t ’.,: ■'; ■■. ■/ . • ■■ • i. » 
 
 * *-. 
 
 * JUL 251911 *) 
 
 Divivsion *• 'BFZl 
 Section 
 
 \ i .\2 
 
 ), 
 

ERRATA. 
 
 Cover Read William T. Shepherd instead of William' Shepherd. 
 Title page “ ‘‘ “ “ of 
 
 Page headings Read William T. Shepherd instead of William Shepherd. 
 
 Page III line 5 “ ii “ of i. 
 
 “ 12 line 25 “ not been “ of been not. 
 
 27 ‘‘ 28 discriminate “ of discrimination. 
 
 “ 33 6 from bottom read experimenters instead of experiments. 
 
 “ 46 ‘‘ 16 read animal instead of animals. 
 
 “ 50 “ 34 “ unanalyzed instead of analyzed. 
 
 “ 53 “ 19 “ page 41 “ of page 42. 
 
 “ 58 " 36 “ p.45ff- “ ofp. 4off. 
 
Digitized by the Internet Archive 
 in 2018 with funding from 
 Princeton Theological Seminary Library 
 
 https://archive.org/details/somementalprocesOOshep 
 
Vol. XII 
 No. 5 
 
 PSYCHOLOGICAL REVIEW PUBLICATIONS 
 
 November, 1910 
 Whole No, 52 
 
 THE 
 
 Psychological Monographs 
 
 EDITED BY 
 
 JAMES R. ANGELL, University of Chicago 
 HOWARD C. WARREN, Princeton University {Index) 
 
 JOHN B. WATSON, Johns Hopkins University (Review) and 
 ARTHUR H. PIERCE, Smith College (Bulletin) 
 
 Some Mental Processes of the 
 Rhesus Monkey 
 
 ^ BY 
 
 William Shepherd, Ph.D. 
 
 From the Psychological Laboratory of the George Washington University 
 
 THE REVIEW PUBLISHING COMPANY 
 
 41 NORTH QUEEN ST., LANCASTER, PA. 
 AND BALTIMORE, MD. 
 
WILLIAMS 
 
 PRESS OF 
 
 A WILKINS COMPANY 
 BALTIMORE 
 
CONTENTS. 
 
 Introduction . i 
 
 Experimental. 6 
 
 Formation of habits . 8 
 
 Releasing fastenings. 8 
 
 Visual discrimination. i 
 
 Brightness. ii 
 
 Color. 17 
 
 Auditory discrimination. 26 
 
 Noise. 26 
 
 Pitch. 27 
 
 Inhibition of habits . 29 
 
 Imitation . 33 
 
 Ideation . 45 
 
 Reasoning . 51 
 
 Adaptive intelligence. S3 
 
 Memory . 59 
 
 General Summary and Conclusions. 60 
 
 Bibliography. 61 
 
4 
 
Introduction^ 
 
 Though one of the newest of the sciences, comparative 
 psychology, or, speaking more strictly, animal psychology, 
 may properly claim as its father that acute observer and many- 
 sided Greek scientist and philosopher, Aristotle. The works 
 of this pioneer in science show him to have been an interested 
 and critical student of the mental, as well as the more strictly 
 zoological (morphological and physiological) side of animal 
 life, of which latter sciences (morphology and physiology) 
 he is the acknowledged father. His observations on the 
 comparative mental and moral traits of man and the lower 
 animals, so striking when we consider the data he had at 
 hand, may well entitle him to the credit of being the founder 
 of comparative psychology. 
 
 Within the past century, the evolution and anecdote 
 schools, represented preeminently by Darwin, Huxley, Rom¬ 
 anes, Lubbock and their co-laborers, have contributed largely 
 to this work and their observations and critical consid¬ 
 erations have thrown much light on the mental capabilities 
 of animals. These men, however, depended too much on a 
 method we now believe to be of little value in comparative 
 psychology, viz., observation uncontrolled by experiment, 
 and their critical considerations have been colored because 
 they were interested chiefly in the theoretical, evolutionary 
 interpretation of their observations. This work, however, 
 has been invaluable in that the broader questions have been 
 set, and especially in that it has given great stimulus to the 
 study of the animal mind. 
 
 It has remained for the new, the conservative, experi¬ 
 mental, school of comparative psychology, to define more 
 
 * The animals used in this research were supplied by Prof. Shepherd Ivory Franz, 
 of the George Washington University, through a grant to him from the Carnegie 
 Institution of Washington. Acknowledgment is, therefore, made to the Carnegie 
 Institution, without whose aid the work would not have been possible. 
 
2 
 
 WILLIAM SHEPHERD. 
 
 accurately the problems of the science, and by more accur¬ 
 ate observations and by carefully controlled experiments to 
 solve those problems. To this more scientific and growing 
 body of workers, to Morgan and Hobhouse in England, to 
 Forel, Bethe and Hatchet-Suplet on the Continent, and to 
 Mills, Thorndike, Yerkes and Watson in America, must be 
 paid the highest tribute. 
 
 As results of the first part of the work of this newer school 
 distinct problems have been more precisely formulated by the 
 investigators. These problems may be divided into two 
 general classes, each of which may be resolved into a number 
 of concrete special problems. The first general problem is: 
 Do animals possess the lower mental powers that man pos¬ 
 sesses? Do they have the same fundamental psychic states 
 that man has; and, if so, how do these states differ qualita¬ 
 tively and quantitatively from those of man? To be more 
 concrete, we may ask: Do the lower animals discriminate 
 sensory qualities, do they discriminate brightness of lights, 
 do they discriminate colors or hues, noises, smells, and tactile 
 stimuli? Do they form and inhibit habits, do they retain 
 impressions and have the elements of memory? In what 
 way, if at all, do these differ from similar mental powers of 
 man? The second general problem is: Do the lower animals 
 have the so-called higher powers? Have they ideas, have 
 they the ability to learn by imitation, have they general 
 notions and reason? From any psychological considera¬ 
 tion we may omit the question of the presence of moral 
 judgments, a subject which is chiefly of ethical and religious 
 interest. 
 
 Despite both the newness and the difficulties of the prob¬ 
 lems and the imperfections of many of the present experi¬ 
 mental methods of study in this field, much work has been 
 accomplished and many valuable results have been obtained. 
 The results are, however, yet too few, and some of those of 
 more general interest have led to unsettled controversies. 
 Some of the results which at present may be considered 
 established are as follows: Some animals, the higher forms 
 at least, discriminate brightness values. Some classes of 
 
INTRODUCTION. 
 
 3 
 
 animals discriminate pitch. All vertebrates and some in¬ 
 vertebrates show an ability to form habits of reaction to 
 stimuli, and they form at least simple associations. Animals 
 retain impressions and have a memory of some sort. Even 
 the most conservative and hypercritical have found this to 
 be the only satisfactory interpretation of the results of ex¬ 
 periments, and it is almost needless to say that the adherents 
 of the so-called anecdote school admit a high degree of mem¬ 
 ory in all mammalia. 
 
 The question on which comparative psychologists are 
 divided are as numerous as may be expected in a growing 
 science. Those of most interest in connection with the 
 present study may be briefly mentioned. Have animals 
 the same sensory equipment as man? Have animals color 
 vision? Or, do they merely appear to discriminate colors 
 because of differences in the brightness values of the stimuli 
 that are used? Do animals learn by inferential imitation? 
 Have the lower animals ideas or mental images? 
 
 In answer to these questions animal psychologists have 
 arrived at directly opposite conclusions, although all the 
 recent investigators believe they follow the law of parsimony 
 enunciated by Lloyd Morgan. In regard to the so-called 
 higher mental powers, recent experimenters have taken a 
 less decided stand than formerly. They cease to deny to 
 animals reason, imitation and other similarly complex proc¬ 
 esses, but they say the case is not proven and demand 
 additional experiments and observations. 
 
 Among experimenters on animals Hobhouse is almost 
 alone in claiming that animals have true general notions. 
 It is true he does this after having made rather important 
 qualifications of the term ‘idea.’ The latter topic has, how¬ 
 ever, not been the subject of many experiments or, rather, 
 it has not been the subject of many reported experiments 
 and observations other than a few illy-controlled ones. The 
 question of the ability of animals to reason has been the 
 subject of numerous tests; but the casual observations of 
 the anecdote psychologists are the mainstay of those who 
 support the view that animals show ability to reason. Hob- 
 
4 
 
 WILLIAM SHEPHERD. 
 
 house has made some experiments which he believes indicate 
 some power or inference in monkeys and apes, but the obser¬ 
 vations that have been made by others of the group of 
 experimental animal psychologists are in the main of a de¬ 
 cidedly negative character. With the exception of Hob- 
 house, it may be said that the consensus of opinion is that 
 the presence in animals of this so-called higher faculty is not 
 proven. 
 
 These points we may summarize by saying that at present 
 comparative psychologists admit the possession by animals 
 of most of the lower powers of intelligence that man possesses. 
 Such powers of animals may, however, differ from those of 
 man quantitatively and in a qualitative manner. Possession 
 by animals of the so-called higher powers is an open question, 
 or set of questions, which have not been so widely studied, nor 
 so clearly determined; and, respecting which the available 
 evidence points toward a negative answer in the form of a 
 verdict of ‘Not proven,’ 
 
 On account of the ease of obtaining the domesticated 
 animals a large amount of the experimental work on animal 
 behavior has been performed upon cats, dogs and chicks. 
 On the other hand, the lower forms (invertebrates and non¬ 
 mammalian vertebrates) have been extensively studied in 
 respect to the simplest powers, such as tropisms. On account 
 of expense, difficulties of care and the apparent complexity 
 and variety of behavior, the higher mammalian types, especi¬ 
 ally anthropoids, have not been studied so consistently. 
 
 However, for studies in comparative psychology the value 
 of using primates instead of the lower vertebrates and inverte¬ 
 brates is almost self-evident. On the one hand, the anatomi¬ 
 cal similarities between man and the monkey are apparent. 
 There is the well-known similarity between both their peripheral 
 and central nervous systems; and the similarity of arrange¬ 
 ment of muscles and bones, especially of the extremities, 
 must not be forgotten. Observations have made it evident 
 that there are also similarities in a physiological way. Move¬ 
 ments and reactions of an apparently human type are known 
 to be present in these higher forms, which are difficult if not 
 
INTRODUCTION. 
 
 5 
 
 impossible to observe in the lower forms. We may there¬ 
 fore expect from the careful studies of the mental states of 
 those animals, admitted to be immediately inferior to that 
 of man, more light on certain problems in the psychology 
 of human consciousness than from similar studies of others. 
 In making such a comparison of values it is not intended in 
 any way to detract from or to belittle the studies on the lower 
 forms. The value of experiments with the lower orders is 
 everywhere admitted. But at present it must also be ad¬ 
 mitted that the latter have received the attention of many 
 more writers, and their reactions have been subjected to a 
 much finer analysis. 
 
Experimental 
 
 The brief historical summary of comparative psychology 
 which we have given will, perhaps, appear as a sufficient 
 justification for undertaking further study in the field; both 
 in regard to the lower and to the higher faculties in the animal 
 mind. This study of the psychology of monkeys was under¬ 
 taken, therefore, with a view to add to the knowledge we 
 now possess of the lower faculties in the animal mind, and if 
 possible, to throw some light upon the question of whether 
 or not monkeys possess some of the higher mental powers, 
 and also to limit or characterize any such higher powers, if 
 they should be found. The experiments have been much too 
 incomplete to be entirely satisfactory to me, and, in many 
 cases, the results are naturally lacking in definite conclusions, 
 though they yielded, to me at least, some valuable indica¬ 
 tions in the field. It must be said, however, that the unsatis¬ 
 factoriness and the indefiniteness are not peculiar to this 
 particular study, but that they apply to any work that is 
 performed in such a complex science. In certain of the later 
 experiments, it will be noted, definite conclusions have been 
 arrived at, and, had time permitted the carrying out of 
 further experiments, it seems probable that other tests might 
 have been concluded in as definite a way. 
 
 As tests of some of the so-called lower faculties of the 
 monkey’s intelligence, experiments were made on brightness 
 discrimination, color discrimination, auditory discrimination, 
 the formation and inhibition of habits, and retentive power 
 (memory). In regard to the higher powers, observations 
 and experiments were made on learning by imitation, on 
 ideation, on reasoning, on adaptive intelligence and on 
 general notions. In brief, the chief aim of the study has 
 been, while profiting by the work of preceding investigators, 
 and with a steady adherence to the law of parsimony, to study 
 some of the lower powers of the animal mind, but especially 
 
EXPERIMENTAL. 
 
 7 
 
 by a modest study of the higher powers to seek for some 
 light, however little, on the subject of mental evolution. 
 Eleven monkeys were used in this work. All were Rhesus 
 (Macacus) obtained soon after their importation from India. 
 So far as we could ascertain they were eight to nine months 
 old when received and, with the possible exception of monkey 
 2, they appeared to be without training of any kind. Monkey 
 I was a rather large, spare male, moderately active and bright 
 looking; in his work he showed only medium capacity; 
 when he could not perform an act and thus get food he showed 
 signs of anger, by jumping about and by shaking the wire 
 sides of the cage; he would sometimes jump at the experi¬ 
 menter and visitors who happened to be in the room, but it 
 must be remembered he was in a cage and could not get at 
 anyone; the movements he made were those of intimidation, 
 so often noticed in the Rhesus and other Macacque monkeys; 
 towards the end of the series of experiments he became 
 cross. Monkey 2 was a medium-sized male; moderately 
 active and of only ordinarily intelligent appearance; he 
 showed a mediocre capacity in work; not friendly; he wore 
 a collar about his neck when received and this would indicate 
 that he may have been accustomed to handling by sailors 
 or by others who previously owned him. Monkey 3 was a 
 small female, apparently the youngest of the eleven, very 
 active, alert; good-natured and friendly for a Rhesus; formed 
 associations the quickest of all the animals. Monkey 4 
 was a medium-sized male, not very active, but bright and 
 cunning looking; good-natured; jumped at experimenter 
 half playfully; of medium capability in his work. Monkey 
 
 5 was a large male, dull looking, not active; slow in learning 
 as compared with all the others; he had a way of whining when 
 left alone; was rather ill-natured; would get angry when he 
 failed to do a required act, and thus fail to get food. Monkey 
 
 6 was a small female; bright and active; good-natured; very 
 cautious in all her work and actions. Monkey 7 was a large 
 female; inactive, not very bright looking, but the most 
 friendly of the eleven animals; proved to be quick in forming 
 habits; very quiet. Monkey 8 was a large male; he was the 
 
8 
 
 WILLIAM SHEPHERD. 
 
 master of the band; of gruff appearance, domineering, active; 
 of not especially intelligent appearance; but was extremely 
 cautious, appeared to like to go through the tests. Monkey 
 9 was a medium large female; rather friendly; quick to form 
 habits. Monkey lo was a large female; not active; quite 
 friendly; fairly capable as indicated in the tests. Monkey 
 II was a large male; wise looking; quiet and friendly; some¬ 
 what timid; of medium capability. 
 
 Formation of habits. 
 
 Releasing fastenings: Since the appearance of Thorndike’s 
 Animal Intelligence in 1898 the puzzle box method has been 
 a favorite one for comparative psychologists in studying the 
 formation of associations’ by the mammalia. Partly as pre¬ 
 liminary to succeeding work on visual discrimination, an 
 experiment similar to those of Thorndike was made with 
 monkeys i, 2 and 3. 
 
