[Illustration: DANCING MICE--SNIFFING AND EATING. ] THE ANIMAL BEHAVIOR SERIES. VOLUME I THE DANCING MOUSE A Study in Animal Behavior BY ROBERT M. YERKES, Ph. D. INSTRUCTOR IN COMPARATIVE PSYCHOLOGY IN HARVARDUNIVERSITY The Cartwright Prize of the Alumni Association of the College ofPhysicians and Surgeons, Columbia University, was awarded, in 1907, for anEssay which comprised the first twelve chapters of this volume. 1907 IN LOVE AND GRATITUDETHIS BOOK IS DEDICATED TOMY MOTHER PREFACE This book is the direct result of what, at the time of its occurrence, seemed to be an unimportant incident in the course of my scientific work--the presentation of a pair of dancing mice to the Harvard PsychologicalLaboratory. My interest in the peculiarities of behavior which thecreatures exhibited, as I watched them casually from day to day, soonbecame experiment-impelling, and almost before I realized it, I was in themidst of an investigation of their senses and intelligence. The longer I observed and experimented with them, the more numerous becamethe problems which the dancers presented to me for solution. From a studyof the senses of hearing and sight I was led to investigate, in turn, thevarious forms of activity of which the mice are capable; the ways in whichthey learn to react adaptively to new or novel situations; the facilitywith which they acquire habits; the duration of habits; the roles of thevarious senses in the acquisition and performance of certain habitualacts; the efficiency of different methods of training; and the inheritanceof racial and individually acquired forms of behavior. In the course of my experimental work I discovered, much to my surprise, that no accurate and detailed account of this curiously interesting animalexisted in the English language, and that in no other language were allthe facts concerning it available in a single book. This fact, inconnection with my appreciation of the exceptional value of the dancer asa pet and as material for the scientific study of animal behavior, has ledme to supplement the results of my own observation by presenting in thislittle book a brief and not too highly technical description of thegeneral characteristics and history of the dancer. The purposes which I have had in mind as I planned and wrote the book arethree: first, to present directly, clearly, and briefly the results of myinvestigation; second, to give as complete an account of the dancing mouseas a thorough study of the literature on the animal and long-continuedobservation on my own part should make possible; third, to provide asupplementary text-book on mammalian behavior and on methods of studyinganimal behavior for use in connection with courses in ComparativePsychology, Comparative Physiology, and Animal Behavior. It is my conviction that the scientific study of animal behavior and ofanimal mind can be furthered more just at present by intensive specialinvestigations than by extensive general books. Methods of research inthis field are few and surprisingly crude, for the majority ofinvestigators have been more deeply interested in getting results than inperfecting methods. In writing this account of the dancing mouse I haveattempted to lay as much stress upon the development of my methods of workas upon the results which the methods yielded. In fact, I have used thedancer as a means of exhibiting a variety of methods by which the behaviorand intelligence of animals may be studied. As it happens the dancer is anideal subject for the experimental study of many of the problems of animalbehavior. It is small, easily cared for, readily tamed, harmless, incessantly active, and it lends itself satisfactorily to a large numberof experimental situations. For laboratory courses in ComparativePsychology or Comparative Physiology it well might hold the place whichthe frog now holds in courses in Comparative Anatomy. Gratefully, and with this expression of my thanks, I acknowledge myindebtedness to Professor Hugo Münsterberg for placing at my command theresources of the Harvard Psychological Laboratory and for advice andencouragement throughout my investigation; to Professor Edwin B. Holt forvaluable assistance in more ways than I can mention; to Professor WallaceC. Sabine for generous aid in connection with the experiments on hearing;to Professor Theobald Smith for the examination of pathological dancers;to Miss Mary C. Dickerson for the photographs of dancing mice which arereproduced in the frontispiece; to Mr. Frank Ashmore for additionalphotographs which I have been unable to use in this volume; to Mr. C. H. Toll for the drawings for Figures 14 and 20; to Doctors H. W. Rand and C. S. Berry for valuable suggestions on the basis of a critical reading ofthe proof sheets; and to my wife, Ada Watterson Yerkes, for constant aidthroughout the experimental work and in the preparation of this volume. R. M. Y. CAMBRIDGE, MASSACHUSETTS, August, 1907. CONTENTS LIST OF ILLUSTRATIONS LITERATURE ON THE DANCING MOUSE CHAPTER I CHARACTERISTICS, ORIGIN, AND HISTORY Peculiarities of the dancing mouse--Markings and method of keeping recordof individuals--The dancer in China and Japan (Kishi, Mitsukuri, Hatai)--Theories concerning the origin of the race: selectional breeding; theinheritance of an acquired character; mutation, inheritance, andselectional breeding; pathological changes; natural selection--Instancesof the occurrence of dancers among other kinds of mice--Results ofcrossing dancer with other kinds of mice. CHAPTER II FEEDING, BREEDING, AND DEVELOPMENT OF THE YOUNG Methods of keeping and caring for dancers--Cages, nest-boxes, andmaterials for nest--Cleansing cages--Food supply and feeding--Importanceof cleanliness, warmth, and pure food--Relations of males and females, fighting--The young, number in a litter--Care of young--Course ofdevelopment--Comparison of young of dancer with young of common mouse--Diary account of the course of development of a typical litter of dancers. CHAPTER III BEHAVIOR: DANCE MOVEMENTS Dancing--Restlessness and excitability--Significance of restlessness--Forms of dance: whirling, circling, and figure-eights--Direction ofwhirling and circling: right whirlers, left whirlers, and mixed whirlers--Sex differences in dancing--Time and periodicity of dancing--Influence oflight on activity--Necessity for prolonged observation of behavior. CHAPTER IV BEHAVIOR: EQUILIBRATION AND DIZZINESS Muscular coordination--Statements of Cyon and Zoth concerning behavior--Control of movements, orientation, equilibration, movement on inclinedsurfaces, climbing--The tracks of the dancer--Absence of visualdizziness--Comparison of the behavior of the dancer with that of thecommon mouse when they are rotated in a cyclostat--Behavior of blindeddancers (Cyon, Alexander and Kreidl, Kishi)--Cyon's two types of dancer--Phenomena of behavior for which structural bases are sought: dancemovements; lack of response to sounds; deficiency in equilibrationalability; lack of visual and rotational dizziness. CHAPTER V STRUCTURAL PECULIARITIES AND BEHAVIOR The functions of the ear--Structure of the ear of the dancer as describedby Rawitz, by Panse, by Baginsky, by Alexander and Kreidl, and by Kishi--Cyon's theory of the relation of the semicircular canals to spaceperception--Condition of the auditory organs--Condition of theequilibrational organs--Condition of the sound-transmitting organs--Thebearing of the results of anatomical investigations upon the facts ofbehavior. CHAPTER VI THE SENSE OF HEARING Experiments on hearing in the dancer made by Rawitz, by Panse, by Cyon, byAlexander and Kreidl, by Zoth, and by Kishi--Hearing and the voice--Methods of testing sensitiveness to sounds--Results of tests with adults--Importance of indirect method of