the snail, is added to that of the food on the hook and response is prolonged. AN ANALYSIS OF THE LEARNING PROCESS IN THE SNAIL 37 That the physiological state of the animal is in part responsible for fluctuations in response is indicated by the no-response shown in Tables II and III, pp. 8 and 13. Some of these were known to be due to mechanical disturbance and are marked x in the tables. Others were probably due to disturbance. Others again may be attributed to the physiological state of fatigue, although It is believed that the method of testing the snails alternately practically eliminated this element. It is noteworthy that a large percentage of the no-response trials occurs at the end of the day's series. This may be attributed to fatigue or adapta- tion— in either case a changed physiological state. With a more delicate apparatus, and particularly^ with the elimination of external disturbances such as the vibrations caused by sudden jars beyond the limits of the experimental room, responses of more constant value might be obtained. It is quite possible that responses obtained by more refined methods would equal those of the salivar}^ reflex method in constancy. Throughout each of the tables it is to be noted that the duration of the individual reaction is reasonably uniform. It fluctuates mainly between 1.0 and 1.25 seconds. The fluctuations may be attributed to the method of timing with a stop watch. Were a better method available for the purpose, much greater uni- formity in the reaction duration would doubtless appear. The uniformity actually found is striking and indicates that the use of the single reaction as a unit for measuring the response is justified. The unit is probably as accurate as the drop of saliva. That the method described in this paper would yield results if used in the study of discrimination in the snails the writer does not doubt. It seems to oft'er for this purpose whatever advantage the Pawlow method affords although possibly in a less degree. It is also available in working out many of the details of the association process in a statistical way. The data' presented in this paper show that the behavior of the snail is modified by two processes, adaptation and learning. Adaptation appears throughout the experiments and needs no extended discussion. It is enough to cite the following in- stances. Others will doubtless appear from a study of the tables. (a) The snails became " tame." When first brought into the laborator}^ they expel the air from the lung and drop from 38 ELIZABETH LOCK WOOD THOMPSON the surface film when in the least disturbed. After a time they no longer drop from the surface film under any moderate stimu- lation. To secure the expulsion of air from the lungs it is then necessary to use mechanical means. (b) The snails recover from the inhibition of the mouth response due to manipulation. This appears in Table III, p. 13. Here inhibition is more marked in the first 60-80 trials, after which the snails recover and the number of reactions per response increases as the trials proceed until it reaches a m.aximum toward the middle of the series. (c) A waning of response occurs in the series of trials shown in Table III, p. 13. Following a maximum number of reactions per response toward the middle of the series, the number grad- ually diminishes to the end of the series (figs. 2-7, Plate II). The snails are becoming adapted to the stimulus which is not fol- lowed by its wonted reward. The same thing occurs in the pressure response of trained snails shown in Table IV, p. 25. Here response is entirely lacking toward the end of the series. (d) The absence of the response in the final trials on many days' series appears in Table III, p. 13. This may be due to fatigue, but is more likely the result of adaptation. Other cases of adaptation in the snail appear in the paper of Dawson (1911) and in the work of Nagel (1894). Learning, modifiahility by association. vSnails which gave no mouth response to pressure on the foot, were so modified by the simultaneous application of pressure to the foot and food to the mouth, that they then gave the mouth response to pressure on the foot. The effect of training with the simultaneous stimuli persisted for ninety-six hours after the cessation of training. Ob- jectively this persistence of the training effect simulates memory. Its ps^^chological implication is not discussed here. The training process was accompanied by a shortening of the average duration of the individual reactions, in other words in an increase in their rate, as compared with the r^te obtaining when the food stimulus alone was used. This acceleration is no doubt due to the unwonted stimulus applied to the foot and may be taken as evidence (if any is needed) that this stimulus was effective. The training process was accompanied by a reduction in the average number of reactions per response and AX ANALYSIS OF THE LEARXIXG PROCESS IX THE SXAIL 39 this again may be attributed to the unwonted foot-stimulus and may be accepted as evidence of its effectiveness. Notable in the training series of trials of Snails 4 and 5 is the increase and subsequent decline in the average number of reac- tions per response after the initial inhibition. The average number of reactions per ten trials rises to a maximum in the neighborhood of the 150th trial and then declines. An examina- tion of the graphs for these two snails (figs. 2, 3, 8 and 9, Plates II-III) suggests that the training has produced its maximum effect in the neighborhood of the 150th trial. If this were true a longer response to pressure only should be obtained from snails that had been trained for 150 trials than from those of longer training and this response might persist for a longer period. Whether this is actually the case could only be determined by comparing the response to pressure alone in a considerable num- ber of snails with training periods of different lengths. A com- parison of the graphs for Snails 4 and 5 with those for 6, 7, 8 and 14 shows that the latter are quite irregular. They show numerous maxima one of which commonly lies in the region of trial 150, but they do not support the view of a maximum train- ing effect at a particular region in the series of trials. Thus an interpretation that seems wholly tenable from a study of two snails becomes untenable when the study is extended to a larger number of individuals. The maxima in the graphs for Snails 4 and 5 probably have their explanation in some undiscovered varia- tion in the physiological state of these individuals. That taken by itself, it is interpretable in terms of learning may serv^e as a warning against generalizations based on experiments on one or two individuals. Nevertheless the relation of length of training and training effect in the snail may be worthy of further investigation. II. MODIFIABILITY STUDIED BY THE LABYRINTH METHOD 1. Introduction The evidence of learning in the snail, PJiysa gyrina Say, obtained by the use of the method of simultaneous stimuli suggested the query whether the snail could solve a simple labyrinth. Failure to solve the labyrinth would be evidence that this requires a capacity beyond that of forming simple associations. The labyrinth method has been used extensively 40 ELIZABETH LOCKWOOD THOMPSON in nearly all groups of vertebrates, less extensively among invertebrates. Notable among the work on the latter group is that of Yerkes (1902) on the crayfish Carcinus granulatus, Yerkes and Huggins (1903) on habit formation in the crayfish Camharus af finis, Fielde (1901) and Shepard (1916) on ants, and Yerkes (1912) on earthworms. Just as the natural habits of rats in running about in narrow horizontal passages suggested to Small (1899) a series of such passages for experimental work, so the structure and habits of Physa suggested an upright or vertical path. Physa breathes air, but remains under water indefinitely when the so-called lung is full of air. When a new sup|:»ly of oxygen is required it commonly seeks the surface of the water, although bubbles below the surface are sometimes " tapped " for air by the snail (Dawson, 1911). In order to gain the surface it ordin- arily crawls up the stems of water plants growing in its natural habitat. It seemed then as though the tests might be placed on a purely natural basis by mechanically depriving the snail of air and placing it at the base of an upright path leading to the surface of the water. The animal's need for air would furnish the motive necessary to action. This need also suggested a natural punishment and reward for a correct or incorrect choice of paths, that is, the attainment of air at the end of the path or failure to get it. 2. Apparatus and Material* The labyrinth used had the form of a rectangular U on an upright stem (fig. 14, Plate IV). One arm of the U reached the surface of the water while the other was so much shorter that the snail could not reach the surface from its tip. Various materials were tried out in the construction of the apparatus, such as cardboard coated with paraffin to protect it from the water, metal, mica and celluloid. The latter was finally selected because the surface of this material is like that of glass upon which snails crawl readily and also because it could be easily cut and manipulated should it be decided to alter the form of the path. Transparent celluloid 2.5 mm. in thickness was used in constructing the path (fig. 14, Plate IV). The upright * The author is indebted to Professor George R. La Rue for advice in the mechanical construction of the labyrinth. AN ANALYSIS OF THE LEARNING PROCESS IN THE SNAIL 41 arms A and B were 8.5 mm. by 4.9 cm., the base of the U, C was 8.5 mm. by 4.25 cm. This rested on an upright stem D, 8.5 mm. by 3.6 cm., the base of which was cemented into a block of lead E, 2.1 cm. by 1.6 cm. by 6 mm. The celluloid part of the apparatus was carefully squared to the lead base E, before the cement was allowed to harden. Physa in crawling up anything but a broad surface has a tendency to crawl in a spiral direction, as may be readily demonstrated by allowing it to crawl up a glass rod placed upright in the aquarium. Should the animal ascend the stem of the labyrinth in a spiral path it would have the long or correct arm sometimes on its right and sometimes on its left. In order that the snail might have a choice of right and left paths on the labyrinth it was necessary to keep it confined to one side of the apparatus. In an attempt to accomplish this, the celluloid path was backed with galvanized, 0.5 inch-mesh wire netting, so cut that the entire path was bordered by a " picket fence " of wire, the pickets extending outward in the same plane as the path rather than at right angles to it. By cutting out every other " picket " the re- maining wires were just far enough apart to make it mechanically impossible for the snail to pull its shell between them. This device proved an effective but not an absolute barrier, inas- much as the animal occasionally managed to crawl over the end of a " picket," but the number of crossings became so reduced by its use, that this source of error was practically eliminated. The wire was cemented to the path by means of celluloid dis- solved in acetone. The wires themselves and the lead base were also coated with this material, so that all parts of the completed apparatus with which the snail could come in contact were either made of celluloid or coated with it. This apparatus when in use was placed in the exact center of the experimental tank, thus bringing it directly under the centered Hght, so that the rays falling on each branch of the path A and B (fig. 14, Plate IV) were of equal intensity and thus the possibility of pho- totropism directing the course of the snail was avoided. Tap water at room temperature was used in the experimental tank. This temperature varied within the limits of a degree or two but it will appear later in the records of the experiments that the variations in temperature were the cause of no irregularities in the results. 42 ELIZABETH LOCKWOOD THOMPSON In order to conduct the experiments under controlled lighting conditions and thus avoid any possible phototropic influence it was deemed advisable to create special conditions governing the light and keep them uniforrri throughout the experiments (fig. 15, Plate V). An experimental tank consisting of a battery jar 6 inches high and 4 inches in diameter was painted black on the outside and a level floor was obtained by means of a false bottom made of weighted paraffin. The center of the bot- tom was marked in the paraffin. This tank, during experimenta- tion, was placed in the center of a " house " constructed from gray beaver-board fastened together with strips of gummed paper and painted dead black on the inside. The walls were 26 cm. by vS8 cm. with a strip 19 cm. by 5 cm. cut from the lower edge on three sides for ventilation. For this purpose also small holes uniformly spaced were punched 2 cm. from the upper edge of the four sides. The fourth wall, the one next the operator, had a horizontal slit 2 cm. wide cut in it at 10 cm. from the top. The part of this wall below the slit was made removable. With the lower part of the fourth wall removed the operator could prepare experiments without taking the tank from the " house." When the movable part of the fourth wall was re- placed there remained a slit 2 cm. in width, through which observations could be made. The top of the house was ground glass above which at a distance of 4 cm. was an 8 candle power electric light. This was centered over the house and held in a stationary position l)y means of a wooden frame. Two thick- nesses of ordinary medium weight white paper were placed on top of the ground glass to further weaken and diffuse the light. The room in which the apparatus was located was darkened during experimentation. This apparatus was used without change throughout the experiments. The labyrinth itself was varied to conform with the requirements of each set of experi- ments. All snails used were tamed, individually numbered and kept under the same conditions as in the experiments with simultaneous stimuli . 3. Experiments on Preference Two different kinds of preference tests were carried out. The first was to find whether a tendency to turn to the right or left existed. The asymmetry of the animal suggested the possibility. AN ANALYSIS OF THE LEARNING PROCESS IN THE SNAIL 43 The second test was to find any possible preference for an ascending or descending path. A. Experiments to test right or left preference. In the first series an unpainted glass dish of beaker type, containing tap water at room temperature was placed in the centre of the experimental " house " and the ventilating spaces at the bottom of the walls of the " house " were blocked to exclude light. The air was forced from the lung of the snail by pressing its body gently back into its shell. It was then placed against the side of the dish at the four points of the compass in turn. It was allowed in each case to crawl up the side of the dish for air. The path was sketched and its right or left divergence from a perpendicular noted. The snails were handled with forceps and always placed on the horizontal bottom of the dish at right angles to the surface to be traveled. The mucous trail left on the sides of the dish was broken up after each trial by brushing the traveled surface with a clean camel's hair brush. Three snails, Nos. 15, 16 and 17, were used and four tests given each on two consecutive da3'S. The record follows. TABLE VII Showing the deviation from a perpendicular shown by three individuals of Physa gyrina Say, in following an unconfined path up the side of a cylindrical glass dish in search of air. Trial Snail Snail Snail No. 15 16 17 First day 1 Left Right Right 2 Right Left Right 3 Left Right Right 4 Left Left Right Second dav 1 Left Left Right 2 Left Left Left 3 Right Right Right 4 Right Left Right Total 5L 3R 5L 3R IL TR Total number of trials 24 Total number of deviations to the right 13 approximately 54 Sc Total number of deviations to the left 11 approximately 4o% This experiment was based upon the normal behavior of the snail in its search for air; its path from the bottom of the dish to the surface was not limited in any way. Besides the record just given, three points of general interest were observed. 44 ELIZABETH LOCKWOOD THOMPSON 1. At no time during these experiments did the snail appear to make any effort to obtain air except at the surface of the water. Inasmuch as the depth of the water in the experimental tank was greater than that in the home aquarium, the observa- tions conflict with those of Dawson (1911) which refer, how- ever, to young snails. 2. The touching of the tentacles to the surface of the dish, at intervals, on either side of the path of the snail appeared in no case to alter the direction of the path. 3. There appeared to be no effort on the part of the snail under observation in these experiments, to re-distribute its weight by shifting the shell from side to side. Such shifting of the shell might exert a pull upon the body of the animal, strong enough to turn it in one direction more than another. In the second scries of preference tests the U-shaped labyrinth was used. Both of the upright arms reached the surface of the water so that a choice of either did not result in punishment. Six snails were used. They were divided into two groups of three each and tested on alternate days. This division was made because six snails could not be worked each day during the time at the disposal of the operator, and it was deemed advisable to have records from more then three snails. Each snail in each group of three was given ten trials per day. Enough fresh tap water at room temperature was placed in the experi- mental tank to reach the top of the arms of the U apparatus resting in the center of the dish. The snail was taken from the bacteria dish, in which it was kept at room temperature, and held in the hand while the body was gently pressed back into the shell by means of a dull metal seeker. In this way the snail was compelled to expel air from its lung. The animal, now negatively geotropic, was placed on the base of the U, E (fig. 14, Plate IV) by means of long forceps. The apex of the shell rested on base E while the opening of the shell touched the stem D. Thus when the snail came out of its shell, which it usually did almost immediately, it at once started to crawl up the stem D to the surface. At the top of D it turned either to the right or left on the cross bar C and then moved up either arm A or B to the surface of the water where it was permitted to obtain air. It was then lifted off with the hand as gently as possible and returned to its individual di.sh where it was AN ANALYSIS OF THE LEARNING PROCESS IN THE SNAIL 45 allowed to feed undisturbed while the other two snails of the group were tested in a similar way. After each test the celluloid apparatus was removed to a separate dish of water and thoroughly cleaned with a stiff brush in order to remove the mucous trail left by the tested snail and thus make it impossible for the next snail to follow the trail. A complete record was made for each test given to each individual snail. Yerkes' method (1912), modified to suit the experiments, was adopted in making the records. Outlines of the labyrinth were drawn on separate slips of paper and the course chosen by each snail traced on one of these. The time required for the trip was taken with a stop-watch. The additional data of date, time of day, and temperature were added to the slip, and a record was also made of the eggs, if any, which had been deposited within twenty-four hours previous to the tests. It was thought that a later analysis of the records might show some relation between the physiological condition of the animal caused by ovulation, and the results of the experiments. The following table shows the results obtained from a series of forty tests on each of six snails. . TABLE VIII Showing the right and left preference of six individuals of Physa gyrina Say, which had been deprived of air and placed at the base of a vertical U-shaped path resting on an upright stem. Both arms of the U reached the surface. The horizontal of the U necessitated a choice of paths (fig. 14, Plate IV) No. of No. of No. of No. of Snail days trials left right No. tested per day choices choices Eggs 15 1 10 8 2 0 16 10 7 3 0 17 10 7 3 0 24 2 10 2 8 0 25 10 1 9 1 26 10 2 8 1 15 3 10 5 5 0 16 10 5 5 1 17 , 10 8 2 1 24 4 10 3 7 0 25 10 5 5 0 26 10 4 6 0 15 5 10 6 4 0 16 10 1 9 0 17 10 5 5 0 46 ELIZABETH LOCKWOOD THOMPSON TABLE VIII— Continued No. of No. of No. of No. of Snail days trials left right No. tested per day choices choices Eggs 24 6 10 3 7 1 25 10 2 8 1 26 10 3 7 1 15 7 10 8 2 0 16 10 6 4 0 17 10 9 1 1 Total number of trials 240 Total number of left turns 116 approximately 48% Total number of right turns 124 approximately 52% B. Experiments to test prejerence for an ascending or descending ■path. If the cross bar of the U should not be horizontal, the snail after ascending the stem would have a choice of a descending path on one side and an ascending path on the other. In a set of preference tests an attempt was made to determine whether the animal showed preference for either of these paths. In order to tilt the labyrinth at various angles the experimental tank was placed on a small board, 8 1-2 in. by 10 in. by 1-4 in., made of three layers to insure an unwarped surface. One end of the board was pierced by two screw eyes set one inch from the edge. By turning these screws the board, together with the apparatus resting upon it, could be readily adjusted to any angle. The angle used in each experiment was measured with a Sterret spirit level. The slant of the right and left sides of the path was frequently and irregularly reversed in order to avoid any possible right and left habit formation during experi- mentation. These tests were performed in the same way as those already described and the records were similar. The tabulated results follow. AN ANALYSIS OF THE LEARNING PROCESS IN THE SNAIL 47 TABLE IX Showing the tendency of six individuals of Physa gyrina Say to follow an ascending or a descending path when the apparatus was tilted. The U-shaped apparatus was used and the angle made by the cross bar of the U with a horizontal measure. Both arms of the U reached the surface. Snail Angle of inclination to horizontal of cross bar of U-shaped No. labyrinth 0° -1° l°-2° 3°-4° 5°-6= 7o_8o up down up down up down up down up down 15 37 23 3 7 7 3 5 5 8 2 16 23 27 2 8 4 6 5 5 6 4 17 38 22 6 4 5 5 6 4 6 4 24 36 24 6 4 5 5 6 4 10 0 25 32 28 4 6 4 6 7 3 8 2 26 32 28 4 6 6 4 6 4 5 5 Totals . 208 152 25 35 31 29 35 29 43 17 58% 42% 42% 58% 52% 48% 58% 42% 72% 28% Table IX shows that when six individuals were given a choice of an ascending or a descending path, both of which ended at the surface, and when the angle of deviation from a horizontal at the point of selection was not greater than 8 degrees, the ascending path was chosen in 342 trials out of 600 or in 57% of the trials. There appears to be a strong tendency to follow the upward path only at the angle 7-8 degrees. In further experiments the cross bar of the U was kept horizontal. 4. Experiments on Learning the U-shaped Labyrinth; Choice of Right and Left Path; Punishment Failure to Get Air Inasmuch as the snails showed a slight tendency to turn toward the right in the preference tests, the right arm of the U-shaped path was now shortened by cutting 2 cm. from the upper end. A choice of this arm would lead to punishment (failure to get air) and would be a wrong choice. When a wrong choice was made the snail had to retrace its path to the point of choice where it usually continued its course along the horizontal bar and up the long arm to the surface of the water. This was the only modification made in the apparatus for the learning experiments. The method of handling the animals and the apparatus was the same as in the preceding tests. Three snails were used in these experiments. They were at first tested every other day. The three snails were worked in turn and as many trips given each as was possible in the time at the disposal of 48 .ELIZABETH LOCK^YOOD THOMPSON the experimenter. The number of trips per day varied con- siderably because it was found that sometimes individual snails refused to start up the labyrinth. At such times they were frequently allowed to remain at the base of the stem for a considerable length of time in the hope that the tests might be continued in regular order. Sometimes, however, work on the individual snail so affected had to be discontinued for the day. When a snail lost its hold on the labyrinth and dropped to the bottom of the tank during its test, it was immediately started up again and this second trip was given the same serial number as the previous one with a small letter after it. The same method of recording the observations was used as that described for the preference tests. A record was also made of the " latent " period, that is, the time between the placing of the animal at the base of the stem and the actual start on the upward path. Although not noted in the tabulations, this record appears to have no bearing on the results of the experiments, since fluctuations were probably due in part to the fact that the snail still retained a small amount of air in its lung, so that its need for a fresh supply was less urgent. In constructing curves based upon these records the percentage of errors or wrong choices for each snail per day was used. An analysis of the records appears to show no relation between the latent periods, oviposition, the time required for the tri]j and the number of errors. TABLE X Showing the paths followed by three individuals of Physa gyrina Say in attempt- ing to reach the surface of the water for air. U-shaped labyrinth used. The right arm was so short that the snail could not obtain air by following it (punish- ment), while air might be obtained at the top of the left arm (reward). The apparatus was cleaned between trials. Sn.'^il No. 1 No. of No. of No. of No. of No. of trial trial correct incorrect days for for or left or right tested series day choices choices *1 1 1 . 1 0 2 2 1 0 3 3 0 1 4 4 1 0 5 5 1 0 6 6 1 0 tals.... 6 5 (83%) 1 (16%) AN ANALYSIS OF THE LEARNING PROCESS IN THE SNAIL 49 No. of days tested Totals . TABLE X — (Continued) Snail No. 1 — (Continued) No. of trial for series 7 8 9 10 11 12 13 14 Totals . 15 16 17 18 19 20 21 No. of trial for dav No. of correct or left choices (75%) No. of incorrect or right choices (25%) (85%) 0 0 0 0 0 0 1 1 (14%) Totals . 22 23 24 25 26 27 28 Totals . 29 30 31 32 33 34 1 2 2a 2b 3 4 5 7 I 2 3 4 5 5a 6 (85%) (14%) (83%) 0 0 0 0 1 0 1 (16%) Totals . 35 36 37 38 39 40 41 42 1 2 3 4 4a 4b 4c 5 8 (50%) 0 0 0 1 1 1 0 1 4 (50%) 50 ELIZABETH LOCKWOOD THOMPSON TABLE X— {Continued) No. of days tested Snail No. 1— {Continued) No. of No. of No. of trial trial correct for for or left series day choices Totals. 43 44 45 46 47 48 49 50 1 2 3 4 5 6 6a 6b 1 0 1 1 1 0 0 1 5 (62%) No. of incorrect or right choices (37%) Totals. 51 52 53 54 55 56 57 (71%) (28%) Totals . 58 59 60 61 62 1 1 1 1 1 5 (100%) 10 Totals. 63 64 65 66 67 (60%) (40%) ai Totals. 68 69 70 71 72 73 74 1 2 3 4 5 6 6a 7 (57% 0 0 0 1 1 1 0 3 (42%) n2 Totals . 75 76 77 78 79 80 81 (28%) 1 1 1 1 0 1 0 5 (71%) AX ANALYSIS OF THE LEARXIXG PROCESS IN THE SNAIL 51 No. of days tested 13 Totals. *14 Totals. •■IS Totals . 16 Totals . 47 Totals . 18 TABLE 1^— {Continued) Snail No. 1 — {Continued) No. of No. of No. of trial trial correct for for or left series day choices 82 83 84 85 86 87 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 ToUls. 119 120 121 122 123 1 2 3 4 5 6 7 7 1 2 3 3a 3b 3c 3d 3e 4 5 6 11 1 2 3 3a 4 1 2 3 4 5 6 7 7a 8 1 2 3 4 5 6 7 7 1 2 3 4 5 5 (28%) (9%) (25%) (75%) (57%) No. of incorrect or right choices (71%) (60%) 0 1 1 1 1 1 1 1 1 1 1 10 (90%) 1 1 1 0 3 (75%) 0 0 0 0 0 1 1 0 2 (25%) 0 1 0 1 1 0 0 3 (42%) 0 0 0 1 1 2 (40%) 52 ELIZABETH LOCKWOOD THOMPSON TABLE X— {Continued) No. of days tested 19 Snail No. 1 — (Continued) No. of trial for series 124 125 126 127 128 No. of trial for day Totals . No. of correct or left choices 0 0 0 - 0 1 1 (20%) No. of incorrect or right choices 1 1 1 1 0 4 (80%) 20 129 130 131 132 133 134 Totals . 1 la 2 3 4 5 6 0 1 1 0 1 1 4 (66%) 1 0 0 1 0 0 2 (33%) *21 135 136 137 138 139 140 Totals. 1 2 2a 3 4 5 6 (66%) (33%) *22 141 142 143 144 145 146 Totals . 1 2 3 4 5 5a 6 (83%) (16%) 23 Totals. 24 147 148 149 150 151 152 153 154 155 156 157 158 159 160 Totals. 1 2 3 4 5 6 7 7a 1 2 3 3a 3b 4 6 1 0 1 1 1 1 0 1 6 (75%) 0 0 0 0 1 0 1 (16%) (25%) (83%) AN ANALYSIS OF THE LEARNING PROCESS IN THE SNAIL 53 No. of days tested 25 Totals . 26 Totals . *27 Totals . 28 Totals . 29 Totals. 30 Totals . 31 Totals. 32 TABLE y.— {Continued) Snail No. 1 — {Continued) No. of trial for series 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 Totals . 193 194 195 196 197 No. of trial for dav 1 2 3 4 5 5 1 2 3 3a 3b 5 1 2 3 3a 4 5 1 2 3 4 5 5 1 2 3 4 4 1 2 3 3 1 2 3 4 5 5 1 2 3 4 5 5 No. of correct or left choices 0 0 0 0 0 0 0 1 0 0 1 2 (40%) 1 0 0 0 0 1 (20%) 0 0 0 0 0 0 0 1 0 0 1 (25%) 0 0 0 0 (60%) 0 0 1 1 1 3 (60%) No. of incorrect or right choices 1 1 1 1 1 5 (100%) (60%) (80%) 0 1 1 1 1 4 1 1 1 1 1 0 (100%) 1 0 1 1 3 (75%) 1 1 1 3 (100%) 0 0 0 1 1 2 (40%) 1 1 0 0 0 2 (40%) 54 ELIZABETH LOCKWOOD THOMPSON TABLE X— (Continued) No. of days tested *33 Snail No. l^ [Continued No. of trial for series 198 199 200 201 202 No. of trial for day Totals . 1 la 2 3 4 5 No. of correct or left choices No. of incorrect or right choices 1 1 1 1 1 5 (100%) 34 Totals . 203 204 205 206 207 1 la 2 2a 3 5 (20%) (80)% 35 Totals . 208 209 210 211 1 0 1 1 3 (75%) 0 1 0 0 1 (25%) *36 Totals . 212 213 214 215 216 (40%) (60%) 37 Totals . 217 218 219 220 221 1 2 2a 3 4 5 (20%) (80%) *38 Totals . 222 223 224 225 226 1 la 2 3 4 5 0 1 1 0 0 2 (40%) (60%) 39 Totals. 227 228 229 230 231 1 2 3 3a 4 5 1 0 0 1 1 3 (60%) (40%) AN ANALYSIS OF THE LEAR-NING PROCESS IN THE SNAIL 55 No. of days tested 40 Totals. 41 Totals. 42 Totals. *43 Totals . 44 Totals . 45 Totals. *46 TABLE X— {Continued) Snail No. 1 — {Continued No. of trial for series 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 Totals . 263 264 265 266 267 268 269 270 No. of trial for day 1 2 3 3a 4 5 1 2 3 4 5 5 1 2 3 3 1 2 3 3a 4 5 6 6a No. of correct or left choices (40%) (40%) (40%) (60%)' 1 1 1 0 0 1 0 0 4 (50%) 0 1 0 1 (33%) 0 1 0 0 0 0 0 0 1 (12%) No. of incorrect or right choices (60%) (60%) (60%) (40%) (50%) 1 0 1 2 (66%) 1 0 1 1 1 1 1 1 7 (87%) 56 ELIZABETH LOCKWOOD THOMPSON No. of days tested 47 Totals . 48 TABLE X— {Continued) Snail No. 1 — {Continued) No. of No. of No. of trial trial correct for for or left series day choices 271 272 273 274 275 276 Totals . 277 278 279 280 281 282 1 2 2a 2b 3 4 6 1 2 3 4 5 6 6 1 0 0 0 0 0 1 (16%) 1 0 0 1 1 1 4 {66%) No. of incorrect or right choices (83%) (33%) Summary — Snail No. 1 No. of Correct Incorrect days No. of or left or right tested trials choices choices *1 *2 *3 *4 5 *6 7 8 *q 10 ni *12 13 *14 ns 16 m 18 19 20 *21 22 23 24 25 26 *27 28 29 30 7 5 5 7 7 7 11 4 8 7 5 5 6 6 6 8 6 5 5 5 5 4 3 1 2 1 1 1 4 3 2 0 2 3 5 5 10 3 2 3 2 4 2 2 1 2 5 5 3 4 5 3 3 AN ANALYSIS OF THE LEARNING PROCESS IN THE SNAIL 57 TABLE X— {Continued) Summary — Snail No. 1 — {Continued) No. of Correct Incorrect days No. of or left or right tested trials choices choices 31 5 3 2 32 5 3 2 *33 5 0 5 34 5 1 4 35 4 3 1 *36 5 2 3 37 5 1 4 *38 5 2 3 39 5 3 2 40 5 2 3 41 5 2 3 42 5 9 3 *43 5 3 2 44 8 4 4 45 3 1 2 *46 8 1 7 47 6 1 5 48 6 4 9 Total number of trials 282 Total number of correct or left choices 136 approximately 48% Total number of incorrect or right choices 146 approximately 52% Summary — Snail No. 2 No. of Correct Incorrect days No. of or left or right tested trials choices choices 1 5 4 1 *2 8 8 0 *3 7 3 4 *4 5 3 2 5 6 2 4 *6 5 5 0 7 6 3 2 8 6 3 3 *9 5 2 3 10 6 1 5 *11 9 3 6 *12 8 4 4 13 7 4 3 14 7 4 3 *15 3 2 1 16 9 4 5 17 7 4 3 *18 7 2 5 19 5 4 1 20 5 2 3 21 5 3 2 *22 7 1 6 23 7 6 1 58 ELIZABETH LOCKWOOD THOMPSON TABLE y.— {Continued) Summary — Snail No. 2 — {Continued) No. of Correct Incorrect days No. of or left or right tested trials choices choices 24 4 0 4 25 5 2 3 26 4 2 2 *27 5 1 4 28 5 0 5 29 4 1 3 *30 3 0 3 31 4 2 2 32 5 1 4 33 5 0 5 *34 5 1 4 *35 4 1 3 36 5 2 3 37 5 0 5 *38 5 1 4 39 5 1 4 *40 5 0 5 41 5 2 3 42 5 3 2 43 5 4 1 *44 5 0 5 45 8 4 4 46 3 0 3 47 7 4 3 48 6 1 5 49 6 1 5 *50 10 2 8 51 5 1 4 52 7 2 5 53 6 0 6 54 6 1 5 Total number of trials 307 Total number of correct or left choices 117 approximately 38% Total number of incorrect or right choices 190 approximately 62% Summary — Snail No. 3 No. of Correct Incorrect days No. of or left or right tested trials choices choices *1 2 1 1 *2 7 4 3 3 6 5 1 *4 5 0 5 5 6 1 5 6 7 2 5 7 5 2 3 8 6 1 5 9 5 4 1 10 5 2 3 AN ANALYSIS OF THE LEARNING PROCESS IN THE SNAIL 59 TABLE X— {Continued) Summary — Snail No. 3 — {Continued) No. of Correct Incorrect days No. of or left or right tested trials choices choices *11 6 3 3 *12 7 4 3 13 7 2 5 14 7 3 4 *15 3 1 2 16 7 3 4 17 7 3 4 *18 5 1 4 19 5 2 3 20 5 2 3 21 4 2 2 22 4 1 3 *23 8 6 2 24 4 3 1 *25 7 2 5 26 5 2 3 *27 5 2 3 28 5 2 3 *29 5 1 4 30 5 1 4 31 4 3 1 32 5 1 4 *33 5 1 4 34 5 2 3 35 4 -0 4 36 . 5 3 2 37 5 0 5 *38 5 2 3 *39 5 4 1 40 5 1 4 *41 5 4 1 42 5 1 4 43 5 1 4 44 5 1 4 45 8 2 6 46 3 1 2 47 7 6 1 48 5 3 2 49 6 1 5 *50 10 3 7 51 8 0 8 52 7 3 4 53 6 0 6 *54 6 2 4 Total number of trials 299 Total number of correct or left choices". . '. '. '. '. '. '. *. '. '. '. '. '.'.'.'. 113 approximately 38% 1 otal number of incorrect or right choices 186 approximately 62% * Indicates that an egg mass was deposited bv the snail during the 24 hours immediately preceding the tests on the day marked. 