 A box 2 feet long, 7 inches wide and 8 inches deep was made 
 with wooden slats and attached to one end of the cage in 
 which the three animals were kept. A solid wooden door, 
 6 by 4 inches, was arranged on each side of the inner or parti¬ 
 tion part of this box nearest to the cage. These doors had 
 vertical hinges and were kept shut by a turn button on the 
 side away from the animals. The hinged portions were 16 
 inches, the opening or button portions 8 inches apart. A 
 space was left on the sides of the doors next to the button 
 sufficiently wide to enable a monkey to put his hand through. 
 The doors opened inward into the feed box, i.e., outward 
 from the monkey cage. The food was usually so placed on 
 the floor of the food box that the animal could not see it, 
 except by going to the extreme right or left, depending upon 
 
 1 The term ‘ association’ as applied to the intelligence of animals has been loosely 
 used. Two principal usages are: (o) the animal’s simple cerebral (or mental) processes 
 (merely anatomical, or it may be, physiological) such as is shown in a simple adjust¬ 
 ment to a stimulus, e.g., in its learning to turn a button to open a door; {h) as explain¬ 
 ing what are thought by some psychologists to be higher mental processes, e.g., idea¬ 
 tion or reasoning. When used in this paper the word is to be understood to have a 
 connotation similar to the first meaning given above. 
 
RELEASING FASTENINGS. 
 
 9 
 
 the box into which the food was placed. In all experiments, 
 even in those in which no time is recorded in the paper, the 
 times for the performance of acts were taken by a stop watch 
 and recorded. 
 
 By reaching through the space or crack at the side of the 
 door and turning the button, the monkey could open the door 
 and secure the food which was placed in the food box behind 
 one of the doors. In the beginning of the work only one of 
 the doors was used; when the association of turning the 
 button, i.e., of opening one door, was formed, both doors were 
 used. In the latter (two-door) preliminary experiments the 
 food was placed in a chance order behind one of the doors, 
 the animal not knowing which door that might be. 
 
 The records of the three animals used in this first experi¬ 
 ment are as follows: On the fourth day, after 32 trials, 
 monkey i gave evidence of the formation of the habit of 
 opening the door. Previous to this trial he had shown 
 many random movements, going from one part of the cage 
 to another, getting upon a bench in the cage, stopping all 
 work for a time, etc. On the fourth day, however, the habit 
 of going to the door directly was apparently formed, and the 
 method of opening was that of vigorously shaking the door 
 so that the button was gradually moved to permit the open¬ 
 ing of the door. On the fifth day, after 53 trials in all, he 
 began to fumble with the button in addition to shaking the 
 door. He continued to use both of these means with an 
 increasing tendency to rely on the button alone. On the 
 sixth day, after 95 trials in all, he had formed the association 
 of opening the doors by turning the buttons alone. His 
 average time for 10 succeeding trials on this day was 3 seconds. 
 Monkey 2 formed the association of opening the door in much 
 the same manner as Monkey i. He pulled and shook it, 
 and on the third day, after 21 trials in all, managed to get 
 the trick of always opening it in this manner. On the fourth 
 day, after 51 trials in all, he began to turn the button in 
 addition to pulling and shaking the door. Also similarly 
 to the actions of Number i, he continued to use one or both 
 of the means, with an increasing tendency to turn the button 
 
lO 
 
 WILLIAM SHEPHERD. 
 
 alone. On the sixth day, after 102 trials, he always used the 
 button alone as the means of opening the door. His average 
 time for ten successive trials was 1.5 seconds. Number 3 
 on the second day, and after ii trials in all, formed the asso¬ 
 ciation of opening the door by pulling and pressing it. On 
 the third day, after 34 trials in all, she began to turn the 
 button. On the fourth day, after 64 trials in all, she appeared 
 to have perfected the latter association and had inhibited 
 all tendency to use other means {e.g., shaking) for opening 
 the door. Her average time in the last 10 successive trials 
 on this day was approximtaely 1.5 seconds. 
 
 The method of learning in this experiment appeared to be 
 of the Trial and error’ type. As has been mentioned, the 
 animals scrambled about in a general way at first; they pulled 
 and bit at the door and the adjacent parts of the partition; 
 shook the door violently; but, it is to he noted that after the 
 first few experiments the attention of the animal was always 
 directed to the door. Even after an accidental success in 
 turning the button, the association was not at once set, but 
 only after a number of trials, with accidental successes many 
 times repeated, alternating, or rather interspersed, with many 
 errors. All the animals used in this work appeared, however, 
 to recognize a chance success and profited by it more quickly 
 than did raccoons under similar conditions. Another thing 
 of particular notice in their work, in which respect also they 
 differed from the raccoons, was their tendency to show signs 
 of anger when they could not open the door and get the 
 desired food. This was especially noticeable with Monkey i. 
 
 It is also of interest to note that when, preparatory to an¬ 
 other trial, I attempted to close the door after an animal 
 had opened the door and obtained food, the animal would 
 repeatedly attempt to hold the door open, as if the idea was 
 present that ‘door-being-open’ meant food. This reaction 
 is somewhat similar, I judge, to one observed by Thorndike 
 in his experiments with cats. It will be remembered this 
 author found a tendency on the part of the animals to walk 
 into the opened cage from which they had just previously 
 escaped to obtain food. The action, on the other hand, may 
 be merely of the nature of a reflex. 
 
VISUAL DISCRIMINATION. 
 
 II 
 
 Visual discrimination. It was formerly supposed, or rather 
 taken for granted, that animals possess the power of dis¬ 
 crimination of visual qualities, hues, and saturations. This 
 assumption was based upon observations of general be¬ 
 havior, of reactions to objects of different color, etc., but 
 in the past few years, however, comparative psychologists 
 have concluded that such power of discrimination could not 
 be assumed and they have sought by carefully controlled 
 experiments to test the truth of the matter. While the 
 experiments already made have yielded many valuable and 
 apparently positive results, several investigators, among 
 whom may be mentioned Yerkes (4) and Watson (2), 
 await more proof of such discrimination ability. The prob¬ 
 lem has been resolved into two questions. Is there a dis¬ 
 crimination by animals of brightnesses or intensities? Do 
 animals discriminate objects by their color or hue qualities? 
 In the work on visual discrimination to be recorded in this 
 paper nine monkeys were tested for color discrimination and 
 six for brightness (or intensity) discrimination. 
 
 Brightness Discrimination—Simultaneous Expos¬ 
 ures: Some experiments on monkeys i, 2, and 3 were made 
 with the same apparatus that was employed in the prelimin¬ 
 ary experiments on the releasing of fastenings, with the addi¬ 
 tion of two cards, respectively black and white, which were 
 placed above the doors. The cards were 5 inches square, 
 and were placed above the doors. They were, therefore, 8 
 inches apart. A piece of food, usually a half or a whole pea¬ 
 nut was dropped in the food box; the black card was placed 
 above the door behind which the food had been placed, and 
 the white card above the other door. The monkey was to 
 open the door under the black card and obtain the food. 
 The food was placed in the compartments in an irregular 
 order, and the cards were correspondingly placed. At times 
 the food was not placed In the box until the door had been 
 opened by the animal, but often it was dropped In the box 
 immediately before or after the buttons were adjusted. In 
 this way it was possible to prevent the animal knowing which 
 door to open from the sight of food, and since raw peanuts 
 
12 
 
 WILLIAM SHEPHERD. 
 
 were used, the smell component, judging at least from human 
 ability to smell such food was at a minimum. 
 
 These experiments were begun December 4th and con¬ 
 tinued for eleven weeks. Each animal was given from 10 
 to 25 trials every second or third day. In some cases there 
 were longer intervals between the work periods. Each 
 animal was given a total of 1000 trials. At the end of the work 
 all were substantially perfect in opening the door under the 
 black card first. No animal was ever able to inhibit alto¬ 
 gether the tendency to open the door under the white card 
 after the door under the black card had been opened and the 
 food had been secured. The average time for opening the 
 door under the black card in the last 25 trials for these three 
 monkeys was approximately one second. The curves of 
 learning, i.e., the time-experiment curves, correspond to 
 those found by almost all experimenters, but on account of 
 mistakes in opening the second door they cannot have much 
 value in this connection. 
 
 The fact that the animals formed the habit of opening the 
 door under black first would indicate that they discriminated 
 the cards. The quickness with which they finally performed 
 the act would also confirm this conclusion. Moreover their 
 looks and actions at the time appeared to indicate a knowledge 
 that the opened door meant food. When, for example, food 
 had been not placed in the box, after opening the door they 
 would look at me as if they expected their reward. The 
 closeness of the door buttons, and the great activity and 
 ready use of the hands, explain, I believe, the opening of the 
 second door so often. Furthermore, it was noted that while 
 after they had opened the door under the black card, they 
 appeared to expect food, when the door under white was 
 opened, their actions did not indicate that they expected to 
 be fed. 
 
 Brightness—Successive Exposures. For the tests with 
 monkeys 4, 5, and 6, I used a card displayer similar to that 
 used by Professor Cole and myself in brightness and color 
 tests made on raccoons. This was placed outside and about 
 6 inches from the animal cage. The accompanying figure 
 
VISUAL DISCRIMINATION. 
 
 13 
 
 is a diagram of the experimenter’s view of this piece of ap¬ 
 paratus. (See Figure i.) 
 
 The front of the displayer was formed by a board 12 inches 
 high. A pin, ‘P’, on which two levers could be turned, was 
 inserted in a hole at the back and near the lower edge. On 
 this pin two displayers, ‘W’ and 'B’, were arranged so as to 
 be freely movable in the plane of the board. The two cards, 
 black and white, were placed at the upper end of the dis¬ 
 players so that the raising of the displayers showed the black 
 and white cards respectively. The card displayers were 
 I-inch thick so that there was a difference of i inch in the 
 distances of the cards from the animal box. On alternate 
 
 days, however, the cards were changed from the front to 
 the back lever, so the animals should not react merely to the 
 distance position of a card. When the black was exhibited 
 the animal was to go upon a platform arranged inside the 
 cage and was fed, while he was not to go up at white and was 
 not fed. Usually and except at the stimulus, the animal sat 
 on a bench 12 inches from the floor and inside the cage. 
 He was, therefore, in a position to look down at the card 
 apparatus, for the upper portion of the card was about 5 
 inches below the horizontal level of his eyes. When the 
 black card was displayed and the monkey had climbed to the 
 food platform the experimenter rose from his seat and pre¬ 
 sented a piece of food to the animal. It is of interest to note, 
 therefore, that there was no possibility of the formation of 
 an association between the smell of food and food, rather 
 
14 
 
 WILLIAM SHEPHERD. 
 
 than between the appearance of the card and food. Since 
 the food was not placed near the cage or near the food plat¬ 
 form until the appropriate response was obtained the animal 
 had no smell stimulus for its guidance to an appropriate 
 reaction. 
 
 These experiments were begun February 5th and con¬ 
 tinued for a period of seven weeks. Each animal was given 
 from 10 to 25 trials daily, with intervals of two days, in some 
 cases of three or four days, during which intervals no work 
 was done with them on brightness. In all, each was given 
 700 trials. At the close of this work monkeys 4 and 6 ap¬ 
 peared to have the association perfected. The average times 
 for responding by going upon the platform when black was 
 displayed in the last 25 trials were 1.5 seconds for monkey 4 
 and 1.25 seconds for monkey 6. The tendency to climb to 
 the food platform when white was displayed appeared to be 
 entirely inhibited. Monkey 4 was, in this regard, perfect 
 in the last 25 trials, and monkey 6 responded to white only 
 once in the same number of trials. Moreover, at the time 
 of this mistake it was noted that in the trial (the 7th) in which 
 she responded to white, monkey 6 immediately came down 
 from the food platform with all the appearance of having 
 knowledge that she had made a mistake; she did not appear 
 to expect food and did not remain upon the platform for a 
 sufficient time to have food presented to her. 
 
 Monkey 5, at the end of the 700 tests, appeared to be 
 almost perfect. In responding to black he made only one 
 error in the last 25 trials. His inhibition of the tendency to 
 respond to white was not quite perfect, for during the same 
 series he responded to the white card three times. His 
 errors in this latter regard, however, were rather of the type 
 shown by monkey 6. He showed by his actions that he knew 
 he had made a mistake, for without waiting to be fed he 
 immediately returned from the food platform to the bench. 
 
 In the later experiments, and probably for some time 
 previous to the ones just mentioned, the movement of the 
 card displayer appeared to set up in the animals a tendency 
 to react, and the mistakes that were made on the last day 
 
VISUAL DISCRIMINATION. 
 
 IS 
 
 are, in my opinion, undoubtedly due to the lack of inhibition 
 of this reflex tendency to movement rather than to mistakes 
 in ability to discriminate. 
 
 For the next tests in this series small pieces (about one- 
 half inch cube) of white and rye bread were presented simul¬ 
 taneously on a board placed outside the cage but within reach 
 of the animal. The breads were prepared so that the taking 
 of one resulted in a punishment and in this way an 
 attempt was made to have the association formed quickly. 
 The rye bread was soaked in a solution of quinine bisulphate 
 (about I per cent, although no attempt was made to keep 
 the bitterness a constant factor). The white bread con¬ 
 tained no quinine and was presented in a comparatively 
 moist, fresh state, but much dryer than the rye bread that 
 had been soaked in the solution of quinine immediately before 
 the experiments. By taking the white bread and avoiding 
 the rye bread the animals were to show their ability to dis¬ 
 criminate these brightnesses.^ In these experiments and in 
 later ones to be reported it was assumed that monkeys dis¬ 
 liked bitter tastes; this assumption, it will be noted, proved 
 to be true. 
 
 To obviate the objection that the smell of the quinine or 
 the rye bread might enable the animal to differentiate the 
 white from the rye bread, the two pieces of bread were placed 
 in varying positions on the board. At times they were placed 
 at equal distances from the front of the cage and within an 
 inch of each other. In other tests one was placed directly 
 in front of the other, and in other tests the positions were 
 irregularly varied, one being nearer, the other farther from 
 the cage. It might still be objected that the rye bread which 
 was wet from having been soaked in the solution of quinine 
 would give them a clue. That this objection was not met in 
 the conduct of the experiments must be admitted, but the 
 formation of the association, i.e., the positive reaction would 
 give evidence of sharp visual discrimination. 
 
 1 In addition to its darker appearance, there was a hue difference in the rye bread. 
 The latter, however, is the less prominent factor. 
 
i6 
 
 WILLIAM SHEPHERD. 
 
 Nine animals were tested by this method, i, 2, 3, 4, 5, 6 
 7, 8 and 9. This plan of tests was very gratifying in its 
 results. It showed, apparently, complete discrimination 
 of the two pieces of bread. The rapidity with which the 
 habit of taking only the white bread was formed was striking. 
 This is especially noticeable when we compare the records 
 of the same monkeys in this and in the preceding brightness 
 tests. As table I shows, only from i to 14 trials were required 
 for each of the eight monkeys to establish the association. 
 This is excluding the work of monkey 9, which also dis¬ 
 criminated the white and rye bread, but which, being in the 
 same cage with monkey 8, could have seen the latter select 
 the white and reject the rye, and might be said to have learned 
 from or to have imitated monkey 8. 
 
 TABLE I. 
 
 Discrimination of white and rye breads. Twenty-five experiments on each animal. 
 
 ANIMALS. 
 
 Mistakes in Trials 
 
 Total Mistakes. 
 
 First Day. 
 
 Second Day. 
 
 I 
 
 I, 2a, 4b 
 
 I2C 
 
 4 
 
 2 
 
 1,2a, 40,5,6,8 
 
 14^ 
 
 7 
 
 3 
 
 I, 20, 3, 4b, 6b, 8b, 120 
 
 13c 
 
 8 
 
 4 
 
 I, 3a 
 
 
 2 
 
 5 
 
 I, 20 
 
 
 2 
 
 6 
 
 I, 20 
 
 
 2 
 
 7 
 
 I, 2b, 30, 40 
 
 
 4 
 
 8 
 
 I 
 
 
 I 
 
 9 
 
 
 
 
 a Took rye bread after having taken white, smelled or tasted, or both, and dropped 
 without eating. 
 
 b Took rye bread after having taken white, and ate both. 
 
 c First trial at second day; took rye bread but dropped it immediately without 
 smelling or tasting. 
 
 d Second trial of second day; without smelling or tasting. 
 