experimentation--Results of tests withyoung--The period of auditory sensitiveness--Individual differences. CHAPTER VII THE SENSE OF SIGHT: BRIGHTNESS VISION What is known concerning sight in the dancer--Brightness vision and colorvision--Methods of testing brightness vision, the visual discriminationapparatus--Motives for discrimination and choice--Punishment versus rewardas an incentive in animal experiments--Hunger as an incentive--An electricstimulus as an incentive--Conditions for brightness vision tests--White-black vision--Evidence of preference--Check experiments--Conclusion. CHAPTER VIII THE SENSE OF SIGHT: BRIGHTNESS VISION (_Continued_) The delicacy of brightness discrimination--Methods of testing the dancer'sability to detect slight differences in brightness--Results of tests withgray papers--Relation of intensity of visual stimuli to the threshold ofdiscrimination--Weber's law apparatus and method of experimentation--Results of Weber's law tests--Practice effects, the training of vision--Description of the behavior of the dancer in the discrimination boxexperiments--Modes of choice: by affirmation; by negation; by comparison--Evidence of indiscriminable visual conditions. CHAPTER IX THE SENSE OF SIGHT: COLOR VISION Does the dancer see colors?--The food-box method of testing color vision--Waugh's food-box method--Results of tests--Tests by the use of coloredpapers in the visual discrimination box--Yellow-red vision--Blue-orangevision--Brightness vision _versus_ color vision--Brightness checktests--Green-blue vision--Violet-red vision--Conclusions. CHAPTER X THE SENSE OF SIGHT: COLOR VISION (_Continued_) The use of color filters--Testing color vision by the use of transmittedlight--Green-blue vision--Green-red vision--Blue-red vision--Stimulatingvalue of different portions of the spectrum--Does red appear darker to thedancer than to us?--Conclusions concerning color vision--Structure of theretina of the dancer and its significance. CHAPTER XI THE ROLE OF SIGHT IN THE DAILY LIFE OF THE DANCER. Sight and general behavior--Behavior of blinded dancers--Experimentaltests of ability to perceive form--Visual guidance in mazes--Followinglabyrinth paths in the dark--The relative importance of visual, olfactory, and kinaesthetic stimuli--Conditions for the acquisition of a motorhabit--Conditions for the execution of an habitual act. CHAPTER XII EDUCABILITY: METHODS OF LEARNING The modifiability of behavior--Educational value of experimental studiesof modifiability--Methods: the problem method; the labyrinth method; thediscrimination method--Relation of method to characteristics of animal--Simple test of the docility of the dancer--Lack of imitative tendency--Persistence of useless acts--Manner of profiting by experience--Individualdifferences in initiative. CHAPTER XIII HABIT FORMATION: THE LABYRINTH HABIT The labyrinth method--Problems--Preliminary tests--Comparison of thebehavior of the dancer in a maze with that of the common mouse--Evolutionof a labyrinth method--Records of time and records of errors--Simple andeffective method of recording the path--Curves of habit formation--Regularand irregular labyrinths--Points for a standard labyrinth--Values anddefects of the labyrinth method. CHAPTER XIV HABIT FORMATION: THE DISCRIMINATION METHOD Quantitative _versus_ qualitative results--Motives--Precautions--Preference--Results of systematic habit-forming experiments--Curves ofhabit formation--Meaning of irregularity in curve--Individualdifferences--Comparison of curves for discrimination habits with those forlabyrinth habits--Averages--The index of modifiability as a measure ofdocility--Reliability of the index. CHAPTER XV THE EFFICIENCY OF TRAINING METHODS Importance of measuring the efficiency of educational methods--Rapidity oflearning and permanency of modifications wrought by training--Results of astudy of the efficiency of discrimination methods--Comparison by means ofindices of modifiability--Number of tests per series versus number ofseries--Efficiency as measured by memory tests. CHAPTER XVI THE DURATION OF HABITS: MEMORY AND RE-LEARNING Measures of the permanency of modifications in behavior--The duration ofbrightness and color discrimination habits--The relation of learning tore-learning--Can a habit which has been lost completely be re-acquiredwith greater facility than it was originally acquired?--Relation ofspecial training to general efficiency--Does the training in one form oflabyrinth aid the dancer in acquiring other labyrinth habits? CHAPTER XVII INDIVIDUAL, AGE, AND SEX DIFFERENCES IN BEHAVIOR Individual peculiarities in sensitiveness, docility, and initiative--Therelation of docility to age--The individual result and the average--Howaverages conceal facts--Sex differences in docility and initiative--Individual differences of motor capacity which seem to indicatevarieties--Is the dancer pathological? CHAPTER XVIII THE INHERITANCE OF FORMS OF BEHAVIOR Characteristics of the race--Inheritance of the tendency to whirl in aparticular way--Tests of the inheritance of individually acquired forms ofbehavior. INDEX ILLUSTRATIONS Dancing Mice--sniffing and eating _Frontispiece_ FIGURE 1. Color patterns of dancers. Record blanks 2. Double cage, with nest-boxes and water dishes 3. Double cages in frame 4. Photographs of dancers climbing (After Zoth) 5. Tracks of common mouse (After Alexander and Kreidl) 6. Tracks of dancer (After Alexander and Kreidl) 7. The inner ear of the rabbit (Retzius) 8. The membranous labyrinth of the ear of the dancer (After Rawitz) 9. Same 10. Same it. Model of the ear of the dancer (After Baginsky) 12. Ear of the dancer (After Kishi) 13. Ear of the dancer (After Kishi) 14. Discrimination box 15. Ground plan of discrimination box 16. Nendel's gray papers 17. Weber's law apparatus 18. Food-box apparatus 19. Waugh's food-box apparatus 20. Color discrimination apparatus 21. Ground plan of color discrimination apparatus 22. Cards for form discrimination 23. Labyrinth B 24. Labyrinth B on electric wires 25. Labyrinth A 26. Curves of habit formation for labyrinth B 27. Plan of labyrinth C, and path records 28. Labyrinth D 29. Curve of learning for white-black discrimination, twenty individuals 30. Curve of learning for white-black discrimination, thirty individuals 31. Curve of habit formation for labyrinth D 32. Curves of learning and re-learning 33. Plasticity curves LITERATURE ON THE DANCING MOUSE 1. ALEXANDER, G. UND KREIDL, A. "Zur Physiologie des Labyrinths derTanzmaus. " _Archiv für die gesammte Physiologie_, Bd. 82: 541-552. 1900. 2. ALEXANDER, G. UND KREIDL, A. "Anatomisch-physiologische Studien überdas Ohrlabyrinth der Tanzmaus. " II Mittheilung. _Archiv für die gesammtePhysiologie_. Bd. 88: 509-563. 1902. 3. ALEXANDER, G. UND KREIDL, A. "Anatomisch-physiologische Studien überdas Ohrlabyrinth der Tanzmaus. " III Mittheilung. _Archiv für diegesammte Physiologie_, Bd. 88: 564-574. 1902. 4. BAGINSKY, B. "Zur Frage über die Zahl der Bogengänge bei japanischenTanzmäusen. " _Centralblatt für Physiologie_, Bd. 16: 2-4. 1902. 5. BATESON, W. "The present state of knowledge of colour-heredity in miceand rats. " _Proceedings of the Zoölogical Society of London_, Vol. 2:71-99. 1903. 6. BREHM, A. E. "Tierleben. " Dritte Auflage. Saugetiere, Bd. 2: 513-514. 1890. 7. BREHM, A. E. "Life of Animals. " Translated from the third Germanedition of the "Tierleben" by G. R. Schmidtlein. Mammalia, p. 338. Marquis, Chicago. 1895. 8. CYON, E. DE. "Le sens de l'espace chez les souris dansantesjaponaises. " _Cinquantenaire de la Société de Biologie_ (Volumejubilaire). P. 544-546. Paris. 1899. 9. CYON, E. VON. "Ohrlabyrinth, Raumsinn und Orientirung. " _Archiv fürdie gesammte Physiologie_, Bd. 79: 211-302. 1900. 