60 ELIZABETH LOCKWOOD THOMPSON TABLE X-A Summary of Left-Right Discrimination Tests Snail Snail Snail No. No. No. 1 2 3 Total number of trials 282 307 299 Per cent of incorrect choices 52 62 62 Table X shows that in a total of 888 trials on three individuals of Physa, in a series of tests offering a choice of right and left paths, with a reward for the correct choice (left), the incorrect path (right) was followed in 58.8% of the trials. The graphs (figs. 16-18, Plate VI) show very great irregularity in the time distribution of the right and wrong choices. There are days when the record is perfect, other days when all choices are wrong. Between these is every intermediate condition. A slight excess of incorrect choices is to be expected on account of the possible tendency to turn to the right shown in the right- left preference tests. TJicsc records shoiv no indication of learning. 5. Experiments With the Y-Shaped Labyrinth: Roughness of Path as a Warning Sign: Electric Shock Punishment It seemed either that the problem itself was too dififictilt or that the immediate demand for air (punishment) was not great enc^ugh to force its solution. In its natural habitat the snail often depends on plant stems as a means of reaching the surface of the water when a fresh supply of oxygen is needed. The stems must frequently be too short. The problem is then familiar to the snail and should be eas3^ To make the problem still easier it was decided to determine whether or not the snail cotild form a simple association between two stimuh upon a labyrinth that it had failed to solve. The labyrinth was changed as follows: 1. A Y-shaped labyrinth was ttsed so that the snail could make a choice immediately upon reaching the top of the stem of the Y. The Y was made of glass tubing. 2. The snail was not confined to one side of it. Owing to the spiral course followed by the ascending animal, there was no right and left choice. vS. One side of the Y was made rough and the other left smooth. The training series had shown that the snail feels the delicate pressure on the foot of the three-pronged pressure fork of the food-pressure apparatus (fig. 1, Plate I). It should AN ANALYSIS OF THE LEARNING PROCESS IN THE SNAIL 61 therefore feel the difference between a rough and smooth path. 4. For the failure-to-get-air punishment with the U-shaped labyrinth there was substituted the electric shock. Reward was the getting of air. The problem resolved itself then, into an attempt to determine whether or not the snail could form a simple association between two dissimilar stimuli one of which (punishment) followed the other (warning). It has already been shown that the snail can form a weak association between two stimuli simultaneously applied (food-pressure series). Can it then bridge over an interval between two stimuli one of which follows the other almost immediately!* Glass tubing, 5 mm. in diameter, in the form of a Y was used. One arm of the Y, 4.5 cm. in length, reached the surface of the water. The other arm, 3.5 cm. in length, was short enough so that the animals could not reach the surface. The stem of the Y was 3 cm. long. Two (No. 23) silk-insulated copper wires were passed through the stem to the upper end of the short arm, where they connected with two fine platinum wire (No. 26) rings passing around the end of the short arm close enough together so that an electric current could be passed between them, when proper connections were made. The upper ends of both arms were then closed w4th sealing wax to keep out the water. A hollow base, 4 cm. by 2 cm. by 1.5 cm., of lead was made, in order to permit an outlet for the wires. These passed from the lower end of the stem of the Y into a rubber tube which carried them, protected from the water, through the metal base, along the floor of the experimental tank and upward to the outside. A direct current of 0.2 amperes was taken from 220 volt mains through a lamp of 1100 ohms in series with S ohms of other resistance. From the terminals of 8 ohms a second circuit was taken including a rheostat of 1800 ohms, a receiving switch, a key and the Y apparatus. When no current was drawn through this second circuit the voltage between the two platinum wires would be 1.6 volts. The maximum current which could be drawn through the circuit by short-circuiting the two platinum wires was slightly less (actually 16-18) than 0.001 amperes or 1 milliampere*. It was *The author is indebted to Professor A. W. Smith for measuring the electric current used in the experiments. 162. ELIZABETH LOCKWOOD THOMPSON found that this current caused the snails to draw backward into their shells for a little way. They usually resumed their normal movements in a short time. Rarely they retracted so much that they lost their hold on the glass tubing and dropped to the bottom of the tank. The short arm of the Y from the point of divergence to within a millimeter or so of the platinum wires was roughened by means of a band of block tin, 2 mm. in width and 1 mm. thick, passing around it in a spiral. The metal was then insulated with a coat of liquid celluloid and further roughened by a coat of coarse cutting sand imbedded in the celluloid before it hardened. These precautions appeared desirable because it seemed possible that enough mucus might be secreted by the snail in crawling over sand alone, to fill in the spaces between the grains and thus make a smooth surface upon which to travel. This hardly seemed possible over the ridges of metal used, inasmuch as they were far enough separated so that the snail had to sink down into the space between, in passing over them. If the snail turned toward the short arm •at the top of the stem it immediately came in contact with the rough ridges and a little later with the bare electric wires, at which point the punishment was administered by closing the 'circuit. No attempt was made to keep the snail on one side ■of the apparatus. It sometimes crawled in an irregular spiral but frequently kept on one side. The whole apparatus was made shorter than the celluloid labyrinth previously used, in ^order to shorten the time used in each test and thus make a 'greater number of tests possible in a given time. The method of manipulation and cleaning the apparatus was the same as in the previous tests, and in order to avoid any possible tendency •of the snail to follow the same path in consecutive trips, the labyrinth was frequently turned so that the short arm was as ■often toward the left as the right. The " map " method before described was adopted in making the records. Four snails were used and a total of 664 trials given. The curves (figs. 19-22, Plate VII) were based upon the percentage of errors for each snail per day. AN ANALYSIS OF THE LEARNING PROCESS IN THE SNAIL 63 TABLE XI Showing the number of correct and incorrect turns made by three individuals of Physa gyrina Say, in their attempts to reach the surface of the water on the Y apparatus. Error was punished by an electric shock. The warning stimulus was roughness. The apparatus was cleaned between trials. No. of days tested 1 Totals. No. of trial for series Snail No. No. of trial for day 1 2 3 3 No. of correct choices 1 0 0 1 (33%) 0 No. of incorrect choices 0 1 1 2 (66%) Totals . Totals . *5 Totals . Totals . Totals . 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 Totals. 1 2 3 3 1 2 2a 2b 3 5 1 la 2 3 3a 5 1 la 2 3 4 1 2 2a 2b 3 3a 4 7 1 2 3 3a 3b 4 6 1 1 1 3 (100%) 1 0 0 0 0 1 (20%) 0 1 0 0 0 1 (20%) 0 1 0 0 1 (25%) (42%) 1 0 0 0 1 1 3 (50 f (80%) (80%) 1 0 1 1 3 (75%) 0 1 1 0 1 1 1 4 (57%) (50%) 64 ELIZABETH LOCKWOOD THOMPSON TABLE XI— {Continued) Snail No. 9 — {Continued) No. of days tested No. of trial for series 35 36 37 38 39 40 41 No. of trial for day Totals. 1 2 3 4 5 5a 5b 7 No. of correct choices 1 1 1 0 0 0 1 4 (57%) No. of incorrect choices (42%) 10 Totals. 42 43 1 1 2 (100%) 11 Totals. 44 45 46 47 1 2 2a 3 3a 5 5 (100%) 12 49 50 51 52 53 54 55 56 1 2 3 4 4a 4b 5 5a Totals. (37%) 0 0 1 1 1 0 1 1 5 (62%) 13 Totals. 57 58 59 1 1 1 3 (100%) 14 Totals. 60 61 62 63 1 la 2 2a 4 0 1 0 0 1 (25%) (75%) 15 64 65 66 67 Totals. 1 2 2a 2b 3 5 1 0 0 1 1 3 (60%) (40%) AN ANALYSIS OF THE LEARNING PROCESS IN THE SNAIL 65 TABLE XI— (Continued) No. of days tested 16 Snail No. 9 — (Continued) No. of No. of trial trial No. of for for correct series day choices Totals. 69 70 71 72 73 74 1 2 2a 2b 2c 3 6 1 0 0 0 1 1 3 (50%) No. of incorrect choices (50%) 17 Totals . 75 76 77 78 79 1 2 3 4 4a 5 (60%) 0 0 0 1 .1 2 (40%) 18 Totals. 19 Totals. 20 Totals . 21 Totals. 22 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 Totals, 101 102 103 104 105 106 1 la 2 3 4 4a 5 7 1 2 3 3a 3b 5 1 2 2a 2b 4 1 la 2 3 4 5 1 2 3 3a 4 5 6 (42%) "0 °°°'" 1 0 0 1 2 (40%) (50%) (60%) (66%). 1 0 1 0 1 0 1 4 (57%) '!'"■ 0 1 1 0 3 (60%) 0 1 0 2 (50%) 1 0 0 0 1 "2140%) 1 0 1 0 0 0 ■a<33%) 66 ELIZABETH LOCKWOOD THOMPSON No. of days tested 23 Totals. Totals. 24 25 Totals. 26 Totals . 27 Totals. 28 TABLE XI— (Continued) Snail No. 9 — (Continued) No. of No. of trial trial No. of for for correct series day choices Totals. 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 1 la 2 3 3a 3b 4 7 1 la lb Ic 2 5 1 la 2 3 4 5 6 6a 6b 6c 10 1 2 3 4 5 5a 5b 7 1 2 3 3a 3b 3c 3d 4 8 1 la lb Ic 2 2a 2b 7 (57%) (40%) (60%) (71%) (37%) No. of incorrect choices (42%) (60%) (2?%: 1 0 0 0 0 0 1 1 1 0 4 (40%) 0 0 0 0 1 1 0 2 (28%) 1 0 1 1 1 1 0 0 5 (62%) 1 1 1 1 1 1 0 .5 171%) AN ANALYSIS OF THE LEARNING PROCESS IN THE SNAIL 67 No. of days tested 29 Totals . 30 Totals . 31 Totals . 32 Totals . 33 Totals . 34 TABLE XI— {Continued) Snail No. 9 — {Continued) No. of No. of trial trial No. of for for correct series day choices 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 Totals . 187 188 189 190 191 192 1 la lb Ic 4 1 la 2 3 3a 4 5 7 1 2 3 4 4a 4b 4c 7 1 la lb 2 2a 3 4 5 6 6a 10 1 la lb Ic 2 3 4 4a 1 2 2a 2b 2c 2d 6 0 0 0 1 1 (25%) 0 1 1 0 0 1 1 4 (57%) 0 1 1 0. 0 0 0 2 (28%) 0 0 1 0 1 1 1 1 0 0 5 (50%) (50%) No. of incorrect choices 1 1 1 0 3 (75%) 1 0 0 1 1 0 0 3 (42%), 1 0 0 1 1 1 1 5 (71%) 1 1 0 1 0 0 0 0 1 1 5 (50%) (50%) (33%) 0 1 1 1 1 0 4 (66%) 68 ELIZABETH LOCKWOOD THOMPSON No. of days tested 35 Totals . 36 Totals . 37 Totals . 38 TABLE Xl— {Continued) Snail No. 9— {Continued) No. of No. of trial trial No. of for for correct series day choices 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 Totals . 211 212 213 214 215 1 la lb 2 2a 2b 6 1 la lb Ic Id 5 1 la lb Ic Id le 2 7 1 2 2a 2b 2c 5 (33%) 0 0 0 0 1 1 (20%) (14%) (40%) No. of incorrect choices (66%: (80%) (85%) (60%) No. of \ days tested Summary — Snail No. 9 1 2 3 4 *5 *6 7 8 9 10 11 12 13 No. of trials Correct choices 1 0 3 1 1 1 3 3 4 2 0 3 3 Incorrect choices 2 1 0 4 4 3 4 3 3 0 5 5 0 * Indicates that an egg mass had been deposited by the snail during the 24 hours immediately preceding the tests on the day marked. AN ANALYSIS OF THE LEARNING PROCESS IN THE SNAIL 69 TABLE Xl— {Continued) SuM^ lARY— Snail No. 9— {Com 'imied) No. of days No. of Correct Incorrect tested trials choices choices 14 4 1 3 15 5 3 2 16 6 3 3 17 5 3 2 18 7 3 4 19 5 2 3 20 4 2 2 21 5 3 2 22 6 4 2 23 7 4 3 24 5 2 3 25 10 6 4 26 7 5 2 27 8 3 5 28 7 2 5 29 4 1 3 30 7 4 3 31 7 2 5 32 10 5 4 33 8 4 5 34 6 2 4 35 6 2 4 36 5 1 4 37 7 1 6 38 5 2 3 Total number of trials 215 Total number of correct choices 95 approximately 44 % Total number of incorrect choices 120 approximately 56% Summary — Snail No. 10 No. of days No. of Correct Incorrect tested trials choices choices 1 6 1 5 2 3 2 1 3 6 3 3 4 1 1 0 5 3 1 2 6 3 1 2 7 6 3 3 8 5 1 4 9 2 1 1 10 5 2 3 11 5 3 2 12 9 3 6 13 10 3 7 14 6 1 5 15 6 2 4 16 8 4 4 70 ELIZABETH LOCKWOOD THOMPSON TABLE XI— (Continued) Summary — Snail No. 10 — (Continued) No. of days tested 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 Total number of trials 214 Total number of correct trials 75 appro.ximately 35% Total number of incorrect trials 139 approximately 65% No. of Correct Incorrect trials choices choices 11 3 8 9 5 4 5 1 4 14 4 10 9 1 8 7 3 4 5 1 4 3 2 1 11 2 9 6 3 3 15 5 10 5 4 1 8 2 6 7 1 6 5 2 3 6 3 3 4 1 3 Summary- -Snail No. 11 No. of days No. of Correct Incorrect tested trials choices choices 1 4 3 1 2 6 1 5 3 5 4 1 4 9 4 5 5 ■ 7 3 4 6 5 0 5 7 8 1 7 8 7 5 2 9 5 5 0 10 3 2 1 11 9 7 2 12 6 4 2 13 9 3 6 14 6 4 2 15 5 1 4 16 8 5 3 17 7 0 7 18 10 3 7 19 7 3 4 20 9 2 7 21 7 2 5 22 7 1 6 23 7 3 4 24 8 2 6 25 14 7 7 AN ANALYSIS OF THE LEARNING PROCESS IN THE SNAIL 71 TABLE Kl— (Continued) Summary — Snail No. 11 — (Continued) Total number of trials 178 Total number of correct trials 75 approximately 42% Total number of incorrect trials 103 approximately 58% Summary— Snail No. 27 No. of days No. of Correct Incorrect tested trials choices choices 14 2 2 2 3 12 3 5 1 4 4 6 15 5 5 14 6 4 0 4 7 4 2 2 8 7 0 7 9 5 4 1 10 6 2 4 11 5 3 2 12 3 1 2 Total number of trials 57 Total number of correct trials 18 approximately 32% Total number of incorrect trials 39 approximately 68% TABLE XI-A Summary of Rough-Shock-Smooth Tests Snail Snail Snail Snail No. No. No. No. 9 10 11 27 Total number of trials 215 Per cent of incorrect choices 56 214 65 178 58 57 68 Table XI shows the trials of four snails on the rough-shock- smooth Y-shaped labyrinth. These cover periods of 38, 35, 25, and 12 days and involve 215, 214, 178, and 57 trials, or a total of 664 trials. Of these 401 or 60.3% give wrong choices. The curves (figs. 19-22, Plate VII) show great irregularity, especially in the first third of their courses. Nearly all the days on which the choices w^ere either all right or all wrong fall in this part of the curves. After this the curves fluctuate less froin the fifty per cent level, but rise slightly as they proceed. In this series the snails climb a labyrinth of cylindrical glass and are free to move on any side of it; there is no question of right or left. The choice is made between rough and smooth. No rough- smooth preference tests were made, 'but there is evident in this 72 ELIZABETH LOCKWOOD THOMPSON series a possible tendency to prefer the rough side. The punish- ment had no apparent effect upon this tendency. It seemed as though this might be due to the fact that owing to the con- struction of the apparatus the punishment (shock) was delayed too long after the warning stimulus (roughness) had been received, 6. Experiments With Y-Shaped Labyrinth: Mechanical Stimulation AS Warning Signal The results treated in Table XI and in the group of figures (19-22, Plate VII) show no indication of learning. It is believed that the large percentage of error might have been lowered in a greater number of tests, but this would probably not account for the type of curve obtained. It was then suggested by Dr. John F. Shepard that while the snails were unable to solve the labyrinth they might nevertheless be able to form a simple association under the conditions of the labyrinth experiments. They might show on the labyrinth a capacity equal to that developed under tests with simultaneous stimuli. A series of tests was therefore carried out, using the Y apparatus with the electric shock punishment but modified as follows: 1. In place of roughness on the shock side a mechanical warning stimulus was used which could be observed to affect the snail. 2. The wrong arm of the labyrinth was shortened so that the shock followed immediately on the warning stimulus. The time rela- tion of the two stimuli was similar to that in Yerkes' experi- ments (1912), with earthworms and approached that of the method of simultaneous stimuli. The Y apparatus just described was used in a slightly modified form. All traces of the sand and metal were removed. The long arm of the Y was shortened to 3 cm. in order to bring the reward nearer the point of diver- gence of the paths, while the short arm was reduced to 1.5 cm. in length in order to bring the punishment nearer to the point of the path which necessitated choice on the part of the snail. The snail was placed at the base of the path in the usual way. If at the top of the stem the short arm was chosen, as soon as the turn was made the tentacles and head were stroked with a hair. This was fastened to the end of a small stick so that it could be handled by the operator. This tactile stimulus was immediately followed by the usual electric shock. The fact that the tentacles were withdrawn and that the animal recoiled AN ANALYSIS OF THE LEARNING PROCESS IN THE SNAIL 73 slightly when touched with the hair, showed conclusively that the tactile stimulus was effective. A series of check tests in which this stimulus alone was used showed that the irritation thus caused was not severe enough to make it serve as a punish- ment. Six snails were used and the usual method of recording and curve plotting was followed. TABLE XII Showing the number of correct-by-chance, correct-by-association, and incorrect trials in a series of tests in which an electric shock was used as punishment in connection with a mechanical warning stimulus. The Y apparatus was used with the arms shortened to 3 cm. and 1.5 cm. In case of error in choice the snail was stroked across the dorsal head region with a hair and then given an electric shock. The apparatus was cleaned between tests. Four individuals of Physa gyrina Say were used. Snail No. 9 No. of No. of Choices Choices No. of trial trial correct correct days for for by by Incorrect tested series day chance association choices Totals. Totals. Totals. 10 11 12 13 14 15 16 17 18 19 20 21 22 23 Totals . 24 25 26 27 28 29 30 1 2 3 3a 4 5 6 7 1 2 3 4 5 6 6 1 2 3 3a 4 5 5a 6 8 1 2 3 4 5 6 7 7 0 1 0 0 0 0 0 1 0 2(22' o) 0 0 1 0 0 1 2 (33%) 1 1 0 0 0 0 0 1 3 (37%) 0 1 0 1 1 1 1 5(71%) 0 0 0 0 0 0 0 0 0 0 0 1 0 1 1 0 3 (50%) 0 0 0 1 0 0 0 0 1 (12%) 1 0 1 0 0 0 0 2(28%), 1 0 1 1 1 1 1 0 1 7 (77%) 1 0 0 0 0 0 1 (16%) 0 0 1 0 1 1 1 0 4 (50%) 0 0 0 0 0 0 0 0 74 ELIZABETH LOCKWOOD THOMPSON TABLE Xll— {Continued) Snail No. 9 — {Continued) No. of days tested No. of trial for series 31 32 33 34 35 36 37 Totals. No. of trial for day 1 2 3 4 5 6 7 ■ 7 Choices correct by chance 0 0 0 1 0 1 0 2(28^ o) Choices correct by association 0 0 0 0 0 0 1 1 {l^7o) Incorrect choices 1 1 1 0 1 0 0 4 (57%) 38 39 40 41 42 43 44 45 Totals. 0 1 1 1 0 1 1 1 6 (75%) 1 0 0 0 0 0 0 0 1 (12%) 0 0 0 0 1 0 0 0 1(12%) 46 47 48 49 50 51 52 53 Totals . (50%) 0 0 0 0 0 0 0 1 1 (12%) 1 1 0 0 0 1 0 0 3 (37%) Totals . 54 55 56 57 58 59 60 61 62 1 2 3 4 4a 5 6 7 1 0 1 0 1 1 0 0 1 5(55f 0 0 0 0 0 0 1 1 0 2 (22%) 0 1 0 1 0 0 0 0 0 2 (22%) Totals . 63 64 65 66 67 68 69 70 1 la 2 2a 3 4 4a 4b 0 1 0 1 0 0 0 0 2 (25%) 1 0 1 0 1 1 1 1 6 (75%) AN ANALYSIS OF THE LEARNING PROCESS IN THE SNAIL 75 TABLE XU—iContimied) No. of days tested 10 Totals . 11 Totals . 12 Totals . 13 Totals . 14 Totals. 15 No. of trial for series 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 Totals . 103 104 105 106 107 108 109 Snail No. 9 — {Continued) No. of Choices trial correct for by day chance 1 2 3 3a 4 5 6 7 1 2 3 4 5 6 6a 7 1 2 3 4 5 6 6 1 2 3 4 5 6 6 1 2 3 4 5 6 6 1 2 3 4 5 6 7 7 Choices correct by association Incorrect choices 0 0 0 0 0 0 0 0 1 1 0 0 1 0 0 3 (42%) 1 1 1 1 0 1 5 (83%) 1 0 0 1 1 1 4 (66%) 0 1 1 0 1 0 3 (50%) 0 1 1 1 0 0 1 4 (57%) 0 1 0 1 1 1 1 5(71%) 0 0 0 0 0 0 0 0 0 0 0 0 1 0 1 (16%) 0 0 1 0 0 0 1 (16%) 0 0 0 1 0 1 2 (33%) 1 0 1 0 0 0 0 2 (28%) (57%) 0 0 0 0 0 0 0 0 1 0 0 0 0 1 (16^ \) (42%) 1 0 0 0 0 0 1 (16%) 0 0 0 0 0 0 0 0 76 ELIZABETH LOCKWOOD THOMPSON No. of days tested 16 No. of trial for series 110 111 112 113 114 115 Totals. TABLE XII- Snail No. 9- No. of trial for day 1 2 3 4 5 6 6 -{Continued) -{Continued) Choices Choices correct correct by by Incorrect chance association choices 0 0 1 0 1 1 3 (50%) 1 1 0 1 0 0 3 (50%) 17 Totals. 116 117 118 119 120 121 122 123 0 1 1 0 1 1 1 1 6(75^ o) 1 0 0 1 0 0 0 0 2 (25%) 18 Totals . 124 125 126 127 128 129 130 1 1 0 1 1 1 1 6 (85%; 0 0 1 0 0 0 0 1 (14%) 19 Totals. 131 132 133 134 135 136 0 0 1 1 1 1 4 (66%) 1 1 0 0 0 0 2 (33%) 20 Totals. 137 138 139 140 141 1 la 2 3 4 5 0 0 1 1 0 2 (40%) 0 1 0 0 1 2 (40%) 1 0 0 0 0 1 (20%) 21 142 143 144 145 146 147 Totals. 1 2 3 4 5 5a 6 (66%) 0 0 1 0 0 0 1 (16%) 0 0 0 0 1 0 1 (16%) AN ANALYSIS OF THE LEARNING PR(3CESS IN THE SNAIL 77 TABLE Xll— (Continued) Summary— Snail No. 9— (Continued) Choices Choices No. of correct correct days No. of by by Incorrect tested trials chance association choices 1 9 2 0 7 o 6 2 3 1 3 8- 3 1 4 4 7 5 2 0 5 7 2 . 1 4 6 8 6 1 1 7 8 4 1 3 8 9 5 2 2 9 8 2 0 6 10 7 0 5 2 11 7 3 0 4 12 6 5 1 0 13 6 4 1 1 14 6 3 2 1 15 7 4 3 0 16 6 3 3 0 17 8 6 2 0 18 7 6 1 0 19 6 4 2 0 20 5 2 2 1 21 6 4 1 1 Total number of trials 147 Total number correct by chance 75 approximately 51% Total number correct by association 34 approximately 23% Total number incorrect 38 approximately 26% Summary— Snail No. 10 Choices Choices No. of correct correct days ^ No. of by by Incorrect tested trials chance association choices 1 7 3 0 4 2 5 5 0 0 3 11 7 0 4 4 10 4 0 6 5 7 3 0 4 6 6 3 0 3 7 8 5 0 3 8 5 3 0 2 9 9 4 2 3 10 10 2 7 11 7 4 2 12 7 5 1 13 6 2 3 14 9 4 4 15 9 4 2 3 16 8 4 0 4 17 7 5 1 1 18 10 2 0 8 78 ELIZABETH LOCKWOOD THOMPSON TABLE Xn— {Continued) Summary — Snail No. 10 — (Continued) No. of No. of No. of trial trial No. of No. of days for for correct incorrect tested senes day choices choices 19 7 5 2 0 20 6 2 4 0 21 6 1 . 4 1 22 7 3 1 3 23 6 4 1 1 . Total number of trials 173 Total number correct by chance 84 approximately 48% Total number correct by association 22 approximately 13% Total number incorrect 67 approximately 39% Summary — Snail No. 13 Choices Choices No. of correct correct days No. of by by Incorrect tested trials chance association choices 1 5 0 0 5 2 5 3 0 2 3 10 6 0 4 4 5 3 0 2 5 6 3 0 3 6 5 4 0 1 7 10 5 1 4 8 5 4 0 1 9 10 5 1 4 10 7 2 1 4 11 7 3 1 3 12 8 4 0 4 13 8 6 0 2 14 9 5 2 2 15 8 3 4 » 1 16 6 3 2 1 17 7 3 2 2 Total number of trials 121 Total number correct by chance 62 approximately 51% Total number correct by association 14 approximately 12% Total number correct by chance 45 approximately 37% Summary — Snail No. 19 Choices Choices No. of correct correct days No. of by by Incorrect tested trials chance association choices 1 9 2 0 7 2 8 6 2 0 3 8 4 3 1 4 6 1 3 2 5 6 3 1 2 AN ANALYSIS OF THE LEARNING PROCESS IN THE SNAIL 79 TABLE XII— {Continued) Summary— Snail No. l9~{Continued) Choices Choices No. of correct correct days No. of by by Incorrect tested trials chance association choices 6 5 4 1 0 7 6 2 2 2 8 5 4 1 0 9 7 5 1 1 10 8 3 3 2 11 7 5 1 1 12 7 6 0 1 13 8 4 1 3 14 8 4 3 1 15 6 3 2 1 16 6 5 1 0 17 6 4 2 0 18 6 4 2 0 19 6 2 2 2 20 6 4 1 1 21 7 3 2 2 Total number of trials 141 Total number correct by chance 78 approximately 55% Total number correct by association 34 approximately 24% Total number incorrect 29 approximately 21% Summary— Snail No. 21 Choices Choices No. of correct correct days No. of by by Incorrect tested trials chance association choices 1 6 5 0 1 2 8 6 0 2 3 8 3 0 5 4 7 2 1 4 5 6 2 0 4 6 8 2 2 4 7 7 5 0 2 8 8 6 1 1 9 8 8 0 0 10 5 1 0 4 11 4 3 0 1 12 5 3 2 0 13 6 3 2 1 14 5 5 0 0 15 6 3 2 1 16 7 6 0 1 17 7 4 0 3 18 7 6 0 1 19 6 3 1 2 20 6 4 0 2 21 5 . 4 I 0 22 6 4 1 1 23 8 4 0 4 80 ELIZABETH LOCKWOOD THOMPSON TABLE XU— (Continued) Summary — Snail No. 21 — (Continued) Choices Choices No. of correct correct days No. of by by Incorrect tested trials chance association choices 24 6 2 2 2 25 5 2 2 1 26 8 4 1 3 27 6 3 2 1 Total number of trials 174 Total number correct by chance 103 approximately 59% Total number correct by association 20 approximately 12% Total number incorrect 51 approximately 29% Summary — Snail No. 22 Choices Choices No. of correct correct days No. of by by Incorrect tested trials chance association choices 1 6 1 0 5 2 8 2 1" 5 3 7 4 0 3 4 6 3 1 2 5 8 1 0 7 6 8 5 1 2 7 8 3 1 4 8 12 1 1 10 9 7 2 1 4 10 • 5 0 0 5 11 5 4 0 1 12 6 4 0 2 13 7 4 0 3 14 5 2 0 3 15 9 3 2 4 16 7 4 0 3 17 9 4 1 4 18 6 5 0 1 19 7 4 1 2 20 7 2 0 5 21 5 2 1 2 22 5 1 4 0 23 7 3 2 2 24 6 3 2 1 25 6 4 1 1 26 6 4 1 1 27 6 4 1 1 Total number of trials 174 Total number correct by chance 79 approximately 45% Total number correct by association 22 approximately 13% Total number incorrect 73 approximately 42% AN ANALYSIS OF THE LEARNING PROCESS IN THE SNAIL 81 TABLE XII-A Summary of tests in which a mechanical stimulus was used in connection with electric shock. Snail Snail Snail Snail Snail Snail No. No. No. No. No. No. 9 10 13 19 21 22 Total number of trials 147 173 121 141 174 174 Per cent of choices correct by chance 51 48 51 55 59 45 Per cent of choices correct by association 23 13 12 24 12 13 Per cent of incorrect choices. . 26 39 37 21 29 42 Table XII shows that in a total of 930 trials on six individuals of Physa, 51.2% were correct by chance, that is, the correct path was chosen without contact with the warning stimulus. In 15.6% of the trials the correct path was chosen after contact with the warning stimulus, but without punishment, showing that a weak association had been formed between the two stimuli. In 32.5% of the trials, the warning stimulus was dis- regarded and punishment received. This shows that the behavior of the snails tested was modified during the experiments inas- much as they apparently learned to recognize the warning stimulus and turned back before receiving the electric shock. The snails profited by experience. The slightly downward tendency of the jagged ctirves based on these records (figs. 23-28, Plate VIII) shows an erratic decrease in error, weak asso- ciation and limited associative memor3^ 7. Summary of Observations 1. All observations and experiments were made on the snail Physa gyrina Say while in tap water at room temperature and under uniform lighting conditions. 2. When the air is pressed from the lung of Physa and the snail is placed in the water at the base of a U-shaped labyrinth supported on an upright stem, it crawls up the stem and along the arms of the U to the surface of the water. Fig. 14, Plate IV. 3. Experiments to test right or left preference : first series. — In a series of 24 trials in which three individuals of Physa were deprived of air and placed on the bottom near the wall of a glass dish filled with water, and allowed to crawl unhindered to the surface of the water, the path chosen deviated from a 82 ELIZABETH LOCKWOOD THOMPSON perpendicular toward the right in 54.1% of the trials and toward the left in 45.8% of the trials. This showed that the natural path chosen by the three snails tested, in traveling over a flat unrestricted substratum to the surface of the water for air, in a comparatively small number of trials deviated somewhat to the right. Table VII, p. 43. 4. Experiments to test right and left preference : second series. When the U apparatus (paragraph 1) is bordered by a " picket fence " of wires, the snail is forced to travel on one side. When it reaches the top of the stem it must turn toward either the right or left end of the horizontal bar of the U to reach the surface of the water. Fig. 14, Plate IV. 5. In a series of 240 trials in which the two arms of the U were identical and both reached the surface of the water, the three snails tested followed the right arm in 51.6% of the trials and the left arm in 48.3%, showing a slight right preference for the snails tested. Table VIII, p. 45. 6. Experiments to test preference for an ascending or descending path. — In order to determine whether any preference existed for an ascending or a descending path, the U-shaped labyrinth was so tilted that the bar of the U deviated from a horizontal at angles varying from 0-1 degree to 7-8 degrees. In 6()() trials the three snails tested chose the descending arm in 43%) of the trials and the ascending arm in 57%, showing a slight preference for an ascending path. Table IX, p. 47. 7. In subsequent experiments in which the U labyrinth was used the horizontal bar was kept level. 8. In the experiments to test learning, complete records are presented for two individuals in each series of tests and sum- maries of results are given for additional snails tested. 9. Experiments on learning the U-shaped labyrinth. — In a series of 888 tests on three individuals of Physa, using the U-shaped labyrinth with the right arm shortened so that a snail could not reach the surface of the water from its upper end, thus making failure to obtain air serve as a punishment incident to the choice of this arm, 58.8% of the turns were toward the right, that is, incorrect, and 41.2% toward the left or correct. There is then shown no evidence of ability in the snail to distinguish right or left in connection with failure to AN ANALYSIS OF THE LEARNING PROCESS IN THE SNAIL 83 obtain air (punishment) and the attainment of air (reward). Table X, p. 48. 10. The curves based on these trials show irregularity at the beginning and a slightly upward tendency toward the end, indi- cating that the possible right preference was covered in the early part of the series by disturbance such as manipulation and the change in apparatus. Figs. 16-18, Plate VI. 11. Experiments with Y-shaped labyrinth: roughness of the- path as a warning stimulus; electric shock as punishment. — In a series of 664 tests a Y-shaped path of glass tubing, unlimited by wires, was used. One arm of the Y gave access to the surface- of the water and the other was shortened and supplied with wires over which an electric current could be sent. The space between the point of divergence of the two arms and the wires was roughened. The four snails tested chose the short arm or punishment in 60.3^^ of the trials. This series of tests then,, shows no indication of learning. Table XI, p. 63. 12. The curves based on these trials show less irregularity than those in the preceding tests (paragraph 10). The latter part hes slightly above the 50% level indicating a possible preference for the rough side, masked by disturbance in the first part of the series, combined with punishment too long delayed after the warning stimulus had been received. Figs. 19-22, Plate VII. 13. Experiments with Y-shaped labyrinth: mechanical stimu- lation as warning signal. — In a series of 930 tests the Y apparatus was used but instead of roughness in connection with the electric shock, the snail was stroked on the tentacles and dorsal head region with a hair, preceding the shock, if a wrong choice were made. Check tests showed that the irritation caused by the warning stimulus was not sufficient to cause it to serve as a punishment. The six snails tested chose the short arm and punishment in 32.5% of the trials, the long arm, by chance in 51.2% and the long arm through association in 15.6%. Modification of behavior through the formation of an association between the two stimuli, one serving as punishment, the other as a warning signal, is thus shown. Table XII, p. 73. 14. The curves based on these trials show irregularity through- out but a downward tendency toward the latter end, indicating 84 ELIZABETH LOCKWOOD THOMPSON the formation of a weak association between the two stimuU used. Figs. 23-28, Plate VIII. 15. The last series of experiments (paragraph 13) shows results comparable to those obtained in the first part of the present work which indicated an association formed between two dissimilar stimuli, food and pressure, in a series of tests based on the Pawlow salivary method. It does not show ability to learn the labyrinth. 8. Discussion The foregoing attempt to learn whether the snail can solve a simple vertical U- or Y-shaped labyrinth falls into three series. 1. In the first of these there was used a U-shaped labyrinth supported on a stem and bordered by a paling of wires so that the snails could crawl up but one side of it. At the top of the stem they had a choice of a right or left path. If the right path were chosen it led to the right vertical bar of the U from the tO}) of which the snail could not reach the surface of the water. If the left path were chosen it led to the vertical bar of the U, longer than the right, from the top of which the snail could fill its lung with air. The upward course of the snail is in search of air. Choice of the right path resulted in failure to get air which was assumed to act as punishment. Such a choice was regarded as incorrect. Choice of the left path led to the attainment of air, which was regarded as reward and such choice was registered as correct. Preliminary tests showed that the snails have a slight pref- erence for the right path, a preference possibly related to the asymm.etry of the animal. It was to be expected that this right preference would appear at the beginning of the series of learning tests, but that it would be soon corrected by the result of the learning process. If thd snails actually learned and a curve were plotted showing their percentage of errors from each succes- 'sive day's experience, this curve would start somewhere above the 50% level on account of the right i^reference, which would result in an excess of incorrect choices. The curve would then gradually descend, through learning, toward the zero level with decrease in error present. If the snail did not learn and were in no way affected by its experiences with the labyrinth the ■curve of errors might be expected to follow the 50% level or somewhat above it, while fluctuating above and below. AN ANALYSIS OF THE LEARNING PROCESS IN THE SNAIL 85 An inspection of the graphs (Plate VI) shows that the curves are quite irregular and with the progress of time there is a slight tendency to rise above the 50% level. The snails not only do not learn but their percentage of errors apparently increases as the series of trials lengthens. Nothing has been detected in the physical environment to direct the snail to either side of the path. It is in a symmetrical environment as to structure of labyrinth, surrounding tank and house, lighting and temperature. It seems probable that the disturbance due to manipulation and the change in conditions incident to the experimental work probably caused irregularities in the first part of the series which covered the slight tendency to turn to the error (right) side of the path, evident in the latter part of the curves. The graphs show then, a probable masking of the " right " preference in the early part of the experiments, due to disturbance rather than increase in this preference as the series lengthens. There is no evidence of learning. 