 It is of some interest to note the individual reactions to 
 the two stimuli that were presented. At the first trial each 
 animal took both pieces of bread, placed them in the mouth 
 and began to eat. In many cases the bitterness had a retard¬ 
 ing or inhibiting effect, for the animal would take the rye 
 
F/5 UA L DISCRIMINA TION. 
 
 17 
 
 bread from the mouth, look at it, smell it, and then either 
 reject or reinsert it in the mouth. 
 
 Monkey i, in addition to eating the rye bread on the first 
 trial, took the piece of rye bread in the second trial, smelled, 
 tasted and dropped it. The bitter bread was not taken on 
 the third trial, but on the fourth the animal took first the white 
 which it ate, then the rye, which it also ate. The rye bread 
 was not taken on the next seven trials on that day. On the 
 twelfth test (the first test on the second day) the animal first 
 took the white bread and ate it, then took the rye bread and, 
 without smelling or tasting, dropped it immediately. On 
 this and on the following days all other tests were perfect in 
 that only the white bread was taken and the rye bread not 
 even handled, although, as has previously been noted, at 
 times the rye bread was placed closer to the cage than the 
 white bread. 
 
 Similar results were obtained with animals 2 and 3 as is 
 indicated in the table. The rapidity of learning is remark¬ 
 able in all animals, but particularly so in 4, 5, 6, 8 and 9. 
 In each of these animals two tests were sufficient to inhibit 
 altogether the tendency to take the darker bread. This find¬ 
 ing is so at variance with the results of comparable tests on 
 other animals that there would be ample opportunity to use 
 the results as an indication in monkeys of some form of 
 reasoningor of a marked activity to form practical judgments. 
 
 Color discrimination—white and red. The same 
 general plan was followed as in the white and rye bread tests. 
 Rice, cooked to such a consistency as to be stiff, was, however, 
 used instead of bread. It was cut into small pieces of approxi¬ 
 mately the same size. Some of these pieces were used for the 
 white stimulus, and others when colored with Congo red for 
 the red. The red pieces were soaked in a quinine solution 
 and used while wet. The white contained no quinine and 
 was in a comparatively dry state. The precautions noted in 
 the previous test were taken to prevent an animal obtaining 
 a clue from either the relative positions or the smell of the 
 pieces of food. 
 
 The nine monkeys which had been used in the white and 
 
i8 
 
 WILLIAM SHEPHERD. 
 
 rye bread experiments were tested in this experiment. The 
 rapid learning to take the white rice and to reject the red rice, 
 as in the white and rye bread test, was notable. From i to 
 3 trials were sufficient for the different animals to form the 
 association. Animal i took the red rice only the first trial, 
 and a similar result was obtained with monkeys 2, 4, 5, 6, 
 7, 8 and 9. Monkey 3 took the red rice only three times, on 
 the first, second and fifth trials. Not only is the rapidity 
 of the formation as striking here as in the case of the white 
 and rye bread, but the fact that all but one of the animals 
 formed the association, or, rather, inhibited the tendency to 
 take the red rice, after having experienced its effect only once, 
 may be taken as an indication of some form of reasoning. 
 
 Color discrimination—pink and green. A plan was 
 followed in the next test similar to that employed in the 
 white and red experiment. A smaller amount of coloring 
 matter was used and some of the rice was colored a light pink. 
 The pink rice also contained quinine. Another portion of 
 rice was colored green, and this portion contained no quinine. 
 The pink coloring was obtained, as has been said, by using 
 a weak solution of Congo red, the green was made by using 
 a solution of Malachite green. Both kinds of rice were used 
 while wet, and in this experiment, it is not probable that 
 discrimination could take place on account of difference in 
 the light reflecting qualities of the two kinds of rice. Care 
 was taken to have the two colors as nearly as possible of the 
 same approximate brightness. To this end the two colors 
 were mixed and compared to the colors pink and green of the 
 same approximate brightnesss on a color scale and the mixed 
 colors as well as those of the color scale were also tested by 
 the minimal perception method in a dark room to insure 
 their being of approximately the same relative brightness 
 {i.e., to the human eye). 
 
 In mixing the rice the coloring materials were not meas¬ 
 ured, nor was the relative amount in proportion to the water 
 of the solutions kept constant, nor the relative proportions 
 of the two colors to each other, nor to the amount of rice. 
 The color solutions were usually made up anew each day 
 
VISUAL DISCRIMINATION. 
 
 19 
 
 and there was inevitably a greater or less variation in satura¬ 
 tion and in brightness of the colors in different tests. These 
 remarks hold true in all the color tests. While criticism 
 may be made on any apparent discrimination on the basis 
 of hue under such conditions, to the writer discrimination 
 of color would be indicated by proper reaction from the very 
 fact of the variation of intensities. This will be pointed 
 out again in the discussion of the color experiments. As in 
 the preceding tests the same control precautions were taken 
 to prevent the animals from taking a cue either from the 
 position or from the smell of the rice. 
 
 In this test the monkeys formed the habit of taking the 
 green and rejecting the pink with even greater rapidity than 
 in the white-red test. Table II gives the records of nine 
 animals which were tested. 
 
 Color discrimination—pink and yellow. In this test 
 the same plan was followed as in the preceding experiments. 
 The pieces of pink rice contained quinine, while the yellow 
 contained no quinine. As in the preceding tests, both the 
 pink pieces and the yellow pieces were used while wet. The 
 pink coloring was produced by the same means as in the last 
 mentioned experiment, while the coloring of the yellow pieces 
 of rice was produced by the addition of lead chromate. The 
 test was controlled by the same means as in the pink-green 
 experiment to secure the same approximate degree of bright¬ 
 ness in the colors pink and yellow, and to prevent the animal 
 from receiving any cue from either the smell of the two 
 pieces of rice, or from their relative positions or relative dis¬ 
 tances. The same animals were used as in the preceding 
 tests, monkeys i, 2, 3, 4, 5, 6, 7, 8 and 9. 
 
 1 The colors were mixed and compared to the colors pink, yellow, green of the same 
 approximate brightness on a color scale, and the mixed colors on the color scale tested 
 in a dark room for quality of brightness. In the different experiments, the pieces of 
 rice were also placed at varying distances from the animal to prevent his taking a cue 
 either from the smell or the positions of the two pieces. 
 
20 
 
 WILLIAM SHEPHERD. 
 
 The following is the record for the animals for 25 trials each: 
 
 No. 
 
 1 took yellow in each trial; took pink trial i 
 
 2 9 
 
 3 
 
 4 
 
 5 
 
 6 
 
 7 
 
 8 
 
 9 
 
 none 
 
 none 
 
 I 
 
 none 
 
 none 
 
 none 
 
 none 
 
 Color discrimination—pink, yellow and green. The 
 same method was employed in this experiment as in the 
 preceding ones. In this case the three colored pieces of 
 rice were simultaneously exposed. The pink and yellow 
 contained quinine, the green contained no quinine. The 
 
 TABLE II. 
 
 Discrimination of pink and green rice. Twenty experiments on each animal. 
 
 MONKEYS. 
 
 MISTAKES IN TRIALS. 
 
 TOTAL MISTAKES. 
 
 I 
 
 3 
 
 I 
 
 2 
 
 3 
 
 I 
 
 3 
 
 I 
 
 I 
 
 4 
 
 0 
 
 0 
 
 5 
 
 0 
 
 0 
 
 6 
 
 0 
 
 0 
 
 7 
 
 L 2, 3, 4, 5, 6, 7, 
 
 14 
 
 
 8, 9, 10, II, 12, 
 
 
 
 13, 14, 15, 16, 
 
 
 8 
 
 0 
 
 0 
 
 ■coloring for the pink, yellow and green was the same as used 
 in the preceding tests, each was used while wet. Exactly 
 the same means to control the tests were employed as above. 
 The same monkeys were tested as in the white and rye bread 
 test, the white-red, the pink-green, and the pink-yellow experi¬ 
 ments. Here again the rapidity of the rise of the habit of , 
 response is striking. Table III gives the record of each 
 monkey for 25 trials. 
 
VISUAL DISCRIMINATION. 
 
 21 
 
 TABLE III. 
 
 Discrimination of pink, yellow and green rice. Twenty-jive experiments each animal. 
 Ten experiments first day, fifteen on second day. y, took yellow; p, took pink. 
 
 MONKEYS. 
 
 MISTAKES IN TRIALS. 
 
 TOTAL MISTAKES. 
 
 I 
 
 I y, 
 
 I 
 
 2 
 
 I y, 
 
 I 
 
 3 
 
 I y,iiy 
 
 2 
 
 4 
 
 I y 
 
 I 
 
 5 
 
 I y, 2 y, 3 y, 4 y, ii yp, 12 y 
 
 6 
 
 6 
 
 I y, 2 y, 3 y, 4 y, 5 y, 6 y. 
 
 10 
 
 
 II y, 12 y, 13 y, 14 y. 
 
 
 7 
 
 I y, 2 y, 3 y, 4 y, s y, 6 y, 7 y, 
 
 16 
 
 
 8 y, 9 y, 10 y, n y, 12 y, 
 
 
 
 13 y, 16 y, 17 y, i8 y, 
 
 
 8 
 
 I y, 2 y, 3 y, 4 y, 5 y, 6 y, 
 
 10 
 
 
 7 y II y, 12 y, 13 y, 
 
 
 We see from the records and tables that the nine animals 
 quickly learned to select the rice that contained no quinine 
 and to leave the quinine rice alone. The learning to respond 
 quickly was much more rapid than, so far as I am aware, 
 that of any other monkeys which have been studied for 
 visual discrimination. In many cases, one trial was sufficient 
 to inhibit any tendency to take the quinine rice, and the 
 rapidity in inhibition of the wrong response was undoubtedly 
 due to the use of a punishment or to the association of a 
 disagreeable sensation, i.e., to the quinine in the rice. 
 
 Another factor in the rapid formation of the habit of 
 avoiding the quinine rice, in the case of pink in the pink- 
 green rest, was probably the knowledge or memory of dis¬ 
 agreeableness in red rice in the white-red test immediately 
 preceding. In the pink-yellow test and in the pink-yellow- 
 green test, knowledge of quinine in pink in the preceding 
 tests, was also probably a factor in their rapid discrimina¬ 
 tion. In the pink-yellow-green test the experience of the 
 preceding tests of pink-bitter, yellow-good, green-good helps 
 to explain the number of mistakes in yellow rice. But, with 
 all these allowances, we still have evidence of rapid formation 
 of the habits of selecting one color and of rejecting another, 
 and thereby apparently discriminating the colors red, pink, 
 
22 
 
 WILLIAM SHEPHERD. 
 
 yellow and green. That discrimination has taken place 
 cannot be doubted, but the question arises: did the animals 
 discriminate the colors as hues, or as merely brightness 
 values? 
 
 Watson (2), in very carefully controlled tests made on 
 three monkeys in 1908, failed to find evidence satisfactory 
 to him that his animals discriminated colors as hues. From 
 later tests (3) also, on monkeys, he is still not prepared to 
 affirm whether such discrimination is of color or of mere 
 brightness. Yerkes (4), from the results of his well-controlled 
 experiments on the dancing-mouse, takes a view of the matter 
 similar to that of Watson. Both condemn the use of cards, 
 filters, etc., in experimenting on color discrimination. They 
 believe such methods are too loose to have much value 
 and they urge the use of more exact methods of determining 
 the matter. 
 
 In the experiments on color discrimination reported in 
 this paper, as has already been stated, the criticism that the 
 method used is a loose one may be urged. The exact propor¬ 
 tion of the different coloring materials to the bread or rice 
 and to the amount of water in making the solutions was not 
 determined. Furthermore, it may be said that as the color¬ 
 ing solutions were mixed anew almost daily, it would hardly 
 be possible not to have some variation in the intensities of 
 the colors at different times. This is true, notwithstanding 
 that means were taken to control the experiments, viz., com¬ 
 parison with certain color standards. Shall we infer that the 
 experiments were thereby so vitiated as to have little or no 
 value as a test of color discrimination? Such will doubtless 
 be the view of some, at least on first thought, but, to the 
 writer, these inaccuracies in method point to a different 
 conclusion. 
 
 It is admitted that the method and apparatus used in the 
 experiments herein reported appear loose and crude when 
 compared to the ingenious methods and complicated appara¬ 
 tus used in the experiments of Watson and others. It may 
 also be urged that my experiments take no adequate account 
 of the question of ‘monochromatic bands’ in the problem 
 
VISUAL DISCRIMINATION. 
 
 23 
 
 of color vision. To this we may reply that in the very loose¬ 
 ness and naturalness of the tests is to be found perhaps the 
 strongest evidence of the discrimination of color by the 
 animals which I tested. On the other hand I would point to 
 the highly artificial character of the methods used by some 
 experimenters. The complication of apparatus with its 
 reflectors, electric shocks and other appendages is artificial 
 in the extreme and and must result in an artificial attitude on 
 the part of the animal. The tests, with complicated appara¬ 
 tus, conducted in a dark room, bring about another artificial 
 and unusual situation, viz., the necessity for dark adaptation 
 by the animal. The method employed in these experiments 
 leaves to the animal a large amount of freedom and places 
 the animal in a position as natural as is possible in such work. 
 If, as has been urged by some, the experiments with colored 
 cards and filters may only mean that the animals react to 
 brightness or intensity, and not to hue relations, we should 
 expect an animal to react to a definite relation of brightness 
 unless we admit in animals some complex form of the feel¬ 
 ing of relation or a certain amount of inferential reasoning. 
 Such an explanation {i.e., brightness discrimination) may be 
 justified when the red-white test be considered alone. But 
 how may we explain the results in the pink-green test? Half 
 of the animals used in the latter test made no mistakes. 
 From the beginning of this experiment pink rice was avoided, 
 although it is not possible that the animals had any experi¬ 
 ence with red or pink rice previous to the time of these 
 experiments. Four of the animals, therefore, reacted prop¬ 
 erly to a hue of an intensity or brightness, considered 
 from the human standpoint, very different from that to which 
 the animals had learned to react. Do not these positive 
 reactions indicate rather clearly that somehow or in some way 
 the pink rice has been taken to be equivalent to or mean some 
 thing similar to red rice? Do not these results rather lead 
 to the conclusion that it is not a difference in brightness or 
 in intensity which has led to the appropriate adjustment, 
 but rather a difference in hue? Furthermore, the fact that 
 in the pink-green test the animals took the green and avoided 
 
24 
 
 WILLIAM SHEPHERD. 
 
 the pink regardless of the difference in the amount of color 
 is evidence that color as such was an important, and probably 
 the only, factor in the discrimination. This view receives 
 added weight from the similar results under similar condi¬ 
 tions (of approximate brightness and color) in the pink- 
 green-yellow tests. 
 
 It may, however, be urged that the carrying over of a 
 habit from a red to a pink indicates merely a dullness in dis¬ 
 crimination, that the red and pink may have been sensed or 
 perceived as approximate equivalents. It must be admitted 
 that this may be so, but the wide differences in the intensity 
 of the red and pink would indicate rather clearly that the 
 discrimination (or, comparison, if you will) has been due not 
 to simple intensity relations but to hue or color similarities. 
 For the discussion of the results in the pink-green tests, we 
 may make two assumptions, that the two kinds of rice were 
 of equal or unequal brightness. If we assume an equal 
 brightness, the experiments must, it seems to me, be con¬ 
 sidered to show that discrimination has taken place because 
 of difference in hue. If we assume the two kinds of colored 
 rice to have been of unequal brightnesses, we must, remember¬ 
 ing the variations in the experiments on different days, con¬ 
 sider that the discrimination has taken place in spite of this 
 variation. The results then lead to the conclusion that the 
 discrimination has been due to a factor different to that of 
 intensity on the physical side, or that of brightness on the 
 mental. The only other factor which, in man, would pro¬ 
 duce such a reaction, is that of color. 
 