10. CYON, E. DE. "Presentation de souris dansantes japonaises. " _Comptesrendus du XIII Congrès International de Paris, Section dephysiologie_, p. 160-161. 1900. 11. CYON, E. VON. "Beiträge zur Physiologie des Raumsinns. " I Theil. "NeueBeobachtungen an den japanischen Tanzmäusen. " _Archiv für die gesammtePhysiologie_, Bd. 89: 427-453. 1902. 12. CYON, E. DE. "Le sens de l'espace. " Richet's "Dictionnaire dephysiologie, " T. 5: 570-571. 1901. 13. DARBISHIRE, A. D. Note on the results of crossing Japanese waltzingmice with European albino races. _Biometrica_, Vol. 2: 101-104. 1902. 14. DARBISHIRE, A. D. Second report on the result of crossing Japanesewaltzing mice with European albino races. _Biometrica_, Vol. 2:165-173. 1903. 15. DARBISHIRE, A. D. Third report on hybrids between waltzing mice andalbino races. _Biometrica_, Vol. 2: 282-285. 1903. 16. DARBISHIRE, A. D. On the result of crossing Japanese waltzing withalbino mice. _Biometrica_, Vol 3: 1-51. 1904. 17. GUAITA, G. V. "Versuche mit Kreuzungen von verschiedenen Rassen derHausmaus. " _Berichte der naturforschenden Gesellschaft zu Freiburg i. B_. , Bd. 10: 317-332. 1898. 18. GUAITA, G. V. "Zweite Mitteilung uber Versuche mit Kreuzungen vonverschiedenen Hausmausrassen. " _Berichte der naturforschendenGesellschaft zu Freiburg i. B_. , Bd. 11: 131-138. 1900. 19. HAACKE, W. "Ueber Wesen, Ursachen und Vererbung von Albinismus undScheckung und über deren Bedeutung für vererbungstheoretische undentwicklungsmechanische Fragen. " _Biologisches Centralblatt_, Bd. 15: 44-78. 1895. 19a. HUNTER, M. S. "A Pair of Waltzing Mice. " _The Century Magazine_, Vol. 73: 889-893. April, 1907. 20. KAMMERER, P. "Tanzende Waldmaus und radschlagende Hausmaus. "_Zoölogische Garten_, Bd. 41: 389-390. 1900. 21. KISHI, K. "Das Gehörorgan der sogenannten Tanzmaus. " _Zeitschriftfür wissenschaftliche Zoölogie_, Bd. 71: 457-485. 1902. 22. LANDOIS, H. "Chinesische Tanzmäuse. " _Jahresbericht des WestfähschenProvinzial-Vereins_, Munster, 1893-1894: 62-64. 22a. LOSE, J. "Waltzing Mice. " _Country Life in America_, September, 1904. P. 447. 23. PANSE, R. Zu Herrn Bernhard Rawitz' Arbeit: "Das Gehörorgan derjapanischen Tanzmäuse. " _Archiv für Anatomie und Physiologie_, Physiologische Abtheilung, 1901: 139-140. 24. PANSE, R. "Das Gleichgewichts- und Gehörorgan der japanischenTanzmäuse. " _Münchener medicinische Wochenschrift_, Jahrgang 48, Bd. I: 498-499. 1901. 25. RAWITZ, B. "Das Gehörorgan der japanischen Tanzmäuse. " _Archiv fürAnatomie und Physiologie_, Physiologische Abtheilung, 1899: 236-243. 26. RAWITZ, B. "Neue Beobachtungen über das Gehörorgan japanischerTanzmäuse. " _Archiv für Anatomie und Physiologie_, PhysiologischeAbtheilung, 1901, Supplement: 171-176. 27. RAWITZ, B. "Zur Frage über die Zahl der Bogengänge bei japanischenTanzmäusen. " _Centralblatt für Physiologie_, Bd. 15: 649-651. 1902. 28. SAINT-LOUP, R. "Sur le mouvement de manège chez les souris. "_Bulletin de la Société Zoölogique de France_, T. 18: 85-88. 1893. 29. SCHLUMBERGER, C. "A propos d'un netzukè japonais. " _Memoires de laSociété Zoölogique de France_, T. 7: 63-64. 1894. 30. WELDON, W. F. R. Mr. Bateson's revisions of Mendel's theory ofheredity. _Biometrica_, Vol. 2: 286-298. 1903. 31. ZOTH, O. "Ein Beitrag zu den Beobachtungen und Versuchen anjapanischen Tanzmäusen. " _Archiv für die gesammte Physiologie_, Bd. 86: 147-176. 1901. 32. ANONYMOUS. "Fancy Mice: Their Varieties, Management, and Breeding. "Fourth edition. London: L. Upcott Gill. No date. CHAPTER I CHARACTERISTICS, ORIGIN, AND HISTORY The variety of mouse which is known as the Japanese dancing or waltzingmouse has been of special interest to biologists and to lovers of petsbecause of its curious movements. Haacke in Brehm's "Life of Animals" (7p. 337)[1] writes as follows concerning certain mice which were brought toEurope from China and Japan: "From time to time a Hamburg dealer inanimals sends me two breeds of common mice, which he calls Chineseclimbing mice (Chinesische Klettermäuse) and Japanese dancing mice(Japanische Tanzmäuse). It is true that the first are distinguished onlyby their different colors, for their climbing accomplishments are notgreater than those of other mice. The color, however, is subject to manyvariations. Besides individuals of uniform gray, light yellow, and whitecolor, I have had specimens mottled with gray and white, and blue andwhite. Tricolored mice seem to be very rare. It is a known fact that wealso have white, black, and yellow mice and occasionally pied ones, andthe Chinese have profited by these variations of the common mouse also, tosatisfy their fancy in breeding animals. The Japanese, however, who are noless enthusiastic on this point, know how to transform the common mouseinto a really admirable animal. The Japanese dancing mice, which perfectlyjustify their appellation, also occur in all the described colors. Butwhat distinguishes them most is their innate habit of running around, describing greater or smaller circles or more frequently whirling aroundon the same spot with incredible rapidity. Sometimes two or, more rarely, three mice join in such a dance, which usually begins at dusk and is atintervals resumed during the night, but it is usually executed by a singleindividual. " [Footnote 1: The reference numbers, of which 7 is an example, refer to thenumbers in the bibliographic list which precedes this chapter. ] As a rule the dancing mouse is considerably smaller than the common mouse, and observers agree that there are also certain characteristicpeculiarities in the shape of the head. One of the earliest accounts ofthe animal which I have found, that of Landois (22 p. 62), states, however, that the peculiarities of external form are not remarkable. Landois further remarks, with reason, that the name dancing mouse is illchosen, since the human dance movement is rather a rhythmic hopping motionthan regular movement in a circle. As he suggests, they might moreappropriately be called "circus course mice" (22 p. 63). Since 1903 I have had under observation constantly from two to one hundreddancing mice. The original pair was presented to the Harvard PsychologicalLaboratory by Doctor A. G. Cleghorn of Cambridge. I have obtainedspecimens, all strikingly alike in markings, size, and general behavior, from animal dealers in Washington, Philadelphia, and Boston. Almost all ofthe dancers which I have had, and they now number about four hundred, werewhite with patches, streaks, or spots of black. The black markingsoccurred most frequently on the neck, ears, face, thighs, hind legs, aboutthe root of the tail, and occasionally on the tail itself. In only oneinstance were the ears white, and that in the case of one of the offspringof a male which was distinguished from most of his fellows by thepossession of one white ear. I have had a few individuals whose markingswere white and gray instead of white and black. The method by which I was able to keep an accurate record of each of mydancers for purposes of identification and reference is illustrated inFigure 1. As this method has proved very convenient and satisfactory, Imay briefly describe it. With a rubber stamp[1] a rough outline of amouse, like that of Figure 1 A, was made in my record book. On thisoutline I then indicated the black markings of the individual to bedescribed. Beside this drawing of the animal I recorded its number, sex, [2] date of birth, parentage, and history. B, C, and D of Figure 1represent typical color patterns. D indicates the markings of anindividual whose ears were almost entirely white. The pattern varies somuch from individual to individual that I have had no trouble whatever inidentifying my mice by means of such records as these. [Footnote 1: For the use of the plate from which this stamp was made, I amindebted to Professor W. E. Castle, who in turn makes acknowledgment toDoctor G. M. Allen for the original drawing. ] [Footnote 2: I have found it convenient to use the even numbers for themales and the odd numbers for the females. Throughout this book this usageis followed. Wherever the sex of an individual is not specially given, thereader therefore may infer that it is a male if the number is even; afemale if the number is odd. ] All of my dancers had black eyes and were smaller as well as weaker thanthe albino mouse and the gray house mouse. The weakness indicated by theirinability to hold up their own weight or to cling to an object curiouslyenough does not manifest itself in their dancing; in this they areindefatigable. Frequently they run in circles or whirl about withastonishing rapidity for several minutes at a time. Zoth (31 p. 173), whomeasured the strength of the dancer in comparison with that of the commonmouse, found that it can hold up only about 2. 