2. In the second series of labyrinth tests the U-shaped laby- rinth was discarded and in its place was used a Y-shaped labyrinth of cylindrical glass tubing. Owing to the snail's tendenc}^ to crawl in a spiral such a labyrinth offers no choices of right or left and the snail is free to move in any direction in order to reach the top. One arm of this labyrinth was made rough while the other remained smooth. Thus roughness served as a means of discrimination between the two. At the upper end of the rough arm the snail received an electric shock. The roughness served as a warning signal of the approaching shock. No rough-smooth preference tests were made. The curves (Plate VII) again show much irregularity, but not so much as those with the U-shaped labyrinth. There are again days on which all choices are correct and other days on which all are incorrect. Nearly all of these days, however, fall in the first third of the curves, with the result that this part is much more irregular than that which follows. After their first , thirds the curves remain nearer the 50%: level, but on the whole tend to lie above it. In other words, the percentage of error again appears to increase as the series lengthens. It seems possible that a greater number of trials might have brought the curves again to the 50% line or even slightly below it, indicating that the upward tendency was due to a wave of error and that a 86 ELIZABETH LOCKWOOD THOMPSON corresponding wave showing fewer errors might follow. It is more probable however, that the greater irregularity evident in the first thirds of the curves was due in part to disturbance following the change of apparatus and that the slight increase in error toward the end was the result of a possible rough preference linked with a too long delayed punishment. 3. The tliird scries of labyrinth tests was planned to test the capacity of the snail to form associations on the labyrinth that it had failed to solve. The Y labyrinth was used, but all traces of roughness were removed. The short arm was shortened still more making the interval between the warning stimulus and the punishment less than in the preceding tests. The warning stimulus adopted was the irritation caused by stroking the dorsal head region and tentacles of the snail with a hair. That this stimulus was an irritation was evidenced by the fact that its use caused the snail to withdraw its tentacles,, and at times to partially retract its head toward the edge of its shell. Check tests, however, showed that the irritation was not severe enough in itself to act as a punishment. The labyrinth as before offered no choice of right or left. The curves (Plate VIII) of error percentage show much irreg- ularity but a slight downward tendency. In 15.6% of the total of 930 trials, the snails changed their course from the wrong to the right path after contact with the warning stimulus but before the shock (punishment) was received. Since the warning stimu- lus was known to irritate the animal, there could have been no preference for the wrong side of the labyrinth to offset the associative effect of the shock (punishment). The 15% of correct-by-association trials, then, represents the strength of the association formed between the warning stimulus and the punishment, no part of which could have been masked through preference for the signal stimulus used. This series of tests thus shows the formation of a weak association between two stimuli, one used as a warning of the punishment to follow if the course be not changed. Physa then, profits by experience, but cannot be said to solve a labyrinth. The work with the modified apparatus is merely confirmatory of that by the method of simultaneous stimuli. It shows no greater capacity on the part of the snail than was made evident by that method except that to form an association between stimuli separated by a small time interval. AN ANALYSIS OF THE LEARNING PROCESS IN THE SNAIL 87 Observational work on mollusks has-been reported, but little experimental work has been done on this group of animals, none along the lines of the present work. Kollmann (1877) attributed " consciousness " to Octopus vulgaris, basing his con- clusions, however, on observation rather than experiment. In his analysis of this report, Schneider (1909) explains the behavior observed as reflexive, not " conscious," and hence typical of any of the lower animals. Pieron (1909) concludes that the octopus is capable of " learning," and although Kafka (1914) reports this work he does not analyze it. Modifiability of behavior then, may be attributed to members of the Cephalopod group. Pieron (1909), (1911) experimented on the waning of the effect of stimuli inducing reflexive action in Limnaea stagnalis and Littorina, which he terms memory. Associative memory as observed in the present work was believed to be an added proof of learning, but no attempt was made to determine its exact duration nor the way in which it developed. The work of Pieron is here of interest only in that it deals with so-called memory (adaptation) in closely related forms. III. CONCLUSIONS 1. The present work on the snail, Physa gyrina Say, gives much evidence of capacity for adaptation — the getting-used-to- the-stimulus. This is quite apart from evidence for fatigue which appears in some places. 2. In addition to adaptation its behavior is modifiable by experience in another way. As the result of prolonged training with two simultaneous stimuli it is found to have acquired a response to one of these stimuli which it did not give before. It has formed an association between the two stimuli. 3. This definite acquired response is retained for four days beyond the completion of the training (memory). 4. When put through a long series of tests with the simplest form of labyrinth, under controlled conditions it shows no capacity to solve the labyrinth. 5. On a similar labyrinth it gives evidence of the capacity to form associations. 6. The capacity to form associations does not then suffice for the solution of the simplest labyrinth, that is, " selective " ability is apparently lacking. 88 ELIZABETH LOCKWOOD THOMPSON IV. LITERATURE CITED 1. Dawson, J., 19U. The biology of Physa. Behavior Monogr., I, 4. 2. FlELDE, A. M., 190L A study of an ant. Proc. Philadelphia Acad. Nat. Set. vol. 53 p. 425. 3. " '1901. Further study of an ant. Ibid., vol. 53, p. 521. 4. Jennings, H. S., 1906. Behavior of the lower organisms. New York. 5. KoLLMANN, J., 1877. Aus dem Leben der Kephalopoden. Vierteljahrssch. iviss. Philos., Bd. I, S. 136. 6. Kafka, G., 1914. Einfiihrung in die Tierpsychologie. Leipzig. 7. MoRGULis, S., 1914. Pawlow's theory of the function of the central nervous system and a digest of some of the more recent contributions to this subject from Pawlow's laboratory. Jour, of An. Behav., vol. 4, no. 5, p. 362. 8. ■ 1914. The auditory reactions of the dog studied by the Pawlow method. Ibid., vol. 4, no. 2, p. 142. 9. Nagel, W. a., 1894. Beobachtungen tiber den Lichtsinn augenloser Muscheln. Biol. Cent., Bd. 14, S. 385. 10. NicoLAi, G. F., 1907. Die physiologische Methodik zur Efforschung der Tierpsyche. Jour. f. Psych, ii. Neur., Bd. 10, S. 1. 11. , 1908. Das Lernen der Tiere (auf Grund von Versuchen mit Pawlowscher Spiechelfistel). Cent. f. Physiol., Bd. 22, S. 362. 12. Oberli, L. a., 1909. Reflexes conditionnels du cote de I'oeil chez le chien. Arch, de Sci. Biol., t. XIV, p. 31. 13. Parker, G. H., 1912. The mechanism of locomotion in gastropods. Join. of Morph., vol. 22, p. 155. 14. Pawlow, I. P., 1904. Sur la secretion psychique des glandes salivaires. Arch. intern, de Physiol., t. I, p. 119 15. , 1906. The scientific investigation of the psychical faculties or processes in the higher animals. (Huxlev Lecture for 1906.) Brit. Med. Jour., 1906, p. 871 (Pub. in part). Lancet, 1906, p. 911 (Pub. in full). Science, N. S., vol. 24, p. 613 (Pub. in part). 16. Pii'iRON, H., 1909. Sens de I'Orientation et memoire topographique de la limnee. Compt. rend. Acad. Sci. Paris, 148, p. 530. 17. , 1909. La loi d'evanouissement des traces mnemoniques en fonction du temps chez la limnoe. Ibid., 149, p. 513. 18. , 1911. Les curbes d'evanouissement des traces mnemoniques. Ibid., 152, p. 1115. 19. , 1911. Sur la determination de la periode d'etablissement dans les acquisitions mnemoniques. Ibid., 152. p. 1410. 20. Schneider, K. C., 1909. Vorlesungen uber Tierpsychologie. Leipzig. 21. Shepard, J. F. Unpublished, referred to by permission. 22. Small, W. S., 1899. An experimental study of the mental processes of the rat, I. Atn. Jour, of Psych., vol. 11, p. 80. 23. Spaulding, E. G., 1904. An establishment of association in hermit crabs, Eupagurus longicarpus. Jour, of Comp. Neur. and Psych., vol. 14, p. 49. 24. Washburn, M. F., 1908. The animal mind. New York. 25. Watson, J. B., 1914. Behavior. An introduction to comparative psychology. New York. 26. Yerkes, R. M., 1902. Habit formation in the green crab, Carcinus granulatus. Biol. Bull., vol. 3, p. 241. 27. , 1912. The intelligence of the earthworm. Jour, of An. Behav. vol. 2, p. 332. 28. Yerkes, R. M., and Huggins, G. E., 1903. Habit formation in the crawfish, Cambarus affinis. Harvard Psych. Studies, vol. 1, p. 565. 29. Yerkes, R. M., and Morgulis, S., 1909. The method of Pawlow in animal psychology. Psych. Bull., vol. 6, no. 8, p. 257. AN ANALYSIS OF THE LEARNING PROCESS IN THE SNAIL 89 LIST OF ILLUSTRATIONS PAGE Plate I. Figure 1. — Apparatus used for the simultaneous application to the snail of food and pressure and pressure alone. Plate II. Figures 2-7. — Curves showing the distribution of response and re- actions per 10 trials with the food-pressure stimulus. Plate III. Figures 8-13. — Curves showing the distribution of responses and reactions per 50 trials with the food-pressure stimulus. Plate IV. Figure 14. — U-shaped labyrinth used in the right-left discrimina- tion tests. Plate V. Figure 15. — Apparatus used in an attempt to control light condi- tions, the Y-shaped labyrinth, and the reostat used to control the electric current employed in administering the shock. Plate VI. Figures 16-18. — Curves showing the distribution of error in the right-left discrimination tests. Plate VII. Figures 19-22. — Curves showing the distribution of error in the rough-shock-smooth tests. Plate VIII. Figures 23-28. — Curves showing the distribution of error in the tests in which a mechanical stimulus was used in connection with the electric shock. Plate I Ay/ /}jt?i?ar(7/^i/5 iy3ec//>? exper/>??e/7/s fyj^/c/? reari//rec/ //?e 3/mi///a/7eoi/3 (7^//ccr//o/7 /o //?e jr7(7// o/' /bod a/?c/ Dre^3ure or2c/jore^SL/re o/one. a, /?(7r?c//e; b. ^hee/^ of /?7e/(7/; c j/ee/ r?ee(y/e; c/ /^//r //pro^/? e/e o/z^eetTy^' e, /o(xy/?ooA, f.jDreJ3iyre ZbrA, a. Jr?a//: A, Ji/r/aceoffyo'/e/r. Plate II ^■^ ^ 1 i Si % 5} r T 1 1 ^' ~ ■"]■ "T— ■■ ^1 ?}' - -1 ! - ^ /■ ' ii / '/ ^ ---. , i ~ ~ \ y\ ^-^ ? 1 ~ ^ N, \. L i ~ •ij ?» '-^■ 1 ~ \ ^ 1 Or >' ^ i ' : ^S; ■< N, f! / » If? s \ ^ V S ^ 1 '^ J ^ r ^ 'n f i r- 1^ ^.--1 4^ s ^ v^ \^ \ ^. \ X ■-; Si 1 ? ^ \ f N^ k ^^ i i- / i i ' \ J* 1 i 5; - -' f ii ^ i \ \ T S' ^ 1 % ^J ^ j / ^ ! V ^ 9 ^ \ N ^ ^ "ii: <; C ,^1-' "^-v'^ ^ ir ^ \% \; ■ ' s ' 5 ^ N X \ i| r IC / ^ 1 / j \ S 1 l\ 1 'N ^ I |:^? ^ — 1 ,' ^'; tA \ •^ . ■' _ -i ' j P S' ^ --.^'-, ~- — • k & / > ' J ^ -:! /j "--^ / f ^ V \ 1 /^ y / ^ J J - JS "^ ^J- L , _. "^is 1 1 M 1 1 1 s^ss§^^$ ^1 «5^ ^ §^ ,^ 1 ^1 ft H ' ' 1 1 1 1 1 1 1 1 L J U _ Plate III ' r ■ \ - ^ \ 1 5 \' V \ \ " y, \l ; .^ "— ■~_ 1 i '1- %' , S ^ § S ^ ^ ^ ' ^# il^ " >■ +^ ■— ' \ 1 1 N. * - § ^ s % 5 ^ ^ 1 V s — "■ ^ -^ ■^ s ^ ?l 1 ^ ? / Up, ^.^ \ ' 1. \ 1 'r ^ ^ ^ '^' ^ ^ ^ ^ 1 .1 -^ N^ ^ ^^ .% I t'^ \ ' ii ^ ^- \ -- " ,' 4-- ^- -" / — - -.__ — \ \ -- ^ 1 § s S s % ^ ^ ^ k^ ^ •cj ^ 1 § 1 ^ ^£ / ', 11^^ / -T'^- ^--- j "'' — ^ ■ -^-rj- -----, — — V — -- -- ^i- -- — — =" 'N — — — — <; "*T - 14 -s^ 1 5 J 1 ^ S 5J S § !« ^ 1 t X ^i 1 ± _ _ _ _ — — — — M. 1 1 M 1 Plate IV ^iftll a*— > /yd./4:- ^j)/?<7^ec/ /(7^r//7/AH'//A//j^p/c/fe/' 6//?/oer e/7(/ o/ fy'A/cA a/r cocy/c/ 770//:^ oA/o//7e(/; C. Aoa-/- ^ 0/7/(7 / /><7r of //?e I// J?^ ujor/'o/?/ ^/c/ti c/pon h/A/cA /Ae 6^roj/ea^; z5, /eoo/ So'je h^A/cA o/7cAoroa /Ae /c76/r//7/A /o /Ao Ao//o/?7 0/ /Ae ^W. Plate V Plate VI ~ "~ "" ~ ^ 5^ , " y ^ *= -_ ^ ^^ ■^ ^ > «^ •5^ > *i 1 " -> ^ % _- f ^ / --\ - ^ ^ ^ ^ "--■ ^ . __ — / ■^ ■N ' — / ,^'^' \ . " ^ — — ^ y "^ — — t ^ ^ y — — ^ 5 ^ ^ -^ " — ■— .^ ^> ^ Z^ < ^ — ■ 2^^ H ^ ^ - , -- ^ ^ ^ ^ ~ ~~ , ^^^ .^1^ ~~I ~— ^ — — ^ " f y ■— ^ 7^ / y y ~~ — . ^ y^ N 'J- ^1 r- ■J . — — " ~-^ " -^ "^ --■ ^ — 1 '<. ^ ^ < ^ ;^^ ■^ ■^ ^ ^ ~^ \ / V ^ / ^ J "~-J ■— — -_ ^ ,? ^ '1 \ ' " — ' r~ K, ^ —■ <5 ^ y , — ' ^ x^ 1 ■— / S ^ N --' «\ ^ > ^ ^ "^ ■^ ^ -- ^ U- !: T^ 1 - ^ — ^'i [K \ \ ^^ ,1 S- v^ [— s ^" ^ [> / ~" ■■ / r" -»- s \ ~ <^ / y ~^ ^ \' 1 ■-- ' ^ ~ "-■ V / / .^ ^^ \ ^ / / <;> ^ y -— ! ^ ■~> __ — ■ — ^^ ^' / \ ^ y ~ _^ / \ ^ y ' 1 ~~ — — - ' ■^ ^ ^ i! ~ ~ -\ ^ / 2 ~ 1 ^ 1 \ -^ L§S^S^§^4 ^ ^_j ^S S^^^^^^^^ ? ^ss^§^%^^^^l i ^ ^ ' _ ii _ L fh Plate VII ■S r 1 1 1 r — |— ~' - — ^- 5, _ ;{- ^ - ^ — !^ - !? - *) — ^- 5 — ^ — * — «a — ^ — Tl — ^ — ^ — \ — If ^^ 111 ^^^^ 1 \ / - -^ ^ "~ -, ^ — - r- 1 — - - ^ ^ ^ ■^ f> x^ ^ ^ ^ ^! 4 S;' .^ ? --5 \ ^ L^ '— <^ s --. ^ ^ ^ J^ -* ^ — ■ 10 ^ ~ -, - - — Ol / — -- — ^ ^ ~- 5 >■ - \ — "^ ^ ^ "< — .r, ? — — ^ ~ — ___ ^ ? \ ^•^^ >^^§^^^^$^ k' N •n ^ ^ ^^ §^ ^^^^^^!^§^ <<> - ;Oi \ \ 1 s^ •<3 "- '^ ^ ^ / r! -^ ^ ^ _ ^ - ^ " ;' — ■^ S \ •^ \ ~~ ~- , — ^ ^- - ^^- _^^ = 11: \ - ^ " ^ \ \ / ^- \ ^ ~— ..^ •^ ■v. 0, IS ■ — ^ \ \ \ / ^ -^ ^ ,^ / "^ ^ ,s / — 1" ^ ' / k 1 " ~^ ^ V > ^ / •^ \ ' r ^ ^ " - y 1 — — ~~ ^ ' \ — — ;^ \ ^ -1 "C; — — - ""J f -^ ■> ^ ^ ? r '1 2 ) I 3 4 f 1 , 1 1 , 1 1 1 1 1 E £ 1 1 5 L - ') "£ . - 1 - - - ^ ; 1 1 s- i 1 1 G 1 1 1 1 ^ nnnj 7 j — - - -> _J /^ tiUA EAST Figure I.— The four mazes used, IB, IIB, lA, and IIA. The heavy lines mark the division between the B- and the A-mazes. Dotted lines across bUnd alleys show position of the shutters. E and E' are the entrances to the mazes. GENERAL SCHEDULE OF EXPERIMENTS Rats used Practice distributions Maze IB Mu Group 7 females (9, 11, 12, 13, 14, 16, 18) and 1 male (10). Two tests, or runs, daily for ten days, then four daily for three days, then three daily for each rat until eight runs out of ten were correct. (Rat 12 did not complete the habit in time available.) S/ Group 4 males (2, 3, 4, 6), trained on Maze IIA. Three runs daily for three days, then an intermission for six- teen days (see explanation on next page), then bv the intensive method the rats were run Aug. 29th and 30th each three times at the following periods of day: 9-9:20, 9:40-9:55, 3-3:15, 3:30- 10 JOSEPH PETERSON 3:40, and 8:30-8:40,— total runs, twenty-four for each rat. All records were left incomplete, but all rats were equally practiced to the point of discontinuance. All animals were eager and active. Maze IIB Nil Group 8 females (15, 17, 19. 20, 21, 22, 23, 24). Rat 20 was blind in left eye. Distribution of practices pre- cisely same as for Group Mii, in Maze IB, same days. R/ Group 4 males (1, 5, 7, 8) trained in lA. Distributions of practice same as for St, Maze IB, same days, Practice periods, Aug. 29th and 30th: 9:25-9:35, 9:55-10:10. 3:15-3:30, 3:40-3:55, and 8:45-9. All records left incomplete; rats eager and active. Maze IA Ru Group 4 males (1, 5, 7, 8). Practice distributions same as for M», Maze IB. N/i Group 4 females (15, 20, 21, 22), trained on Maze IIB. Three runs daily for each rat until habit was completed, eight runs of ten correct. M/i Group 3 females (9, 13, 14). trained on Maze IB. By intensive method : the three animals were given three runs each, alternating with short periods of rest, during the forenoon of Aug. 28th. Rat 14 completed habit in twenty-eight runs, or trials: rat 9, in twenty-four runs; rat 13, in forty-one runs, eleven of which were made early the morning of the following day. All rats were eager and active, except 13 on the last run of first day, when it took sixteen seconds following two runs of two seconds each. Maze IIA Su Group 4 males (4, 6, 3, 2). Practice distributions same as for the Ru Group in Maze IA. N/o Group 4 females (17, 19, 23, 24), trained on Maze IIB. Practice distributions same as for N/i in Maze IA. M/i Group 3 females (11, 16, 18), trained on Maze IB. Practice distributions same as for M/i, and alternating after each three trials with them. Each rat completed the habit, getting eight out of ten runs correct, in a total of twelve runs. All were very active and eager throughout. This schedule is given as actually carried out, not exactly as originally planned. It will be noted that the programs for the two B-mazes are precisely alike, and that the same is true of the A-mazes. This affords means of control of a number of factors which otherwise might favor one or the other of the control groups. Temperature conditions changed considerably; it was also necessary to modify occasionally, to suit the time at the disposal of the experimenter, the number of runs per day by each animal. At the early stages of the learning there was not enough time to give each animal more than two runs EFFECT OF LENGTH OF BLIND ALLEYS ON ^L\ZE LEARNING 11 daily. Later four runs daily were tried, but the eagerness of the animals seemed in one or two cases to diminish in the last run. Three runs a day proved to be very satisfactor3\ It was originally supposed that each rat could learn both one of the B- and one of the A-mazes during the time available for the experiment — July 18th to August 30th — but a difficulty arose, which had been underestimated in the planning. When the male rats had finished their more simple problems — the A- ma?es — and were started on the B-maze problems, signs of trailing the females appeared. To prevent this possibility the male and the female groups of animals had been made to, occupy the same cages alternately in successive days. It was imprac- ticable to wash the maze thoroughly before each experiment for each group. The first day that the S^ and the Kt male rats were run in the B-mazes, after the runs of the females, there was no difficulty. On the second and the third day, however, there seemed to be evidences of trailing and of excite- ment, and some of the rats deposited urine drops in the maze from the second to the fifth blind alleys. This seemed to influence, as a guide, later members of the same groups (i. e., also males), and to stimulate them to make similar deposits along the trail. Thorough washing of the entire maze with soap water and Creolin-Pearson, a disinfectant, did not change the behavior materially. Consequently, after the third day the practices of these males were discontinued for sixteen days, until the females had completed their problem. This experience with the males seemed in only one (questionable) case to influence in the least the runs of the females whose habits had been already reduced to the stage of proprioceptive control. The mazes, moreover, had been carefully washed after the second and the third day of the experience with the males alread}^ described. The postponement of the experiment with the males in the B-mazes made it necessar}^ to run them by the intensive method described in the schedule, if at all. It was found that if each rat was given three runs and then put back into the cage without feed it could again be rim soon after with no loss of eagerness. In fact, the method worked surprisingl}^ well. The fact that the records had nevertheless to be left incomplete on this maze so far as these rats were concerned does not affect the data so 12 JOSEPH PETERSON far as they go, as the two comparable groups had identical experiences. Since the A-mazes were cleared earlier than the B -mazes, it was possible to put into them, as indicated in the schedule, some of the females — four on each A-maze — which first com- pleted their original problem." Finally, — leaving out the female 12, which did not complete its original problem until the maze was taken over for tie males, and the small male 10, which had been running with the female group Mm — three females were practiced on each of the A-mazes by the intensive method. The results of these two groups are, for obvious reasons, strictly comparable only to the twelfth trial inclusive, when the rats in the II A maze had completed their problem. All comparable, or control groups were then run on the same days, the same number of times, and as nearly as possible the same time of day. Moreover, to give no possible advantage of trailing to either group — and aside from the cases noted, not between control groups, no such behavior was observed — the group which was practiced first one day was second the next. Both time and error results were noted. The experimenter devised a system of short signs with which to record the com- plete gross behavior of each animal. Returns in the maze were noted as accurately as possible; only minor ones not reaching cul de sacs or corners of the various m.aze alleys before being corrected were left out of the records. Entrances into the blind alleys were all classified by means of appropriate signs, into three classes, — complete entrances, entrances about half way in, and beginning entrances bringing the animal's head and fore part into the blind alley while the hind feet remained in the true path. In the table of results these entrances constituting the last class are in the column headed " Start." It was also noted, as the animal emerged from the ciil de sac, whether it continued forward toward the food or turned back toward the place of beginning. Hesitancies were also noted. Of these a peculiar and amusing kind was frequent. Occasionally, an " To be sure that the two 1 16- 25 Ru 4 4 4 4 1 1 3 3 N/i 1 1 1 1 1 1 1 3 1 2 M/i 1 4 1 3 5 5 Su 2 6 6 Nt, Mh 1 26- 35 Ru 3 4 > 1 1 1 Nh 1 1 Mh 1 1 Su 3 1 1 1 4 4 36-45 R» 2 1 1 S. 1 46- 55 Ru . 1 1 Su 56- 65 Ru Su Totals lA 68 29 20 59 28 3127 9 18,51 8 43 Totals IIA 1 45 20 9 15 8 711 3 8|24 1 23 Total number of entrances into cul de sacs lA, 117 137 Total number of er itrances intr ) cul de sacs IIA, 74 50 20 JOSEPH PETERSON TABLE II — Continued General Summary of Results in the A-Mazes Blind alley Third Fourth Degree of entrance Compl. ] Half Start. Compl Half Start. Direction E R F E R F E R F E R F E R F E R F Runs Group 1- 2 (Rn lA N/i 4 1 3 6 3 3 12 8 4 2 -> 7 5 2 3 1 2 2 1 1 15 9 6 IM/i 1 1 1 1 1 1 4 1 3 (Su IIA N/, 2 2 3 1 2 1 1 1 1 5 2 3 2 1 1 2 C) [Mh 1 1 1 1 1 1 1- 5 Ru 6 1 5 7 3 4 14 9 5 5 5 11 6 5 Nh 6 2 4 3 1 2 18 9 9 3 1 2 3 3 Mh 8 6 2 2 2 2 210 1 9 1 1 1 1 Su 7 3 4 2 2 12 6 6 3 1 2 5 1 4 N/o •) 2 1 1 3 3 5 2 3 9 1 1 2 2 M/2 1 1 1 1 9 2 6-15 Ru 15 6 9 5 5 4 4 20 11 9 3 1 2 2 2 Nh 7 2 5 5 5 4 4 6 1 5 1 1 6 6 M/i 10 3 7 2 2 1 1 7 2 5 1 1 6 6 Su 6 3 3 6 6 1 1 8 3 5 4 4 8 8 Nt-2 1 1 5 5 1 1 1 1 4 1 3 M^2 1 1 16-25 Ru 5 1 4 5 5 6 6 7 1 6 2 1 1 6 6 Nh 3 2 1 1 1 6 6 Mh 1 1 6 6 4 1 3 1 1 1 1 1 1 Su 1 1 5 1 4 10 10 2 2 4 4 Nl2 2 2 1 1 M/, 26-35 Ru 3 3 4 4 4 4 1 1 3 3 N/i M/i 2 2 2 1 1 1 1 Su 5 3 2 1 1 8 8 1 1 4 4 6 6 36-45 Ru 3 1 2 4 4 4 4 1 1 1 1 Su 1 1 1 1 8 8 46-55 Ru 1 1 1 1 3 3 1 1 1 1 Su 1 1 1 1 3 3 56-65 Ru Su 1 1 Totals lA 64 25 39 35 35 43 6 37 90 35 55 21 11 10 47 6 41 Totals II A 25 10 15 14 1 13 ?? 33 28 12 16 18 2 16 49 9 10 Total number of er it ranees into ciil de sacs, I A 142 158 Total number of ei itra nces mtc ) cttl de sacs, IIA 72 88 EFFECT OF LENGTH OF BLIND ALLEYS ON MAZE LEARNING TABLE 11— Continued General Summary of Results in the A-Mazes 21 Blind alley Fifth Sixth Degree of entrance Compl. Half Start. Compl. Half Start. Direction ERF ERF ERF ERF E R FE R F Runs Group 1- 2 fRu Ia\ Nh [M/i fSii IIA N/, 2 2 1 1 2 2 4 4 6 1 5 1 1 3 3 1 1 2 2 2 2 3 3 1 1 1 1 1- 5 Rm Sm M/2 3 3 6 6 2 2 10 10 9 2 7 2 2 1 1 4 4 1 1 8 8 2 2 1 1 4 4 1 1 1 1 6-15 Rm Nil Sm N/2 3 3 1 1 4 4 1 1 3 3 2 1 1 4 4 1 1 2 1 1 4 4 3 2 1 1 1 1 1 16-25 Rm Nh Sm N/2 M/2 1 1 1 1 26-35 Rm N/x Sm 1 1 1 1 1 1 36-45 Rm Sm 1 1 46-55 Rm Sm 1 1 1 1 1 1 1 1 56-65 Ru Sm Totals lA 6 6 Totals IIA 9 9 4 426 2 24 10 1 9i 5 1 4 11 11 14 3 11 1 16 6 3 3 Total number of entrances into ail de sacs, lA, 36 Total number of entrances into cul de sacs, IIA, 24 18 17 22 JOSEPH PETERSON RESULTS Tables I and II give in a condensed form the main results of the entire experiment. In the separate larger divisions are given the reactions to the several blind alleys. These reactions are classified in a manner most easily made clear by taking up a concrete case. In table I the words " First," " Second," etc. at the top stand for the blind alleys of the B -mazes of the corresponding numbers. The results of the first blind alley, for example, are then divided into three parts, " Complete," " Half," and " Start," meant to designate the degree of entrance by the rats into the blind alleys, as already explained. Complete entrance means going entirely to the end of the blind alley, or so near the end that the animal might reach the end by means of the vibrissae. Frequently the rats ran against the end with considerable force. Half entrance means approximately half way, or all entrances between complete and beginning. Those marked " Start " include cases in which the animal either just put the head in or entered Mdth the fore half of the body. In such cases the hind feet of the animal usually remained in the true path, so that the general orientation was not completely given up as in the other two cases. The three columns coming again under each of these rubrics show respectively, the number of entrances into the blind alley in question, E, the number of returns toward the place of starting in the maze on the rats' emerging from the blind alley, R, and the number of times the animals kept the general orientation, i. e., continued toward the food box, F. The totals for R and F must therefore equal the number under E. The figures in the left column of the table indicate the number of the run, or of the test, of the animals, while the letters Mm, S/, etc. stand for the group, as Group M wwtrained. Group S trained, and so on. The description of each group and of its practice distributions are given in detail in the schedule, pages 9 and 10, to which frequent reference is advisable. Now, to illus- trate in a concrete case, in the first line of the data, giving results of the first two trials of the animals, we find that Group Mm (eight rats, untrained, running in Maze IB) made nineteen complete entrances into the blind alley No. 1, with ten returns and nine cases in which the rat continued forward toward the food box. There were two entrances half way, with two forward runs and no returns; thirteen beginning entrances, with four EFFECT OF LENGTH OF BLIND ALLEYS ON ^L\ZE LEARNING 23 returns and eight cases of the animal keeping its general forward orientation; and so on, through the results for all the blind alleys in order. Note that the figures for groups 'Mii and St are in bold face for blind alley 1, and not for 2, and that these relations are just reversed for the groups of animals running in Maze I IB. The bold face designates full length cul de sacs, and the figures not in bold face indicate that the blind alley was shortened. The amount of shortening in any case is shown in figure I, as already explained. Careful attention to all these matters will greatly aid the reader in getting quickly and conveniently the general results of numerous reactions. Without such attention the tables are meaningless. The results cannot so well be effectively and accurately shown in graphs. The totals at the foot of the columns must not be taken too seriously, as will be evident in subsequent discussion. These are totals only of changing comparative quantities. For this reason the results of the experiments have been classified for different periods of the training. The results of the first two trials are given separately — and are not added in the totals because they are again included in the data for the 1st to 5th trials — as they are least affected by the animals progressive training. They show us approximately whether mere chance, or probabilit}^ laws, can explain the direction that an animal beginning in the maze takes on emerging from a cul de sac, whether it returns or continues forward keeping its general orientation, — not accurately, however, for learning begins from the very first experiences in the maze. The progression of the learning in the case of each particular cul de sac is shown by a gradual decrease in entrances in the summaries of the 1st to 5th, the 6th to 15th, 16th to 25th, etc. trials; also by the gradual decrease in returns and the increase, correspondingly, in the number of cases of keeping the general forward orientation. These two kinds of changes are very interesting and illuminating toward showing, in a manner not hitherto done with data on learning, just how the cul de sacs are eventuall}" eliminated. '^ This is our main concern in this paper. " Professor Carr has pointed out that the ex'tent of entrance to cul de sacs grad- ually decreases, as well as the number of entrances. (Hicks, V. C, and_Carr, H. A. Human Reactions in a Maze. Jour. Animal Behav., 1912, 2, 98-125. See par- ticularly page 116.) 24 JOSEPH pp:terson Three important features of the resuhs are to be noted. The first is the rapid decrease in the proportion of the returns to forward runs, on the rat's emergence from bhnd alleys. With the exception of blind alleys 2 and 5 in the B -mazes few such returns were made after the 15th trial, though the animals continued to enter some of the blind alleys beyond the 75th, some even beyond the lOOth. These cul de sacs, noted as AVERAGE PER TRIAL PERIOD t-5 6-15 16-25 36-45 56-G5 TRIAL PERIODS Figures II and III. — CE is the curve of decrease in average number of complete entrances per trial to Blind Alley 1 full length, Maze IB; R indicates the decrease in returns, and 2R twice the returns, from this citl de sac. C'E', R' and 2R' show corresponding data for Blind Alley 1 shortened. Maze IIB. Eight untrained rats in each case. EFFECT OF LENGTH OF BLIND ALLEYS ON ^L\ZE LEARNING 25 exceptions, have directions such as to favor returns in the case of a rat emerging from them. This more rapid decrease in returns than ' in entrances to cul de sacs is least comphcated, and also shown most emphatically, in the case of the complete entrances to cul dc sac 1, which is encountered before the rat could be confused by running into any other blind alleys. Figure II shows the matter graphically. Curve CE represents the number of complete entrances of eight untrained rats to the first blind alley at full length, as in Maze IB; curve R, the returns; and 2R, twice the returns. 2R is a better curve for comparison with CE because originally, i. e. before an animal is at all practiced, about half of the entrances are followed by returns; twice the returns, therefore, gives a number initially about equal to the total number of entrances. Figure III gives corresponding curves, C'E', and 2R', respectively, for the same number (eight) of untrained rats in cul de sac 1, shortened from 22 inches to 8.5 inches. Here the same result is evident: while the elimination of entrances is far more rapid than in the case of the longer blind alley, the returns are still more rapidly reduced as shown by the 2R' curve. It may also be noted here under our first point that the returns in both the B-mazes persisted longer in the cases of the blind alleys farther from the food box than of those near it. That is, returns from blind alleys first encountered were less easily eliminated, as were also entrances, than from those further along the true path. This is true even in cases of blind alleys nearer the food box that were comparatively long, as 7 and 8, even though, as in the case of 8, the direction of movement in emerging from the cul de sac favored returns. It is barely possible that an odor factor may have entered in case of 8. The mazes I A and II A are not so well adapted to show these relationships, as there are few^er blind alleys of various individual differences of complexity, but the same conclusions as those given for the B-mazes may also be made for them. A second important point to note is, that the nature of the response to a blind alley gradually changes with practice, as well as the relative number of entrances into it. This change in the nature of the response is more marked in longer than in shorter blind alleys, particularly in those whose elimination was most difficult. It is illustrated best in the data from cul de sac 26 JOSEPH PETERSON 1 of Maze IB. Many of the entrances to 2, as the observation of the animals in their responses and also their individual records showed, are clearly due to confusions resulting from entrances to 1. The experimental notes supply many evidences. As a rule the rat in the early stages of response to such a blind alley runs rapidly into it the entire distance, usually coming into L-5 6-15 16-35 36-45 56-65 TRIAL PERIODS 96-105 116-125 Figures IV and V. — C shows decrease in average number of complete entrances per trial to Blind Alley 1 full length in the B-mazes; H same for half-way entrances, and S for beginning entrances. C, H', and S' show corresponding results for the blind alley shortened. Eight untrained rats in each case. EFFECT OF LENGTH OF BLIND ALLEYS ON MAZE LEARNING 27 contact one way or another with the end; but with succeeding trials the entrance is less and less complete, until finally the impulse to enter is wholly inhibited. Thus in the records of responses of two groups each of eight untrained rats to the first blind alley in the B-mazes (table I) the large numbers in the E-columns shift gradually from, the " Complete " through the " Half " to the " Start " column. This shift is graphically shown in figure IV and figure V for first blind alley of mazes IB and IIB, respectively. C and C are the curves representing the rate of elimination of complete entrances, H and H' of half entrances, and S and S' of beginning entrances. Note that while the C-curves fall rapidly from the first, especially the one (C) from the shortened cul de sac, there is a decided rise in the H- and the S-curves. Specifically, in the case of Maze IB (the cul de sac long), C falls gradually, with two minor excep- tions, all the way at a nearly uniform rate; H rises almost uniformly to the 35th trial, then it keeps almost a uniform height to the 65th trial, and finally gradually declines; and S, after a rapid initial decline, gradually rises again until the 65th trial is reached, when it gradually declines and reaches zero before the other two curves. In the case of Maze IIB {cul de sac shortened) the same relationship between these respective curves is shown, though all these curves drop earlier in the process than with the longer blind alley, except that in this case the S' curve holds out longer than either of the other curves. A cursory examination of the data for other blind alley records shows that this type of transition from complete to only partial entrance and then to final elimination is a general feature of the results for the different groups of animals in the various mazes. A few exceptions only, in cases of very short cul de sacs, are noticeable. This is a phenomenon of learning in the maze to which little attention has previously been given, and which seems to the writer to be inexplicable on the basis of mere fre- quency and recenc}^ laws. Several impulses working together, some facilitating others inhibiting one another, gradually result in the survival of the most consistent, or complete acts. No hesitancies in the rats' behavior in these cases were present, such as might be secured from persons in similar circumstances. The rats evidently did not have time, nor adequate sense organs and conscious memories as a person would have, to recognize 28 JOSEPH PETERSON and take note of external stimuli, but resembled automatic machines in the quickness and uniformity of their responses. This change appears more significantly in the results of most of the individuals than in those of all averaged. Here are some examples. Rat 18, of Group Mw made entrances to the first bhnd alley (full length) in this order: 12c (complete), Ih (half), Ic, Ih, 5c, Is (started), 4c, Ih, 2c, Ih, 5c, Ih, 4c, Ih, Ic, 4h, 2c, 3h, 3s, Ic, Ih (total 55 entrances). Rat lO's record, same group, is Is, Ic, Ih, 9c, Ih, 7c, 2h, 3c, Ih, Ic, 2h, Is, 2c, Is, 2c, Ih, Is, Ic, Ih, 2c, Ih, Ic, Ih, Ic, Ih, Ic, 3h, Is, Ic; the next time on passing this cid de sac there was a momentary pause with three very rapid in and forward vibrations of the head, causing a confusion in which the animal made eleven errors in the other nine citl de sacs none of which it had entered, with but one exception, for twelve trials; then Is, 2c, Is, 2c (after another such vibrating pause before the cul de sac), 2s, Ic, Ih, Is, 2c, Ih, Ic, Ih, Ic, Is (total 70 entrances). These results are typical. In numerous cases when the habit of avoiding the cul de sac was nearly complete, so that the animal usually made the " s " type of entrance, the peculiar rapid vibration of the head noted above took place. The pause was, however, but for an instant. This response seems to indicate that the impulse to go forward at the critical place is still partly checked or impeded l:)y one to enter, not quite eliminated. It is important to note, more- over, that when finally the rat does succeed in passing the cid de sac, even when this hesitant, vibrating behavior does not take place, it very frequently runs headlong into some neighbor- ing cul de sac which had long since been inhil:)ited, and thereby gets considerably confused. Frequently, after such an experience it makes a complete entrance into the cul de sac in question the next trial, just as a child " speaking a piece " must bow again and start over when she goes wrong. This is one reason why a few complete entrances continue to occur. More than once an animal which had successfully passed cid de sac 1 for several trials would, without any hesitancy, run into it with great speed and against the closed end with terrific impact. In one such case the animal's whole maze habit, just on the finishing stage, seemed to have been temporarily jolted wholly out of gear, its next trial being much like that of a beginner. All this makes it very plain that maze habits are not to be explained on the EFFECT OF LENGTH OF BLIND ALLEYS ON MAZE LEARNING 29 basis of individual, disparate " acts," following in their occurrence some law of chance. "^On the contrary, the various impulses in the random activity of the early trials are gradually and collectively woven into one matrix of successive responses, each setting off the next succeeding one, and all shaped by the whole lyj ifljii 9 /O 3 BLIND 4 S /ALLEYS Figure VI. — Heavy columns, double lines, single lines, and discontinuous lines show, respectively, total entrances by all animals to full length and shortened blind alleys in the B-mazes, and to full length and shortened blind alleys in the A-mazes. Figures above columns give the totals represented. 