 A comparison of the results obtained in the different color 
 tests strengthens this view. The consideration of the results 
 in the white-red, pink-green and pink-yellow tests is instruc¬ 
 tive. Monkeys 4, 5 and 8 which had one experience in tast¬ 
 ing bitter red rice did not take the pink rice in the succeed¬ 
 ing tests. Now, it must be remembered that these monkeys 
 had no previous experience with pink rice, and their avoid¬ 
 ance of it can be accounted for only on the assumption that 
 something in the pink rice gave a clue to the animal. 
 
 The differences (to us humans, of course) between the red 
 
VISUAL DISCRIMINATION 
 
 25 
 
 and the white rice are three-fold: color, intensity, and, akin 
 to the latter, reflecting quality or sheen. The last named 
 was due to the differences in moisture, the red rice having 
 been soaked in the quinine solution only a few minutes before 
 the tests were begun. The sheen of the pink, green and yel¬ 
 low rice was the same, for all were equally moist at the time 
 of the experiments. There were, therefore, only two possible 
 differences between these colors, viz., hue and intensity. 
 
 If the avoidance of pink is to be explained, we must admit 
 that it was due either to hue or to intensity. If it were due 
 to a feeling of intensity difference, to an inference (taking the 
 human standpoint, of course) that darkness means bitter 
 and lightness means sweet we must account for a transfer 
 from the red (very dark) to the pink (slightly dark). We 
 must also account for the transfer in connection with the 
 other stimuli (green) which from tests was found to have an 
 approximate intensity equal to that of pink. So far as I 
 can see at present, the only possible intensity explanation 
 of this transfer is that red has a low brightness effect in com¬ 
 parison with other colors. However, we know that the inten¬ 
 sities of the red and the pink rice varied greatly from each 
 other, and we are forced to account for the phenomenon of 
 transfer on the basis of some quality not so markedly changed. 
 This, I believe, is the hue. 
 
 To sum up we may say that the evidence is in favor of the 
 conclusion that monkeys have the ability to discriminate 
 colors as such. So far as is known of the structure of the 
 retina and of the remainder of the visual apparatus there is 
 nothing to indicate any difference to the human mechanism, 
 and, from this point of view, no reason exists why the monkey 
 family may not be able to differentiate the four colors red, 
 green, yellow and blue. The rapidity of formation of the 
 association between the light bread and agreeableness and 
 that between the dark bread (rye) and disagreeableness, as 
 evidenced by their selection of the former and their avoidance 
 of the latter, is indicative of sharp visual (brightness or 
 intensity) discrimination. The selection of one color and 
 the rejection of another in spite of differences in intensities 
 
26 
 
 WILLIAM SHEPHERD. 
 
 speaks strongly for the discrimination of hue. The transfer 
 of the habit from red to pink is indicative that the animals 
 had some form of feeling of relation or of similarity between 
 the colors as such, or that the hue (in itself and independent 
 of intensity) acted as the appropriate stimulus to inhibit the 
 tendency to take the particular food. 
 
 Auditory Discrimination — Noise. The apparatus used was 
 a wooden box 22 x i8 x lo inches and a small board or slat 
 18 X 3^ X I inches arranged to strike the box and thus make a 
 noise. One end of the board or slat was fastened to the top 
 of the box by a leather hinge. By raising the free end of the 
 slat and suddenly letting it go, it struck the top of the box 
 and made a sound varying in loudness with the force with 
 which it struck. To give two sounds of different degrees 
 of intensity or loudness two small sticks, one 3 inches in 
 length, the other 5 inches in length, were separately used to 
 be placed perpendicular to the box and under the free end 
 of the board. By pressing slightly on the slat near the hinge, 
 and suddenly removing the shorter stick, the board would 
 strike the box and produce a noise of a noticeable intensity, 
 and by pressing on the board as before, and withdrawing the 
 larger stick that had been placed at the free end, the board 
 would strike the box and produce a much louder noise. The 
 same pressure, as nearly as possible, was exerted by the 
 experimenter on the board in both cases. By going upon 
 a platform arranged inside the cage when the louder noise 
 was made, the animals were to show their discrimination of 
 the louder and lesser noises. They were fed on the platform 
 when the louder noise was made, and were not fed when the 
 lesser noise was made. The noise apparatus was manipu¬ 
 lated near the closed side of the cage in which the monkeys 
 were kept. It was out of sight and it was not possible that 
 the reactions were made to stimuli other than the sounds. 
 The louder and lesser noises were made in an irregular order. 
 Three animals, 4, 5 and 6, were tested by this method. 
 
 Monkey 4 formed the habit of responding to the louder 
 sound and not to the lesser in eight days of 10 trials each, 
 i.e., 10 trials with each sound) or in 80 trials in all. In the 
 
A UDITOR Y DISCRIMINA TION. 
 
 27 
 
 first day’s trials he responded to the louder sound once in the 
 10 trials given him, and to the lesser sound twice in the same 
 number of trials. In 10 trials on the eighth day he always 
 responded to the louder noise correctly and to the lesser 
 noise only once in the same number of trials. On the ninth 
 day he made no errors in the 10 trials given him. Monkey 
 6 formed the association in eleven days of 10 trials each. 
 On the first day she responded to the louder sound three times 
 and to the lesser four times. On the eleventh day she re¬ 
 sponded to the louder sound nine times and to the lesser 
 sound once. She made no errors in the 10 trials for each 
 sound on the twelfth day. In the time devoted to this work 
 with number 5, he did not form a perfect association. In 
 thirteen days’ experiments of 10 trials each he finally came 
 to respond to the loud sound in about 75 per cent of the trials. 
 His inhibition of response to the lesser sound was less perfect, 
 or about 50 per cent. Had the tests been continued it seems 
 likely he would have become perfect in discriminating the 
 two sounds. It is of some interest to note the apparent 
 inferior ability of 5 in comparison with 4 and 6 in this experi¬ 
 ment. The work of 5 appeared typical of all his work re¬ 
 ported in this paper. Inhibition was apparently his weak 
 point, for he responded to no matter what form of stimulus. 
 The difference illustrates the individual variations in mental 
 capacity of the different animals. 
 
 Sound discrimination, Pitch. Kalischer (5, p. 204 ff.) has 
 reported experiments on sound discrimination with dogs, 
 which animals show an ability to discrimination pitch. Al¬ 
 though interested in the matter more from a physiological 
 than a psychological standpoint, he incidentally obtained 
 satisfactory evidence of discrimination of pitch by his animals. 
 His method was to sound a certain note on an organ or har¬ 
 monium as a sign that the animal should react in a certain 
 way, such as snapping at a piece of meat. When a different 
 note was sounded the dog was not to react and was not fed. 
 Selionyi using a form of the Tawlow method’ on dogs has 
 also lately obtained evidence of discrimination by them of 
 the tones of an organ, organ pipes and of two whistles. He 
 
28 
 
 WILLIAM SHEPHERD. 
 
 also was chiefly interested in sound discrimination from a 
 physiological standpoint. 
 
 In experiments on auditory discrimination in raccoons 
 made by the present writer in 1906, reported by Cole (16, 
 p. 230) evidence of pitch discrimination by those animals 
 was obtained. In the experiments on pitch to be reported 
 in this paper I used the same plan that I had employed with 
 the raccoons, and which is similar to the method Kalischer 
 has used. An ordinary German mouth harp or harmonica 
 A was used. When I sounded the higher note, A 3, the 
 monkey was to go upon the platform used in the preceding 
 noise tests, and was fed there when the note was sounded. 
 When the lowest note, A i (two octaves lower), was sounded 
 he was not to go up and was not fed. The notes were sounded 
 in an irregular order so the animal might not react in a rhythm 
 to the sounds. Care was taken to sound the notes with the 
 same degree of intensity, as nearly as possible. I took the 
 usual precautions that the animal should not obtain a cue 
 from my looks, motions or in any other manner, and react 
 to these stimuli rather than the tones. 
 
 The records for the different animals are as follows: 
 
 Monkey 4 formed the habit of responding to the high note, 
 A 3, and not responding to the low note, A i, in three days 
 tests of 10 trials each.^ On the first day monkey 4 responded 
 to the high note six times in 10 trials, and not at all to the 
 low note. On the third day he responded to the high note 
 ten times and to the low note three times. On the fourth 
 day of the experiment (three days after the experiments just 
 noted) he responded to the high note ten times in 10 trials 
 and to the low note not at all in the same number of trials. 
 WTen tested on the fifth day, two days later, he was perfect 
 in 10 trials. 
 
 Monkey 6 formed the association in four days, in 40 trials 
 in all. On the first day she responded to the high note twice 
 in ten trials and to the low note not at all. On the fourth 
 day she responded to the high note nine times and to the 
 
 1 i.e., ten trials of the high note and ten trials of the low note. 
 
INHIBITION OF HABIT. 
 
 29 
 
 low note once in 10 trials of each note. In the 10 trials on 
 the fifth day she was perfect. 
 
 Monkey 5 did not form the habit of correct response in 
 six days’ tests of 10 trials each. In the 10 trials of high and 
 low pitch on the sixth day he responded to the high note six 
 times and to the low note only once. 
 
 If we may generalize from the work of the animals tested 
 in the above experiments we may conclude that Rhesus 
 monkeys discriminate quantitative differences in noises. They 
 also discriminate musical notes of widely different pitch. 
 These experiments also indicate that monkeys learn to dis¬ 
 criminate pitch with considerably more facility than do 
 raccoons in similar tests. 
 
 Inhibition of habit. 
 
 The activities or modes of behavior of animals, including 
 man, are of five kinds: {a) reflex actions; (6) instinctive actions; 
 (c) habitual actions; (d) intelligent actions;'(e) rational 
 actions. These different actions, in varying degrees, charac¬ 
 terize the different orders of animals. Generally speaking, 
 the lower the order of the animal the more the a form of 
 action is present, and the higher the order of the animal the 
 more the d and e forms of action are present. The lower 
 forms are, however, basal for man as well as for the amoeba. 
 
 Habitual actions are seen not only in the behavior of 
 animals of the higher orders but also in the activities of man. 
 Many apparently intelligent actions of man are of the nature 
 of habit. To say that habit dominates man’s actions to a 
 very large degree, and reason to a very small degree is only 
 to state the simple truth. 
 
 Previous observations have indicated to comparative psy¬ 
 chologists that monkeys and other mammals have a marked 
 tendency to form habits of action. This was well shown 
 by Lloyd Morgan and many others. A certain perceived 
 stimulus is followed by an agreeable or by a disagreeable 
 result, and an animal soon forms the habit of reacting in the 
 appropriate manner when the stimuli are given. The for- 
 
30 
 
 WILLIAM SHEPHERD. 
 
 mation of habits is, however, only one way of meeting the 
 conditions of the environment. Habits once acquired often 
 become useless, and sometimes have to be replaced by actions 
 opposite in character. In other words the inhibition or the 
 replacement of a habit is often necessary to the well being 
 of an animal or of man, and the ability to inhibit definite 
 modes of reaction, including reflex tendencies, is important 
 for advancement. 
 
 To test in some measure the ability of monkeys to inhibit 
 a recently formed habit four experiments were made. Three 
 monkeys which had learned to discriminate white (normal) 
 from rye (quinine) bread were reviewed on this work daily 
 for seven days in order that the habit be firmly established. 
 A rest period of seven days was given, and then the memory 
 tested. The results of the latter test showed a perfect reten¬ 
 tion by all the animals. On the succeeding day each monkey 
 was presented with some pieces of rye bread that had not 
 been soaked in quinine. Then an experiment was performed 
 in which the rye bread was not made bitter. These tests 
 were continued for seven days, lo tests on each day. The 
 results are given in Table IV. It will be noticed that on the 
 first day monkeys 6 and 8 disregarded the rye bread, and 
 monkey 8 continued to disregard it for two more days. 
 Monkey 7 on the fourth trial on the first day took the white 
 bread first and then the rye. Five times thereafter she repeated 
 this, and the habit of refusing the rye bread when it was 
 simultaneously exposed with the white was broken. There¬ 
 after, on the six succeeding days of the experiment she left 
 the rye bread only once. 
 
 After this series a second test was made in which the 
 white bread was made bitter with quinine. Both pieces of 
 bread were presented simultaneously on a board, as in the 
 previous experiments. The results of the tests, 10 experi¬ 
 ments a day for seven days, are given in Table V. It will 
 be seen that the animals .soon learned to avoid the bitter 
 bread. The mistake made by monkey 6 on the second day 
 was only a partial mistake for the animal picked up the white 
 bread, smelled it and immediately dropped it. The inhibi- 
 
INHIBITION OF HABIT. 
 
 31 
 
 tion in the case of monkey 7 was fixed on the first day after 
 four mistakes had been made, 
 
 A third series was then made in which the small pieces of 
 bread (white, bitter; rye, normal) were presented successively. 
 In these experiments the animal had a choice of taking or 
 of leaving the single piece, white or rye. The results are 
 given in Table VI. It will be seen that the three animals 
 avoided the white bread on the first day of the experiments, 
 but that on the second all took the white at least once. On 
 the second day monkey 6 took the white bread six times, 
 and monkey 8 four times. After these experiments both ani¬ 
 mals disregarded the white bread when it was presented to 
 them on the board, even though they were very hungry. Only 
 once during the remainder of these tests did monkey 6 touch 
 the white bread. On the fourth day, when the white bread 
 was presented, she put her hand through the wire of the cage 
 and swept the piece off the board. 
 
 The fourth series of experiments was begun after the 
 seven days of series 3. In this last series the- breads were 
 presented to the animal by the experimenter, to see if the 
 inhibition effect, or the new habit of leaving the white bread 
 was associated with the presentation of the bread upon the 
 board outside the cage. Each piece of bread was taken by 
 the experimenter in his fingers and held just within the wire 
 netting of the cage. Even with this added inducement 
 monkey 6 refused to take the white bread, but both of the 
 other animals finally, and monkey 7 repeatedly, took it. 
 In the 70 trials with monkey 7, she took the white bread 
 fourteen times. At the sixth trial on the fifth day monkey 
 7 struck at me when I offered her the white bread, as if angry 
 that it should be presented to her. Monkey 8 took the white 
 bread twice on the second day, but after he had smelled it he 
 dropped it. On the fifth day he once reached toward the 
 white bread that was offered him, but drew back before he 
 had touched it. On the sixth day he struck at the white 
 bread or at me just as he had monkey 7 on the previous day. 
 
 These experiments show, on the part of the monkeys 
 investigated, a rapid inhibition of a previously formed habit. 
 
32 
 
 WILLIAM SHEPHERD. 
 
 and a rapid acquisition of a habit opposed to the original 
 habit. The acquisition of the new habit, however, took a 
 much longer time than that of the original habit, as can be 
 seen from a comparison of the results. In the later experi¬ 
 ments so many new factors, especially those of tempting the 
 animal by single presentations, were introduced that the 
 curves of learning cannot be directly compared. It would 
 be fair, however, to compare the results in Table V with those 
 in Table I. When this is done it is seen that the replacement 
 of the association is almost, if not equally, as rapid as the 
 formation. Should this result be true for other activities 
 of the monkey, it would indicate a rather high degree of 
 adaptivity, which goes far towards the production of appar¬ 
 ently intelligent actions. 
 