8 times its own weight, whereas the common white mouse can hold up 4. 4 times its weight. No otheraccurate measurements of the strength, endurance, or hardiness of thedancer are available. They are usually supposed to be weak and delicate, but my own observations cause me to regard them as exceptionally strong incertain respects and weak in others. [Illustration: FIGURE I. --Typical markings of dancers. A, blank outline ofmouse for record. B, markings of No. 2 [symbol for male], born September7, 1905, of unknown parents, died March 30, 1907. C, markings of No 43[symbol for female], born November 10, 1906, of 212 and 211. D, markings of No. 151 [symbol for female], born February 28, 1906, of 1000and 5, died February 26, 1907. ] What the Japanese have to say about the dancing mouse is of specialimportance because Japan is rather commonly supposed to be its home. Forthis reason, as well as because of the peculiar interest of the factsmentioned, I quote at length from Doctor Kishi (21 p. 457). "The dancingmouse has received in Europe this name which it does not bear in its ownhome, because of the fact that the circular movements which it makes aresimilar to the European (human) dance. Sometimes it is also called theJapanese or Chinese mouse; originally, however, China must have been itshome, since in Japan it is mostly called '_Nankin nesumi_, ' the mouse fromNankin. When this animal came from China to Japan I shall inquire at alater opportunity. There were originally in Japan two different species ofmouse, the gray and the white; therefore in order to distinguish ourdancing mouse from these it was necessary to use the name of its nativecity. "In Japan, as in Europe, the animal lives as a house animal in smallcages, but the interest which is taken in it there is shown in quiteanother way than in Europe, where the whirling movements, to which thename dancing mouse is due, are of chief interest. For this reason inEurope it is given as much room as possible in its cage that it may danceconveniently. In Japan also the circular movements have been known for along time, but this has had no influence upon our interest in the animal, for the human fashion of dancing with us is quite different from that inEurope. What has lent interest to the creature for us are its prettiness, its cleverness in tricks, and its activity. It is liked, therefore, as anamusement for children. For this purpose it is kept in a small cage, usually fifteen centimeters square, sometimes in a somewhat broader woodenbox one of whose walls is of wire netting. In this box are built usually atower, a tunnel, a bridge, and a wheel. The wheel is rather broad, beingmade in the form of a drum and pierced with holes on one side throughwhich the animal can slip in and out. Running around on the inside, themouse moves the wheel often for hours at a time, especially in theevening. Moreover, there are found in the box other arrangements ofdifferent kinds which may be set in motion by the turning of the wheel. Nospace remains in the box in which the animal may move about freely, andtherefore one does not easily or often have an opportunity to observe thatthe animal makes circular movements, whether voluntarily or involuntarily. This is the reason that in its home this interesting little animal hasnever been studied by any one in this respect. " It is odd indeed that the remarkable capacity of the dancer for theexecution of quick, graceful, dextrous, bizarre, and oft-repeatedmovements has not been utilized in America as it has in Japan. The miceare inexhaustible sources of amusement as well as invaluable material forstudies in animal behavior and intelligence. Concerning the origin and history of this curious variety of mouse littleis definitely known. I have found no mention of the animal in scientificliterature previous to 1890. The fact that it is called the Chinesedancing mouse, the Japanese dancing mouse, and the Japanese waltzing mouseis indicative of the existing uncertainty concerning the origin of therace. Thinking that Japanese literature might furnish more information bearingon the question of racial history than was available from Europeansources, I wrote to Professor Mitsukuri of the University of Tokyo, askinghim whether any reliable records of the dancer existed in Japan. Hereplied as follows: "I have tried to find what is known in Japan about thehistory of the Japanese waltzing mice, but I am sorry to say that theresults are wholly negative. I cannot find any account of the origin ofthis freak, either authentic or fictitious, and, strange as it may seem toyou, no study of the mice in a modern sense has been made, so you mayconsider the literature on the mouse in the Japanese language asabsolutely _nil_. " In explanation of this somewhat surprising ignorance ofthe origin of the race in what is commonly supposed to be its native land, Professor Mitsukuri adds: "The breeders of the mice have mostly beenignorant men to whom writing is anything but easy. " In response to similar inquiries, I received the following letter, confirmatory of Professor Mitsukuri's statements, from Doctor S. Hatai ofWistar Institute, Philadelphia: "If I remember rightly the so-calledJapanese dancing mouse is usually called by us _Nankin-nedzumi_. _Nankin_means anything which has been imported from China, and _nedzumi_ meansrat-like animal, or in this case mouse, or Chinese mouse. I referred toone of the standard Japanese dictionaries and found the followingstatement: 'The _Nankin-nedzumi_ is one of the varieties of _Musspiciosus_ (_Hatszuka-nedzumi_), and is variously colored. It was importedfrom China. These mice are kept in cages for the amusement of children, who watch their play. ' _Mus spiciosus_, if I remember correctly, is verymuch like _Mus musculus_ in color, size, and several othercharacteristics, if not the same altogether. " In Swinhoe's list of the mammals of China, which appeared in the_Proceedings of the Zoological Society of London_ for 1870, _Mus musculusL_. Is mentioned as occurring in houses in South China and in Formosa. Itis further stated that black and white varieties which are brought fromthe Straits are often kept by the Chinese (p. 637). The statements of Kishi, Mitsukuri, and Hatai which have been quoted, taken in connection with the opinions expressed by various Europeanscientists who have studied the dancer, make it seem highly probable thatthe race appeared first in China, and was thence introduced into Japan, from which country it has been brought to Europe and America. Acceptingfor the present this conclusion with reference to the place of origin ofthe dancer, we may now inquire, how and when did this curious freak, asProfessor Mitsukuri has called it, come into existence? Concerning thesematters there is