30 JOSEPH PETERSON circumstance of the maze enviromnent.'^ This seems to iriiply that the effect of one stimulus holds over into and conditions effects of later stimuli. ■ The third point to note in our results is that when any given cul de sac is shortened it is eliminated more readily than when left at full length. That is to say, other things equal, and within certain limits, a long cnl de sac is eliminated less readily than a short one. This statement is amply borne out in our data both from the A- and the B-mazes. The general results of all our experiments are shown roughly in the accompanying diagram, figure VI, representing the total number of entrances to each of the blind alleys in the various mazes. The heavy black columns and the double lines represent the totals for the full length and the shortened blind alleys, respectively, in the B-mazes; the single continuous and the broken lines stand for the corresponding totals for the A-maze blind alleys. In the B-mazes the total entrances to the full length ctd de sacs is 1311, while the total number of entrances to the same ctd dc sacs when shortened, by an equal number of animals under the same conditions, is 929, a decrease of 29%. This decrease would doubtless be considerably greater but for the fact that confusions by the long blind alleys resulted in random behavior which increased the totals for the shortened cul de sacs. For instance, table I shows that more entrances were made into 2 short than into 2 full length. This was very clearly due to the fact that as long as the habit to avoid 1 was incomplete the animals in the confusion also entered 2. It will be recalled that in the A-mazes the cid de sacs were all full length in the one and all shortened to about half their length in the other case. Here we do not have the confusion noted in the B-mazes. The shortened blind alleys were entered 47% fewer times than those of full length. This bears out the conclusions drawn from the B-mazes. The effect of shortening the ciil de sacs was most noticeable in the case of 1 in the B-mazes, which was by all means the most difficult to eliminate. Being the first to encounter, it was '^ On this point our results agree with some aspects of those by Peckstein, L. A. Whole vs. Part Methods in Motor Learning: a comparative Study. Psych. Man., Ser. No. 99, 1917. " Each aspect of the course is no doubt associated with and located in reference to all the details of the course and to the entire objective en- vironment as well." P. 30. EFFECT OF LENGTH OF BLIND ALLEYS ON MAZE LEARNING 31 likely the least complicated by the results of entering other blind alleys. Table III and figure VII show comparatively the rate of elimination of all entrances to this cul de sac full length and shortened, 22 and 8.5 inches respectively, by the two groups of rats, Mu and Nw. While the two curves start near together the one, S, representing the entrances to the shortened cul de sac drops rapidly after the 15th trial; the other one, L, after the initial decline keeps nearly the same height to the 55th trial. The percentage eliminations are shown for the long and for the shortened cul de sac, respectively, by curves E and E'. TABLE III Elimination of All Entrances to Blind Alley 1. Two Groups of Eight Rats Each. Mazes IB and I IB Trials 1-2 1-5 6-15 16-25 26-35 36-45 46-55 Av. No. of entrances to blind alley 1, long. . . 16.5 11.1 7.7 7.8 6.6 6.4 6.0 Per cent 100.0 67.3 46.6 47.3 40.0 38.8 36 4 Av. No. of entrances to blind alley 1, short 'd. 13.0 8.8 7.2 4.1 1.8 1.1 1.1 Per cent 100.0 67.7 55.4 30.8 13.8 8.5 8 5 TABLE III— Continued Elimination of All Entrances to Blind Alley 1. Two Groups of Eight Rats Each. Mazes IB and IIB Trials 56-65 66-75 76-85 86-95 96-^105 106-115 116-125 Av. No. of entrances to blind alley 1. long. . . 5.0 3.5 2.5 .8 .6 .6 0 Per cent 30.3 21.2 15.2 4.8 3.6 3.6 n Av. No. of entrances to blind alley, 1 short'd. 2 .1 .1 0 0 0 0 Per cent 1.2 .8 .8 C 0 0 0 The results from cul de sac 2 are 221 entrances to the full length (40 inches) and 234: to the shortened form. This would appear to contradict our general conclusion. However, it must be remembered that the rats for which 2 was shortened made 32 JOSEPH PETERSON 346 more entrances to 1 (long) than did the control animals for which 2 was left full length; they also made 18 more returns to the starting place in the maze. This not only required that 2 shortened be passed more times than 2 long, but also with greater probability of entrance for each time. It was noted that rats entering 1 were likely thereby to be thrown out of Av. number Per Period 96-i05 116-135 Figure VII. — L shows rate of elimination of all entrances to ciil de sac 1 long, Maze IB, and S same for 1 short, Maze I IB. E' and E show corresponding percentage eliminations. orientation and to make other errors. There can therefore be little doubt that if 2 had been the only blind alley in the maze, it would have formed no exception to the general rule. The greater number of entrances to 4 shortened than to 4 long (86 to 69) is due to the fact that rats emerging from 5 had a strong tendency to run into 4. A glance at the maze will show why this is to be expected. There were 32 more returns from 5 long than from 5 short; furthermore, entrance to 5 long had the greater tendency to disorient the animal so that entrance to 4 would be an increased probability. Just why 8 short should have been entered 51% more than 8 long is not easy to EFFECT OF LENGTH OF BLIND ALLEYS ON ^L\ZE LEARNING 33 determine. There were, moreover, 14% more returns from the entrances in the former than in the latter case. It is possible that the rats entering 8 at full length, which runs along side the food box, had time and opportunity to get sufficient odor from the food to influence them against returning. Accidental factors may have been the cause in part; half of the entrances were made in the first two trials, and the total numbers are too small to indicate with much probability the actual trends. On the whole there can be no c^uestion that, other things equal, entrances to short cul de sacs are more easily eliminated than entrances to long ones. The results from both types of mazes used in this experiment (see tables I and II) show that on the whole cul de sacs first encountered in the maze were entered more frequently, and that the impulses to enter them were overcome with more difficulty, than were those occurring further along the true path, or nearer the food. In this respect our results are in agreement with those of Miss A'incenfi'' and contrary' to those of Miss Hubbert.'" While in the present experim.ent, not intended especially to test this point, the bearing of the results is necessarily complicated by an inequality of the lengths of the various blind alleys, there is no evidence to show that results would have been different with ctd de sacs of equal lengths and of equal direction difficulties. In the B-mazes. for example, 6 and 7 were much less troublesome than 3 and 4, in many respects similarly located with respect to the correct path, and all of equal length. By all means the most difficult cnl de sac to avoid entering was 1, even when shortened to 8.5 inches. The total entrances to 6 and 7 long are 101, against 142 to 3 and 4 long; to 6 and 7 shortened 83, against 130 to 3 and 4 shortened. The total number of entrances to 1 short are 275, whereas the totals to 6, 7, 8, 9, 10 full length amount only !o 192. It seemed that the rats got rather firmly registered in their proprio-ceptive system of controls the tendency to make two successive turns of 90 degrees each to the right, beginnino- 18 Vincent, Stella B. The White Rat and the Maze Problem —IV. The Number and Distribution of Errors: a Comparative Study. Jour. Animal Beliav., 1915, 5, 367-374. " The final members of the C2tl de sacs were entered less frequently'- and eliminated first." P. 374. 1" Hubbert, Helen B. Elimination of Errors in the Maze. Jour. Animal Behav 1915, 5, 66-72. 34 JOSEPH PETERSON at the corner of the maze before cul de sac 1, and that since the turns were so close together they tended very persistently to fuse together into a single turn of 180 degrees, thus taking the rat into the blind alley. It was very interesting to see certain rats continue to run into 1 with almost monotonous regularity for three weeks, three trials each day, while other errors, errors of entering other cul de sacs, occurred ver^' seldom. Thus from the 10th to the 79th trial, inclusive, rat 9 made 60 errors of entering 1 with only 11 entrances to all the other nine blind alleys ; rat 1 1 from the 24th to the 83rd trials made corresponding errors of 47 to 15. In the A-mazes cul de sacs 5 and 6 were likewise entered fewer times and eliminated more easily than 1, 2, and 3, all of length equal to that of 5 and shorter than 6. It is, of course, not contended here that the two sets of blind alleys compared are of equal difficulty in all respects other than that here con- sidered. At the same time, they may be approximately equal; that is a matter which can be determined only empirically. The accompanying table (table IV) shows that not only is the number of entrances to blind alleys first to be passed along the true path greater than that nearer the food box, but also that the percentage rate of elimination is greater in the latter. This is shown by comparing the number of entrances to the different groups of cul de sacs in question for different successive periods in the learning process from the first to the last trial. In the first five trials of all the animals, trained and untrained, the average number of entrances per trial into cul de sacs 1-4 of the B-mazes is twice that of entrances into 6-10. Calling these numbers for the first period (the average of the 1st to the 5th trial) 100% each, to get a common basis for comparison, we find that there is a much more rapid percentage drop of elim- ination of entrances in the case of the blind alleys nearer the food box. Since the trained rats discontinued the experiment with the 25th trial without finishing the habit, the percentages for the two groups in the B-mazes are not correct after the 25th trial, though they are strictly comparable. An additional line is given, in the case of each of these groups, of the accurate percentages of elimination of entrances for the untrained rats (eight in each group) alone. It will be noted that in the case of the five cul de sacs nearest to the food box the percentage EFFECT OF LENGTH OF BLIND ALLEYS ON MAZE LEARNING 35 of elimination is considerably more rapid than in that of the first four blind alleys encountered. In the case of the A-mazes the percentage elimination is considerably greater for cul de sacs 5 and 6 than for 2 and 3. Figures VIII and IX represent graphically the data of Table IV. There is no room to doubt that the blind alleys first to be passed along the true paths in the mazes used are both more frequently entered and more slowly eliminated than are those further along the trail. TABLE IV Pe riods of trials 1-5 6-15 1&-25 26-35 36-45 46-55 (V fBlind allev 1 Blind allev 2 134 147 79 89 210 76 11 16 166 68 5 19 84 39 6 8 75 33 2 6 71 24 Blind alley 3 7 Blind allev 4 6 Totals Av. per trial 449 89.8 100.0 100.0 313 31.3 34.7 27.6 258 25.8 28.6 25.5 137 13.7 15.2 19.7 116 11.6 12.9 16.7 108 10 8 Per cent, 24 rats 12 0 Per cent, 16 untr'd rats 15.6 fBlind alley 6 Blind allev 7 Blind allev 8 42 28 62 49 36 28 33 15 7 14 9 18 5 4 3 1 10 3 5 1 2 6 2 3 1 2 2 3 N < Blind allev 9 Blind alley 10 2 1 S Totals Av. per trial Per cent, 24 rats 217 43.7 100.0 100.0 97 9.7 21.3 15.6 39 3.9 9.0 6.8 20 2.0 4.6 4.0 14 1.4 3.2 2.8 10 1.0 2 3 Per cent, 16 untr'd rats 2.0 fBlind alley 2 84 51 65 73 28 49 8 25 3 13 1 < Blind alley 3 3 ^l ci s Totals Av. per trial Per cent 135 27.0 100.0 138 13.8 46.9 77 7.7 26.5 33 3.3 11.2 16 1.6 5.6 4 .4 1 4 < Blind alley 5 Blind alley 6 32 23 21 9 2 0 1 2 1 0 3 1 Totals Av. per trial Per cent 55 11.0 100.0 30 3.0 27.3 2 2 1^8 3 o . ij 2.7 1 .1 .9 4 .4 3 6 36 JOSEPH PETERSON TABLE IV—Cflntiniied Period of trials 56-65 66-75 76-85 86-95 96-105 106-115 116-125 CQ 2 1 .1 .2 .2 4 .4 .9 .8 0 0 0 0 0 0 0 0 0 0 0 0 It may be that the odor of the food is a factor that at least partly explains the more easy elimination of the ciil de sacs nearer the food box. However, there is very little, if any, real evidence that such is the case. A crucial test would be to use anosmic rats, though other means of controlling the odor factor are easily possible. Some facts in the present experiment count against the influence of odor as suggested. For example, errors of entrance into cul de sac 10 are nearly as numerous as those of entrance into 9, although to get to 10 the animal had to pass a short alley of 8.5 inches leading directly into the food box. Moreover, all the rats, with occasional exceptions, i-* ran so rapidly after the first trial that it is improbable that food odor had any immediate direct influence in the behavior in the maze. There was no evidence in the behavior of the animals that they were attracted to the food box by such odors. '» In the cales of supposed trailing, already noted, the animal which appeared 1* Occasionally, without any apparent external condition to explain the behavior, an animal would sneak slowly and cautiously all the way through the maze. In a few cases such activity seemed to be due to recent fights with other rats or to noises from fights between other animals. '^ An exception should be made here of the case of returns from cul de sac 8, already discussed. The floor of the food box was covered with paper (double thickness) during the feeding each day, and during the experiment the food was kejDt in a dish in the extreme comer of the food box away from cul de sacs 8, 9, and 10. EFFECT OF LENGTH OF BLIND ALLEYS ON MAZE LEARNING 37 to be following a scent of any kind moved perceptibly more slowly, holding the nose continuously or frequently to the floor. The writer does not believe that the more rapid elimination of the cmI de sacs nearer the food is to be explained on the basis of scenting the food. The matter, however, needs further test. i-5 6-J.5 16-25 36-45 58-65 TRIAL PERIODS 96-105 U6-L-25 Figures VIII and IX.— F and L show percentage elimination of all entrances to cul de sacs 1-4 and 6-10 combined, respectively, by twenty-four rats in the B-mazes; UF and UL the corresponding data for sixteen untrained rats. F' and L' show the percentage elimination, respectively, of all entrances to blind alleys 2 and 3, and 5 and 6 combined, by twenty-two rats in the A-mazes. Do pure probability laws govern the returns of the rat on emergence from blind alleys' In the tables of results (tables I and II) the totals of the first two trials have been kept separate so that the percentage of returns from blind alleys toward the starting place in the maze could be found for a period Httle 38 JOSEPH PETERSON influenced by the effects of training. The following table (table V) classifies for easy comparison the results of all the rats on the first two trials. The entrances to cul de sac 1 in the A- mazes are not included as all emergences frorh this blind alley brought the rat to the place of the entrance to the maze. TABLE V Full Length Cul De Sacs Rats Maze Comp . Ent. Half Ent. Start. Ent. % Ret. of all E's. Ent. Ret. Ent. Ret. Ent. Ret. 8 untr'd 4 tr'd B B 170 44 76 16 35 7 6 0 54 14 16 3 38 29 4 untr'd A A 51 33 24 13 7 4 5 2 46 12 16 3 42 7 tr'd 36 Average per cent retur ns 1 ' ' ' 33.75 TABLE V— Continued Shortened Cul De Sacs Rats Maze Comp . Ent. Half Ent. Start. Ent. % Ret. of all E's. Ent. Ret. Ent. Ret. Ent. Ret. 8 untr'd 4 tr'd B B 182 42 72 13 27 4 11 2 48 20 18 5 39 30 4 untr'd A A 23 11 10 2 6 6 3 1 12 10 0 0 32 7 tr'd 11 Average per cent retun IS 23 It will be noted that in the B-maze the per cent of returns from the shortened cul de sacs are practically equal to those from the full length ones, both for the trained and for the un- trained rats. The returns for untrained rats are not far short of 50%. The shortage is mostly due, no doubt, to the small degree of learning that took place in the process of the first EFFECT OF LENGTH OF BLIND ALLEYS ON MAZE LEARNING 39 two trials, during which there was considerable random activity and reduction of excess movements. It would seem that at first — before any learning has taken place — the chance of a rat's returning on emergency from a blind alley is about one to one. There may be a greater tendency to go forward, keeping the general orientation rather than to return; if so, the excess for- ward tendency is but slight. The returns from cul de sacs first to be passed seem slightly to exceed in percentage those from blind alleys further on toward the food box. In the B -mazes the returns from cul de sac 1 (both full length and shortened) are 44% of the total number of entrances; the corresponding percentages for the other blind alleys in order from 2 to 10 are 55, 31, 32, 48, 33, 50, 34, 13, 33. These figures are taken, of course, only from the records of the untrained rats, sixteen alto- gether. Those most favorably situated for returns, so far as the rat's keeping the general direction on emergence from the blind alle}^ is concerned, are 2 and 5. This judgment is sup- ported by the data. It is not clear why the returns from 7 should run so high. The percentage of returns by the eight untrained rats in the A-mazes are, for the 2nd to the 6th blind alley, in order: 36, 33, 67, 0, and 0. The large number for 4 was to be expected. The greater number of returns from the cul de sacs first encountered is likely due to the fact that the animals had already learned something of keeping the general orientation before the other blind alleys were entered. In the B -mazes there appears to be a slight decrease in the returns of the first two trials by the trained rats as compared with the untrained. This seems to be due to a sort of " transfer of training." It is likely, as the writer suggested in the earlier article already referred to, due to a tendency of animals with experience in mazes to proceed with less whole-souled response into cul de sacs. Let us suppose that as an animal enters a cul de sac it also receives certain stimuli of various kinds from the true path from which it departs. These stimuli may produce a weak partial response, or tendency to response, which does not immediately fade away. If this tendency persists until the rat emerges from the ctil de sac it will, of course, enhance the impulse to take the true path and thus increase the probability of continued forward movement. It is not inconceivable that a trained animal may have developed a habit of keeping the 40 JOSEPH PETERSON correct general orientation by some such means as this. Such habits would then have common factors for all cul de sacs, and in mazes of different kinds. It would seem, too, that on some such basis as this the returns would be eliminated more readily than entrances to the blind alleys, as has already been shown to be the case. This explanation may involve an interaction of sensory and motor impulses in the nerve fibres — each sys- tem, sensor}' and motor, interacting upon and stimulating the other — in such a manner as to make comprehensible how the effect of stimuli may be carried over into later responses and partly condition them as suggested below. Possibly the animals also learn with training to utilize better such factors as vague visual stimuli of the closed end of the cul de sac. Certainly the speed of the rat running into the blind alley would make one cautious in assuming that such factors are explicitly reacted to by the animal. That there was a real transfer of some kind is, in any event, a conclusion which also finds support in the results of the A-mazes. For the full length and the shortened cul de sacs, the per cent returns for untrained rats are 42 and 32, respectively, agreeing rather closely with the B-maze results, whereas the corresponding per- centage returns by the trained rats — seven in each A-maze — are 36 and 11, a decrease from that of the untrained animals of 14% for the full length and of 66% for the shortened cul de sacs. In the B-mazes the percentage returns from the full length cul de sacs by trained rats is 24% less than that by un- trained rats; for the shortened blind alleys the percentage re- turns by the trained rats is 21% less than that by untrained rats. THE SIGNIFICANCE OF THE RESULTS It may be urged by the reader that the more rapid elimination of entrances to the shortened cid de sacs than to the full length ones is due to the fact that the rat, in the case of the short blind alleys at least, sees the closed end and thereby avoids entering so frequently, or so completely. In one sense this begs the whole question. Seeing is not some thing that stimulates or directs the animal; it is only a mode of being stimulated. Its possibility in the present study is not at all denied. The whole question with which we are concerned is: How do all possible EFFECT OF LENGTH OF BLIND ALLEYS ON ^L\ZE LEARNLNG 41 kinds of stimuli operate, directly or indirectly, toward the learning to avoid entering cul de sacs'" That the rat is not wholly blind has been demonstrated in a number of cases, ^o but there is no clear evidence to show that the presence of such visual factors as are possible to the rat could operate on the principle of frequency, recency, or intensity, or all combined, in such a manner as to eliminate the impulses to enter the cul de sacs under the conditions of the present prob- lem. They might, of course, aid the rat in getting to the food at any one time, but how could they operate toward cutting short the random processes in successive trials, i.e., in bringing aVjout what is called learning? A brief review of the work on visual controls in the rat's behavior is to be found in Miss Vin- cent's paper. Waugh found-' that though the mouse could perceive the distance of objects " within a range of 15 cm.," it nevertheless seemed not to make use of the " visual percep- tion of depth " in getting past two partitions each from oppo- site sides reaching half way across the problem box, the one being nearer than the other. In the present experiment, it will be recalled, the interior of the maze was stained black, and even if it be granted that the rat could see the ends of some of the shortest cul de sacs there would be but little difference in the visual stimuli between the " blind " and the open alleys, in as much as both were obstructed alike in the further end and the side opening of the latter was not directly visible. Differences in brightness would be irregular and but slight, as the room was lighted from three sides — south, west, and slightly from the north — and an electric light was directly over the maze. It should be said that no difference in behavior between the rat blind in the left eye — No. 20 — and the other rats was noticeable though the experimenter kept watch for such difference. More careful visual controls are of course desirable. But the real question is how any stimulus, visual or otherwise, must operate together with other stimuli so as to inhibit unsuccessful acts and to cause to survive those acts which bring 2" See Richardson, Florence. A Studv of the Sensorv Control of the White Rat. Psychol. Mori., Ser. No. 48, 1909. \incent, Stella B. The White Rat and the Maze Problem — I. The Introduction of a Visual Control. Joitr. Animal Behav., 19 15, 5, 1-24. -' Waugh, K. T. The Role of Vision in the Mental Life of the Mouse. Jour. Camp. Neurol, and Psychol, 1910, 20, 549-599. 42 JOSEPH PETERSON success, in this case those acts which bring the animal to the food box. The results follow the series of stimuli and responses which take the animal through the maze. How can the result work backwards? The writer believes that in the foregoing pages he has presented plausible reasons and data to show the absolute inadequacy of frequency and recency laws as the direct- ing factors in maze learning. Frequency fails to give any basis not only for this kind of learning in general but particularly for the specific kinds of results obtained in the experiments considered. In a complex situation like this, frequency explains only how within a certain probability the rat will finally reach the food, but it fails to explain why subsequent trials should be improvements on the first one. It is not clear how recency, as ordinarily understood, can aid the learning. The principle of intensity needs re-interpretation. When several stimuli act on an animal bringing about a series of responses as in this case, the final one of which is the successful one, it appears that somehow, not well understood yet, the various effects of these stimuli hold over into that of the final stimulus and that all together simultaneously act to direct the energy of the animal into the most consistent channels. In the large, these channels offer the least resistance and afford the most complete response. It is in this sense that the successful acts are more intense than others, and thus their effect is greater toward shaping the neural pathways for their repetition and for the gradual elimination of the more inconsistent and tentative responses leading up to them. On this assumption it becomes somewhat comprehensible why the maze is learned to a large extent " as a whole," so that small errors may throw the animal out completely, or at some other part of the maze, when the habit is nearly per- fected. The specific results of the present experiment are also intelligible. These various hold-over effects in the extero- and the proprio-ceptive systems afford the basis of imagery in human behavior, and supply the " large situation " to which one reacts ideally. They may function, so far as we can know, wholly unconsciously or with but vague consciousness in the case of the rat. In the human being habits of responding to separate groupings of these factors may be acquired, and such exciting factors may be aroused indirectly by association. Nothing is gained in psychological explanation by assuming " ideas " to EFFECT OF LENGTH OF BLIND ALLEYS ON ^L\ZE LEARNING 43 explain behavior, unless in such cases we understand how the ideal dispositions themselves are acquired. The use of the term idea in the higher forms of behavior is justified then only on the basis of simplicity of statement. There is none but questionable evidence thus far that ideational behavior is different in any way but degree from sensori-motor, or the well known trial and error, behavior. " Ideas " can function only when the somewhat detachable dispositions, of which they are the imperfect, subjective aspects, have been built up by experience, and such dispositions require a rather complex nervous mechan- ism. It is needless to say that no evidence of ideational behavior has been found in the white rat. While, as has been pointed out in the foregoing, there are likely some hold-over effects of stimuli in the case of the rat, these likely operate more or less mechanically and en masse so that the animal enjoys little independence of action and is subject rather completely to the dominance of the group of stimuli present or immediately past. That is to say, the animal can respond only to present situations though with a considerable number of random variations, until the most consistent responses to that situation have fixed them- seUes to the exclusion of all others, after many repetitions of tiials. Then the response becomes uniform and mechanical to a high degree. The more advanced behavior as we see it in the case of man — ideational behavior — differs from the lower forms illus- trated in the present study in that it is less fixed and less dependent upon immediate situations. Stimulus-response organ- izations, or tendencies, are more detachable in their separate smaller functional components; and the latter have richer pos- sibilities of combinations among themselves, on the one hand, and on the other there is less dependence for their functioning upon direct or immediate stimulation. Various indirect and vicarious stimuli come to serv^e adequately. Thus various systems of stimulus-response mechanisms may become organized into inconceivably complex relationships about certain symbolic stimuli, such as written or spoken words, various kinds of gestures and attitudes of the stimulating individuals, associated objects, sounds, contacts, and so on. It then becomes practically impos- sible to predict which of the various aspects of the situation will succeed in calling out its particular response. We shall 44 JOSEPH PETERSON not here enter into further consideration of this complex behavior, except to point out that when the various stimulus-response mechanisms have become sufficiently well associated with certain muscular strains or neural excitations, the revival of the latter by favorable stimuli will call out the acts themselves. Thus a stimulus may have entirely ceased to play upon the sense organs from without and long periods of time may have elapsed, and yet, because of this acquired organization, the recurrence of an^^ significant aspect of the outer situation, even such as a sound associated with it, may revive the crucial exci- tation and thus call out the act. Something of this kind — stimulated, however, by the original situation minus the light when the animal is allowed to respond — likely takes place in the delayed reactions of animals, though this assumption leaves entirely open the question as to whether or not the animals have ideas, a rather infertile question for science, it must be confessed. More elaborate systems of acquired associations make possible the continual thinking of absent situations which we know that ive ourselves experience. In these more *:,dvanced forms of behavior groups of response systems may come so to interact upon one another by associations and by stimulation from the inward bodily conditions that rehearsal of a problem mentally may take place long after actual practice has ceased, thus changing behavior materially between practices. It is yet c^uestionable whether there are any such cases in animal