 TABLE IV. 
 
 Inhibition of habits. All monkeys had learned to avoid quinine {rye) bread; no quinine 
 in either bread for this test. W = took white, R = took rye. T en trials each animal 
 daily. 
 
 MONKEYS AND 
 SERIAL DAYS. 
 
 I 
 
 2 
 
 3 
 
 4 
 
 s 
 
 6 
 
 7 
 
 6 
 
 W 
 
 lO 
 
 w 
 
 8 
 
 w 
 
 lO 
 
 W 
 
 lO 
 
 w 
 
 lO 
 
 W 
 
 lO 
 
 W 
 
 10 
 
 
 R 
 
 o 
 
 R 
 
 8 
 
 R 
 
 lO 
 
 R 
 
 lO 
 
 R 
 
 lO 
 
 R 
 
 lO 
 
 R 
 
 lO 
 
 7 
 
 W 
 
 lO 
 
 W 
 
 lO 
 
 W 
 
 lO 
 
 W 
 
 lO 
 
 W 
 
 lO 
 
 W 
 
 lO 
 
 W 
 
 lO 
 
 
 R 
 
 5 
 
 R 
 
 lO 
 
 R 
 
 lO 
 
 R 
 
 9 
 
 R 
 
 10 
 
 R 
 
 lO 
 
 R 
 
 lO 
 
 8 
 
 W 
 
 lO 
 
 W 
 
 lO 
 
 W 
 
 lO 
 
 W 
 
 lO 
 
 W 
 
 lO 
 
 W 
 
 lO 
 
 W 
 
 lO 
 
 
 R 
 
 o 
 
 R 
 
 o 
 
 R 
 
 o 
 
 R 
 
 4 
 
 R 
 
 8 
 
 R 
 
 lO 
 
 R 
 
 lO 
 
 TABLE V. 
 
 Inhibition of habits. All monkeys had been practised on preceding tests {Table IV). 
 Quinine in white bread, rye bread in natural stale. Simultaneous presentation. R 
 took rye, W = took white bread. Ten trials daily. 
 
 MONKEYS AND 
 SERIAL DAYS. 
 
 I 
 
 2 
 
 3 
 
 4 
 
 s 
 
 6 
 
 7 
 
 
 W 
 
 I 
 
 W 
 
 I* 
 
 w 
 
 I 
 
 W 
 
 I 
 
 W 
 
 lO* 
 
 W 
 
 lO 
 
 W 
 
 lO 
 
 6 
 
 R 
 
 lO 
 
 R 
 
 lO* 
 
 R 
 
 lO 
 
 R 
 
 lO 
 
 R 
 
 lO* 
 
 R 
 
 lO 
 
 R 
 
 lO 
 
 
 W 
 
 4 
 
 W 
 
 o 
 
 W 
 
 o 
 
 W 
 
 o 
 
 W 
 
 o 
 
 W 
 
 o 
 
 W 
 
 o 
 
 7 
 
 R 
 
 lO 
 
 R 
 
 lO 
 
 R 
 
 lO 
 
 R 
 
 lO 
 
 R 
 
 lO 
 
 R 
 
 lO 
 
 R 
 
 lO 
 
 
 W 
 
 3 
 
 W 
 
 I 
 
 W 
 
 o 
 
 W 
 
 o 
 
 W 
 
 o 
 
 W 
 
 o 
 
 W 
 
 o 
 
 8 
 
 R 
 
 lO 
 
 R 
 
 lO 
 
 R 
 
 lO 
 
 R 
 
 lO 
 
 R 
 
 lO 
 
 R 
 
 lO 
 
 R 
 
 lO 
 
IMITATION. 
 
 33 
 
 TABLE VI. 
 
 Inhibition of Habits. All animals had been previously practised on experiments recorded 
 in Tables IV and V. Quinine in white bread, rye bread in natural state. Succes¬ 
 sive presentations. W = took white, R = rook rye bread. Ten times daily. 
 
 MONKEYS AND 
 SERIAL DAYS. 
 
 I 
 
 2 
 
 3 
 
 4 
 
 s 
 
 6 
 
 7 
 
 
 W 
 
 I 
 
 W 
 
 6 
 
 w 
 
 0 
 
 W 
 
 0 
 
 W 
 
 0 
 
 w 
 
 0 
 
 W 
 
 0 
 
 6 
 
 R 
 
 10 
 
 R 
 
 10 
 
 R 
 
 10 
 
 R 
 
 10 
 
 R 
 
 10 
 
 R 
 
 10 
 
 R 
 
 10 
 
 
 W 
 
 0 
 
 W 
 
 I 
 
 W 
 
 I 
 
 W 
 
 S 
 
 W 
 
 I 
 
 W 
 
 I 
 
 W 
 
 2 
 
 7 
 
 R 
 
 10 
 
 R 
 
 10 
 
 R 
 
 10 
 
 R 
 
 10 
 
 R 
 
 10 
 
 R 
 
 10 
 
 R 
 
 10 
 
 
 W 
 
 0 
 
 W 
 
 4 
 
 W 
 
 0 
 
 W 
 
 0 
 
 W 
 
 0 
 
 W 
 
 0 
 
 W 
 
 0 
 
 8 
 
 R 
 
 10 
 
 R 
 
 10 
 
 R 
 
 10 
 
 R 
 
 10 
 
 R 
 
 10 
 
 R 
 
 10 
 
 R 
 
 10 
 
 TABLE VII. 
 
 Inhibition of Habits. All animals had been previously practised on experiments recorded 
 in Tables IV, V, and VI. Quinine in white bread, rye bread in natural state. Suc¬ 
 cessive presentations through wire netting of cage instead of outside. W = took white, 
 R = took rye bread. Ten trials daily. 
 
 MONKEYS AND 
 
 SERIAL DAYS. 
 
 I 
 
 2 
 
 3 
 
 4 
 
 s 
 
 6 
 
 7 
 
 
 W 
 
 0 
 
 W 
 
 0 
 
 W 
 
 0 
 
 W 
 
 0 
 
 W 
 
 0 
 
 w 
 
 0 
 
 w 
 
 0 
 
 6 
 
 R 
 
 10 
 
 R 
 
 10 
 
 R 
 
 10 
 
 R 
 
 10 
 
 R 
 
 10 
 
 R 
 
 10 
 
 R 
 
 10 
 
 
 W 
 
 0 
 
 W 
 
 0 
 
 W 
 
 3 
 
 W 
 
 5 
 
 W 
 
 2 
 
 W 
 
 3 
 
 W 
 
 4 
 
 7 
 
 R 
 
 10 
 
 R 
 
 10 
 
 R 
 
 10 
 
 R 
 
 10 
 
 R 
 
 10 
 
 R 
 
 10 
 
 R 
 
 10 
 
 
 W 
 
 0 
 
 W 
 
 2 
 
 W 
 
 I 
 
 W 
 
 0 
 
 W 
 
 I 
 
 W 
 
 0 
 
 W 
 
 0 
 
 8 
 
 R 
 
 10 
 
 R 
 
 10 
 
 R 
 
 10 
 
 R 
 
 10 
 
 R 
 
 10 
 
 R 
 
 10 
 
 R 
 
 10 
 
 Imitation. 
 
 Whether or not the higher animals learn by imitation is a 
 mooted question among comparative psychologists. Involv¬ 
 ing as it does the presence of ideation or of ‘transferred 
 association’ in the animal mind, it is of the utmost impor¬ 
 tance that this mental function be carefully considered. 
 Most experiments in the field of comparative psychology are 
 inclined to deny the higher forms of imitation to animals. 
 Morgan (7), Thorndike (8, 9) Yerkes (4), and Watson (10), 
 take this ground, while Kinnaman (ii), Hobhouse (12), 
 Berry (13), Haggerty (14) and others believe they have 
 found evidence of imitation, apart from mimicry which is 
 
34 
 
 WILLIAM SHEPHERD. 
 
 sometimes spoken of as imitation and which may be considered 
 a reflex adaptation to certain stimuli. 
 
 It is necessary first to define precisely the term imitation 
 as it is applied in comparative psychology. The term imita¬ 
 tion is used in a number of different senses. First it is used 
 as a synonym for what may be more accurately called instinc¬ 
 tive imitation. This is the form of imitation shown by all 
 animals in the performance of certain necessary acts, such as 
 that of a chick when it pecks at a bit of food on seeing 
 another chick do so. A second use of the word is in relation 
 to mass activities and we may therefore speak of it as gre¬ 
 garious imitation. This form of imitation is the performance 
 of an act similar to that performed by another animal of a 
 herd or flock, the latter act being due to a definite stimulus 
 but the act of the imitator, although similar to that of the 
 imitatee, is not produced by this same stimulus but by the 
 stimulus of seeing or of hearing the first animal. This form 
 of imitation is illustrated by a herd of buffaloes running off 
 in alarm or a flock of sheep following their leader and jump¬ 
 ing over an imaginary obstacle. Lastly, there is inferential 
 or reflective imitation, where one individual sees another 
 perform an act, realizes or understands the consequences of 
 that act, and thereupon performs a similar act with the idea 
 of getting the same results. 
 
 Of the above forms of imitation the first two are of a low 
 order, and perhaps no comparative psychologist will deny 
 them to many of the higher animals below man. The con¬ 
 troversy, however, arises regarding the presence and the 
 amount of the higher form of imitation, the imitation that 
 has been defined as inferential. An attempt was made to 
 test the presence of this form of imitation, and three separate 
 kinds of tests were made. 
 
 An apparatus and plan similar to those already employed 
 by Hobhouse (12, chap, x), and by Watson (10, p. 175), 
 with which the former obtained positive and the latter nega¬ 
 tive results, were used by me in this work. This apparatus 
 consists of a glass tube or cylinder, 15 inches long and |-inch 
 in diameter, and a plunger of wood less in diameter but 
 
IMITATION. 
 
 35 
 
 3 inches longer than the tube. The diameter of the wooden 
 plunger was such that it could be readily inserted in the tube 
 and was freely movable. A piece of food, usually banana or a 
 peanut, placed crosswise in the tube, was inserted in the 
 tube and pushed down about half way. A piece of food was 
 always selected sufficiently large not to drop through' the 
 tube when the latter was held in a vertical position, and yet 
 sufficiently small that it could be readily displaced when the 
 plunger was used to push it through the tube. This appara¬ 
 tus was presented in turn to eight monkeys and the imita¬ 
 tion tests were begun after each had failed to manipulate 
 it three times. The tube and the stick were placed in 
 the cage with the animal to see if he could push out the food 
 after his attention had been attracted and he had been shown 
 by the experimenter how the stick was to be manipu¬ 
 lated for obtaining the banana or peanut. I endeavored co 
 have the monkeys see me perform the act at least three 
 times before each trial, and to this end I watched carefully 
 each animal while the exhibition was being given. If there 
 was in my mind any doubt that the animal had paid close 
 attention to the performance of the act by me, I repeated it. 
 Following the three demonstrations the animal was given 
 the apparatus for one minute, and this amount of time was 
 allowed him in which to perform the act. In the case of 
 some animals these tests were made daily and in the case 
 of others there were intervals of one or two days. The varia¬ 
 tion in the time routine did not appear to have any influence 
 upon the learning. The number of trials for each animal 
 was also variable, from 24 to 96, each having seen the act 
 performed three times before each trial, i.e., having seen the 
 act from 72 to 288 times. 
 
 All the monkeys failed to show any signs of imitation, and 
 I was unable to verify on these animals results like those 
 reported by Hobhouse. At the end of the tests all animals 
 acted in much the same way as at the beginning of the tests 
 and the first notes may be used to indicate the general charac¬ 
 ter of their activity. They bit at the food in the tube, looked 
 into the end of the tube at the banana, jerked the tube 
 
36 
 
 WILLIAM SHEPHERD. 
 
 around, often took up the stick and bit it; sometimes picked 
 up the stick and threw it away, dropped the tube, as if they 
 gave it up. While I was exhibiting the mechanism an 
 animal would watch the operation closely, and at the moment 
 the food appeared on the end of the tube within the cage, he 
 seizdd it. Sometimes they attempted to reach into the tube 
 to get the food. In this experiment I also tried Hobhouse’s 
 plan of ‘suggestion.’ When an animal was wrestling with the 
 tube I pointed to and moved the stick towards him to call 
 his attention to it. Thereupon, sometimes, an animal took 
 up the stick, but only to throw it away after a few seconds. 
 In brief, throughout the three weeks of this experiment not 
 one of the animals appeared to show any understanding of 
 the problem, and any ability to cope with it. It may be 
 urged that a greater number of trials than 24, the number 
 some of the animals received, would have brought success 
 for some of them, but, were imitation of this kind an impor¬ 
 tant and a constant condition in monkeys, this number of 
 trials would have been sufficient for them to demonstrate 
 its presence. 
 
 A second experiment on imitation was then performed. 
 This experiment was also similar to one made on monkeys 
 by Hobhouse (12, chap, x) and by Watson (10, pp. 173, 174) 
 A T-rake, consisting of a light handle 18 inches long, with a 
 cross piece 4 inches long nailed at one end, and at right angle 
 to the handle, was used. This was placed with the handle 
 end through the wire side of the cage and with the T-end 
 extending outward on the floor of the room. A piece of 
 banana or a peanut was placed on the floor outside the cage 
 but within the sweep of the rake when handled by an animal 
 inside the cage. In this experiment I wished to see if the 
 monkeys could grasp the situation of the use of the rake. 
 Two separate factors of adjustment enter into this; first, 
 the use of the rake for the use of grasping or reaching out 
 towards the food, and, secondly, the use of the implement to 
 draw the food nearer the cage and within reach of the hand 
 itself. 
 
 Monkeys i, 2, 3, 4, 5, 6, 7 and 8 were tested in this manner. 
 
IMITATION. 
 
 37 
 
 I showed an animal, by hooking the T around the food and 
 pushing the rake forward, but necessarily from the outside 
 of the cage, three times before each trial. I then let the 
 animal try the apparatus for two minutes. Each of the 
 eight animals was given from 24 to 120 trials in the four weeks 
 these experiments were continued. Each animal had failed 
 to use the rake three times, without being shown, before the 
 imitation tests were begun. 
 
 Monkeys i, 2, 4, 5, 6, 8, failed as completely as in the pre¬ 
 vious experiment. Each seized the end of the rake handle 
 within the cage, jerked it around, bit it, tried to reach the 
 food with their hands through the wire of the cage. Some¬ 
 times they became angry and threw the rake aside; in some 
 cases they finally gave it up. I noted, however, that all soon 
 came to pull the rake into the cage, up to the T, which was 
 as far as the wire of the cage would permit. So far as I could 
 tell, each animal watched me closely while I was showing 
 them, appeared to have much interest in the proceedings, and 
 seized the food through the wire of the cage the moment it 
 was within reach. But what the six monkeys did not do was 
 to push the rake out and hook the cross piece around the 
 food. In all of the trials with these six monkeys so far as I 
 could see, none gave a sign of using the rake as a tool with 
 which to draw the food toward the cage. Here again it 
 may be urged that a greater number of trials than 24 should 
 have been given to all animals, but the simplicity of the 
 apparatus was such that in a child of the same age one or 
 two trials are sufficient for a successful imitation. 
 
 After several days tests with two of the animals, 3 and 7, 
 there appeared some evidence of imitation. Monkey 3 
 learned to push out the rake, and, with much slashing about 
 of the T end, would draw in the food. The securing of the 
 food by hooking the T around it was, however, very awkward 
 work for her. Sometimes she would knock the peanut away 
 in her effort to pull it in. On its being replaced by me she 
 usually succeeded in drawing it in roughly. The best that 
 can be said of her performance is that she appeared to learn to 
 imitate me perfectly in pushing the rake out, and awkwardly 
 in pulling the food in with the rake. 
 