wide divergence of opinion. Haacke (6 p. 514), as quoted in Brehm's "Tierleben, " says that an animaldealer with whom he discussed the question of the possible origin of thedancer maintained that it came from Peru, where it nests in the fullcotton capsules, arranging the cotton fibers in the form of a nest byrunning about among them in small circles. Hence the name cotton mouse issometimes applied to it. Haacke himself believes, however, that the raceoriginated either in China or Japan as the result of systematicselectional breeding. Of this he has no certainty, for he states that hefailed to find any literature on the "beautiful mice of China and Japan. "Whether Haacke's description of the dancing mouse was published elsewhereprevious to its appearance in Brehm's "Tierleben" I am unable to state; Ihave found nothing written on the subject by him before 1890. Zoth (31 p. 176) also thinks that the race was developed by systematic breeding, or inother words, that it is a product of the skill of the Asiatic animalbreeders. Another account of the origin of the race is that accepted by Kishi (21 p. 481) and some other Japanese biologists. It is their belief that the formsof movement acquired by the individual as the result of confinement innarrow cages are inherited. Thus centuries of subjection to the conditionswhich Kishi has described (p. 6) finally resulted in a race of mice whichbreed true to the dance movement. It is only fair to add, although Kishidoes not emphasize the fact, that in all probability those individuals inwhich the dancing tendency was most pronounced would naturally be selectedby the breeders who kept these animals as pets, and thus it would comeabout that selectional breeding would supplement the inheritance of anacquired character. Few indeed will be willing to accept this explanationof the origin of the dancer so long as the inheritance of acquiredcharacters remains, as at present, unproved. Still another mode of origin of the mice is suggested by the followingfacts. In 1893 Saint Loup (28 p. 85) advanced the opinion that dancingindividuals appear from time to time among races of common mice. Thepeculiarity of movement may be due, he thinks, to an accidental nervousdefect which possibly might be transmissible to the offspring of theexceptional individual. Saint Loup for several months had underobservation a litter of common mice whose quick, jerky, nervous movementsof the head, continuous activity, and rapid whirling closely resembled thecharacteristic movements of the true dancers of China. He states thatthese mice ran around in circles of from 1 to 20 cm. In diameter. Theyturned in either direction, but more frequently to the left, that is, anticlockwise. At intervals they ran in figure-eights ([Symbol: figureeight]) as do the true dancers. According to Saint Loup these exceptionalindividuals were healthy, active, tame, and not markedly different ingeneral intelligence from the ordinary mouse. One of these mice produced alitter of seven young, in which, however, none of the peculiarities ofbehavior of the parents appeared. In view of this proof of the occurrence of dancing individuals amongcommon mice, Saint Loup believes that the race of dancers has resultedfrom the inheritance and accentuation of an "accidental" deviation fromthe usual mode of behavior. It is scarcely necessary to say that thisopinion would be of far greater weight had he observed, instead ofpostulating, the inheritance of the peculiarities of movement which he hasdescribed. It might be objected, to the first of his so-called facts, thatthe litter resulted from the mating of mice which possessed dancer blood. Until the occurrence of dancers among varieties of mice which are known tobe unmixed with true dancers is established, and further, until theinheritance of this peculiar deviation from the normal is proved, SaintLoup's account of the origin of the dancing mouse race must be regarded asan hypothesis. The occurrence of dancing individuals among common mice has been recordedby several other observers. Kammerer (20 p. 389) reports that he found alitter of young wood mice (_Mus sylvaticus L_. ) which behaved much as dothe spotted dancers of China. He also observed, among a lot of truedancers, a gray individual which, instead of spinning around after themanner of the race, turned somersaults at frequent intervals. It isKammerer's opinion, as a result of these observations, that the black andwhite dancers of China and Japan have been produced by selectionalbreeding on the basis of this occasional tendency to move in circles. Among albino mice Rawitz (25 p. 238) has found individuals which whirledabout rapidly in small circles. He states, however, that they lacked therestlessness of the Chinese dancers. Some shrews (_Sorex vulgaris L_. )which exhibited whirling movements and in certain other respects resembledthe dancing mouse were studied for a time by Professor Häcker of Freiburgin Baden, according to a report by von Guaita (17 p. 317, footnote). Doctor G. M. Allen of Cambridge has reported to me that he noticed among alarge number of mice kept by him for the investigation of problems ofheredity[1] individuals which ran in circles; and Miss Abbie Lathrop ofGranby, Massachusetts, who has raised thousands of mice for the market, has written me of the appearance of an individual, in a race which shefeels confident possessed no dancer blood, which whirled and ran about insmall circles much as do the true dancers. [Footnote 1: Allen, G. M. "The Heredity of Coat Color in Mice. " Proc. Amer. Academy, Vol. 40, 59-163, 1904. ] Although it is possible that some of these cases of the unexpectedappearance of individuals with certain of the dancer's peculiarities ofbehavior may have been due to the presence of dancer blood in the parents, it is not at all probable that this is true of all of them. We may, therefore, accept the statement that dancing individuals now and thenappear in various races of mice. They are usually spoken of as freaks, and, because of their inability to thrive under the conditions of life ofthe race in which they happen to appear, they soon perish. Another and a strikingly different notion of the origin of the race ofdancers from those already mentioned is that of Cyon (11 p. 443) whoargues that it is not a natural variety of mouse, as one might at firstsuppose it to be, but instead a pathological variation. The pathologicalnature of the animals is indicated, he points out, by the exceptionallyhigh degree of variability of certain portions of the body. According tothis view the dancing is due to certain pathological structural conditionswhich are inherited. Cyon's belief raises the interesting question, arethe mice normal or abnormal, healthy or pathological? That the questioncannot be answered with certainty off-hand will be apparent after we haveconsidered the facts of structure and function which this volume presents. Everything organic sooner or later is accounted for, in some one's mind, by the action of natural selection. The dancing mouse is no exception, forLandois (22 p. 62) thinks that it is the product of natural selection andheredity, favored, possibly, by selectional breeding in China. He furthermaintains that the Chinese dancer is a variety of _Mus musculus L. _ inwhich certain peculiarities of behavior appear because of bilateraldefects in the brain. This author is not alone in his belief that thebrain of the dancer is defective, but so far as I have been able todiscover he is the only scientist who has had the temerity to appeal tonatural selection as an explanation of the origin of the race. Milne-Edwards, as quoted by Schlumberger (29 p. 63), is of the opinionthat the Chinese dancer is not a natural wild mouse race, but