behavior. -2 In the foregoing pages we have called into question the prin- ciples of frequency, recency, and intensity of stimulation as usually understood in relation to the fixing of associations, so far as their value in explaining learning is concerned. They do not seem to account for the change in successive trials called learning. This seems to be true at any rate for maze learning; probably it holds for all kinds of learning. All that these fac- tors do is, likely, to make more and more easy any associations and acts brought about by the real directing factors. That is, they tend to fix any series of acts in the order that they are gone through, not to change the order of the acts. Some other directing factors and some vis a tergo must be found to account 22 Cf. Yerkes, R. M. The Mental Life of Monkeys and Apes: A Study of .Idea- tional Behavior. Bchav. Mon., 1916, 3, No. 1. Yerkes thinks Julius, an orang- utan, solved a problem ideal ionally; see particularly pp. 68 and 131. EFFECT OF LENGTH OF BLIND ALLEYS ON MAZE LEARNING 45 for the changes in behavior which gradually make response more and more direct and which gradually eliminates the use- less random acts. We must not forget that the numerous internal life processes, e.g., the contractions of the muscles of the stomach with hunger, serve as the motivation to activity. They determine the stimulating value, as do also modifications in the proprio-ceptive system by past behavior, of various outside factors. The organism continues to respond by var\'- ing behavior until successes are attained which modify these internal conditions and change the inner motivating factors. But the failures also change the organism. The directing fac- tors of the response seem to be the inner organic processes and the total combination of stimuli from external conditions and from muscular contractions, all these overlapping in their several effects as has been suggested. The neural channels involved in the most consistent acts become the most opera- tive through the compelling effects of all these factors, and these acts, or directions of response, in time survive over all others and gradually acquire an ease and automaticity of func- tioning characteristic of habits. The stimuli to action even in as simple an organism as a rat are infinitely more complex than usually imagined in our " neural explanations." Mere contin- gency in the combinations of acts of a rat brought about in the maze, or in other problem boxes, for that matter, cannot be regarded as the important factor that it has sometimes been supposed to be. It is true that some useless acts may occa- sionally survive with the more consistent ones by chance asso- ciations, but such acts are really not vital parts of the s^'stem of learned acts. The precise nature of the hold-over effects of various stimuli posited in the explanations of learning here suggested must be left to physiology and neurolog^^ There is undoubtedly a close connection between sensor}- and motor impulses. Sensor^^ stim- uli bring about responses which in their different stages of expres- sion set up new afferent impulses, or either facilitate or tend to inhibit old ones; these again modify the motor tendencies. We are a long way yet from a satisfactory^ knowledge of nerve im- pulses and their effects upon one another, — Are they periodic or continuous? What relations obtain between stimulus changes and nerve impulse changes"' What is the nature of inhibition 46 JOSEPH PETERSON and of facilitation? These and many other problems not yet solved have important bearings upon our knowledge of the learn- ing act. But psychologists cannot wait for the solution of these problems before attempting to formulate more satisfactory con- ceptions of the processes with which they must deal at every turn. It must be apparent that chaos now reigns with respect to this matter. Some writers invoke imiiation to explain most modifications in behavior; others use pleasure and pain for the same purpose; while ideas, purposes, the effects of random acts, and so on, are freely used directly or indirectly by most writers. All of these factors may have real parts to play in the learning process, in some one or more of its various aspects, but they are all more or less vaguely conceived and frequently erroneously referred to, almost as some sort of original or spontaneous causes, rather than complex aspects of the very thing that is to be better understood and analyzed. Popular, educational, and sociologi- cal writers may be forgiven for their own sins in this part'cular so long as psychologists have nothing more satisfactory to offer than at present. The great problem of how learning takes place is yet largely unsolved. For the best progress, experiments in behavior modification must go hand in hand with physiological investigations into the nature of the nerve impulse. A few rather suggestive studies have been carried out by psychologists upon the mutual effects of successive acts on one another. It appears that while one particular kind of act is being learned a second contrary one is inhibited by it more than after the first has been completed." The extensive investigations of Professor T. G. Brown, -* on the physiological side, have shown a summation of successive liminal stimuH (facilitation) of intervals up to about ten seconds. Such neural overlapping effects may well function to bring about a ^^ Pillsbury, discussing experiments on associative inhibition by MtiUer and Schumann (1894), concludes that " where several things are to be learned in the same connection, it is found that inhibition ceases to be effective if the first is thoroughly learned before the second is begun." Fundamentals of PsycIiolo-ing prices. Same publishers.) Journal of Animal Behavior— Cambridge, Mass.: Emerson Hall. Subscription $5. 450 pages annually. Founded 1911. Bi-monthly. Robert M. Yerkes, Managing Editor. The Behavior Monographs— Cambridge, Mass.: Emerson Hall. Subscription $o. 450 pages per volume. Edited by John B. Watson. Published without fixed dates, each number a single research. Psychoanalytic Review— New York: 64 West 56th Street. Subscription $5. 500 pages annually. Psychoanalysis. Qusr'erly. Founded 1913. Ed. by W. A. White and S. E. Jelliffe. Journal of Experimental Psychology— Princeton, N. J.: Psychological Revievv Comnanv. 480 pages annually. Experimental. Founded 1916. Subscription $3.25 Bi-monthly. Edited by John B. Watson. Journal of Applied Psychology— Worcester, Mass.: Florence Chandler. Subscription $4. 400' pages annually. Quarteriy. Founded 1917. Edited by G. Stanley Hall, John Wallace Baird and L. R. Geissler. THB BaANDOW I^RINIING COMFANT AI.BANY. N. Y 57- S'7 Behavior Monographs Volume 3, Number 5, 1919 Serial Number 16 Edited by JOHN B. WATSON The Johns Hopkins University An Experiraental Study of Transfer of Response in the White Rat BY HARRY H. V/YLIE Published at Cambridge, Boston, Mass. HENRY HOLT & COMPANY 34 West 33d Street, New York G. E. STECHERT & CO., London, Paris and Leipzig. Foreign Agents [Publication temporarily suspended] The Journal of Animal Behavior An organ for the publication of studies concerning the instincts, habits and intelligence of organisms The Journal contains a Department of Notes in which appear brief accounts of especially interesting and valuable observations of behavior. Published bi-monthly at Cambridge, Boston, Mass., by Henry Holt and Company, New York. Each volume contains at least 450 pages with plates and text-figures. The subscription price is $5.00 per volume (foreign subscription $5.50) postpaid. Manuscripts, subscriptions, and all business correspondence should be addressed to THE JOUENAL OF ANIMAL BEHAVIOR Emerson Hall, Cambridge, Massachusetts For the publication of studies in behavior and intelligence which are too lengthy or too expensive for acceptance by The Journal of Animal Behavior Published at irregular intervals at Cambridge, Boston, Mass., in connection with the Journal of Animal Behavior, by Henry Holt and Company, New York. Each volume contains approximately 450 pages with plates and text-figures. The subscription price is $5.00 per volume (foreign subscription $5.50) postpaid. Monographs may be purchased separately at prices varying with the cost of manufacture. Manuscripts and inquiries concerning terms of publication should be addressed to the Editor of the Behavior Monographs, JOHN B. WATSON, The Johns Hopkins University, Baltimore, Md. Subscription to The Journal of Animal Behavior and the Behavior Monographs should be sent to Emerson Hall, Cambridge, Massa- chusetts. Behavior Monographs Volume 3. Number 5. 1919 Serial Number 16 Edited by JOHN B. WATSON The Johns Hopkins University An Experimental Study of Transfer of Response in the White Rat BY HARRY H. WYLIE Published at Cambridge, Boston, Mass. HENRY HOLT & COMPANY 34 West 33d Street, New York G. E. STECHERT & CO.. London. Paris and Leipzig. Foreign Agenu ACKNOWLEDGMENT As in every work of this sort, the real source of inspira- tion, of guidance and of criticism lies in the Instructors of the Department in which the work was done. In this par- ticular case, the author wishes to acknowledge his indebted- ness and gratitude to Professor Angell and Dr. Carr. Their helpfulness has extended far beyond the immediate assist- ance given in this thesis. No one can be under their tute- lage without being influenced for the better by their ideals and practice. In particular, Dr. Carr's constant and patient oversight and criticism was a large element in carrying this present work to a successful issue. Acknowledgment is also due to Mrs. Wylie for assist- ance in performing the experiments and in preparation of the manuscript, Harry H. Wylie TABLE OF CONTENTS Page Section One 1 Introduction 1 Section Two 10 General Features of Experiments 10 A. Animals Used 10 B. The Two Types of Experiment — Positive and Negative 10 C. The Positive, (1) Apparatus 11 (2) Method of Procedure 13 D. The Negative, (1) Apparatus 14 (2) Method of Procedure 15 Section Three 18 Features Bearing on Learning , 18 A. Positive, (1) Light, (2) Sound, (3) Summary 18 B. Negative, (1) Light, (2) Sound, (3) Pain, (4) Summary 30 Section Four 51 Features Bearing on Transfer 51 A. Positive, (1) Light to Sound, (2) Sound to Light 51 B. Negative, (1) Light to Sound, (2) Light to Pain, (3) Sound to Light, (4) Sound to Pain, (5) Pain to Light, (6) Pain to Sound 52 C. Summary 64 SECTION ONE INTRODUCTION The experiments reported in this paper were inspired by the conviction that almost all of the work previously done on the problem of the transfer of response or the transfer of learning lacked exact definition and control of the fac- tors involved. But in spite of this lack two generally accepted conclusions have been fairly well established: (1) Transfer, negative as well as positive, in varying degrees has been found; (2) A satisfactory theory of explanation must find its basis in the " something in common " or the " identical elements " in the problems or situations between which the transfer occurs. However, to get at the actual conditions under which transfer, both positive and nega- tive, takes place, and to state the explanatory principle in a definite and usable way, it is necessary to define the two problems between which transference is tested accurately and completely in terms of each other as to their similarities and diferences. One can then find by concomitant varia- tion what factors and conditions are correlated with posi- tive and negative transfer and the relative infiuence of these factors and conditions upon one another. This, of course, has been attempted. One needs but to glance at Coover's(l) whole page of " identical elements " to be con- vinced of this. But to facilitate this analysis two condi- tions are favorable: (1) Simplicity of problems. This, too, has been attempted. It is the keynote of most attacks on the problem. Instead of trying the effect of a certain course of study upon practical success in some chosen business or profession, the experimentation has taken up simple cases of memory, discrimination, etc. But these are not sufficiently simple. One element of complexity is the central or subjective factors involved in what goes under the names " attention," " general ideas," " ideals," etc. Hence comes our second condition suited to facili- 2 HARRY H. WYLIE tate analysis, — i. e. (2) the reduction of the problem to perceptible and definitely measurable objective terms — sejisory stimulus and motor response. This condition has not been obtained and can not easily be obtained with human subjects. It can presumably be best obtained with animals. Without attempting any exhaustive survey of cases, let us notice briefly the problem as studied in animal psychology. It is well to bear in mind at this point that whatever study has been given this problem up to the present time by those experimenting with animals has been largely of an incidental character. Yet the incidental results have often been interesting and suggestive. The problem has presented itself to most of the animal psychologists in this form: What is the effect of previous habits upon the formation of new habits? Yerkes(2) study of the Dancing Mouse has furnished some data bearing on the question as stated. Animals which had not been previously trained on a simple labyrinth (C) succeeded in making the trip correctly for the first time on the average after 19.7 trials. Animals which had previously learned the labyrinth (B) succeeded in making the first correct trip in (C) after 7 trials. The learning of (B) without previous training required 8.2 trials. After previous training on (C), the learning of (B) required 5 trials. Thorndike(3) reports in his experiments with cats in problem boxes, that " previous experience makes a difference in the quick- ness with which the cat forms the associations. After get- ting out of six or eight boxes by different sorts of acts the cat's general tendency to claw at loose objects within the box is strengthened and its tendency to squeeze through holes and bite bars is weakened; accordingly it will learn associations along the general line of the old more quickly." He cites three animals, Nos. 10, 11, 12, which had learned the problem box A (0 at front) and were afterwards taught the problem box Bl (0 at back). He remarks, " It nat- urally takes a cat much longer time to accidentally claw a loop in the back than in the front, yet a comparison of these curves with those— shows the opposite to have been the case with 10, 11 and 12." A parallel series of controls TRANSFER OF RESPONSE IN THE WHITE RAT 3 is needed here, however, to make sure the statement. Richardson's(4) experiments on white rats also furnish some data. Two sets of rats were obliged to learn to respond correctly to box III. One set had previously learned to respond correctly to boxes I and II. The curves for the two sets show a decided advantage for the trained rats. Yoakum's(5) experiments with squirrels show that training is sometimes a disadvantage. Thus animals trained on box I and box II took a longer time to learn to open box III than animals which had had no training with boxes I and II. The responses required in the first two boxes, such as scraping, butting up the latch with the nose, were car- ried over to box III where a different response was required. Hunter(6) offers two sets of experiments bearing on this question, only one of which will be mentioned here, the other being discussed later. In his work on pigeons he found that there was no difference as to number of trials required for trained and untrained animals in learning a given maze where training had been on another maze. The excess time and errors during early trials were greater for the trained, yet the excess effort was eliminated in a smaller number of trials by the trained. Bogardus and Henke(7) also made some tests upon the question of transfer in the case of rats learning a maze. The animals were taught a certain maze. Then by use of doors for blocking the true pathway at certain points and for opening up the pathway at new points, the correct pathway was altered in certain respects. Then the ani- mals were taught to run the new pathway. An effort was made to make the maze different, each time any alteration was made, in but one respect. These were arranged ac- cording to what was judged to be their relative difficulty in order I, II, IV, III, V, number I being the original maze. The actual order of difficulty in learning these five proved to be, as shown by the table of results, V, IV, III, I, II. This leads to the conclusion that the disadvantages of the old habits rather overshadow their advantages, that is, that responses to the previously learned maze were carried over to the new maze and actually proved a hin- drance to the formation of the new habit. In Watson's 4 HARRY H. WYLIE work on Animal Education where animals were taken from one problem to another, the results were generally found to be advantageous. Whatever advantage, however, is shown by previous training in any of these cases can be explained in part at least by saying that the animal carries over the food seek- ing reaction, that fear and emotional disturbances have been eliminated, and that a definite manner of attacking the problem has carried over. In so far as these factors enter into the new problem positive transfer was shown. Likewise where the responses have some similarities pos- tive effects obtain. But we also find negative effects in varying degrees. In Bogardus' work the negative over- balanced the positive effects. In his work the responses were similar in some respects and different in others. It was attempted however, to make the maze different in but one respect each time a change was made. But upon examination we discover perhaps three important changes which entered in each time: (1) a change in the direc- tion or amount of turning at the critical point, (2) a change in the length of the runway at the critical point, (3) a change in the order of succession of the runways of different lengths at the critical point. Of these perhaps the first would be most important, although the third has been shown by Watson to be very important also. In fact the maze reaction has never been carefully enough analyzed to be used successfully in transfer tests. At least it is too complex to serve for pioneer tests. So our conclusion must be that in none of the experiments men- tioned has there been a careful analysis of situation and response into their various factors and then a careful control and variation of those factors. Both situation and response have been varied in a lump or grossly. This is just the criticism that has been offered on a previous page upon the experiments on the same problem in human psychology Thus while the taking over of the problem from the human field into the animal field has secured simplicit}'- in as large a measure as seems possible by the elimination of the central or subjective factors, yet there is still too TRANSFER OF RESPONSE IN THE WHITE RAT 5 much complexity of objective factors and a lack of accurate analysis and definition of the relation of the two problems in terms of similarities and differences. It seems to the writer, therefore, that if the experiments upon this problem are to have any finality and secure for us the solution of the problem they will have to begin with such conditions as will either keep the response as a whole constant and vary the stimulus or situation one factor at a time, or keep the stimulus or situation as a whole con- stant and vary the response one factor at a time. From some such simple beginning we could then pass, as experi- ence and technique indicate, to tests where more than one factor of the stimulus or situation was varied while the response remained constant, or to tests where more than one factor of the response was varied while the stimulus or situation remained constant, or to tests where one or more factors in the stimulus or situation and one or more factors in the response were varied while all others in the stimulus and in the response remained constant. Most experiments which have been performed would presumably fall under the fourth class mentioned, although no careful analysis, especially of the response side, has been made in any of them, and in many no careful analysis of the stimu- lus or situation. We have already cited Bogardus and Henke's tests. For further example, how much, and what change is made in the stimulus and how much and what change is required in the response in passing from one of the boxes used by Yoakum to another, or from one of Hunter's mazes to another? Shall we, following Thorn- dike's principle of identical elements, decide upon the amount of likeness or difference of the two problems by the amount of transfer shown when one is learned after the others? The true course would require first such an analysis and control of factors as is suggested above and a long series of experiments in which control was present, in order to furnish the only sound basis for Thorndike's principle, or any other. It appears to the writer, moreover, that after all, the problem back of the transfer of learning finds one of the best points of attack, as suggested above, in the form in 6 HARRY H. WYLIE which the response remains constant while the stimulus, is varied under proper control. Evidently such a situa- tion is in mind when the problem of transfer is spoken of as the problem of general training, of generalized response, of generalized habit. Is there such a thing as generalized response? Will a response which has been learned for one stimulus or situation be found to carry over in any degree for a stimulus or situation which is different in a definite way from the first stimulus or situation? If so, under what conditions does such a transfer take place? The experi- ments reported in this paper were designed to answer partially these questions. How successfully they have done so will appear in the issue. They represent an attempt to secure both the greatest simplicity possible and an accur- ate definition of the relation of the two problems. Both have the same motor response, both present the same sen- sory situation except in one definite particular and both pre- sumably have a minimum of central factors. Before pre- senting the writer's experiments, however, let us notice a, few theoretical considerations and some further experi- ments that bear upon the question in the form in which this paper expects to deal with it. Theoretically it is possible for a " generalized response " to occur under two widely different conditions: (1) under circumstances in which discrimination of the stimuli is not possible for the organism called upon to make the response, or if possible, has not yet been learned; (2) under circumstances where discrimination of the stimuli is pos- sible and is known to be made by the organism. Thus under number one would fall such cases as those where an animal makes the same response to all colors owing to an apparent inability to discriminate differences in colors, making the response wholly on the basis of brightness. Also the case of the child calling all men " papa " might be mentioned. These are to be looked upon not as cases of " transfer of response " or " generalized response " be- tween previously discriminated particulars, but as cases merely of the absence of discriminated particulars. They represent a primitive and unspecialized stage, not an ad- vanced state. Under number two would fall what psy~ TRANSFER OF RESPONSE IN THE WHITE RAT 7 chologists are pleased to call, in their advanced stage in human psychology, " concepts," " generalizations," and " abstractions." Psychologists have not hesitated to apply the term " generalized response " to what is included, on the behavior side, under the terms mentioned. Thus to quote Angell(8) : " Concepts of adults may also be considered as forms of generalized motor activity." Gore(9) sets forth something of the same idea. The idea presented is that the essence of such types of behavior after all lies in the fact that one response serves for a multitude of stimuli, situations, objects. The word " man," for instance, serves as a proper and efficient response to a whole group of objects and to each member of that. group. The question •before us in this paper is, when looked at from this point of view, whether we can get experimental data bearing upon this second type of " generalized response " or " trans- fer." Can we actually get well controlled situations where a response learned for one stimulus is transferred to another different stimulus? If so, what are the conditions of such transfer? We find simple experimental illustrations in human psychology in the reaction experiments. Here one learns to make a simple response, pressing a key, upon the recep- tion of a certain stimulus, say a visual stimulus. The same response is required upon the reception of a second stimulus of a different kind, say an auditory stimulus. So likewise with a touch stimulus. While such experiments furnish us little data on the conditions of transfer, yet they do suggest methods of procedure. In animal work we find also suggestive data. In Hunter's work(lO) on the auditory sensitivity of the white rat he reports that his animals reacted properly (a given, required, controlled response in all cases the same) when either of the following noises were substituted for the standard whistle: the rush of air through the whistle; sound of the rush of air made with lips, and clapping hands. He found also that animals trained to react to hand clapping reacted successfully to the following noises when these were substituted: rattling of paper, dropping sunflower seed on tin, scratching on wood, drum- ming on table with fingers, etc. Here it may be contended 8 HARRY H. WYLIE that the response is a " generalized " one, but one belonging to the first type mentioned above. We have no definite ground for believing that the stimuli mentioned — all be- longing to the group called noises — -were in any way actually different stimuli for the animals tested, so we can not maintain that this furnishes an illustration of transfer between " different " stimuli, the second type mentioned above. Better illustrations, perhaps, can be found among cases which are said to illustrate " substitution." Here one feature of the situation being the dominant one in arousing the response to the situation comes to be replaced by an- other as the dominant element. Watson(ll) in his book on Behavior defines substitution as follows: " By substitution" we mean that a stimulus which originally did not call out a given response comes later to call it out." He cites two examples. One is from the Pawlow type of experiment on dogs, in which a green light came to call out the salivary secretion in large amounts when the sight of the food had been the dominant stimulus for that reaction. The other is the experiment with cats which Thorndike(12) offered as a tentative proof of the presence of ideas in animals. Here the clapping of hands came to be the dominant stimulus in calling out the desired reaction, although the laying of fish on the top of the cage had been the dominant stimulus. Further examples might be taken from experiments with rats learning the maze. In the work of Bogardus and Henke(13) it was shown that rats make considerable use of tactual stimuli in learning the maze, although when once thoroughly learned the response was largely in kinaesthetic and organic terms, as Watson contended. It is evident then that here we have a substitution of kinaesthetic and organic stimuli for tactual as dominant in securing the required response. In the later work of Vincent(14) we have admirable illustrations of the same sort of thing. She has showed that the maze is probably learned in tactual, visual, and olfactory terms, when the maze is so constructed as to call these into play. Later the kinaesthetic and organic come to be substituted for them. Yet the kinaesthetic never becomes accurate enough to maintain correctness but needs almost constantly some sort of cutaneous guidance. TRANSFER OF RESPONSE IN THE WHITE RAT 9 Cases of " substitution " are, in the opinion of the writer, after all cases of " transfer " of a certain sort. The essential characteristics of one type of transfer are all present, i. e., a certain definite, required response which is kept constant and a definite variation in the nature of the stimuli or situa- tion calling out the response. While it can not be main- tained that in the maze problem certain stimuli were re- moved and others, which had not been present before, put in their place, yet there is evidence that as far as effective- ness of the stimuli goes, such was actually the case. In the experiments to be reported in this paper the conditions were so made that one stimulus which was shown to be the dominant one was actually removed and another, which had not been present before, was substituted for it so that it became the dominant one in place of the former. These dominant stimuli were taken from the different sense fields, light, sound, and pain, so that a stimulus from one sense field was replaced by a stimulus from a different sense field. In this way it was hoped to secure results which were based on stimuli which were simple, but which were also as dif- ferent as it was possible to secure. Of course these dom- inant stimuli had for their background or setting the entire situation composed of various stimuli from the box, room, etc. All these were conditioning stimuli, it is true, but not the dominant ones. All these conditioning stimuli were kept constant, the dominant stimuli alone being changed. The response also was kept the same. By such a method we hoped to secure some reliable and important results bearing on the possibility of one and the same response serving for two or more different sorts of stimuli, or on the possibility of a " generalized " response of the second type mentioned above. SECTION TWO GENERAL FEATURES OF THE EXPERIMENTS A. Animals Used. The animals were white rats of the same species as used by Watson, Carr, Vincent, and others in work done in the same laboratory in previous years. The rats were for the most part young animals, beginning the work when some three or four months old on the average. When young animals were not used the fact will be found indicated in the detailed reports of the experiments. In all, close to one hundred and seventy-five animals were used. They were divided into groups of five or six or seven to the group, and the record for such a group was taken as representing what the normal animal would do on the average under like conditions. The food used was sunflower seeds. This allowed better control of amount secured at each trial than other types of food at hand. B. The Two Types — Positive and Negative Two series of experiments were planned and carried out. In the first series the attempt was made to teach the ani- mals a definite positive response to one sort of sense stimu- lus. Then a second sort of stimulus was substituted for the first. The purpose was to see whether and to what extent such a response would carry over from the first to the second. This series falls into two groups: (a) where the response was first learned to the light stimulus, and then the sound stimulus was substituted for the light; (b) where the response was first learned to the sound stimulus, and the light stimulus substituted for the sound. In the second series a negative response was required in all cases. The purpose was the same as in the positive series. This series falls into six groups: (a) where the response was first learned to the light and the sound sub- stituted, (b) where the response was first learned to the TRANSFER OF RESPONSE IN THE WHITE RAT 11 light and the electric shock substituted, (c) where the response was first learned to the sound and the light sub- stituted, (d) where the response was first learned to the sound and the electric shock substituted, (e) where the response was first learned to the electric shock and the light substi- tuted, (f) where the response was first learned to the electric shock and the sound substituted. C. Positive (1) Apparatus. — The apparatus consisted of a wooden box whose plan is shown in figure I. The base LMNO upon which the box rested was approximately two feet by three. The food box (F) was 13 inches wide and 13 inches long on one side but longer on the other because of the entrance of the two return alleys as shown in the figure. ED is a door which leads from the food box to the main part of the box (B) through the entrance alley (EA). This entrance alley was four inches in length. X and X' are the entrances to the return alleys (RA and R'A'). RD and R'D' are doors from return alleys to foodbox. S and S' are sounders hung just outside of the box close to three-quarter inch holes through the side of the box into the heads of the return alleys. Lt and Lt' are two small electric light bulbs just inside the wall of the box and just above the holes mentioned above. K and K' are two keys which control the left and right sounders and the left and right lights respectively. Sw and Sw' are two switches which allow the use of the same keys (K and K') for both the sounders and the lights, either singly or simultaneously. Thus the key (K) could be used to work the sounder (S) or the light(Lt) or both the sounder (S) and the light (Lt) simultaneously. The sounders were small electric buzzers so fixed that with one pressure of the key but one click was produced, instead of the rapid succession of clicks ordinarily produced by such an instrument. The lights were small electric bulbs of approximately one and one half candle power. The current necessary was produced by a cell of dry batteries indicated in the figure by the letters DB. The box and all partitions were six inches deep. The alleys were five inches wide. All the alleys were 12 HARRY H. WYLIE Figure II TRANSFER OF RESPONSE IN THE WHITE RAT 13 uncovered excepting the heads of the alleys where the sound and light stimuli were presented. The covered portions are indicated in the figure by shading. (2) Method of Procedure, — The animals were first tamed by being fed in the experiment boxes, by