38 
 
 WILLIAM SHEPHERD. 
 
 Following is an account of the tests on the third day. It 
 will be remembered that the method of securing the food by 
 the use of the rake was demonstrated three times before each 
 test. On this day, however, this demonstration was omitted 
 after the seventh test. Trial i was a failure; the animal did 
 not attempt to take hold of the rake after being shown three 
 times. On the second trial, however, after the three dem¬ 
 onstrations, the animal grasped the rake, and moved it 
 about in the direction of the food. She succeeded in bring¬ 
 ing in the nut sufficiently far so that she reached it and ate 
 it; time, 85 seconds. In this first successful attempt she 
 grasped the rake awkwardly and did not at first seem able 
 to manage it, but eventually she managed to swing it around 
 so that the nut was caught and, as has been said, pulled in 
 the nut so that she secured it by reaching through the bars 
 of the cage. In the next trial the movements of handling 
 the rake were about the same as those in the preceding trial, 
 although she had acquired some facility in the use of the tool, 
 and managed to secure the nut in 29 seconds. On the 
 fourth trial, during the demonstration she took hold of the 
 rake and pushed it outwards toward the food. After this 
 demonstration she pushed the rake outwards in an apparently 
 purposeful manner and secured the nut in 20 seconds. The 
 fifth trial was similar to the fourth, but with an increasing 
 ease in adjustment and with apparent greater facility in the 
 use of the rake. In 8 final trials on this day the demonstra¬ 
 tions were omitted but the actions of the animal were similar 
 to those mentioned in connection with the fourth and fifth 
 trials. On the sixth and tenth trials the manipulation of the 
 rake caused the nut to roll further away from the cage and 
 each time I replaced it. The times for these 8 trials were 
 respectively: 15, 15, 9, 12, 24, 14, 4 and 8 seconds. 
 
 Three days later, the fourth day of the tests, demonstra¬ 
 tions were given in the first and second tests. The results, 
 however, were similar to those on the third day, but with 
 shortened times for the performance of the act. The times 
 were as follows: 10, 8, 3, 4, 2, 6, 7, 9, 5, 4, 3,9,2, and 2 seconds. 
 On the seventh trial she managed to get the nut only part 
 
IMITATION. 
 
 39 
 
 way with the first pulling movement, but then, with apparent 
 intention, she gave another pull and brought the nut within 
 reach. On the thirteenth trial the nut rolled away when 
 touched with the rake but the animal pushed the rake out 
 farther beyond the nut and managed to secure it in the usual 
 way. 
 
 Monkey 7 had failed to show evidence of imitation on six 
 days of 3 trials each when I manipulated the rake and at¬ 
 tempted to exhibit the mechanism of securing food. It was 
 thought that the operation of another animal might be imi¬ 
 tated, and for this reason I placed her with monkey 3, which 
 at that time had learned to use the rake with facility. In 
 these experiments I carefully watched monkey 7 in order 
 to be reasonably certain that she observed monkey 3 perform 
 the act with the rake. Only after I was sure the animal 
 had been looking in the direction of monkey 3 while the latter 
 performed the act three times did I begin a test of monkey 
 7. At these times also I called attention to the tool by 
 pointing to it and to the food. In each trial she was per¬ 
 mitted two minutes to perform the imitation act, and if it 
 was not performed in that time I counted the test a failure. 
 The first five tests were failures, and likewise the seventh 
 and tenth. On the sixth trial she pushed the rake outwards 
 awkardly, but with such movements that she could not 
 secure the food. Finally, however, after 79 seconds, she 
 managed to hook the cross pieces about the food and immedi¬ 
 ately pulled it in. The eighth and ninth trials were similar 
 to the sixth but the food was secured in 20 and 8 seconds 
 respectively. The tenth trial I have counted a failure, for 
 in this case the animal pushed out the rake and pulled it in 
 but did not manage to get the cross piece hooked around the 
 food. In the two minutes allowed for the performance of 
 the act, she did not manage to secure the food. 
 
 On the second day of these tests she failed to show by her 
 actions any evidence of imitation in the second, third, sixth, 
 ninth and tenth trials. Only ten seconds were required for 
 the proper performance of the act on the first trial, for she 
 immediately manipulated the rake so that it caught the nut 
 
40 
 
 WILLIAM SHEPHERD. 
 
 and she pulled it inwards. The fourth, fifth and seventh 
 trials were similar, 12, 15 and 9 seconds respectively. Al¬ 
 though the eighth trial was counted a failure the monkey 
 did manipulate the rake properly, but in pushing it out¬ 
 wards the nut was hit and rolled farther way. In the final 
 trials of this day the animal did not make many efforts to 
 use the rake, although two minutes were allowed. In these 
 experiments the movements of pushing the rake outwards 
 were well performed, but those of pulling inwards were very 
 badly executed. This was also noted on the following day, 
 when in 6 trials, she managed to secure the food four times, 
 in 17, 9, II and 5 seconds. In subsequent trials she con¬ 
 tinued to be incoordinate in the pulling in movements, but 
 accurate in pushing out the rake. 
 
 It may be said that monkey 7 did not make the mental 
 connection between seeing me manipulate the rake and the 
 idea of the acquisition of food, but that some connection or 
 association was formed between seeing monkey 3 perform 
 the act and such an idea (or what corresponds to an idea in 
 the monkey mind). The impulse to handle the rake, to 
 manipulate it and to use it in connection with food may prop¬ 
 erly be said to be due to her having seen monkey 3 obtain 
 food by the use of this tool. This impulse or association 
 may be explained as A, ideal, or B, imitative. In the present 
 state of comparative psychology it is best to consider it the 
 latter, and we conclude that the use of the rake by monkey 
 7 has been due to an imitative impulse, the tendency to per¬ 
 form the same or a similar act performed by another. 
 
 In a third experiment to test the ability to imitate the 
 apparatus used in a previous test of analogical reasoning was 
 again employed. A piece of banana was suspended from a 
 pole that extended across the room. The food was placed 
 about 4 feet from the floor, high enough to be beyond the 
 reach of the animal when it stood upright on the floor. A 
 light pole, 9 feet long, 1.5 inches in diameter, was loosely 
 attached at one end by means of a pivot to a support and 
 extended 7 feet across the open space of the room to a hori¬ 
 zontal supporting board (C), which was 2.5 feet from the 
 
IMITATION. 
 
 41 
 
 floor. Figure 2 shows the arrangement. On the horizontal 
 board the free end of the pole was easily moved, and by 
 shifting it to the position P-X the pole was brought under 
 the suspended food. To bring about this change it was 
 necessary to move the free end of the pole about 3 feet, with 
 corresponding decreases in the amount of movement the 
 nearer to the pivot the pole was grasped. In the position 
 P-X the animal was able to reach the food if it climbed upon 
 the pole. The problem for the animal was to slide the free 
 end of the pole sufficiently far on the horizontal board or 
 
 Fig. 2. Diagram of apparatus used in tests of reason and of imitation. A, sus¬ 
 pended banana; B, sliding pole in original position; PX, position to which pole had 
 to be moved in order that the animal might reach the food; C, supporting slide. 
 
 slide, to bring the pole under the suspended food: to go upon 
 the pole and get the food by reaching upwards about 10 inches. 
 
 Monkey 3 had been tested with this apparatus in a series 
 of experiments on analogical reasoning (see p. 52 ff.) and 
 had failed to manipulate it. In this first test she was not 
 shown how to manipulate it, but was shown when the ex¬ 
 periments on imitation were begun. Before each test by the 
 animal I moved the pole from its original position to that 
 of P-X and permitted the animal to obtain the food each 
 time. After having been shown in these three trials that the 
 moving of the pole was essential for the securing of food I 
 allowed two minutes in which the animal might perform the 
 
42 
 
 WILLIAM SHEPHERD. 
 
 act. If the act was not accomplished in that time I called 
 the test a failure and repeated the demonstrations. On the 
 first day of this experiment she succeeded in performing the 
 complex adjustment of pole, etc., four out of eight times. 
 On the first trial she moved the pole awkwardly, but suffi¬ 
 ciently far that she was able to reach the food; 30 seconds. 
 On the second trial she moved the pole only a trifle at first, 
 but returned to it and succeeded in moving it the required 
 distance; 25 seconds. On the next trial she changed the 
 movement of pushing to that of lifting the pole and pushing 
 at the same time. In this her movements were awkward 
 but she managed to get the food in 45 seconds. The next 
 three trials were counted as failures, although in all of them 
 the animal manipulated the pole and showed signs of recogni¬ 
 tion of the use to which the pole was to be put. On the fourth 
 trial after having moved the pole a short distance and not 
 obtaining food, she tried to get the food by going to the cross 
 piece from which the food was suspended. On the fifth 
 trial she pushed the pole off the slide, and could not get the 
 food. On the sixth trial she moved the pole a short distance, 
 and refused to complete the act. On the seventh trial she 
 moved the pole sufficiently far to obtain the food, 20 seconds; 
 but on the eighth trial she did not try to move the pole or to 
 obtain the food. 
 
 On the following day she was not shown how to manipu¬ 
 late the apparatus the first three trials, two of which were 
 successful, but was shown on the other seven trials. At first 
 she pulled the pole part way, but it appeared to be difficult 
 work, and she obtained the food in 20 seconds. The second 
 trial she pulled the pole about 6 inches and stopped; I took 
 hold of the pole and helped her to move it a few inches, and 
 then she made a great effort and pulled the pole sufficiently 
 far to obtain the food. The third trial she did not attempt 
 to move the pole. The following seven trials were, it has 
 been mentioned, preceded by three demonstrations each. 
 In only the fourth trial did she take hold of the pole, at other 
 times she did not attempt any manipulation. On the third 
 day of the tests, six days following that just mentioned, she 
 
IMITATION. 
 
 43 
 
 failed to manipulate the apparatus in six trials, although she 
 was shown three times before each trial. Only once on this 
 day did she touch the poll. Two days later, she failed on 
 the first six trials. On the seventh she helped me to push 
 the pole when I was demonstrating it, and following this 
 she managed to get the pole moved by her own efforts in 68 
 seconds. On the eighth trial she moved the pole part way 
 and obtained the food in 30 seconds. The ninth trial was 
 a failure in two minutes. Seven days later she was given 
 three trials and was not shown how to manipulate the appara¬ 
 tus. She did not exhibit any signs of ability to manipulate 
 the apparatus and was then shown. The fourth trial was a 
 failure; on the fifth, she took hold of the pole at the time I 
 was moving it, and after the demonstrations she immediately 
 attempted to move the pole, which she managed in 36 seconds. 
 On the sixth trial she managed to get the food in 26 seconds. 
 The seventh trial, 18 seconds, and the eighth trial in 9 seconds 
 were successful. In the remaining 17 trials on this day 
 she gradually lowered the time for the .performance of 
 the act although her actions were always about the same. 
 Two days later there were no failures in 20 trials, but the 
 actions of the animal in 2 of the trials were noteworthy. 
 In these trials the monkey moved the pole part way, then 
 stopped, and appeared to be observing the amount of the 
 movement and making a judgment regarding the possibility 
 of reaching the food from the pole in the position in which 
 she had placed it. Then, not liking the position she moved 
 the pole farther and climbed upon it and obtained the food. 
 
 In these tests there was a gradual learning to eliminate 
 unnecessary movements and to perform the necessary move¬ 
 ments in a satisfactory manner. The experiment does not, 
 however, wholly belong to the type of learning of trial and 
 error, for, as has been remarked, the monkey moved the pole 
 properly the first time she attempted to do so. There was 
 no previous groping for a something, no fumbling with any¬ 
 thing but the pole. The attention, so far as shown in action 
 was directed to the pole. There was an immediate grasp 
 of the situation, and this coming after she had been shown 
 
44 
 
 WILLIAM SHEPHERD. 
 
 that the food was to be reached by changing the position of 
 the pole is taken as evidence of imitation. The fact that in 
 later tests she failed to pay attention to the pole and to 
 attempt to move it does not mean that she could not imitate, 
 but it is well known that the monkey is extremely variable in 
 its actions, and its attention is held with difficulty. We know 
 that in the production of habit, as many observers have 
 pointed out, there is not a gradual shortening of the time of 
 reaction as in the case of cats and dogs, but that the time for 
 the performance of an act is extremely variable. It is this 
 factor of variability due to elements such as lack of hunger, 
 etc., which seems to me to account for the lack of attention 
 to the problem. It is further to be noted that in the success¬ 
 ful trials she used a variety of means for the accomplishment 
 of the end; sometimes she pushed the pole unhesitatingly 
 through the required distance; at times she pulled it, and at 
 times she pulled and pushed it. The variety of action indi¬ 
 cates that the performance of the act is not like that of a habit 
 formation, but that of attempting to accomplish an object 
 in any manner that this could be done. The performance 
 of the act was awkward, but there appeared to be imitation 
 of a relatively high order. 
 
 In the experiments that have been described evidence of 
 imitation by monkeys 3 and 7 was found in the second experi¬ 
 ment of the series. These two monkeys and the other six 
 failed to show signs of imitation in the first experiment and I 
 failed to find any signs of imitation on the part of the six 
 monkeys in the second experiment. When these results are 
 combined, I think they indicate in a general way that some 
 monkeys may and do learn by imitation. The amount of 
 imitation is not shown, but the fact that in so many simple 
 experiments negative results were obtained indicates that 
 these animals do not imitate to the extent that has been 
 ascribed to them. We are, I believe, justified in concluding 
 that imitation is a mental function of the monkey or of some 
 monkeys, although the results of my experiments may be 
 interpreted to mean that imitation does not play a very 
 important part in their learning process. 
 
IDEA TION. 
 
 45 
 
 Ideation . 
 
 The evolution and anecdote schools attribute ideation of 
 a comparatively high order to the animals with brains similar 
 to that of man, and they hold that ideas similar to those of 
 man play an important part in animal behavior. They base 
 this belief, however, upon uncontrolled observation and upon 
 flimsy and circumstantial evidence and not upon controlled 
 experiments and unequivocal facts. The evolutionists also 
 appear to believe it necessary to attribute ideas to animals, 
 else the doctrines of a progressive mental development would 
 not coincide with that of the physical development. On the 
 other hand, others, mainly experimenters, hold that the 
 casual observations do not indicate animal ideation and the 
 same facts are interpreted differently. Special tests have 
 given negative results and from these and from the careful 
 observation of animals in laboratory surroundings they con¬ 
 clude that animal ideation is not proven, and that the present 
 evidence tends to indicate an absence or a lack of ideas in 
 the animal mind. Some, however, are willing to admit that 
 animals may have ideas, although in small number, but are 
 forced to conclude that ideas are a very unimportant element 
 in minds below that of man. It is true that for the most 
 part the experiments of these men have not been special 
 tests for ideation, but tests for other reactions in which idea¬ 
 tion, if present, should or might have been exhibited. 
 
 Before passing to the evidence of ideation, it seems well 
 to define the term ‘idea’ and thus to have a precise notion 
 of what we should look for in animal behavior. For the pur¬ 
 pose of comparative psychology an idea may be defined in 
 the following ways: a, an image or picture of a visual object 
 which is formed by the mind; h, a general notion or concep¬ 
 tion; c, a plan or purpose of action or an intention; d, idea in 
 the sense of an understanding of a certain relation or situa¬ 
 tion (as of sensible objects). If, therefore, an animal ex¬ 
 hibits any of the mental conditions noted above we may 
 conclude that ideation is present. 
 