instead theproduct of rigid artificial selection. And in connection with thisstatement Schlumberger describes a discovery of his own which seems tohave some bearing upon the problem of origin. In an old Japanese woodcarving which came into his possession he found a group of dancing mice. The artist had represented in minute detail the characteristics of themembers of the group, which consisted of the parents and eight young. Thefather and mother as well as four of the little mice are represented aswhite spotted with black. Of the four remaining young mice, two areentirely black and two entirely white. The two pure white individuals havepink eyes, as has also the mother. The eyes of all the others are black. From these facts Schlumberger infers that the dancer has resulted from thecrossing of a race of black mice with a race of albinos; the two originaltypes appear among the offspring in the carving. Experimental studies of the inheritance of the tendency to dance are ofinterest in their bearing upon the question of origin. Such studies havebeen made by Haacke (19), von Guaita (17, 18), and Darbishire (13, 14, 15, 16), and the important results of their investigations have been wellsummarized by Bateson (5). By crossing dancing mice with common white mice both Haacke and von Guaitaobtained gray or black mice which are very similar to the wild house mousein general appearance and behavior. The characteristic movements of thedancers do not appear. As the result of a long series of breedingexperiments, Darbishire (16 pp. 26, 27) says: "When the race of waltzingmice is crossed with albino mice which do not waltz, the waltzing habitdisappears in the resulting young, so that waltzing is completelyrecessive in Mendel's sense; the eye-color of the hybrids is always dark;the coat-color is variable, generally a mixture of wild-gray and white, the character of the coat being distinctly correlated with characterstransmitted both by the albino and by the colored parent. " When hybridsproduced by the cross described by Darbishire are paired, they producedancers in the proportion of about one to five. Bateson (5 p. 93, footnote), in discussing the results obtained by Haacke, von Guaita, and Darbishire, writes: "As regards the waltzing character, von Guaita's experiments agree with Darbishire's in showing that it wasalways recessive to the normal. No individual in F1 [thus the first hybridgeneration is designated] or in families produced by crossing F1 with thepure normal, waltzed. In Darbishire's experiments F1 x F1 [first hybridsmated] gave 8 waltzers in 37 offspring, indicating 1 in 4 as the probableaverage. From von Guaita's matings in the form DR x DR the totals offamilies were 117 normal and 21 waltzers.... There is therefore a largeexcess of normals over the expected 3 to 1. This is possibly due to thedelicacy of the waltzers, which are certainly much more difficult to rearthan normals are. The small number in von Guaita's litters makes it verylikely that many were lost before such a character as this could bedetermined. " Bateson does not hazard a guess at the origin of the dancer, but merelyremarks (5 p. 86) that the exact physiological basis of the dancingcharacter is uncertain and the origin of this curious variation inbehavior still more obscure. "Mouse fanciers have assured me, " hecontinues, "that something like it may appear in strains inbred from thenormal type, though I cannot find an indubitable case. Such an occurrencemay be nothing but the appearance of a rare recessive form. Certainly itis not a necessary consequence of inbreeding, witness von Guaita's longseries of inbred albinos. " (von Guaita (17 p. 319) inbred for twenty-eightgenerations. ) From the foregoing survey of the available sources of informationconcerning the origin and history of the race of dancing mice thefollowing important facts appear. There are four theories of the origin ofthe race: (1) origin by selectional breeding (Haacke, Zoth, Milne-Edwards); (2) origin through the inheritance of an acquired character(Kishi); (3) origin by mutation, inheritance, and selectional breeding(Saint Loup, Kammerer, Cyon); (4) origin by natural selection, andinheritance, favored by selectional breeding (Landois). Everythingindicates that the race originated in China. It is fairly certain thatindividuals with a tendency to move in circles appear at rare intervals inraces of common mice. It seems highly probable, in view of these facts, that the Chinese took advantage of a deviation from the usual form ofbehavior to develop by means of careful and patient selectional breeding arace of mice which is remarkable for its dancing. Even if it should beproved that the mutation as it appears among common mice is not inherited, the view that slight deviations were taken advantage of by the breederswould still be tenable. The dancing tendency is such in nature as to unfitan individual for the usual conditions of mouse existence, hence, in allprobability human care alone could have produced and preserved the race ofdancers. In answer to the question, how and when did the race of dancers originate, it may be said that historical research indicates that a structuralvariation or mutation which occasionally appears in _Mus musculus_, andcauses those peculiarities of movement which are known as dancing, hasbeen preserved and accentuated through selectional breeding by the Chineseand Japanese, until finally a distinct race of mice which breeds true tothe dance character has been established. The age of the race is notdefinitely known, but it is supposed to have existed for severalcenturies. CHAPTER II FEEDING, BREEDING, AND DEVELOPMENT OF THE YOUNG In this chapter I shall report, for the benefit of those who may wish toknow how to take care of dancing mice, my experience in keeping andbreeding the animals, and my observations concerning the development ofthe young. It is commonly stated that the dancer is extremely delicate, subject to diseases to an unusual degree and difficult to breed. I havenot found this to be true. At first I failed to get them to breed, butthis was due, as I discovered later, to the lack of proper food. For threeyears my mice have bred frequently and reared almost all of their young. During one year, after I had learned how to care for the animals, when themaximum number under observation at any time was fifty and the totalnumber for the year about one hundred, I lost two by disease and one by anaccident. I very much doubt whether I could have done better with anyspecies of mouse. There can be no doubt, however, that the dancer isdelicate and demands more careful attention than do most mice. In March, 1907, I lost almost all of my dancers from what appeared to be anintestinal trouble, but with this exception I have had remarkably goodluck in breeding and rearing them. My dancers usually were kept in the type of cage of which Figure 2 is aphotograph. [1] Four of these double cages, 70 cm. Long, 45 cm. Wide, and10 cm. Deep in front, were supported by a frame as is shown in Figure 3. The fact that the covers of these cages cannot be left open is ofpractical importance. A similar type of cage, which I have used to someextent, consists of a wooden box 30 by 30 cm. By 15 cm. Deep, without anybottom, and with a hinged cover made in part of 1 cm. Mesh wire netting. Such a cage may be placed upon a piece of tin or board, or simply on anewspaper spread out on a table. The advantage of the loose bottom is thatthe box may be lifted off at any time, and the bottom thoroughly cleansed. I have had this type of cage constructed in blocks of four so that asingle bottom and cover sufficed for the block. If the mice are being keptfor show or for the observation of their movements, at least one side