being handled freely by the experimenters, and by being allowed free run of the box until all emotional factors were removed as far as possible. The regular day's work for any animal was a series of twenty trials. However, when just beginning to learn, the animal was given fewer trials and gradually brought up to the full stint. The standard adopted in this series as indicating that the animal had learned the problem was 95 correct responses out of 100 chances. Since a daily series consisted of 20 trials, the standard required the completion of five successive series of 20 trials each with not more than one error on the average to each series of 20 trials. Normal records were secured first for the sake of com- parision. These served to indicate how many trials and series of trials were necessary for the average untrained rat to learn the response to each of the various sorts of stimuli used in the experiments. Then rats which had learned to respond to one stimulus, say light, for example, were required to learn the same response to another, sound, for example. Then this ' sound ' record was compared with the normal 'sound' record to see whether there was any advantage or disadvantage in first learning the response to the light stimulus. Also, in some cases, after the response to one stimulus had been learned, this stimulus was presented simultaneously with another sort to determine the effect of simultaneity of presentation upon the learning of the response to the second sort. In regard to 'controls,' the simple expedient was adopted of omitting the stimulus and observing the effect upon the behavior. The series of 20 trials in which the stimulus had been omitted was compared with the series in which the stimulus had been presented, and the effect of the presence of the stimulus inferred from the comparision. Two sorts of stimuli were used in the positive series — light and sound. The particular method of procedure with 14 HARRY H. WYLIE any one animal was to place it in the food box (F), allow it to secure a morsel of food, open the door (ED) into the alley (EA), then press the key K or K', as the case demanded, and thus require the animal to choose the lighted or the noisy return alley. In case the animal entered partly or wholly into the unlighted or the quiet return alley it was accounted an error. The door from the unlighted or the quiet return alley to the food box was kept closed. The light stimulus was not always given as a continuous stimulus in any one trial until the choice of correct return alley had been made. The stimulus was given sometimes continuously and some- times brokenly. Care was taken that the action of the key (K or K') was noiseless. In case of the sound stimulus it was at first attempted to sound the clicks regularly until the correct choice had been made. This plan, however, was not followed in all the work reported. Whenever a different course was followed it will be indicated. The following order of choices between right and left return alleys was required in each series of 20 trials: L, R, L, R, L, L, R, R, L, R, L, R, L, L, R, R, L, L, R, R. D. Negative (1) Apparatus, — The apparatus for the negative response consisted of a box approximately two feet by three, MNOP in figure II. F represents the food box approximately 12 inches by 14 inches. This food box had three openings or doors, ED being the entrance door from food box into main alley, MA; RD and R'D' being return doors from return alleys, RA and R'A', to food box. All the alleys had a uni- form width of approximately five inches and a uniform depth of six inches. From the main alley, MA, two winding alleys led back to the food box. These are marked RA and R'A' in the figure and are spoken of as return alleys. PP are partial partitions obstructing the view of the return doors and causing the animal to take a devious course back to the food box. C, C, C, etc., are brass plates practically covering the entire width of the bottom of the alleys and are joined to batteries and an induction coil, IC in figure. S, S, S, etc., are small electric sounders of the same type as used in the positive experiments. They are here hung just inside the TRANSFER OF RESPONSE IN THE WHITE RAT 15 walls of the box close to the partial partitions, excepting the one found on the wall of the food box. These sounders are also joined to the batteries. L, L, etc., are small electric lights of the same size as used in the positive experiments. They are attached to the partial partitions near the sound- ers and are also joined to the batteries. In one box used for these experiments a switch was attached to the side of the box, marked Sw in the figure. This switch was used to turn the current into the right or left sounders or lights as the case demanded. A later box had a slightly different arrangement. Two keys were used on the battery and induction box, one for the right and one for the left. This allowed of more rapid and less noticeable action on the part of the experimenter when giving the stimulus or combination of stimuli required in any particular case. The sounders, lights, and brass plates were so wired as to allow them to be used singly or in com- bination. Thus the lights on any one side could be used alone or in combination with the sounders of that side or in combination with the brass plates for pain stimuli. Also it was arranged at first so that one, two or all three of the sounders might be used at any one time. This was found to be unnecessary, however, and so it was finally arranged that all of the sounders on any one side would come into play by pressing the key controlling the sounders on that side. The sounder found on the wall of the food box was the first one used in attempting to teach the negative response to the sound, but it was found necessary to place the sound- ers along the side of the box as indicated above before the animals would pay sufficient attention to them to learn the response required in any reasonable number of trials. So the sounder in the food box was abandoned and the sound- ers along the side were substituted. Since the wiring re- quired was rather intricate no detailed description of it will be given here. Any electrician can make the proper con- nections to secure the working combinations of sound, light and pain required in the experiments. (2) Method of Procedure, — The animal was put through the entrance door (ED) into the main alley (MA) and was required to go up the main alley and back to the food box through either the right or the left return alley (RA or R'A')- 16 HARRY H. WYLIE When the animal had entered either of these return alleys the stimulus was given, that is, the light or sound or electric shock, or a combination of any two of them as the case might require, and the animal was required to go back and come down the other return alley to the food box. In fifteen out of every series of twenty trials the proper stimulus was given and the animal was required to turn back and enter the food box by the other alley. In the other five trials the animal was allowed to continue on its chosen way and return to the food box, no stimulus of any kind having been given. These five where no stimulus was given were scattered irregularly through the series of 20 to prevent any possibility of the animal's learning a definite fixed order in the series. The purpose was to train the animal to stop at a given stim- ulus, to turn around, and to go back to the food box through the other return alley but, when given no stimulus, to pro- ceed on to the food box through the alley first chosen. When the animal was first learning, the stimulus was given more or less irregularly as to loudness and continuity from the time the animal had fully entered the return alley until it had turned back, even if the animal persisted in its course until it found the door at the end of that alley closed. The fact is that the usual course was for the animal at first to ignore com- pletely the stimulus other than a short delay for examina- tion and to proceed clear down to the closed door at the end of the alley. After the first strangeness had worn itself away, which happened after a few trials, there was little indication, if any, to show that the stimulus was sensed at all until actual learning of the required reponse began. However, as accurate records as possible were kept of hesitations as well as of the actual times of turning back, when the hesitations were apparently due to the effect of the stimulus. Some- times also in the learning process the speed of the reaction was apparently much lessened by the disturbing effect of the stimulus. This was particularly true when the light stimulus was used, so a record of these was also kept. Both doors, RD and R'D', were kept closed until a choice had been made, when the proper one was opened as noiselessly as pos- sible. An attempt was made to prevent the experimenter's movements from being so evident, as to become cues for the TRANSFER OF RESPONSE IN THE WHITE RAT 17 animal. A response was counted correct provided the ani- mal had fully entered one or the other of the return alleys and turned back, when the stimulus was given, before it had reached a point where there was any possibility of its seeing the closed doors of the alley. After the animal had begun to learn it was given the stimulus when at various places to prevent the formation of position habits. Thus some- times the stimulus was given just when the animal had entered the alley, sometimes when it was half way down the alley, and sometimes when it had come almost to the point where the closed doors could be seen. By such a variation the animal would not learn to stop at any particular place but only when the stimulus was given. The omission of the stimulus five times out of twenty trials also prevented the formation of the turning habit regardless of the presenta- tion of stimulus and served as an effectual control. The standard adopted for the negative series was 100 correct responses out of 105 chances to respond to the stimulus given. Since the stimulus was given but 15 times out of each series of 20 trials it required the completion of seven such series with but five errors all told to meet the standard. SECTION THREE FEATURES BEARING ON LEARNING A. Positive (1) Light— Two groups, I and II, were trained with the light stimulus and were then transferred to the sound. The following tables give the results as to the number of trials and series of trials necessary for learning the light sufficiently well to meet the standard of 95 correct choices out of 100 chances. Group I Animal Trials Series No. 15 680 34 No. 16 880 44 No. 22 640 32 No. 23 460 23 No. 26 340 17 Average 600 30 Group II Animal Trials Series No. 17 920 46 No. 18 580 29 No. 19 580 29 No. 20 600 30 No. 21 540 27 No. 25 560 28 Average 630 31 .5 Average for 11 animals, 616.36 trials, 30.81 series. This average for eleven animals can be taken as a normal record for learning a positive reaction to a light stimulus under the conditions of the experiment. It will be noticed there is great variation in the number of series and trials required for the individual animals to learn the response. Thus number 26 needed only 17 series or 340 trials, while number 17 took 46 series or 920 trials. For this reason group curves would not show the character- istics of the learning process. So two typical individual TRANSFER OF RESPONSE IN THE WHITE RAT 19 curves have been given instead. The curves for number 22 of Group One, (Figure III) and for number 25 of Group Two (Figure IV) are given. These curves show not only the learning of the response to the light stimulus but they also indicate the record for the control series in each case. These curves indicate neither a rapid initial rise nor a long period of final perfecting of response. They rather show a gradual, though very irregular, rise from the beginning to the end of the learning process. The unusually large number of trials required by Numbers 16 and 17, for learning the light finds its explanation, apparently, largely in terms of distracting olfactory stimuli. The males in these two groups were kept in a cage by them- selves and were given their daily stint of work immediately after the females of the two groups. The two individuals mentioned, being unusually vigorous animals, were con- stantly attracted apparently by the odors left by the females and would often wander about the box sniffing here and there to the neglect of the light stimulus. So even after the response had been apparently learned, it took a greater number of series to meet the required standard. The important thing about learning the light was, of course, to be as sure as possible that the animals were re- sponding to the light and not to other stimuli. The order of presentation of right and left pathways followed in the regu- lar and control series has been given already on page 29. It will easily be seen that position habits might account for a rather large percentage of correct choices in such a series. Of the position habits most likely to be learned the most advantageous would be a double alternation between the right and left pathways. Such a position habit, if acquired, would account for a record in which 80% of the responses were correct. The next most advantageous position habit would be a simple alternation between right and left. That, if acquired, would account for a record in which 70% of the responses were correct. Chance, of course, in any series where two possible choices must be presented an equal number of times, will account, in the long run, for 50% of the responses being correct. A position habit of always going to one side would also give a record in which 50% of 20 HARRY H. WYLIE ^00- Figure III " 1 i g r -_U § T! 3rr ^ r"t -f , 1 ; ; ^^ ir -z=n P 3 =fi £=^- ^^"f - 1 ^ 1 ^ E ^ f 1 1 f i^ 1 — h :":x 1 — -- r 1 k i 1 ^ i ^i E ~: izT^i^ 1^ V ^ zr - B ---- s ;::|l|: E E S^: fr— Fv~ ~ - ^ H± - — - ^ , i 1 i - ' 1 1 ' : : p: :::|::|: = = t t^ --^ jT t ^ = 1 ^^ T +ti^4+ w |: -f-|- ffr E -ii— = H' — +- ■^-H ':': ■"I :| 1 III :: i 1 li i nw +- fe " H Il'j +f4-- 1 1 in I i-::: 1U# 1 I :x:: 1 i: i 5 ^#4^ i -t— — ' 3 i 1 — 3 i it:- 75- Figure IV Curve 22 (Figure III) is an individual curve from Group One. Curve 25 (Figure IV) is an individual curve from Group Two. The horizontal lines represent the per- centage of correct responses. The vertical lines represent the number of series of 20 trials each. In lettering, — A is the end of the learning process with light, B is the first control series, C is the second control series, D is the beginning of training with the sound, S (in Curve 25) is the beginning of simultaneous light and sound. TRANSFER OF RESPONSE IN THE WHITE RAT 21 the responses would be correct. From a study of the daily records it was found that there is no evidence in any case for believing that the animals at any time used the double alternation position habit. A study of the records does show, however, that at first the choice of response was practically a matter of chance. In a very few instances the position habit of always going to one side was picked up after a series or two but was not maintained for longer than one day at a time. Then in most cases there seems to be a gradual shift to the simple alternation position habit, but in none of the cases in this list of eleven animals was such a habit actually learned and consistently maintained while learning the light. A careful study of the daily records did not reveal a single series where a perfect simple alternation was car- ried clear through the series while learning the light. This was not true, however, after the animals had learned the light and had been transferred to the sound. There the animals seemed to be working largely on a basis of simple alternation but not with any degree of consistency. Out of 215 series where sound was given as a stimulus after light had been learned only 20 showed a perfect simple alternation. From a study of the records, however, one must conclude that there is a strong tendency toward using simple alternation in a series of the type used, but that there is little ability to maintain it consistently. In order to make it as certain as possible that the animals were actually responding to the light stimulus when the required response was considered learned, first, a high standard of excellence was chosen and second, control series were introduced. The standard adopted as indicating that the problem had been mastered was 95 correct choices out of 100 chances or five series in which there was not more than one mistake to a series on the average. The actual fact was that most of the animals met even a higher standard than that, as a glance at the sample individual curves will show. It was not considered probable that the animals used would be able to develop a position habit of sufficient complexity to en- able it to respond correctly, 95 times out of 100 even with a series apparently as simple as the one adopted. As indicated above the results justified this assumption. But to be cer- 22 HARRY H. WYLIE tain control series were used. After the animal had apparently- learned the problem, it was given one or two series without the light stimulus. The result was a drop in efficiency of from 25% to 70% with an average drop of 40%, leaving an average of 60% correct. This is not much above what chance would enable the animal to do. Upon the light being restored again an immediate return to the required efficiency appeared again. This was taken as conclusive evidence that the light stimulus was the dominant stimulus in making the correct response when the percentage of efficiency had reached 95. (2) Sound— Five groups of animals. III, IV, V, VI, VII, were trained with the sound stimulus preparatory to trans- ferring them to the light. In all 32 animals were trained in these groups. The response required was the same simple positive response required in the case of the experiments with light. Groups III and IV used the sounder mentioned in the description of the apparatus. Group V was trained with the telephone receiver as sounder in place of the regular electric sounder. Groups VI and VII were trained with the telephone receiver as sounder and with pain as punishment for a wrong choice. Groups III and V were older animals than those used in the other groups and had had training on the maze. They were all practically full grown animals. The results can be summarized very briefly. All the mem- bers of group III were given 85 series or 1700 trials without learning the response. All the members of group IV received 60 series or 1200 trials without learning the response. All the members of group V received 40 series or 800 trials without learning the response. Three members of group VI received 40 series or 800 trials, one 35 series or 700 trials, and two 30 series or 600 trials. None learned the response. Four members of group VII received 60 series or 1200 trials, two 55 series or 1100 trials, and none learned the response. In order to bring to light the significant features of these results we need to notice the method and results more in detail. In the main, the method with the sound stimulus was the same as with the light. The stimulus was given in a more or less regular way, that is, the sounder was 'clicked' with TRANSFER OF RESPONSE IN THE WHITE RAT 23 more or less regularity until the animal had made a choice. When the choice had been made, whether right or wrong, the stimulus ceased to be given. When the choice was wrong and the animal had returned to the main portion of the box (B in figure I) the stimulus was again given until correct choice was made. However, the animals in groups III, IV, and V, where no punishment for wrong choice was, given, soon learned to work so rapidly that one scarcely had time to give any stimulus, after the animal had been put info the entrance alley (EA in figure I), before a choice had been made. The animal would dash into one or the other of the return alleys (X or X' in figure I) and, after glancing about the turn at the head of the return alley, dash toward the door into the food box if the choice happened to be correct, or if wrong, whirl quickly and dash into the other return alley and into the food box. When the reaction became of this type, the stimulus was given from the time the animal entered EA until the correct return alley had been entered. With group V, beginning with the 26th series, the sfmulus was given irregularly in place of regularly, that is, the number of times the key was pressed and the sounder 'clicked' varied greatly from one unit of time to another during the time the stimulus was being given. This was done with the idea of attracting the animal's attention. A like change of procedure was made w4th the animals in groups VI and VII where pain was given as punishment for wrong choice. Another change of procedure introduced was punishment for wrong choice. The punishment given was an electric shock. The return alleys (RA and R'A') beginning at X and X', had small copper plates placed in them and these plates were attached to an induction coil. Two keys were used to connect the circuits, one for RA and one for R'A'. With the introduction of punishment the behavior of the ani- mals was quite different. Considerable hesitation appeared as a rule before choice of return alley was made. This hesitation gave more time to attract the animal's attention by means of the sounder than when the animal rushed pell- mell into one of the return alleys as soon as placed in the entrance alley (EA). While some improvement appeared, as will be shown 24 HARRY H. WYLIE later, yet the main thing to be noticed in the results is that not a single animal out of the 32 trained learned to choose the 'noisy' alley even though some were given as high as 1600 trials and even though pain was used as punishment for wrong choice in the case of a number. Also as will appear later in our study of Transfer effects, even those animals which had learned the positive response to the light failed to learn the same response to the sound. A number of explan- ations or reasons may be offered for this. First, it may be claimed that the animals could not hear the sound presented. That is disproved, however, by the fact that other animals did learn the negative response to the very same type of sound. Second, it may be that the reaction required lay outside the animal's ability to learn. However, eleven animals did learn the same response, when a light stimulus was used, with much smaller number of trials than were given with the sound stimulus. Third, perhaps the rats were negative to the types of sound used and so would not learn a positive response to such sounds. The records, however, do not give any con- vincing evidence to support this contention. Since the rats had free opportunity to enter either the 'noisy' or the 'quiet' alleys, if they were negative to the sound, we should expect some evidence of it in the early part of the records. The rats should have made many more than 50% of errors until they had become accustomed to the sound. In Group III the averages for the first 200 trials were 46, 49, 51, 53, 50, 51, 48, 54, 49, 56; just about what chance would give in the long run. The records for the other groups show about the same results. No signs appeared in the general behavior of the rats, except in one or two individual cases, that would go to support the contention that they were negative to the sounds presented. Rat number 7 in group III was one of the ex- ceptions. It was, however, an extremely nervous animal and objected to being handled. Its daily record often contains such a statement as the following; " Would not work well today. Seemed frightened." With practically all the animals, however, one could not tell from their gene- ral behavior that they even .heard the sound at all. As TRANSFER OF RESPONSE IN THE WHITE RAT 25 stated above, the responses became so rapid, where no punishment was given for a wrong choice, that one would be led to think that the sound was wholly ignored. Further, if the rats were negative, the omission of the sounds should show an increase in percentage of correct choices. Let us notice the last 100 trials with the sound stimulus in groups III and IV, and the five control series or 100 trials where the sound was omitted. The last 100 trials with the sound in Group III average 68, 66, 66, 67, 67. The 100 trials with- out the sound average 69, 69, 69, 64, 64. A slight advan- tage is shown for those without the sound. In group IV the last 100 trials with sound average 64, 68, 62, 66, 6S. The 100 trials without sound average 67, 68, 63, 68, 63. No advantage is shown for either. Although no convincing evidence can be found in the records that the rats did not learn because they were negative to the sound presented, yet to give the matter a further trial a less metallic sound was sought, if that should prove to be the difficulty in the way of learning. For that reason a telephone receiver was substituted for the electric buzzer used with groups III and IV. A duller, more wooden sound was thus secured. No one of the six rats in Group V gave any evidence of learning the response after 800 trials. Neither the records nor the general behavior showed that the rats were negative to the sound. Groups VI and VII were trained with the same sound as Group V but were given pain as a punishment for wrong choice, but 1200 trials for four of them brought no convincing results. As a last resort a short test with one rat was tried with a sound to which from general behavior we had every evidence for believing the rats to be positive. This sound was the dropping of sunflower seed in a dish. Any one who has handled rats knows that such a sound will cause in a hungry rat a very active hunting for the source of the sound. The method of procedure was as follows: One experimenter was placed behind a screen at the end of the box marked MN in figure I. A dish of sunflower seed was placed at the position of each of the sounders, S and S'. The other ex- perimenter took the regular position in front of the box. When the animal was placed in the entrance alley (EA) the 26 HARRY H. WYLIE experimenter behind the screen would rattle the seeds in the dish at the head of the open return alley. Punishment was given for wrong choice. Rat number 52 in group VII was given 280 trials under such conditions. The record by series runs: 60, 80, 75, 65, 50, 75, 70, 65, 55, 70, 75, 50, 75, 65. Since no signs of improvement were appearing the work was dropped. Because of the above considerations the conviction was reached that the failure to learn the positive response to the sound stimulus was not due to the animals' being negative to the sounds presented. The fourth possible explanation for the failure that sug- gested itself was that the rat could not localize the sounds well enough under the circumstances to enable it to guide its responses by them. This does not mean that the rat can not localize sounds, but that, if it can, the conditions under which the experiments were conducted were not suited to call localization into play. When the experiments began it was assumed that the rat could localize sufficiently well under the conditions presented to enable it to guide its response by the sound after a certain period of training. The records for Groups VI and VII contain some signi- ficant things that require some comment and explanation. It will be recalled that pain for a wrong choice was given in these two groups. The records show that Numbers 42, 46, 47, 48, and 50 gave considerable evidence that they were actually learning the response. The fact is. Numbers 42, 47 and 50 actually met the standard required, that is, 95 correct responses out of 100 chances. Why was it that these records were not accepted as indicating that those animals had learned the sound? The fact was that their general behavior did not indicate definitely enough that they were actually responding to the sound stimuli when they were making enough correct responses to come close to meeting the standard required. The experimenter became convinced that some other factor was present in these cases as the determining factor. To make plain the basis for this conviction let us notice again the method of procedure. The keys which closed the circuit containing the induction coil and controlled the electric current used TRANSFER OF RESPONSE IN THE WHITE RAT 27 for punishment were located just to the left of the keys (K and K') used for working the sounders. The method of procedure had regularly been to use the left hand for working the 'sounder' keys and the right for opening and closing the doors of the box whenever necessary and for handling the animal. To have the left hand free to work the 'sounder' keys the right was also used to work the 'punishment' keys whenever necessary. The various movements made by the experimenter were usually as follows: The left hand remained resting at the 'sounding' keys while the right was used in adjusting the return doors (RD and R'D' in figure I) and in putting the animal into the entrance alley (EA). Then the proper 'sounder' key was manipulated by the left hand and the sound stimulus given until a choice had been made, the right hand in the meanwhile resting upon the edge of the box. If a wrong choice had been made the sound was discontinued and the right hand was stretched across to the 'punishment' key and punishment given. In case of the animals mentioned above as showing evidence by the records of learning the required response the experimenter became convinced from observation of the general behavior that it was this very evident motion of the right hand or perhaps of his whole body that in some way served as a warning when about to enter the wrong alley. So a change was made in procedure. Instead of using the right hand for pressing the 'punishment' key, the right hand was allowed to remain lying on the edge of the box and the left was slipped as unobtrusively as possible from the position of the 'sounder' keys to that of the 'punishment' keys. A com- parison of the records for five series before this change was made and for five series after the change was made will be surprising. Practically every animal which gave evidence that the response was being learned showed a considerable decrease in efficiency of response after this change was made. For the sake of comparison I have gathered them together in the following table: 28 HARRY H. WYLIE Before After No. 39... 65 60 80 80 90 45 85 80 55 50 No. 40... 65 50 55 80 80 55 65 65 70 40 No. 41 . . . 85 75 80 80 95 75 70 70 55 40 No. 42... 100 90 90 95 100 50 60 80 70 65 No. 43 . . . 80 85 80 90 95 75 55 45 60 70 No. 45 . . . 65 45 80 75 65 55 50 45 55 60 No. 46.. . 95 95 100 95 80 50 40 70 60 80 No. 47... 95 100 100 100 100 85 60 65 70 65 No. 48 . . . 80 95 95 90 80 20 40 65 55 60 No. 49 . . . 80 70 95 95 65 50 55 60 70 70 No. 50... 90 85 100 100 100 60 65 75 70 80 No. 51... 85 90 65 80 75 70 60 40 45 65 No. 52 . . . 55 75 70 80 80 35 55 50 These records indicate that it is highly probable that the rats were depending in part at least upon certain movements made by the experimenter rather than upon the sound stimulus which was to serve as the guide in making the choice. This complete failure to get any rats, either trained pre- viously with the light or wholly untrained, either with punishment for wrong choice or without punishment, ta learn the positive response to any sound stimulus presented was one of the most surprising results of the experiments. As suggested above the most plausible reason for this failure seems to be that the rats were unable to localize the sound sufficiently well under the conditions of the experi- ment to enable them to utilize the sound in learning the response required. However, it would not be correct to say that these animals trained with the sound made no improvement at all. The fact is that all of them actually made improvement during the first 300 trials. To what was this improvement due? To answer that question we need first to know the amount of improvement made. We find that Group III in the first 100 trials made 49.8 per cent of correct responses, while in the fourth 100 trials the same group made 61 per cent of correct responses. Group IV in the first 100 trials made 51.2 per cent of correct responses, while in the fourth 100 trials the same group made 60 per cent of correct responses. Group V in the first 100 trials made 53.6 per cent of correct responses, while in the fourth 100 trials the same group made 65.8 per cent of correct responses. Thus roughly for the three groups the gain amounted to about 10 per cent. TRANSFER OF RESPONSE IN THE WHITE RAT 29 We have already called attention to the fact, that with the series of choices required, chance would account for 50 per cent of correct responses and simple alternation between left and right would account for 70 per cent of correct responses. We have also called attention to the fact that the records show a tendency toward simple alternation in the case of the sound stimulus, but that there was no consistency at all in maintaining such an alternation. We are led to conclude, therefore, that whatever gain was made by those animals trained with the sound stimulus was due to a partial perfecting of the simple alternation position habit. In the case of those which learned the positive response to the light stim- ulus, since there is no direct evidence in the records of such a position habit having been acquired while learning the light, we would be led to believe that the light stimulus itself became the increasingly dominant factor right from the beginning of the learning process. However, when those same animals were transferred to sound, the records immed- iately show, as we have pointed out, that such a position habit had been partially formed. Thus the evidence seems conflicting. However, the conclusion seems warranted that chance and simple alternation, imperfectly carried out, will account for 60 per cent of correct responses when using such a series as was used in the experiments with positive response. Since such improvement is made in the first 300 trials any improvement beyond that amount, whether during the first 300 trials or later, must be accounted for in other ways. Thus Group I required only 120 trials with the light stimulus to surpass 60 per cent. Group II needed only 180 trials to surpass 60 per cent. Our conclusion must be that the light stimulus began to be effective in both cases not later than during the second 100 trials and reached its maximum influence at the end of 500 trials. But we still have the question why it required those trained with the light only 120 and 180 trials to reach an efficiency of 60 per cent whereas it required those trained with the sound 300 trials to reach the same point. We have already found that the sound stimulus had no influence since it could not be localized. We have, therefore, to take account of the light only. In. one case it was present 30 HARRY H. WYLIE and in the other case it was absent. Therefore we must account for the rapidity with which the 60 per cent standard was reached in the first case as compared with the second by means of the effect of the light stimulus. Therefore we can conclude that the light stimulus was a factor practically from the beginning in learning the response, although upon the removal of the light the position habit, gained partly through the influence of the light and partly by mere re- petition, still remained effective. (3) Summary, — (a) White Rats do not localize sounds sufficiently well to make use of them as a guide in learning under the conditions of these experiments. (6) Movements of the experimenter can be utilized by white rats as cues which aid materially in the learning process. (c) White rats will not learn and maintain consistently, under the conditions of these experiments, a simple alter- nation position habit even though given 1700 trials. (d) In learning such a series as the one used in these experiments where a simple alternation position habit is of considerable value in getting the correct order of responses, such a position habit is not formed first independently of the light stimulus and then the learning process completed in terms of the light stimulus. Rather there is evidence that the light stimulus is more or less effective in helping to establish whatever position habit enters into the order of responses and is wholly effective in carrying the learning process beyond the point where position habit alone could carry it. (e) The learning process in this sort of problem is evidently very gradual, though somewhat irregular, having neither a rapid initial rise nor a long final perfecting of the response. B. Negative (1) Light, — Six groups of animals, VIII, IX, X, XI, XII, XIII, were taught the negative response to the light stimulus. The results have been summarized in the following tables: TRANSFER OF RESPONSE IN THE WHITE RAT 31 Animal No. 53. No. 54. No. 55. No. 56. No. 57. No. 58. Average. Animal No. 93.. No. 94.. No. 95.. No. 102. No. 99.. Average. Group VIII Trials 255 255 240 195 255 225 237.5 Group X Trials 210 195 180 210 180 195 Series 17 17 16 13 17 15 15.8 Series 14 13 12 14 12 13 Group IX Animal Trials No. 77 210 No. 78 255 No. 79 225 No. 80 285 Average 243.75 Animal No. 84.. No. 85. . No. 86. . No. 91.. No. 97.. No. 101. Average. Group XI Trials 120 165 195 150 165 120 Series 14 17 15 19 16.25 Series 11 13 10 11 152.5 10.2 Group XII Animal Trials Series No. 81 225 15 No. 82. . . . 