 A simple test for ideation was made with animals i, 2 
 
46 
 
 WILLIAM SHEPHERD. 
 
 and 3. A board 20 inches long and 3|-inches wide was 
 placed with one end against the side of the cage and within 
 reach of the animal with the other end extending outwards 
 from the cage and beyond reach. Food was placed at the 
 farther end. The object of the tests was to determine if an 
 animal would understand the situation and be able to secure 
 the food by the indirect method of pulling at the board near 
 the cage rather than by directing its efforts at the food or 
 the food end of the board which was beyond reach. In the 
 first trial monkey 3 after some testing of the openings of the 
 wire netting, seized the board awkwardly near the end with¬ 
 in reach, pulled the board alongside the cage and secured 
 the food, 30 seconds. In this trial the attention for only a 
 few seconds was directed to the food position but was directed 
 mainly to the end of the board that did not contain the food. 
 The actions of the animals in manipulating the board were 
 awkward but only in the method of reaching for and hand¬ 
 ling it. The actions were directed to the board, and not to 
 the food, and it appears that here is an example of an under¬ 
 standing of a situation, a direction of action to an end not of 
 special interest in the situation. In the second and third 
 performances she duplicated her actions, but with shorter 
 times for its accomplishment because the awkwardness 
 largely disappeared and because the preliminary direction 
 of attention to the food was absent, 20 and 10 seconds. In 
 later tests she continued to react with complete success, 
 reducing the time to three seconds in the fifth trial, which 
 time includes that for the manipulation of the board and the 
 securing of the food. Monkey i succeeded In the third 
 trial but showed more misdirected efforts than did monkey 
 3. The third animal also managed to perform the necessary 
 acts to secure the food in the third trial, but the actions were 
 more awkward and poorly directed than those of monkeys 
 I and 3, 
 
 At the beginning of this test there were many unnecessary 
 movements on the part of each animal, but this awkward¬ 
 ness or lack of understanding of how to deal with the situa¬ 
 tion was mainly that of motor adjustment. It was a diffi- 
 
IDEATION. 
 
 47 
 
 culty in dealing with the board through the wire netting of 
 the cage and not a difficulty or a lack of understanding of 
 the problem, viz., of reaching for and of manipulating the 
 board. In the first trial of monkey 3 the animal directed 
 its attention to the board near the cage, and showed that 
 it appreciated the relation between this part of the board 
 and the securing of food. The later reduction in time for 
 the performance of the act, which at first sight makes the 
 experiment appear to be only another instance of learning 
 by trial and error, was mainly that of a proper adjustment 
 of arms and hands to the wire netting and not that of attack¬ 
 ing the board at the proper place. 
 
 These results indicate the presence of ideas of the above 
 described third or fourth class, i.e., a plan of action or an 
 understanding of a situation. It was apparent that from 
 the first the animal understood that the food could not be 
 reached directly but must be obtained by an indirect attack 
 on some other part of the apparatus. There was in this 
 case no general activity such as has been described by numer¬ 
 ous investigators in connection with the puzzle box experi¬ 
 ments; the trying here and the pulling there were notably 
 absent, and the attention of the animal was concentrated on 
 the board and on the getting the food. For a few seconds 
 after the board and food were displayed the hungry animal 
 would reach in vain for the food which was beyond its reach, 
 but this unsuccessful method was given up and the attack 
 directed to the board on which the food was placed. It is 
 also of great importance to note that the attack was directed 
 not to the part of the board on which the food was placed, 
 but to the part of the board away from the food. 
 
 The simplicity of the situation may be urged as an argu¬ 
 ment against the presence of ideation in solving the problem, 
 but I would again call attention to the ordinarily complex 
 character of similar experiments that have given negative 
 results to other observers. In such experiments we must 
 steer clear of both an absolute simplicity and of a complexity 
 abnormal to the animal. We must test the animal under 
 conditions which are within ks mental range and which will 
 
48 
 
 WILLIAM SHEPHERD. 
 
 show the ability to reason. It is equally bad to set impossible 
 conditions and to draw unsupported conclusions, and in 
 both these ways comparative psychologists have attempted 
 to solve the problem of animal reason and ideation. To 
 conclude, for example, that a dog or a monkey possesses 
 no power of ideation or of reasoning because it does not 
 thread a needle when the needle and thread are supplied 
 or because it does not unlock a door when the key is pre¬ 
 sented is to limit the terms ideation and reason to the 
 ability to perform certain activities connected with a certain 
 class of civilized man. The examples cited are, perhaps, 
 extremes but they illustrate the attitude of a certain class 
 of experimentalists. On the other hand, to conclude that 
 a cat reasons because it attracts attention by scratching on 
 the window pane or because it manages to strike an electric 
 button for the opening of the door is to take no account of 
 the possibility of previous training of the nature of trial and 
 error. We should steer between the two extremes and test 
 the so-called higher powers of animals by presenting to them 
 conditions appropriate to the class. In the simple experi¬ 
 ment described in this section and in others to be described 
 in the section on reasoning (p. 52 ff.) the conditions are appro¬ 
 priate to the monkey family. We have presented to the 
 animal a new problem which it must solve, a set of condi¬ 
 tions to be dealt with in order that a resultant pleasure 
 (hunger satisfaction) ensue. In the solution of the problem 
 the animals took a direct path. There was none of the fum¬ 
 bling or groping, no trying here and there, no attempts upon 
 other parts of the cage or its surroundings, but a direct 
 attack upon the board that held the food. 
 
 Somewhat similar actions were observed and noted above 
 in connection with the food box experiments. After the 
 previously ignorant (so far as the food box is concerned) 
 monkeys had been fed from the food box a few times so they 
 might become acquainted with the location of the food, the 
 doors were closed. Then began an attack upon the doors, 
 not upon the wire netting of the sides of the cage. The 
 animals remained active about t3he doors of the food boxes; 
 
IDEATION. 
 
 49 
 
 they did not sit sulkily upon the platform in the cage; the 
 attention was directed to the doors of the food boxes. Thorn¬ 
 dike indicates a belief that the attention of his cats was 
 directed to the string or to the button because the movement 
 in connection with these parts of the appliances was followed 
 by a pleasure. With my monkeys, however, the attention 
 of the animals could not have been directed to the doors of 
 the food boxes on account of a pleasure in connection with a 
 movement, for the animals did not have to perform any 
 movements (except those of taking the food from the boxes) 
 in the neighborhood of the food boxes, and in connection with 
 the doors there were no movements to be performed. In 
 these cases there was no activity in the region of the food 
 boxes or of the doors which could become associated with a 
 pleasure. The attention must then be due to something 
 different from the movement-pleasure association; and it 
 seems probable that there is some form of ideation to account 
 for it. 
 
 In the brightness tests interesting behavior on the part of 
 monkey 3 was observed. As has been explained above, the 
 doors were opened by turning a button and thus food was 
 secured. After several days’ experiments monkey 3 began 
 to close the door of her own accord instead of waiting for 
 me to close it preparatory to another trial. Before I could 
 close the door after she had opened it she would close it. 
 This would expedite the getting of food. It soon became 
 almost an habitual custom for her to close the door, and in 
 the remaining tests she did so from three to ten times each 
 day throughout the experiments. Why should an animal 
 close the door if it was not with some design or intention 
 (perhaps vague) of hurrying the food-getting process? It 
 seems too parsimonious and even inadequate to call this 
 kind of reaction a mere reflex. In the puzzle box experi¬ 
 ment the same animal and two others, monkeys i and 2, 
 often attempted to hold open the door with their hands and 
 this kind of action was recorded many times throughout the 
 series. At first I was inclined to consider this action as a 
 reflex but it is overworking the meaning of the term reflex 
 
50 
 
 WILLIAM SHEPHERD. 
 
 to use it to explain or to describe the actions of the three 
 animals in this particular. Somewhat similar behavior in a 
 different situation was also noted with monkey 8. In test¬ 
 ing his color discrimination with colored rice and breads, the 
 food was placed on a block about a foot away from the cage 
 and then moved close to the cage so that he could readily 
 reach the food through the wire netting. When the food was 
 taken or rejected the block was moved backwards a short 
 distance preparatory to placing food upon it for another 
 trial. During the first few experiments he acted as most 
 of the monkeys do, viz., moved away from the side of the 
 cage as soon as the food was obtained. After he had learned 
 to discriminate the colors, however, it was noted that he 
 began to push the block away after having taken the appro¬ 
 priate food. This he continued to do during the remaining 
 trials on that day. The actions of monkey 8 in this case 
 had all the appearances of an understanding of the situation 
 and a desire to hurry the getting of food. It is of consider¬ 
 able importance in this connection to bear in mind that the 
 performance of this action was sudden and not a gradual 
 growth and not the result of trial and error. 
 
 The actions of the three animals in the situation of the 
 diagonally placed board, the actions of monkey 3 in closing 
 the door of the food box and the actions of monkey 8 in push¬ 
 ing away the block on which food was placed have a similar 
 appearance. Were we not obsessed by the law of parsimony 
 we would immediately say that these actions show the 
 presence of ideas in these animals, ideas of the highest form. 
 We are, however, in keeping with the law of parsimony, quite 
 justified in saying that the actions indicate the presence in 
 the minds of the animals of a something very much like an 
 idea in the human sense, a something that has for the animal 
 a function or a use similar to an idea in man. This ‘some¬ 
 thing’ may be crude and simple, and doubtless it is analyzed 
 by the animal, but it serves practical purposes. These 
 somethings may be termed, as does Hobhouse, practical 
 ideas; they may be partly sensory-motor reactions, and may 
 be partly instinctive, and in part they may be accounted for 
 
REASONING. 
 
 51 
 
 by the superior equipment of the monkey in coordinated 
 vision and in the use of the hands. The evidence of the 
 ability to imitate supports the view of the presence of these 
 practical ideas, and in the section of this paper on adaptive 
 intelligence (p. 54 ff.) will be found additional evidence. 
 The fact that only a few of the animals exhibited anything 
 like the actions described, and each animal only a few, indi¬ 
 cates that these ‘practical ideas’ play a subordinate and unim¬ 
 portant role in the ordinary life of the animal, but they do 
 indicate that ideas may be present and have effect under 
 certain circumstances. 
 
 Reasoning. 
 
 It was formerly held that man alone possesses reasoning 
 power; other animals only instinct. Comparative psy¬ 
 chology has modified the older view of instinct, viz., that it 
 is a mysterious power, perfect at birth, unerring, unchangeable 
 in its working, and radically different from intelligence. It 
 has given the term a more precise definition- by limiting it 
 in various ways, e.g., from reflex action and habit on the one 
 hand and the higher mental processes on the other. But 
 comparative psychology has not materially changed the 
 general view that reason is confined to man. And the 
 generally accepted position is that the observed actions of 
 animals may be explained as the results of simple associations. 
 Some experimentalists have found material facts that indi¬ 
 cate to them reasoning ability on the part of some animals, 
 but this conclusion is not accepted by others. Here again, 
 as in considering the subject of ideation, it is necessary to 
 define what we mean by the term reason. This word has 
 been used in a variety of ways and the following definitions 
 include the most important of the meanings that have been 
 given to the term. A, implied reasoning (Harris), e.g., my 
 recognition of yonder horse; B, inference from particular to 
 particular (James), e.g., the bird which finds bread upon the 
 window one morning comes back the next morning; C, 
 adaptive intelligence, the ability to adapt to our purposes 
 conditions more or less difficult and more or less unfamiliar; 
 
52 
 
 WILLIAM SHEPHERD. 
 
 D, analogical reasoning, which involves construction or 
 creation, e.g., to reach an upper window I utilize a ladder 
 which I find; E, rational thinking (James); F, formal or 
 syllogistic reasoning. 
 
 That the higher mammals possess the ability to reason in 
 the first two senses, probably no comparative psychologist 
 will deny, although the explanation of the process may differ. 
 Implied reason is probably a function of all animals and the 
 ability to infer from particular to particular is well shown in all 
 the experiments in habit production. It is the higher level of 
 constructive analogical reasoning concerning which there is dis¬ 
 pute. Has the animal power to create or to construct? Hob- 
 house in some very interestingexperiments, claims to have found 
 satisfactory evidence of the presence in animals of this class 
 of reasoning. He calls the mental states that lead to this 
 form of reasoning articulate ideas, and he has satisfied him¬ 
 self that these are present in some monkeys and apes. 
 
 With the purpose of confirming the work of Hobhouse I 
 made experiments on three monkeys which had been exten¬ 
 sively used in the previously described work on discrimina¬ 
 tion, etc., and with which I found it most easy to experiment. 
 
 The first experiment was one similar to the box and the 
 chair experiments of Hobhouse. Food was suspended by a 
 string from a long pole reaching across the room, too high 
 for the monkeys to reach or to grasp by standing or by jump¬ 
 ing. A light box was placed near the point of suspension, 
 but sufficiently far away that the animal could not reach 
 the food by standing upon the box. Only when the box was 
 moved and the animal climbed upon it could the food be 
 secured by the animal. Each of the three monkeys were 
 given three trials, in each of which five minutes were allowed 
 for the performance of the act. At the time the experiments 
 were made the animals were hungry and apparently each 
 watched me attentively when I fastened the piece of banana 
 to the end of the string and suspended it. When all had 
 been arranged the animal was permitted to approach the 
 food and to secure it if possible. The actions of the animals 
 were similar. Sometimes the animal tried to reach the food 
 
REASONING. 
 
 S3 
 
 by jumping, sometimes after the unsuccessful attempts the 
 animal would have a puzzled look and finally gave up attempt¬ 
 ing to secure the food. None of the animals seemed to notice 
 the box and none made the attempt to use it as a means of 
 reaching the food. In their actions there was nothing that 
 I could interpret as a sign of deliberation or reflection, and 
 in this experiment I was unable to verify the results of Hob- 
 house. It is possible that the box appeared too heavy for 
 them to move, but I did not notice any indication that it had 
 been observed, or rather observed in connection with the food. 
 That the box was not too heavy was shown by the fact that 
 a similar and equally heavy box was being constantly moved 
 by monkeys in a large adjoining cage. Failure to secure 
 evidence of reasoning in this experiment may have been due 
 to the fact that I did not continue the experiment so long as 
 did Hobhouse. 
 
 A second experiment was then tried. This experiment 
 has been described in connection with the observations on 
 imitation, and the apparatus is illustrated on page 42, fig. 2. 
 The animal had to move a long pole, pivoted at one end, 
 under the suspended food in order to secure the food. In 
 all the experiments of this kind (three for each of the three 
 animals) there was no evidence of ability to grasp the situa¬ 
 tion and to solve the problem. In one trial an animal climbed 
 upon a cage which was near and jumped from it to the food. 
 In general, however, each animal made unsuccessful attempts, 
 most of them directed toward the food, and in each test 
 finally appeared to give up trying. Two months later one 
 of the animals (monkey 3) was again given three trials pre¬ 
 vious to the imitation tests and failed. 
 
 Adaptive intelligence. 
 
 Although no direct evidence of analogical reasoning was 
 obtained in the tests just described, other experiments on 
 reasoning gave interesting results. Those which are recorded 
 in this section deal with adaptive intelligence, the reason as 
 defined in the third class. 
 
54 
 
 WILLIAM SHEPHERD. 
 
 A piece of twine was permitted to hang in front of and 
 12 inches away from the cage, beyond the reach of the 
 longest armed animal which was tested. At the end of the 
 twine a piece of banana was arranged; a thin piece of wood 
 was pushed through the banana and turned so that one end 
 could be grasped by an animal in the cage. By grasping and 
 pulling the stick inwards the food was secured. All the 
 animals were tested in this experiment. The results with 
 all except one of the animals were similar, and a description 
 of the actions of one will suffice to indicate the whole. As 
 soon as the banana and stick were arranged monkey 6 put 
 her arm through the wire of the cage, seized the end of the stick, 
 drew it toward her and secured the banana. This experi¬ 
 ment was repeated a number of times and in all there was a 
 similar immediate characteristic response. There appeared 
 to be a decided adaptation of means to end. No efforts 
 were wasted upon random movements. It did not appear 
 that any preliminary attempt was made to grasp or even to 
 reach for the food, but there was an immediate movement 
 toward the stick. The results for all animals are given in 
 table VIII, in which is shown the approximate time for the 
 performance of the act by each animal in each trial. The 
 absence of hesitation, the direction of the movement away 
 from the food and towards the stick, and the promptness with 
 which the food was secured speak for the presence of adaptive 
 intelligence in ten of the monkeys. It would seem that this 
 is almost always found in these animals. 
 