ofthe cages should be of wire netting, and, as Kishi suggests, such objectsas a wheel, a tower, a tunnel, a bridge, and a turntable, if placed in thecage, will give the animals excellent opportunity to exhibit theircapacity for varied forms of activity. [Footnote 1: This cage was devised by Professors W. E. Castle and E. L. Mark, and has been used in the Zoological Laboratories of HarvardUniversity for several years. ] [Illustration: FIGURE 2. --Double cage, with nest boxes and water dishes. ] The floors of the cages were covered with a thin layer of sawdust for thesake of cleanliness, and in one corner of each cage a nest box of somesort was placed. During the warm months I found it convenient andsatisfactory to use berry boxes, such as appear in Figure 2, with a smallentrance hole cut in one side; and during the cold months cigar boxes, with an entrance hole not more than 5 cm. In diameter at one end. In thenest box a quantity of tissue paper, torn into fragments, furnishedmaterial for a nest in which the adults could make themselves comfortableor the female care for her young. Cotton should never be used in the nestboxes, for the mice are likely to get it wound about their legs withserious results. Apparently they are quite unable to free themselves fromsuch an incumbrance, and their spinning motion soon winds the threads sotightly that the circulation of the blood is stopped. [Illustration: FIGURE 3. --Double cages in frame. ] The cages and nest boxes were emptied and thoroughly cleaned once a weekwith an emulsion made by heating together one part of kerosene and onepart of water containing a little soap. This served to destroy whateverodor the cages had acquired and to prevent vermin from infesting thenests. In hot weather far greater cleanliness is necessary for the welfareof the mice than in cold weather. The animals attend faithfully to theirown toilets, and usually keep themselves scrupulously clean. For water and food dishes I have used heavy watch glasses[1] 5 cm. Indiameter and 1/2 cm. Deep. They are convenient because they are durable, easily cleaned, and not large enough for the young mice to drown in whenthey happen to spin into one which contains water. It is said that mice donot need water, but as the dancers seem very fond of a little, I have madeit a rule to wash the watch glasses thoroughly and fill them with purefresh water daily. The food, when moist, may be placed in the cages in thesame kind of watch glass. [Footnote 1: Minot watch glasses. ] There is no need of feeding the animals oftener than once a day, and asthey eat mostly in the evening and during the night, it is desirable thatthe food should be placed in the cage late in the afternoon. For almost ayear I kept a pair of dancers on "force"[1] and water. They seemedperfectly healthy and were active during the whole time, but they producedno young. If the animals are kept as pets, and breeding is not desired, adiet of "force, " "egg-o-see, "[1] and crackers, with some bird-seed everyfew days, is likely to prove satisfactory. As with other animals, avariety of food is beneficial, but it appears to be quite unnecessary. Toomuch rich food should not be given, and the mice should be permitted todictate their own diet by revealing their preferences. They eatsurprisingly little for the amount of their activity. I have had excellentsuccess in breeding the mice by feeding them a mixture of dry bread-crumbs, "force, " and sweet, clean oats slightly moistened with milk. Thefood should never be made soppy. A little milk added thus to the foodevery other day greatly increases fertility. About once a week a smallquantity of some green food, lettuce for example, should be given. It iswell, I have found, to vary the diet by replacing the bread and "force" atintervals with crackers and seeds. Usually I give the food dry every otherday, except in the case of mice which are nursing litters. One person towhom I suggested that lettuce was good for the dancers lost four, apparently because of too much of what the mice seemed to consider a goodthing. This suggests that it should be used sparingly. [Footnote 1: A cereal food. ] Success in keeping and breeding dancing mice depends upon three things:cleanliness, warmth, and food supply. The temperature should be fairlyconstant, between 60° and 70° Fahr. They cannot stand exposure to cold orlack of food. If one obtains good healthy, fertile individuals, keeps themin perfectly clean cages with soft nesting materials, maintains atemperature of not far above or below 65°, and regularly supplies themwith pure water and food which they like, there is not likely to betrouble either in keeping or breeding these delicate little creatures. Several persons who have reported to me difficulty in rearing the young orin keeping the adults for long periods have been unable to maintain asufficiently high or constant temperature, or have given them food whichcaused intestinal trouble. The males are likely to fight if kept together, and they may even kill oneanother. A male may be kept with one or more females, or several femalesmay be kept together, for the females rarely, in my experience, fight, andthe males seldom harm the females. Unless the male is removed from thecage in which the female is kept before the young are born, he is likelyto kill the newborn animals. When a female is seen to be building a nestin preparation for a litter, it is best to place her in a cage by herselfso that she may not be disturbed. The sex of individuals may be determined easily in most cases, at the ageof 10 to 12 days, by the appearance of teats in the case of females. The period of gestation is from 18 to 21 days. The maximum number born bymy dancers in any single litter was 9, the minimum number 3. In 25 littersof which I have accurate records, 135 individuals were born, an average of5. 4. The average number of males per litter was precisely the same, 2. 7, as the number of females. On the birth of a litter it is well to see that the female has made a nestfrom which the young are not likely to escape, for at times, if the nestis carelessly made, they get out of it or under some of the pieces ofpaper which are used in its construction, and perish. Several times I haveobserved nests so poorly built that almost all of the young perishedbecause they got too far away to find their way back to the mother. It issurprising that the female should not take more pains to keep her youngsafe by picking them up in her mouth, as does the common mouse, andcarrying them to a place where they can obtain warmth and nourishment. This I have never seen a dancing mouse do. For the first day or two afterthe birth of a litter the female usually remains in the nest box almostconstantly and eats little. About the second day she begins to eatravenously, and for the next three or four weeks she consumes at leasttwice as much food as ordinarily. Alexander and Kreidl (3 p. 567) statethat the female does not dance during the first two weeks after the birthof a litter, but my experience contradicts their statement. There is adecreased amount of activity during this period, and usually the whirlingmovement appears but rarely; but in some cases I have seen vigorous andlong-continued dancing within a few hours after the birth of a litter. There is a wide range of variability in this matter, and the only safestatement, in the light of my observations, is that the mother dances lessthan usual for a few days after a litter is born to her. The development of the young, as I have observed it in the cases of twentylitters, for ten of which (Table I) systematic daily records were kept, may be sketched as follows. At birth the mice have a rosy pink skin whichis devoid of hair and perfectly smooth; they are