225 15 No. 89.... 225 15 No. 90. . . . 165 11 No. 98.... 135 9 Group XIII Animal No. 83. No. 87. No. 88. No. 92. No. 96. Average 195 13 Average. . . Average for 31 animals, 198.87 trials, 13.26 series. Trials 195 195 165 210 135 180 Series 13 13 11 14 9 12 To understand these tables it is necessary to recall that with the negative response the stimulus was given in but 15 trials out of each series of 20 trials, the other five trials in each series being used for control tests as explained under method of procedure. The important question here would be whether we can be sure that the responses were actually made to the light. As far as the writer could tell the other possible types of stimuli which might have been functioning were the sight of the experimenter's movements, the mechanical jarring of the apparatus produced by working the keys, and the kin- aesthetic which would show itself in turning habits due to regularity of position and order of the required turning. As to the first mentioned, as much care was exercised as was possible to make such movements inconspicuous. The 32 HARRY H. WYLIE keys were not in the range of the animal's vision and the finger movement or forearm movement necessary to work them was also below the line where vision might function. This result was secured mainly by having the alleys six inches deep, making it practically impossible for the animals to see any extraneous object or movement except where it appeared immediately over the alley. The rest of the ex- perimenter's body was practically immobile during the time in which the reaction was to be made. As to the mechanical jarring of the apparatus, the extremely light touch necessary to work the keys would certainly render the jarring negli- gible. In the writer's opinion the most likely to manifest itself of the three possible types of things mentioned above is the third. When an animal is required to turn back fif- teen times out of every twenty, it is extremely probable that it will acquire a habit of turning back, even when no stimu- lus has been given. As has been stated the series of trials used was particularly arranged to meet such a contingency. The possibility of "order in the series" functioning was eliminated by introducing irregularly the trials when the stimulus was omitted. The possibility of "position in the ■alley" functioning was eliminated by constantly varying the point at which the sound or light stimulus was given. Finally in cases where the animal did turn back and entered the other return alley when no stimulus was given by the experimenter or before such stimulus had been given, the stimulus was given in the other alley and the animal required to return to the alley first chosen to get back to the food box. Since such cases were confined practically to the work with the sound stimulus we shall omit the discussion of them here. These cases were so rare when using the light that they can well be neglected in considering the results. The outstanding feature of the learning process here is the shape or form of the learning curve. To bring out the peculiar features more plainly, a number of curves are presented. These curves show the record not only for learning the light but also for the transfer to sound and pain. See Figures V to XIII inclusive. It will be noticed that these curves show a period at first. TRANSFER OF RESPONSE IN THE WHITE RAT 33 Figure VI Curves 54 (Figure V) and 56 (Figure VI) are individual curves from Group Eight. The horizontal lines represent the percentage of correct responses. The vertical lines represent the number of series of 20 trials each. In lettering.^A is the end of the learning process with light, B is the end of the learning process with the regular sound stimulus, C is the end of the learning process with the bell, D is the end of the learning process with the pipes. 34 HARRY H. WYLIE Figure VII Figure VIII Curves 77 (Figure VII) and 80 (Figure VIII) are individual curves from Group Nine. The horizontal lines represent the percentage of correct responses. The vertical lines represent the number of series of 20 trials each. The dotted lines repre- sent the record for correct responses and for hesitations. In lettering, — A is the end of the learning process with the light, B is the end of the simultaneous series, C is the end of the learning process with the sound. TRANSFER OF RESPONSE IN THE WHITE RAT 35 100. 1S~ SO' 2i- k ^ — - — ' — ¥■ 4-^ . i 1 : . \ — — ' — — i '"1 , 1 } ■ : I 5 ^ \=zz ES ^- — u_- ^: ^ — -: ^ ^_^ ^==^ ^—-m f :^:e 1 — 1 — ' — ' — 1 ' 1 3 T?-. Figure IX Figure X Curves 99 (Figure IX) and 102 (Figure X) are individual curves from Group Ten. The horizontal lines represent the percentage of correct responses. The vertical lines represent the number of series of 20 trials each. The dotted lines represent the record for correct responses and for hesitations. In lettering, — A is the end of the learning process with the light, B is the end of the simultaneous series, C is the end of the learning process with the sound. 36 HARRY H. WYLIE Figure XI Figure XII Figure XIII Curve 91 fFiKure XI) is an individual curve from Group Eleven. Curve 90 ("Figure XII) is an individual curve from Group Twelve. Curve 88 (Figure XIII) is an individual curve from Group Thirteen. The horizontal lines represent the per- centage of correct responses. The vertical lines represent the number of series of 20 trials each. In lettering.^A is the end of the learning of the hght, B, m curve 91, is the end of the learning with the electric shock, and is the end of the simul- taneous series in curves 90 and 88, C is the end of the learning process with the elec- tric shock. TRANSFER OF RESPONSE IN THE WHITE RAT 37 although short, of practically no learning, then a very sudden jump up to perfect response, rather than the sudden initial rise and the final long period of perfecting the process found in many learning curves. In the dotted curves it appears that the steepness of the ascent is modified a little in some cases. These take into account not only the correct responses but also the "hesitations^" and the times of 'slowing up' due to the stimulus presented. But the light was learned so quickly that little difference is made by tak- ing these additional evidences of learning into account. The few series necessary for learning the light also prevents the early period of non-learning from standing out clearly. One individual case in particular, number 84 in Group XI, illustrates the rapidity of learning very strikingly. The first series in this animal's record shows one correct response out of 15 chances, this one being the last of the fifteen. The next six series show a perfect record. This makes its record run as follows, counting in per cent of correct responses: first series, 6 2-3 per cent; second series, 100 per cent; third series, 100 per cent; fourth series, 100 per cent; fifth series, 100 per cent; sixth series, 100 per cent; seventh series, 100 per cent. One could scarcely find anywhere a record in animal learning that would equal it. Since similar phenomena appear in learning the negative response to the other sorts of stimuli used, discussion will be reserved until those results have also been given. (2) Sound,— Six groups of animals, XIV, XV, XVI, XVII, XVIII, XIX, were taught the negative response to the sound stimulus. The results have been summarized in the following tables: Group XIV Animal Trials Series No. 59 345 23 No. 61 465 31 No. 63 540 36 No. 64 435 29 No. 65 465 31 Average 450 30 GROU.f^ XV Animal Trials Series No. 60 450 30 No. 67 375 25 No. 68 345 23 No. 69 300 20 No. 70 345 23 No. 73 405 27 Average 370 24.66 38 Group XVI Animal Trials No. 62 540 No. 66 495 No. 72 285 No. 74 405 No. 75 ,450 No. 76 450 Average ^37.5 HARRY H. WYLIE Group XVIII Series Animal Trials Series 36 No. 129. . . 795 53 33 No. 130. . . 765 51 19 No. 131... 630 42 27 No. 132. . . 795 53 30 No. 133... 795 53 30 — Average. . . 756 50.4 29.16 Group XIX Animal Trials Series No. 152 705 47 No. 153 630 42 No. 154 645 43 No. 155 630 42 No. 156 705 47 Average 663 44.2 Average for 27 animals, 525.5 trials, 35.03 series. The record for Group XVII is not included in the average with the remaining groups since we were compelled to make some changes in the method of procedure with that group. It was the first group to which we attempted to teach the negative response to sound and by means of it a final method was worked out. We gave the seven members of this group on the average fifty series of trials with the same sort of series as was used in the positive response previously re- ported and with the single sounder located in the food box and marked S in figure II. After the complete failure of all seven animals to show any evidences of learning in the fifty series of twenty trials each except such as could be ascribed to position habits, the three sounders along each side of the experiment box were introduced and the type of series adopted which was used in all the experiments with the negative response. After these changes were made the record for this group in learning the sound is practically the same as the normal record for sound secured later with the other groups reported. To bring out the main features of the learning process when the sound was used, I have inserted a few typical curves. These curves show the same general features as those illustrating the learning with the light. In these curves, however, some of the distinctive features show up TRANSFER OF RESPONSE IN THE WHITE RAT 39 much more plainly than in those for the light stimulus. These curves show the complete record for the animals, both while learning the sound and when transferred to the light and pain. Figures XIV to XVII. The significant features of the learning process with the negative response to the sound stand out most clearly in the individual curves presented. These features have already been mentioned — the rather long period of no pro- gress at first, the sudden and rapid perfecting of the response when once learning begins, and the few errors after once the response is learned. An apparent exception is the record of number 61 of group XIV. Here the early part of the record is very irregular until the eighteenth series, when it rises regularly and rapidly to perfection at the twenty- fifth series. In contrast with this record stands that of number 67 of Group XV. No sign of learning was shown there until the thirteenth series and learning was prac- tically complete in the twenty-first series. Or still more striking is the contrast shown by the record for number 154 of Group XIX. While there is some slight evidence from " hesitations " that the sound stimulus was being noticed from the very first, not until the thirty-fifth series did good evidence of learning appear, and learning was practically complete by the thirty-eighth series. The record for number 129 of Group XVIII shows the most gradual rise of any, although for seventeen series, there is scarcely any sign of learning. For this rat it took thirty-one series to complete the learning after it began to exhibit pro- gress whereas the usual number with the sound is much less, one even falling as low as four. It is the only case where the learning can be truly spoken of as a gradual process. (3) Pain, — Seven groups of animals, XX-XXVI in- clusive, were taught the negative response to pain. All of these represent normal records except Group XXII. Its record, therefore, has been omitted. This group was the first to be trained with the pain stimulus and served in part to work out the particular method for giving the pain stimulus. At first the stimulus was given at one point only as the animal came down the alley, the point being 40 HARRY H. WYLIE (00 Figure XIV Figure XV Curve 61 (Figure XIV) is an individual curve from Group Fourteen. Curve 67 (Figure XV) is an individual curve from Group Fifteen. Ho izontal lines represent percentage of correct responses. Vertical lines represent the number of series of twenty trials each. In lettering,— A is the end of the learning process with the sound, B (in curve 61) is the end of the learning process with the light, B (in curve 67) is the end of the simultaneous series, C (in curve 67) is the end of the learning process with the light. TRANSFER OF RESPONSE IN THE WHITE RAT 41 Figure XVI Figure XVII Curve 129 (Figure XVI) is an individual curve from Group Eighteen. Curve 154 (Figure XVII) is an individual curve from Group Nineteen. Horizontal lines repre- sent the percentage of correct responses. Vertical lines represent the number of series of twenty tnals each. Dotted hnes represent the record both for hesitations and for correct responses. In lettering.^A is the end of the learning process with the sound, B is the end of the simultaneous series, C is the end of the learning process with the pam. 42 HARRY H. WYLIE varied from trial to trial. However, later to make the conditions more nearly uniform with the training given with other sorts of stimuli, the stimulus, when once given, was continued until the animal turned back. The follow- ing tables summarize the results: Average. Group XX Group XXI Animal Trials Series Animal Trials Series No. 165. . . 210 14 No. 162... 210 14 No. 166. . 210 14 No. 163. . . 240 16 No. 167... 195 , 13 No. 164... 165 11 No. 170. . . 150 10 No. 168... 165 11 No. 169. . . 210 14 Average. . . 191.25 12.75 Average. . . 198 13.2 Group XXIII Group XXIV Animal Trials Series Animal Trials Series No. 157... 195 13 No. 146. . . 195 13 No. 158... 180 12 No. 147... 150 10 No. 159... 150 10 No. 148. . . 195 13 No. 160. . . 210 14 No. 149... 225 15 No. 161... 240 16 No. 150. . . 150 10 No. 151... 180 12 A rage. . . 195 13 Ave Average. . , 182.5 12.17 Group XXV Group XXVI Animal Trials Series Animal Trials Series No. 141... 180 12 No. 134... 270 18 No. 142... 165 11 No. 135. . . 255 17 No. 143... 150 10 No. 136... 225 15 No. 144... 135 9 No. 137... 195 13 No. 145. . . 180 12 No. 138... 180 12 No. 139. . . 165 11 162 10.8 Average. . Average for 31 animals, 191.13 trials, 12.74 series. 215 14.3 A number of illustrative curves are given below. These curves show the record of these animals not only while learning the response to the pain stimulus, but also when transferred to the other sorts of stimuli. They will be referred to later also when discussing transfer from pain to Hght and sound. See Figures XVIII to XXI inclusive. Here again practically all curves show a period of no pro- gress at the beginning, then a rapid rise to perfection and no long process of perfecting. TRANSFER OF RESPONSE IN THE WHITE RAT 43 100' 1^~ ■1 i ' ™ N = 1 .a_ ^ : 1 1 — - — i 3 SO' c Oz ^'Trr a 1 -yi N ^ ^ 3 SB -— mi^mm ■ . ■ 1 = 5 Figure XVIII Figure XIX Curve 165 (Figure XVIII) is an indivndual curve from Group Twenty. Curve 162 (Figure XIX) is an individual cur\'e from Group Twenty-One. Horizontal lines represent the percentage of correct responses. Vertical lines represent the num- ber of series of twenty trials each. In lettering, — A is the end of the learning process with the pain, B (in curve 165) is the end of the learning process with the light, B (in curve 162) is the end of th3 simi Itaneous series, C (in curve 162) is the end of the learning process with the light. 44 HARRY H. WYLIE Figure XX Figure XXI Curve 157 (Figure XX) is an individual curve from Group Twenty-three Curve 141 (Figure XXI) is an individual curve from Group Twenty-five. Horizontal lines represent the percentage of correct responses. Vertical lines represent the number of series of twenty trials each. In lettering, — A is the end of the learning process with the pain, B is the end of the simultaneous series, C is the end of the learning process with the sound. TRANSFER OF RESPONSE IN THE WHITE RAT 45 Another noticeable feature in these cases is the smooth- ness of the rising curve. It seems that when once the pro- cess is started there is practically no loss, but what is gained in one day's work is retained in the next in practi- cally all cases. This last feature does not stand out so strikingly in the curves for learning the negative response to the light stimulus, and still less in the curves for learn- ing the sound stimulus. Another interesting feature closely related to the one just mentioned is the almost complete absence of marked plateaus. The curves for pain are the freest, those for light next, and those for sound least free. This feature would indicate that for white rats working under the conditions of these experiments with negative response, the electric shock is the most uniformly effective stimulus of the three sorts of stimuli used. The light ranks second, and sound ranks third. We have every reason to believe from their general behavior in the presence of the electric shock that white rats are naturally negative to such a shock. We do not have definite proof that white rats are negative to the light and sound used. From gen- eral observation of their behavior in the presence of the light and sound stimuli used the writer is inclined to believe that they are more likely to prove to be negative to the sound than to the light. The electric shock uniformly produced the greatest emotional disturbance in the ani- mals, calling for the greatest exercise of patience on the part of the experimenter. The electric shock most uniformly compelled the attention of the animal. The first few series of non-learning with the electric shock were not due to lack of attention to the stimulus as was quite evidently the case, in part at least, with the sound and light. Rather they were due to emotional disturbance produced by the strangeness of the stimulus. So these non-learning series might after all be counted as series indicating the acquire- ment of emotional adjustment to the stimulus rather than the specific response adjustment which the rise in the curve is supposed to indicate. Thus it might be claimed that, if an ideal learning process is one in which there is progress practically from the beginning and one which shows no 46 HARRY H. WYLIE relapses nor plateaus, the process of learning the negative response to the electric shock in these experiments repre- sents practically an ideal case. There is no striking difference in the lengths of the non- learning periods for the pain and for the light. Neither is there any important difference in the number of series necessary to complete the learning process with the pain and with the light. The only really significant difference is in the smoothness of the rise of the curve in each case, yet this difference is not so great as to be very striking. It is likely due, as suggested above, to the greater power of compelling attention which the electric shock seems to possess. Its unlikeness, measured from the human stand- point, to anything else falling within the normal experi- ence of the white rat might be the secret of its power. But we still have the question of why the light was prac- tically just as effective as the electric shock while the sound was decidedly less effective, having both a much longer non-learning period, a much greater irregularity of effec- tiveness, and a much longer learning period. Why this advantage of light over sound if the animals are really negative to neither? As suggested above, the animals seemed more inclined to be negative to the sound than to the light, if we are to judge from the general disturbance in their behavior produced by the two. If this were true, one would expect the response to the sound to be learned more quickly than to the light. Just the opposite was the case. Moreover, the animals were evidently negative to the electric shock but not to the light. Yet they learned the response to the light in practically the same way and practically as quickly as to the electric shock. But it might be that the greater emotional disturbance caused by the electric shock offset its greater effectiveness due to the animal's being negative to it. Perhaps the explanation of the comparative effectiveness of the light and sound can be found partly in the general conditions under which the experiments were conducted. It was much easier to control the light from external sources than the sound. So the light when presented represented a more uniform contrast with the general environment than the sound when it was TRANSFER OF RESPONSE IN THE WHITE RAT 47 presented. This more uniform contrast might have in- creased the effectiveness. Then we still have the problem of the comparative intensities of stimuli from the different sense fields. For instance, how intense a sound shall be considered to be equal to a given intensity of light or of electric shock? We have no way of determining this inde- pendent of the reaction of the animals. All that can be said in the present case is that under the conditions of the experiments the electric shock was the most uniformly effective stimulus, the light next, and the sound least, in establishing the required negative response. Another interesting thing not indicated in the curves and tables was the attempt on the part of some of the animals when learning the response to the pain to avoid the stimu- lus by jumping over the places where the shock had been previously received. A few typical cases are those of numbers 114, 115, 124, 128. The record of number 114 is as follows: in the fifth series of trials, it jumped twice; in the seventh series, five times; in the eighth, four times; in the ninth series, once; in the eleventh series, twice. The record for number 115 is as follows: in the fourth series, once; in the fifth series, three times; in the eleventh series, once. The record for 124 is as follows: in the twelfth series, five times; in the thirteenth series, once; in the sixteenth and seventeenth, once each. The record for 128 is: in the eighth series, eight; in the tenth series, twice. All of these " jumps " occurred before the change of procedure took place as mentioned in the description of this series of ex- periments. That is, they all occurred when the pain stimu- lus was given at but one place in the alley rather than all the way down as was the case in the method finally used. It is noteworthy, too, that these " jumps " occurred as a sort of preliminary method of responding to the pain stimu- lus, since after the animal had really begun to choose, or, in other words, to make the correct response no attempt at such avoidance occurred. Not all the animals in this group which learned the pain stimulus attempted to avoid it in this way. The four animals mentioned are the best examples in the records. The fact of " jumping " indicates a more or less definite 48 HARRY H. WYLIE sense of location, although there is no evidence that the " jump " was even generally made at the place where the last shock had been received. Most of them occurred at the first place in the alley where it was possible to receive the shock or at the last place. Some few occurred at every place in the alley, that is, when the animal started down the alley, it jumped at every place where a shock was likely to be received and so complete'y avoided the shock. One would be led to suspect by gross observation that vision was mainly instrumental in guiding the reaction in this jumping. (4) Summary, — If we are to gather together the results for learning the negative response we find the following facts outstanding: (a) The electric shock is the most effective stimulus, both from the point of view of rapidity of learning and from the point of view of smoothness of the learning curve, there being practically no lapses and no plateaus. (b) The effectiveness of the electric shock is probably due to the fact that the animals were negative to the shock. The failure to show progress from the beginning was likely due to the emotional disturbance produced by the strange- ness of the stimulus. (c) The superior effectiveness of the light over the sound was probably due in part but not wholly to the better control of general environmental conditions in the case of the light. (d) *The inferior effectiveness of the sound in com- parison with the light and pain might indicate also that the visual and tactual stimuli play a more important part in the regular life of the white rat than do auditory stimuli. While the tactual has always been assigned a prominent place, there has been a tendency to belittle the importance of the visual. Another result pointing toward and empha- sizing the visual has already been mentioned in discussing the results of the experiments where pain was given as pun- ishment for wrong choice when learning the positive re- sponse to the sound stimulus. Also, if " jumping " just reported is guided by vision, it furnishes an added case in point. Bearing on this point, see, Pearce, Jour. Animal Beliav., Vol. 7, No. 3, p. 177. TRANSFER OF RESPONSE IN THE WHITE RAT 49 (e) Attention has been called to the unusual form of the majority of the individual learning curves of animals learning the negative response. Contrary to the nature of the ordinary animal learning curve the greatest improve- ment does not take place in the earlier trials, but there is at first a period of little or no progress and later a period of rapid improvement with a quick perfecting of the re- sponse. In form these curves come more nearly into agree- ment with Swift's curves on Learning to Toss and Catch Balls. *As far as the writer is able to discover, the curves given in this paper are the first animal curves of this nature to be reported. Going on the assumption that perhaps the difference was due to the fact that all the evidences of learning were not being recorded, an attempt was made to correct this failure by taking note of " hesitations " and " retardations " in the animals' actions which could evi- dently be assigned to the stimulus as cause. But even this did not " correct " the form of the curves. As far as could be determined " errors " or " choices," with modifications of the latter in the form of " hesitations " and " retarda- tions of movement," were the only criteria which could be made use of in measuring the progress of learning or the relative rate of learning. Unless we posit some change in the organism which is produced by the various trials but which does not manifest itself extrinsically in measurable behavior until a certain degree of strength has been reached, and then manifests itself by taking almost complete control of the animal's activity' in that situation, we are at a loss to account for the difference in the rate of progress of learning. That would mean that in certain kinds of prob- lems we might expect learning curves of this kind or, in other words, that the shape of the learning curve depends in part upon the nature or kind of problem. In contrast with these curves for the learning of the * Note: Professor Yerkes, of Harvard, reports in his monograph on the Mental Life of Monkeys and Apes (Behavior Monographs, Volume 3, Number 1, 1916, page 68) that one of his animals (Julius) showed the type of learning curv^e found in the work reported in this paper and claims that never before has a curve of learning like this been obtained from an infra-human animal. However, his work was done subsequent to the work reported in this paper. The work reported here was done during 1913 and 1914, but, owing to unavoidable delay has not been published until now. 50 HARRY H. WYLIE negative response, the curves for the learning of the posi- tive response show none of these pecuHar features. In the positive curves there is evidence of learning from the very beginning, the process is quite gradual until completed, the process is quite irregular, showing both lapses and plateaus, and when learning is once completed, the stand- ard is not so consistently maintained as in the negative. Since the general situations in the two problems were quite alike and the specific stimuli used (light and sound) were quite alike, the different character of the responses required must be largely the cause of the striking differences in the learning curves. It is to be remembered that the negative curves for sound do not stand out in as great contrast as the other negative curves. SECTION FOUR FEATURES BEARING ON TRANSFER A. Positive (1) Light to Sound, — Two groups were trained with the light stimulus and transferred to sound. Group I was transferred directly from the light to the sound when once the response to the light had been learned. With Group II each animal was given ten series or two hundred trials with the light and sound stimulus presented simultaneously, after the response to the light had been learned, and then transferred to the sound. The curves given on page 20 show the general results. These curves show the learning of the response to the light stimulus, but they also show the record for the control series in each case, the record for any simultaneous light and sound stimuli, and the record for the sound stimulus alone which was substituted for the light stimulus. The immediate result of the transfer was that the two groups dropped in efficiency 36 per cent and 26 per cent respectively, or down to that degree of efficiency which can be. accounted for by chance or position habits. Three of Group I were given 25 series or 500 trials with the sound alone and at that time showed an average efficiency of 72 per cent whereas they started the sound with an average of 67 per cent. The other two of this group were given 11 series or 220 trials with the sound, and ended with an average of 70 per cent, whereas they started the sound with an average of 55 per cent. All six members of Group II were given 20 series or 400 trials with the sound alone after the series with sound and light simultaneously and ended the 20 series with an efficiency of 67 per cent whereas they began the sound alone with an average of 73 per cent. As stated above a study of the records shows that the animals all began to depend almost wholly upon alterna- tion when the sound was substituted for the light, although 52 HARRY H. WYLIE they do not maintain it with any high degree of regularity at all. Because no promise of further results was evident and because of the negative results in learning sound with groups already reported the work with these animals was discontinued. (2) Sound to Light, — Since no animals learned the posi- tive response to the sound we had none to transfer from sound to light. This part of our program had to be aban- doned. B. Negative (1) Light to Sound, — -Three groups of animals were first taught the negative response to the light and then were transferred to the sound. Group VIII was transferred directly from light to sound. Group IX was given two series of simultaneous light and sound stimuli after having learned the response to the light and then were transferred to the sound. Group X was given four series of simul- taneous light and sound stimuli after having learned the light and then were transferred to the sound. The results are gathered together in the following tables. These tables indicate the number of series and trials necessary for learn- ing the response to the sound after having learned the response to the light, and after having been trained with, the simultaneous light and sound wherever such training was given. Illustrative curves showing the complete records of certain individuals may be seen on pages 33-35. To understand these tables one must keep in mind that only fifteen chances to respond to the stimulus were given in each series of twenty trials. Group VIII Group IX This group was transferred directly from This group was given two series of light to sound simultaneous light and sound Animal Trials Series Animal Trials Series No. 53 150 10 No. 77 165 11 No. 54 150 10 No. 78 150 10 No. 55 150 10 No. 79 165 11 No. 56 165 11 No. 80 120 8 No. 57 150 10 ■ No. 58 150 10 Average 150 10 Average 152.5 10.16 TRANSFER OF RESPONSE IN THE WHITE RAT 53 Group X This group was given four simultaneous light and sound Animal Trials Series No. 93 105 7 No. 94 105 7 No. 95 105 7 No. 99 105 7 No. 102 105 7 Average 105 7 Summary of Results Trials Series Average for learning sound (negative) shown in normal records . . 