 In this test the results with monkey 5 were decidedly 
 different from those with the other animals, in that he failed in 
 the trials given him. Monkeys i and 4 had considerable 
 difficulty in getting the food in the first and second tests, 
 but there was no gradual acquisition of the method of secur¬ 
 ing food with the other eight animals. The times for solving 
 the problem in the second and third tests were approximately 
 the same as those in the first tests for monkeys 2, 6 and 10, 
 and there was not much difference in the time between the 
 first and the later tests for monkeys 3, 7, 8, 9 and 11. Much, 
 if not all, the difference in time can be accounted for by the 
 
REASONING. 
 
 55 
 
 better adjustment to the wire netting of the cage, the push¬ 
 ing of the hand through in the proper place, etc. 
 
 TABLE VIII. 
 
 Adaptive intelligence, suspended food and stick. Three or six trials each animal. 
 Time in seconds; f = failed. 
 
 TRIALS AND 
 ANIMALS 
 
 I 
 
 2 
 
 3 
 
 4 
 
 s 
 
 6 
 
 I 
 
 60 
 
 10 
 
 3 
 
 — 
 
 — 
 
 — 
 
 2 
 
 4 
 
 3 
 
 3 
 
 — 
 
 — 
 
 — 
 
 3 
 
 20 
 
 6 
 
 9 
 
 — 
 
 — 
 
 —• 
 
 4 
 
 los 
 
 40 
 
 I I 
 
 — 
 
 — 
 
 — 
 
 5 
 
 f. 
 
 f. 
 
 f. 
 
 — 
 
 — 
 
 — 
 
 6 
 
 2 
 
 I 
 
 I 
 
 — 
 
 — 
 
 — 
 
 7 
 
 6 
 
 4 
 
 3 
 
 I 
 
 I 
 
 I 
 
 8 
 
 6 
 
 4 
 
 4 
 
 3 
 
 3 
 
 3 
 
 9 
 
 5 
 
 3 
 
 3 
 
 3 
 
 4 
 
 0 
 
 10 
 
 2 
 
 I 
 
 I 
 
 I 
 
 I 
 
 I 
 
 I I 
 
 5 
 
 5 
 
 3 
 
 3 
 
 3 
 
 3 
 
 A second test of the presence of adaptive intelligence was 
 made as follows: A light wooden lever B, 18.5 inches long 
 was attached by leather hinges at one end to a board A which 
 rested upon the floor The hinge of the lever was 4 inches 
 from the end of the horizontal board, well within reach of 
 the animals. The lever was inclined at an angle of approxi¬ 
 mately 45 degrees from the horizontal, and could be moved 
 forwards in a vertical plane. Fig. 3 illustrates the apparatus 
 in relation to the front of a cage. The apparatus was placed 
 outside a cage, the lower end of the lever being within, the 
 upper being beyond the reach of the animal. A piece of 
 banana or other food was placed at the farther end of the 
 lever, and the problem for the animal to solve was how to 
 secure the food which was beyond direct reach. 
 
 Eleven monkeys were tested with this apparatus, and all 
 with the exception of monkey 4 succeeded in the first test. 
 In the first test the animals usually took a longer time to get 
 the food, but as in the previous experiment this delay was 
 largely one of making the adjustment of hand to the proper 
 opening in the wire netting and not to attacking the apparatus 
 
$6 
 
 WILLIAM SHEPHERD. 
 
 in the proper manner. In the case of monkey 4, he appeared 
 frightened at the time of the first trial and I showed him that 
 food was to be secured by moving the lever against the cage 
 and permitting him to secure the food. In the succeeding 
 trials he immediately attacked the lever and qbtained the 
 food in the same manner as the other animals. Monkey 9 
 was in the same cage as monkey 8 when the latter wasbeing 
 tested and may have taken the opportunity to observe 
 monkey 8. That she performed the trick may have been 
 due to imitation, but I have credited her with having per¬ 
 formed it in the same way as the other animals. It is reason¬ 
 able to suppose that monkey 4 would have been able to 
 
 Fig. 3. Lever apparatus for test of adaptive intelligence C, leather hinges; B., 
 lever; D, food. 
 
 manipulate the apparatus without being shown if sufficient 
 time had been allowed. However, in whichever way it be 
 conceived that the animal performed the trick we have a 
 mental something very much like imitation or like adaptive 
 intelligence. In many of the experiments the animals did 
 
 TABLE IX. 
 
 Adaptive intelligence, lever test. Time in seconds; f — failed. 
 
 TRIALS AND 
 
 ANIMALS 
 
 I 
 
 2 
 
 3 
 
 I 
 
 9 
 
 4 
 
 3 
 
 2 
 
 6 
 
 4 
 
 3 
 
 3 
 
 21 
 
 2 
 
 2 
 
 4 
 
 f 
 
 6 
 
 3 
 
REASONING. 
 
 57 
 
 not use both hands for the moving of the lever, but one hand 
 for the lever and the other for securing the food when the 
 upper part of the lever was within reach. Table IX gives 
 the time records of four monkeys in this experiment, those 
 for the other animals were approximately the same as those 
 noted in the table, although no accurate measurement by 
 watch was taken. 
 
 A third experiment was then made. A stick 22 inches 
 long, with I inch extending into the cage and the other out¬ 
 wards from the cage at a right angle, was arranged so that food 
 could be secured by drawing the stick lengthwise into the 
 cage. Ten monkeys were tested in this manner and all 
 immediately appeared to grasp the situation for they pulled 
 the stick and secured the food within three seconds. In 
 the cases when the food was dislodged the animals immedi¬ 
 ately gave up the stick and turned to the food, an indication 
 that the stick had been recognized as a means of obtaining 
 the food. 
 
 During the progess of some experiments I noticed that 
 monkey 7 attempted at one time to pull toward his cage a 
 small tin bucket in which water was usually carried to the 
 animals which had been unintentionally left near the cage. 
 From this hint the following test was made with nine of the 
 animals. One end of a piece of twine, i yard long, was 
 attached to the bucket and the other end was left lying within 
 the cage. The bucket was placed at the length of the twine 
 away from the cage, and in it was placed a piece of food. 
 In all cases the animals seized the twine immediately, drew 
 the bucket toward the cage and seized the piece of .banana. 
 Table X gives the times for monkeys 7 and 8. These times 
 are similar to those of the other animals. 
 
 TABLE x. 
 
 Adaptive intelligence, pulling bucket inwards by means of twine. Time in seconds. 
 
 TRIALS AND 
 ANIMALS 
 
 I 
 
 2 
 
 3 
 
 4 
 
 s 
 
 6 
 
 7 
 
 8 
 
 9 
 
 10 
 
 7 
 
 6 
 
 4 
 
 3 
 
 3 
 
 4 
 
 3 
 
 3 
 
 4 
 
 3 
 
 3 
 
 8 
 
 0 
 
 3 
 
 3 
 
 2 
 
 2 
 
 3 
 
 2 
 
 2 
 
 2 
 
 3 
 
58 
 
 WILLIAM SHEPHERD. 
 
 A variation of the problem was then made. The end of 
 the piece of twine which had been left lying loosely within 
 the cage was tied to the wire netting 6 inches from the end, 
 leaving that amount of twine within the cage. In this test 
 the animal had to reach beyond the wire netting, seize the 
 twine and pull the bucket. Seven animals were tested in 
 this way by an assistant, and the notes are not so full as I 
 could wish, but in brief the results are as follows: 
 
 Monkey i, first trial, 3 sec.; average for 10 trials, 3 sec. 
 
 Monkey 2, first trial, 20 sec.; average for other 9 trials, 3.5 sec. 
 Monkey 3, first trial, 20 sec.; average for other 9 trials, 3.5 sec. 
 Monkey 4, first trial, 5 sec.; average for 10 trials, 3 sec. 
 
 Monkey 6, first trial, 4 sec.; average for 10 trials, 2 sec. 
 
 Monkey 7, first trial 3 sec.; average for 10 trials, 2 sec. 
 
 Monkey 8, first trial, 120 sec.; average for last 5 trials, 5.5 sec. 
 
 In the case of the animals whose actions were noted, each 
 pulled at the short free end of the twine at first, but soon 
 reached beyond the knot and pulled the bucket inwards. 
 This was done by some of the animals, notably monkey 8, 
 for four or five trials. The appreciation of the knot condi¬ 
 tion and the ability to deal with it is plainly shown by the 
 records given above. 
 
 The results of all the experiments described in this sec¬ 
 tion appear to speak for the possession by monkeys of adap¬ 
 tive intelligence, of a form of reasoning. The times for the 
 performance of the various acts are fairly conclusive, but in 
 addition the appearances of the animals, their actions, etc., 
 especially during the first trial in each experiment indicated 
 an understanding of the problem. There was no fumbling 
 with the apparatus, no appearance of learning by trial and 
 error, but there was instant action following apparently 
 instant understanding of the situation. The reactions of 
 the monkeys in the situations provided for them gave diversi¬ 
 fied and relatively abundant indications of the presence of 
 practical ideas such as have been referred to in the section 
 of this paper dealing with ideation (p. 4ofif.). 
 
MEMORY. 
 
 59 
 
 Memory. 
 
 No one questions the fact of the possession by animals of 
 a sort of memory for acts once or often repeated. Memory 
 in the restricted psychological sense, however, has been 
 denied to be present in animals but it is obviously difficult, 
 if not impossible, to determine this. The nature of reten¬ 
 tion and recall are matters that are disputed, but that animals 
 have an organic memory, a physiological sort of memory, 
 is admitted. This memory acts well for practical purposes 
 and may therefore be called practical memory. Whether 
 or not this be merely organic, or whether or not animal 
 memory contains a representative factor must be left for 
 future investigation. 
 
 The monkeys observed in this study showed good ability 
 of recognition and retention, of the practical memory. After 
 the completion of the tests on visual and auditory discrimina¬ 
 tion no further experiments were performed with the appara¬ 
 tus, until many of the later experiments on imitation, ideation, 
 and reasoning were finished. Memory tests were then made 
 of the animals previously used in the discrimination tests 
 with the following results. 
 
 Visual discrimination of pink, yellow and green rice: 
 monkey i took the green ten times, took the pink the first 
 trial but only smelled it, and on the fourth trial took the 
 yellow after the green but dropped it almost immediately; 
 monkeys 2 and 8 took only green and paid no attention to 
 the pink and yellow rice; monkeys 3 and 7 took green each 
 trial, and in the first trial took pink after the green and smelled 
 but did not eat it. Auditory discrimination of noises after 
 thirteen days. Monkeys 4 and 6 made no mistakes in ten 
 trials each. 
 
 Lever test of adaptive intelligence. Monkey 6, after 130 
 days, showed perfect retention in four trials; monkeys 7and 
 8, after 123 days, were also perfect in four trials. 
 
6 o 
 
 WILLIAM SHEPHERD. 
 
 General Summary and Conclusions. 
 
 I 
 
 Monkeys learn to discriminate brightnesses, but take a 
 long time for this when the stimuli have not a direct relation 
 to the incentive for work; only a few trials are needed when 
 the visual qualities are a part of the objects to which they 
 naturally pay attention. 
 
 Colors are discriminated with accuracy and rapidity when 
 the colors are parts of the food (red, pink, yellow and green). 
 
 Three animals- gave clear indications of the discrimination 
 of different degrees of noise, and also learned to discriminate 
 musical tones. 
 
 The habits are formed rapidly if there be the double incen¬ 
 tive of pleasant food as an inducement to a correct response, 
 and of an unpleasant stimulus to check a wrong response. 
 
 From the experiments recorded in this paper it appears 
 that monkeys learn to inhibit recently formed habits of 
 action with facility. 
 
 As far as the evidenc-e goes, in regard to both the formation 
 and the inhibition of habits monkeys are superior to raccoons 
 and far superior to dogs, cats, elephants, otters and other 
 mammals which have been experimented with. 
 
 Monkeys have a practical memory; they appear to show a 
 good degree of retention; the representative function in mem¬ 
 ory is an unknown quantity. 
 
 Of the higher powers of mind the monkey has only rudi¬ 
 ments. He has a something which corresponds in function 
 to ideas of a low order and which serves practical purposes. 
 This something we call, with Hobhouse, practical ideas. 
 
 Two of the monkeys learned by imitation, but six others 
 gave no indication of imitation ability. It may be said that 
 while monkeys may learn by imitation to a limited extent, 
 imitation as involving ideation is a small factor in their 
 ordinary learning process. 
 
 All the tested animals appeared to reach a generalized mode 
 of action in dealing with problems but there seemed to be no 
 evidence of true general notions. They have an adaptive 
 intelligence, a lower form of reason, or a mental state inferior 
 to true reason. 
 
BIBLIOGRAPHY. 
 
 6 l 
 
 Bibliography. 
 
 (1) Aristotle: Natural History of Animals. 
 
 (2) Watson, J. B.: Some Experiments bearing upon Color Vision in 
 Monkeys. Journ. Compar. Neurol, and Psychol., vol. 19, no. i, 1909. 
 
 (3) Watson, J. B.: Psychol. Bull., vol. 6, p. 295, 1909. 
 
 (4) Yerkes, R. M.: The Dancing Mouse. New York, 1907. 
 
 (5) Kalischer, 0 : Eine neue Horpriifungsmethode bei Hunden u. s. 
 w., Sitz. d. k. p. Akad. d. Wiss., vol. 10, Berlin 1907. 
 
 (6) Selionyi, G. P.: Contributions to the Study of the Reactions of the 
 Dog to Auditory Stimuli. St. Petersburg, 1907. 
 
 (7) Morgan, C. L.: Animal Behavior, London, 1900. 
 
 (8) Thorndike, E. L.: Animal Intelligence, Psychol. Rev. Monog. 
 Siippl., vol. 2, no. 4, 1898. 
 
 (9) Thorndike, E. L.: The Mental Life of the Monkeys. Psychol. 
 Rev. Monog. Suppl., vol. 3, no. 5, 1901. 
 
 (10) Watson, J. B.: Imitation in Monkeys. Psychol. Bull., vol. 5, 
 no. 6, 1908. 
 
 (i i) Kinnaman, A. J.: Mental Life of Two Macacus Rhesus Monkeys 
 in Captivity. Amer. Journ. Psychol., vol. 13, pp. 99, 173, 1902. 
 
 (12) Hobhouse, L. T.: Mind in Evolution. London, 1901. 
 
 (13) Berry, C. S.: The Imitative Tendencies of White Rats./owm. 
 Compar. Neurol, and Psychol., vol. 16, p. 333, 1906. 
 
 (14) Haggerty, M. E.: A Study of Imitation in Monkeys. Journ. 
 Compar. Neurol, and Psychol., vol. 19, no. 4, 1909. 
 
 (15) Wundt, W.: Lectures on Human and Animal Psychology. Trans, 
 by J. E. Creighton and E. B. Titchener. London, 1894. 
 
 (16) Cole, L. W.: Concerning the Intelligence of Raccoons. Journ. 
 Compar. Neurol, and Psychol., vol. 17, p. 211, 1907. 
 

 
 
BF21.P96v.12 
 
 A study of sensory control in the rat, 
 Princeton Theological Seminary-Speer Library 
 
 1 1012 00008 4774