525 .5 35 . 03 Gained by direct transfer 373 24.87 Gained by 2 simultaneous light and sound 375.5 25.03 Gained by 4 simultaneous light and sound 420.5 28.03 The first thing to be noticed in the results is the fact that it is a decided advantage in learning the response to the sound to learn first the response to the light and then substitute the sound stimulus for the light stimulus either directly or with a few intervening series in which the sound and light are simultaneously given. Thus it took on the average 35.03 series for the animals to learn the response to the sound when no previous training had been received in making such a response. But by training first with the light stimulus and then substituting directly the sound stimulus only, 25.96 series were needed for learning the response to both the light and the sound and only 10.16 of these series were used in learning the response to the sound. That represents a saving of 24.87 series in learning the response to the sound by first learning the response to the light. By presenting the two stimuli simultaneously for a few series after the response to the light had been learned, a greater saving was effected. Thus with two simultaneous series a saving of 25.03 series was secured, while with four simultaneous a saving of 28.03 series was secured. This means that by training with the light until the response had been learned and by giving four series with light and sound simultaneously, the animals were trained to respond to the sound alone with a sufficient degree of accuracy to enable them to meet the standard during the next seven series, the shortest possible time in which they could meet the standard set. It also means that the quickest way for 54 HARRY H. WYLIE the animals to learn the response to the sound is not to train them directly with the sound right from the first, but is to train them first with the light, then give a few simultaneous light and sound series, and then give the sound alone. The latter method required only 24 series altogether for the light, for the simultaneous light and sound, and for the sound, while the former method required 35.03 series all told for the sound only. With Group VIII after learning the response to the light and the sound of the regular electric sounder, tests were also made with the sound produced by striking a three inch bell with a rubber hammer. This bell was held by the experimenter in front of the box and below the vision of the animal. After tests with the bell, tests were also made with wooden pipes of 2048, 1024, 512 and 256 vibra- tions. These pipes were sometimes blown by the experi- menter who handled the animals during the experiments and sometimes by an assistant who stood just behind the regular experimenter. In sounding the bell or pipes no attempt was made to make a regular continuous sound. The only attempt was to keep the sound of moderate loud- ness whether given irregularly or continuously. After a few trials with these devices as sounders, it was the usual thing for the animal to turn back at the first sound of the bell or the first " toot " of the pipe, that is, the transfer was practically perfect. It was in these trials with the sound after having learned the light, that there was some evidence of " position in the alley " functioning as a stimulus to cause the return of the animal, that is, the animal would turn back without any sound stimulus having been given. In cases where the animal did turn back, when no stimulus was given by the experimenter or before such stimulus had been given, and entered the other return alley the stimulus was given in the other alley and the animal required to return to the alley first chosen to get back to the food box. Such cases, however, were rare. A few typical records run as follows: Number 77 made one false return out of 540 trials; number 78 made four false returns out of 580 trials; number 79 made three false returns out of 560 trials; number 80 made three false returns out of 580 trials; number 102 made TRANSFER OF RESPONSE IN THE WHITE RAT 55 three false returns out of 500 trials. The number of false returns recorded run from none to four per animal. Since they were so few no record was kept except for a part of the animals used. It must be noted, however, that prac- tically all of the false returns were made after the animal had learned the response to one stimulus and had been transferred to the other. Practically all of them also are to be found in the records where sound was the stimulus being used. Some of the false returns can be accounted for by extraneous sounds from outside the laboratory. These false returns, to the writer's mind, indicate that the response had become considerable of a reflex and any sort of stimulus of a kind similar to the one which had become effective was sufficient to set it off rather than a matter of habitual turning at a particular place in the alley or at a particular position in the series of trials; in other words, the response had become generalized for stimuli of that sort. Another thing which indicates that the response had become generalized for all stimuli of the sound type which could be " sensed " by the animal is the record of Group VIII with the bell and the pipes. As indicated by the record no great decrease in ability to respond was produced by substituting these two sorts of stimuli for the regular sound stimulus. The tests with the bell and the pipes were given as preliminary tests of the rat's ability to hear tones. Elec- trically driven tuning forks were also tried, but it was found to be impossible to control the sound of the tuning forks sufficiently well to suit the type of reaction required or already learned by the animal. After being trained on the light and the regular sound, the animal's reaction took place so rapidly that a mere click of the sounder was suf- ficient to stop the headlong course of the animal and cause it to turn back into the other alley. After placing the animal in the main alley, so quick was the choice that time was scarcely to be had to press the key and give the sound as a warning to turn back. In the case of the tuning forks the time required to get the fork vibrating with sufficient ntensity to make a plainly audible tone (for the human ear) was so long that the animal had made the complete run up the main alley and down one of the return alleys before the fork had begun to give out such a tone. So the 56 HARRY H. WYLIE work was abandoned as far as making it any further test for sensitivity to tonal stimuli was concerned. Hunter's work on Tonal Sensitivity in White Rats, recently published, makes it reasonably clear that the rats, in cases where the bell and pipes were used, were really reacting to the noise element of the stimuli and not to the tonal element. So even if the tuning forks could have been properly con- trolled for such a response as was here required, it is ex- tremely doubtful whether the animal's action would have been affected by it. (2) Light to Electric Shock — Three groups of animals were first taught the negative response to the light and then were transferred to the pain. Group XI was trans- ferred directly from light to electric shock. Group XII was given two series of simultaneous light and electric shock after having learned the response to the light and then were transferred to the electric shock, Group XIII was given four series of simultaneous light and electric shock after having learned the response to the light and then were transferred to the electric shock. The results are gathered together in the following tables. These tables indicate the number of series and trials necessary for learn- ing the response to the electric shock after having learned the response to the light, and after having been trained with the simultaneous light and electric shock whenever such training was given. Illustrative curves showing the complete records of certain individuals may be seen on page 36. To understand these tables one must keep in mind that only fifteen chances to respond to the stimulus were given in each series of twenty trials. Group XI Direct Transfer Animal Trials No. 84 165 No. 85 165 No. 86 165 No. 91 135 No. 97 135 No. 101 180 Average 157.5 10.5 Group XII Transferred after two simultaneous series Animal Trials Series No. 81 150 10 No. 82 180 12 No. 89 150 10 No. 90 135 9 No. 98 165 11 Average 156 10.4 TRANSFER OF RESPONSE: IN THE WHITE RAT 57 Group XIII Transferred after four simultaneous series Animal Trials Series No. 83 105 7 No. 87 105 7 No. 88 105 7 No. 92 105 7 No. 96 105 7 Average 105 7 Summary of Tables Trials Series Average for learning electric shock, Normal Record given 191 . 13 12.74 Gained by direct transfer 33 . 63 2 . 24 Gained by two simultaneous 35 . 13 2 . 34 Gained by four simultaneous 86 . 13 5 . 74 Again in this series of experiments we find decided evi- dence of transfer. The question here would be whether first learning the response to the light helped or hindered the learning of the same response to the electric shock. The results show that by first learning the response to the light a saving of 2.24 series was made in learning the re- sponse to the electric shock. Whereas by introducing two series of simultaneous light and pain after learning the light a saving of 2.34 series was made in learning the pain, and by introducing four simultaneous a saving of 5.74 series was made or, in other words, by first learning the light and by making use of four series of simultaneous light and pain, the animals were able to meet the required standard for the pain in the next seven series, the smallest number possible. One might wonder why only 2.34 series were gained by two simultaneous when 2.24 were gained by direct transfer and 5.74 series were gained by four simultaneous. The explanation may be found in the following fact. The stimulus was always more or less disturbing whenever an animal was introduced to it for the first time. This was particularly true whenever the electric stimulus was given alone. Considerable care and patience had to be exercised with the animals until they got over the emotional dis- turbance. This usually required about four series. The two or four simultaneous series, while not showing this disturbance in a uniformly low percentage of correct re- 58 HARRY H. WYLIE sponses, yet served mainly, in the experimenter's opinion, to overcome the emotional disturbance produced by such a strange stimulus. It is very possible that if the electric stimulus had had as little disturbing effect emotionally as the light apparently had, the transfer effect would have been more in proportion to the number of simultaneous series given, and so more gradual in its increase. Prac- tically the same type of result was found in the case of transfer from light to sound, where two simultaneous series showed very little advantage over direct transfer, while four simultaneous series produced considerable advan- tageous effect. (3) Sound to Light — Three groups of animals were first taught the negative response to the sound and then were transferred to the light. Group XIV was transferred directly from sound to light. Group XV was given two simultaneous series of sound and light before being trans- ferred to light. Group XVI was given four series of simul- taneous sound and light before being transferred to light. The results are gathered together in the following tables. Illustrative curves are given on pages 40-41. Group XIV Transferred directly from Sound to Light. Summary of record for Light Animal No. 59. No. 6L No. 63. No. 64. No. 65. Average. Trials 165 165 150 120 150 150 Series 11 11 10 8 10 10 Group XV Transferred with two simultaneous ser- ies of Sound and Light. Summary of Record for Light. Animal No. 60. No. 67. No. 68. No. 69. No. 70. No. 73. Average. Trials 135 120 120 120 120 105 120 Group XVI Transferred with four simultaneous series of Sound and Light Summary of Record for Light Animal Trials Series No. 62 105 7 No. 66 105 7 No. 72 105 7 No. 74 105 7 No. 75 105 7 No. 76 105 7 Series 9 Average . 105 TRANSFER OF RESPONSE IN THE WHITE RAT 59 Summary Showing Effect of Transfer Trials Series Average for learning Light, Normal Record found 198.87 13.26 Gained by direct transfer from Sound 48.87 3.26 Gained by transfer with two simultaneous series 78.87 5.26 Gained by transfer with four simultaneous series 93 . 87 6 . 26 Little need be said here except to call attention to the fact that even in learning the response to the light which proved to be very easy for the animals, there was an advan- tage in first learning the same response to some other sort of stimulus. As in the series previously reported a few series with the two stimuli simultaneously given overcame almost entirely any loss there happened to be by passing directly from learning the one to learning the other. (4) Sound to Pain — Three groups of animals were trained with the sound and transferred to the pain. Group XVII was transferred directly from sound to pain. Group XVIII was given two series of simultaneous sound and pain after learning the response to the sound and then were transferred to the pain. Group XIX was given four series of simultaneous sound and pain after learning the response to the sound and then were transferred to the pain. The results are gathered together in the tables given below. Illustrative curves are to be found on page 41. Group XVII Group XVIII Transferred directly from sound to pain Transferred with two simultaneous series Animal Trials Series of sound and pain No. 104 105 7 Animal Trials Series No! 105.'.;;;;;;; 120 8 no. 129 No. 108 135 9 No. 130 No. 109 120 8 No. 131 No. 110 135 9 No. 132 No. Ill 135 9 No. 133 No. 113 120 8 Average 141 9.4 120 8 180 12 120 8 135 9 150 10 Average 124.3 8.3 Group XIX Transferred with four simultaneous series of sound and pain Animal Trials Series No. 152 105 7 No. 153 105 7 No. 154 105 7 No. 155 105 7 No. 156 105 / Average 105 60 HARRY H. WYLIE Summary Trials Series Normal Record for learning pain found 191 .13 12 . 74 Gained by direct transfer 66 . 83 4 . 44 Gained by two simultaneous 50. 13 3.34 Gained by four simultaneous 86 . 13 5 . 74 Little comment is necessary here except to call attention to the fact that the direct transfer brought a greater saving in learning the pain series than the transfer with two simul- taneous sound and pain series. Notice was called in the series of experiments where the animals were transferred from light to sound to a somewhat similar case where two simultaneous series produced little advantageous effect over direct transfer. In the present case the two stimuli, when simultaneously given for two series, actually proved a hindrance to transfer when compared with the direct transfer effects with the same two stimuli. This is the only instance in all the experiments performed where simultaneous series proved less advantageous than direct transfer. (5) Pain to Light — Two groups of animals were trans- ferred from pain to light. Group XX was transferred directly. Group XXI was given two series of simultaneous pain and light, after having learned the response to the pain, and then were transferred to the light. The results are given in the tables below. Illustrative curves are to be found on page 43. Group XX Group XXI Transferred directly from pain to light Transferred with two simultaneous series Animal Trials Series Animal No. 165 165 11 No. 162 No. 166 165 11 No. 163 No. 167 135 9 No. 164 No. 170 120 8 No. 168 Trials Series 105 7 105 7 105 7 105 7 105 7 No. 169 Average 146.25 9.75 Average 105 7 Summary Trials Series Average for learning light, Normal Record found 198.87 13.26 Gained by direct transfer 52 . 62 3.51 Gained by two simultaneous 93 . 87 6 . 26 One interesting feature of the results is that even in the case of learning the response to the light, which was very easily acquired on the average, it turned out to be advan- tageous to learn first the response to the pain stimulus. TRANSFER OF RESPONSE IN THE WHITE RAT 61 While two simultaneous pain and light series were given the second group and that number proved sufficient to perfect the transfer effect, yet it is the conviction of the writer from observation of the general behavior of the animals while the simultaneous series were being given that one series of simultaneous stimuli would have been suf- ficient. Since this series of experiments was the last to be performed, time prevented the testing out of that con- viction. Another interesting feature not shown in the regular tables and curves was the attempt to discover how many times an animal would turn back when the stimulus was presented, without being allowed to return to the food box after a correct response had been given. If the animal chose the right alley it was confronted with the stimulus. Then when it returned and entered the left alley it was again met by the stimulus. Then upon returning to the right alley, the stimulus was again given. This was con- tinued until the animal disregarded the stimulus or until it refused to work longer. Number 168 turned back three times before disregarding the stimulus. Number 169 turned back seventeen times and then refused to work any longer. Number 162 turned back nine times before dis- regarding the stimulus; number 163 five times; number 164 eight times, and number 166 thirteen times. These figures indicate that the response had become reflex and also that it was not merely a habit of entering one alley and turning back and returning to the food box through the other. These tests were conducted with both doors to the food box open and some of the returns to the other alley were made when the animal was within sight of the open door to the food box. This test was tried only with the light stimulus and only with those animals which had been transferred from the pain to the light. (6) Pain to Sound — Five groups were trained with the pain and transferred to the sound. Group XXII was transferred directly from pain to sound. Group XXIII was given two series of simultaneous pain and sound. Group XXIV was given four series of simultaneous pain and sound. Group XXV was given six series of simul- taneous pain and sound. Group XXVI was given eight 62 HARRY H. WYLIE series of simultaneous pain and sound. The following tables gather the results together. Illustrative curves will be found on page 44. Group XXII from pain to sound Trials Series 420 28 Transferred directly Animal No. 114 No. 115 No. 116 No. 171 No. 117 No. 118 No. 123 No. 124 No. 125 No. 126 No. 127 No. 128 Average Group XXIV Transferred with four simultaneous Group XXIII Transferred with two simultaneous Animal with two series Trials 525 35 10 11 10 33 30 16 33 14 19 11 No. 157 No. 158 No. 159 No. 160 No. 161 Average 165 , 150 255 165 135 150 225 495 120 i ... 450 240 180 495 1 210 ■ 285 ; : . . 165 2 .. 312.5 20.83 Series 11 17 9 15 12 Animal No. 146. No. 147. No. 148. No. 149. No. 150. No. 151. Average. Trials 225 210 285 180 150 210 210 Series 15 14 19 12 10 14 14 Group XXV Transferred with si.\ simultaneous series Animal No. 141 No. 142 No. 143 No. 144 No. 145 ials Series 135 9 120 8 105 7 135 9 105 7 Average. 120 Group XXVI Transferred with eight simultaneous series Animal No. 134 , No. No. No. No. No. 135. 136. 137. 138. 139. Average. Trials 105 105 105 105 105 105 105 Series 7 7 7 7 7 7 Summary Average for learning sound, Normal Record found . Gained by direct transfer Gained by two simultaneous Gained by four simultaneous Gained by six simultaneous Gained by eight simultaneous Trials 525.5 213. 345.5 315.5 405.5 420.5 Series 35.03 15.2 23.03 21.03 27.03 28.03 TRANSFER OF RESPONSE IN THE WHITE RAT 63 It again stands out clearly that it is an advantage in learning one sort of stimuli to have first learned the same response to another sort of stimuli, and that a few simul- taneous series will render the transfer effect sufficiently perfect for the animal to meet at once the standard of learning set up at the beginning. A glance at the tables will show, however, that a gradually increasing number of simultaneous series does not bring a gradually increasing advantage. Attention to this fact has been called in pre- vious series of experiments. In the present case two simul- taneous series proved more advantageous than four simulta- neous series. Whether a larger number of animals would give an average which would eliminate this discrepancy and the ones previously reported is yet to be tested. Theoreti- cally, one might expect them to be eliminated. In order to get before us all the tabular results bearing on the problem of transfer, we have gathered them together and expressed the results, not only in the number of trials gained, but also in the percentage of gain. This will enable the reader to get at the amount and the conditions of transfer most quickl3^ Light to Sound Trials Normal sound record . . . 525 . 5 Gained by direct 373 88 . 7% Gained by two simul.. . . 375 89. 1% Gained by four simul. . . 420.5 100' t Sound to Light Trials Normal light record 198 . 87 Gained by direct 48 . 87 52% Gained by two simul. . . . 78 . 87 84% Gained by four simul. . . 93.87 100% Light to Pain Trials Normal pain record 191 . 13 Gained by direct 33.63 39 % Gained by two simul — 35 . 13 40 . 7% Gained by four simul. . . 86. 13 100%o Sound to Pain Trials Normal pain record 191 . 13 Gained by direct 66.83 77.5% Gained by two simul 50 . 13 58 . 2% Gained by four simul. . . 86. 13 100% Pain to Light Trials Normal light record 198 .87 Gained by direct 52 . 62 56' Gained by two simul. ... 93 . 87 100' Pain to Sound Trials Normal sound record . Gained by direct Gained by two simul.. Gained by four simul. Gained by six simul.. . Gained by eight simul 525 5 213' 50.6% 345.5 82.1% 315.5 75 % 405.5 96.4% 420.5 100% 64 HARRY H. WYLIE C. Summary of Conclusions — The results of the experi- ments reported in this paper go to support the general contention that " the fact of transfer can not be doubted," that is, of positive or advantageous transfer. It has been found that in every case learning a response to one situa- tion, having a given element or stimulus as the dominant or controlling factor, is a help in learning the same response to the same situation but having a different element or stimulus, not present before, as the dominant or controlling factor; that even in some situations the learning process for one sort of dominant stimulus is actually reduced in length so much by first introducing another dominant stimulus, that the time and effort for both is less than for the one alone. Such was actually the case in learning the negative response to the sound. Looking at the results from the point of view of " gen- eralized response " or " generalized habit " we can say that responses are not always particular, but may become truly general; that is, the same response may serve for situations whose dominant stimuli are as different as can be found, provided the other features of the situation are the same and remain constant. This is really the reverse of the cases usually spoken of under the term " generalized habit." In such cases the dominant features are supposed to remain the same while the minor features vary from one case to another. So " generalized habit " can not always be said to depend upon similarity of dominant features in the situations responded to, or upon lack of discrimination of such dominant features. We can secure " generalized responses " where the dominant features of the situation are actually different for the organisms making the re- sponses, provided we can take difference of sense channel as the basis for discrimination of stimuli. In such cases there would be difference of neural pathways, at least in the sensory portion of such pathways. The " identical elements " in such pathways would be at least partly in the association and motor portions. Complete identity can not be said to be present. According to the results given, one, and the only one, of the conditions of the degree of advantageous or positive transfer clearly shown is the simultaneous presentation TRANSFER OF RESPONSE IN THE WHITE RAT 65 of the two controlling or dominant stimuli after the response to one has been learned. While in practically every case considerable disturbance was produced at the introduc- tion of the new stimulus along with the old, yet the presence of the old helped in every case in learning the new. In no case did it require more than eight series w4th the old and new together to perfect the response to the new. In one case, pain to light, it required only two series. Would it be too much of a hazard to hypothesize thus. — Variations in stimuli allow of positive or advantageous transfer effects, while variations in response, an aspect of the problem which has not been tested in these experi- ments, produce negative effects?* Of course in any case where training of any sort has gone on in any general situa- tion, the general familiarity with the situation will soon give a freedom from emotional disturbance and a general food seeking reaction. Both of these will tend to be carried over to later responses to the same general situation. These may offset any negative effects produced by a simple change in the response, unless, as was done with the experiments reported in this paper, the animals were first made per- fectly familiar with the apparatus and with the ways back to the food box before any experimenting proper began, so that these factors were made as nearly equal in influence as possible. These experiments are not to be regarded, nor are they regarded by the writer, as offering a final or complete solu- tion of the problem of transfer. They are to be taken as offered — a very simple and humble beginning of the solu- tion of that very complex problem. Our results have made clear that advantageous or positive transfer can and does take place in situations the most simple and have shown some of the conditions favoring such transfer. As the* situations studied grow more complex it must be remem- bered that the conditions at work in these simpler situations may still account for many of the results obtained. On the other hand, however, with the introduction of many vary- ing factors, both subjective and objective, the results for the simpler conditions may not hold at all. * Note: Bearing on this suggestion, see Hunter, Joitr. Animal Behav., Vol. 7, No. 1, pp. 49-65, Pearce, Jour. Animal Behav., Vol. 7, No. 3, pp. 169-177. LITERATURE CITED 1. CoovER. Formal Discipline from the Standpoint of Experimental Psychology. Psych. Rev. Man. Sup., Vol. 20, No. 3, pp. 224 ff. 2. Yerkes. Dancing Mouse, p. 262. 3. Thorndike. Animal Intelligence. 4. Richardson. A Study of Sensory Control in the Rat. Psych. Rev. Mon. Sup.. Vol. 12, No 1. 5. Yoakum. Some Experiments upon the Behavior of Squirrels. Jour. Compar. Neurology and Psych., Vol. 19, p. 514. 6. Hunter. Some Labyrinth Habits of the Domestic Pigeon. Jour. Animal Behav., Vol. I. 7. BOGARDUS and Henke. Jour. Animal Behav., Vol. I, pp. 125-137. 8. Angell. Psychology, p. 261. 9. Gore. Article in Dewey's ' Studies in Logical Theory.' 10. Hunter. Auditory Sensitivity in the White Rat. Jour. Animal Behav., Vol. V, pp. 312-330. 11. Watson. Behavior, p. 272. 12. Op. Cit. (No. 3). 13. Op. Cit. (No. 7). 14. Vincent. White Rat and Maze Problem. Jour. Animal Behai., Vol. V, pp. 1-24, 140-158, 175-185. The Behavior Monographs Edited by JOHN B. WATSON The Johns Hopkins University, Baltimore, Md. VOLUME 1 No. 1 The development of certain instincts and habits in chicks. By Frederick S. Breed. Pp. iv + 78, $1.00, postpaid. No. 2 Methods of studying vision in animals. By Robert M. Yerkes and John B. Watson. Pp. iv + 90, $1.25, postpaid. No. 3 An experimental study on the death-feigning of Belostoma (-Zaitha Aucct.) flumineum Say and Nepa apiculata Uhler. By Henry H. P. Severin and Harry C. Severin. Pp. iii + 47, $.65, postpaid. No. 4 The biology of Physa. By Jean Dawson. Pp. iii + 120, $1.50, postpaid. No. 5 The function of the vibrissae in the behavior of the white rat. By Stella Burnham Vincent. Pp. iv + 81, $1.15, postpaid. VOLUME 2 No. 6 The delayed reaction in animals and children. By Walter S. Hunter. Pp. v + 86, $1.15, postpaid. No. 7 The Canada porcupine : a study of the learning process. By Leroy Vv'^alter Sackett. Pp. iii + 84, $1.15, postpaid. No. 8 Audition and habit formation in the dog. By Harry Miles John- son. Pp. iv + 78, $1.00, postpaid. No. 9 Habit formation in a strain of albino rats of less than normal brain weight. By Gardner Chenej'^ Basset. Pp. iv + 46, $.65, postpaid. No. 10 Distribution of effort in learning in the white rat. By John Linck Ulrich. Pp. iii + 51, $.65, postpaid. No. 11 The effect of age on habit foitnation in the albino rat. By Helen B. Hubbert. Pp. v + 55, $.65, postpaid. VOLUME 3 No. 12 The mental life of monkeys and apes: a study of ideational be- havior. By Robert M. Yerkes. Pp. iv + 145, $1.50, postpaid. No. 13 A study of perseverance reactions in primates and rodents. By G. V. Hamilton. Pp. iv + 65. $.75, postpaid. No. 14 An analysis of the learning process in the snail, physa gyrina Say. By Elizabeth Lockwood Thompson Pp. iii + 97, $1.25, postpaid. No. 15 The effect of length of blind alleys on maze learning: An ex- periment on tAvcnty-four white rats. By Joseph Peterson. Pp. iii + 53, $.75, postpaid. No. 16 An experimental study of transfer of response in the white rat. By Harry H. Wylie. Pp. iii +QQ, $1.00, postpaid- Regular subscription $5 per volume, postpaid. THE BEHAVIOR MONOGRAPHS Emerson Hall, Cambridge, Mass. Directory of American Psychological Periodicals American Journal of Psychology — Worcester, Mass.: Florence Chandler. Subscription $5. 600 pages annually. Edited by G. Stanley Hall. Quarterly. General and experimental psychology. Founded 1887. Pedagogical Seminary — Worcester, Mass. : Florence Chandler. Subscription $5. 575 pages annually. Edited by G. Stanley Hall. Quarterly. Pedagogy and educational psychology. Founded 1891. Psychological Review — Princeton, N. J. : Psychological Review Company. Subscription (v\ith Psychological Bulletin) $6. 480 pages annually. Bi-monthly. General. Founded 1894. Edited by Howard C. Warren. Ps:5^chological Bulletin — Princeton, N. J.: Psychological Review Company. Subscription $3.25. 480 pages annually. Psychological literature. Monthly. Founded 1904. Edited by Shepherd I. Franz. Psychological Monographs — Princeton, N. J.: Psychological Review Company. Subscription $4 per vol. 500 pages. Founded 1895. Ed. by James R. Angeil. Published without fixed dates, each issue one or more researches. Psychological Index — Princeton, N. J.: Psychological Review Company. Subscription $1.25 200 pages. Founded 1895. Edited by Madison Bentley. An annual bibliography of psychological literature. Journal of Philosophy, Psychology and Scientific Methods — New York: Sub-Station 84. Bi-weekly. 728 pages per volume. Founded 1904. Subscription $3. Edited by F. J. E. Woodbridge and Wendell T. Bush. Archives of Psychology — Substation 84, New York: Archives of Psychology. Subscription $5. 600 pp. ann. Founded 1906. Edited by R. S. Woodworth. Published without fixed dates, each number a single experimental study. Journal of Abnormal Psychology — Boston: Richard G. Badger. Subscription $5. 480 pages annually. Edited by Morton Prince. Bi-monthly. Founded 1906. Entire field of abnormal psychology. Psychological Clinic — Philadelphia: Psychological Clinic Press. Subscription $1.50. 280 pages annually. Edited by Lightner Witmer. Monthly (9 numbers). Orthogenics, psychology, hygiene. Founded 1907. Training School Bulletin — Vineland, N. J.: The Training School. Subscription $1. 160 pages annually. Ed. by E. R. Johnstone. Founded 1904. Monthly (10 numbers). Psychology and training of defectives. Journal of Race Development — Worcester, Mass.: Louis N. Wilson. Subscription $3. 460 pages annually. Founded 1910. Quarterly. Edited by George H. Blakeslee and G. Stanley Hall. Journal of Educational Psychology — Baltimore: Warwick & York. Subscription $3. 600 pages annually. Founded 1910. Monthly (10 numbers). Managing Editor, J. Carleton Bell. (Educational Psychology Monographs. Edited by Guy M. Whipple. Published separately at varying prices. Same publishers.) Journal of Animal Behavior — Cambridge, Mass.: Emerson Hall. Subscription $5. 450 pages annually. Founded 1911. Bi-monthly. Robert M. Yerkes, Managing Editor. The Behavior Monographs — Cambridge, Mass.: Emerson Hall. Subscription $5. 450 pages per volume. Edited by John B. Watson. Published without fixed dates, each number a single research. Psychoanalytic Review— Washington, D. C: 3617 Tenth St., N. W. Subscription $5. 500 pages annually. Psychoanalysis. Quarteriy. Founded 1913. Ed. by W. A. White and S. E. Jelliffe. Journal of Experimeiital Psychology — Princeton, N. J.: Psychological Review Company. 480 pages annually. Experimental. Founded 1916. Subscription $3.25 Bi-monthly. Edited by John B. Watson. Journal of Applied Psychology — Worcester, Mass. : Florence Chandler. Subscription $4. 400 pages annually. Quarterly. Founded 1917. Edited by G. Stanley Hall, John Wallace Baird and L. R. Geissler. The Brandow Printing Company Albany N. Y. m,M^^ ^ ^v Si r,- •* ■*''|fcPC-j ■ ► '<' « , :& % . . . f^ ::m ^^^ i.>^A- /-