blind, deaf, andirresponsive to stimulation of the vibrissae on the nose. During the firstweek of post-natal life the members of a litter remain closely huddledtogether in the nest, and no dance movements are exhibited. The motherstays with them most of the time. On the fourth or fifth day colorlesshairs are visible, and by the end of the week the body is covered with acoat which rapidly assumes the characteristic black and white markings ofthe race. For the first few days the hind legs are too weak to support thebody weight, and whatever movements appear are the result of the use ofthe fore legs. As soon as the young mice are able to stand, circlingmovements are exhibited, and by the end of the second week they arepronounced. Somewhere about the tenth day the appearance of the teats inthe case of the females serves to distinguish the sexes plainly. Betweenthe tenth and fifteenth days excitability, as indicated by restless jerkymovements in the presence of a disturbing condition, increases markedly;the auditory meatus opens, and, in the case of some individuals, there aresigns of hearing. On or after the fifteenth day the eyes open and theefforts to escape from the nest box rapidly become more vigorous. Aboutthis time the mother resumes her dancing with customary vigor, and theyoung, when they have opportunity, begin to eat of the food which is givento her. They now dance essentially as do the adults. From the end of thethird week growth continues without noteworthy external changes untilsexual maturity is attained, between the fourth and the sixth week. Forseveral weeks after they are sexually mature the mice continue to increasein size. TABLE I DEVELOPMENT OF THE YOUNG NUMBER JERKY REACT IN HAIR TEATS MOVE- EARS TO EYESPARENTS LITTER VISIBLE VISIBLE MENTS OPEN SOUND OPEN M F APPEAR 152+151 5 0 4th day -- 13th day 14th day 14th day 16th day152+151 1 3 4th day 9th day 10th day 12th day 13th day 15th day410+415 4 1 5th day 11th day 14th day 15th day 15th day 17th day410+415 2 4 5th day 10th day 13th day 14th day 14th day 16th day420+425 0 2 4th day 10th day 12th day 14th day 14th day 16th day210+215 4 1 -- -- 17th day 13th day 17th day 15th day210+215 3 3 5th day 11th day 11th day 14th day No 16th day212+211 1 3 4th day 10th day 15th day 14th day No 15th day220+225 2 4 4th day 10th day 16th day 14th day No 15th day220+225 3 3 4th day 10th day 17th day 13th day No 15th day A course of development very similar to that just described was observedby Alexander and Kreidl (3 p. 565) in three litters of dancing mice whichcontained 3, 5, and 7 individuals respectively. These authors, incomparing the development of the dancer with that of the common mouse, saythat at birth the young in both cases are about 24 mm. In length. Theyoung common mouse grows much more rapidly than the dancer, and by theninth day its length is about 43 mm. As compared with 31 mm. In the caseof the dancer. According to Zoth (31 p. 148) the adult dancer has a bodylength of from 7 to 7. 5 cm. , a length from tip of nose to tip of tail offrom 12 to 13 cm. , and a weight of about 18 grams. The movement of thedancer from the first tends to take the form of circles toward the middleof the nest; that of the common mouse has no definite tendency as todirection. When the common mouse does move in circles, it goes first inone direction, then in the other, and not for any considerable period inone direction as does the true dancer. Neither the young dancer nor thecommon mouse is able to equilibrate itself well for the first few daysafter birth, but the latter can follow a narrow path with far greateraccuracy and steadiness than the former. The uncertain and irregularmovements of the common mouse are due to muscular weakness and toblindness, but the bizarre movements of the young dancer seem to demandsome additional facts as an explanation. A brief account of the development of the dancer given by Zoth (31 p. 149)adds nothing of importance to the description given by Alexander andKreidl. As my own observations disagree with their accounts in certainrespects, I shall now give, in the form of a diary, a description of theimportant changes observed from day to day in a normal litter. The litterwhich I have selected as typical of the course of development in thedancer grew rapidly under favorable conditions. I have observed manylitters which passed through the various stages of development mentionedin this description anywhere from a day to a week later. This was usuallydue to some such obviously unfavorable condition as too little food orslight digestive or bowel troubles. According to the nature of theconditions of growth the eyes of the dancer open anywhere from thefourteenth to the twentieth day. This statement may serve to indicate thedegree of variability as to the time at which a given stage of developmentis reached by different litters. On July 14, 1906, No. 151 (female) and No. 152 (male) were mated, and onAugust 3 a litter of six was born to them. The course of the developmentof this litter during the first three weeks was as follows:-- _First day_ The skin is pink and hairless, several vibrissae are visibleon the nose and lips, but there is no definite response when they aretouched. The mice are both blind and deaf, but they are able to squeakvigorously. The mother was not seen to dance or eat during the day. _Second day_. There is a very noticeable increase in size. The vibrissaeare longer, but touching them still fails to cause a reaction. No hairsare visible on the body. The mother danced rapidly for periods of a minuteseveral times while the record was being made. She ate very little to-day. _Third day_. Scales began to appear on the skin to-day. The animals arerapidly increasing in strength; they can now crawl about the nest easily, but they are too weak to stand, and constantly roll over upon their sidesor backs when they are placed on a smooth surface. Because of theirinability to progress it is impossible to determine with certainty whetherthey have a tendency to move in circles. The mother was seen out of thenest dancing once to-day. She now eats ravenously. _Fourth day_. One of the six young mice was found under a corner of thenest this morning dead, and the others were scattered about the nest box. I gathered them together into a nest which I made out of bits of tissuepaper, and the mother immediately began to suckle them. They are verysensitive to currents of air, but they do not respond to light or soundand seldom to contact with the vibrissae. _Fifth day_. When placed on a smooth surface, they tend to move incircles, frequently rolling over. When placed on their sides or backs, they immediately try to right themselves. They do not walk, for their legsare still too weak to support the weight of the body; instead they dragthemselves about by the use of the fore legs. Fine colorless hairs arevisible over the entire body surface. When the vibrissae are touched, thehead is moved noticeably. The mother dances a great deal and eats abouttwice as much as she did before the birth of the litter. _Sixth day_. Certain regions of the skin, which were slightly darker thanthe remainder on the fourth and fifth days, are now almost black. It isevident that they are the regions in which the black hair is to appear. The movement in circles is much more definite today, although most of theindividuals are still too weak to stand on their feet steadily for morethan a few seconds at a time. Most of their time, when they are firsttaken from the nest, is spent in trying to maintain or regain an uprightposition. The hair is now easily visible, and the skin begins to have awhite appearance as a result. _Seventh day_. Although they are strong enough to move about the nestreadily, none of the young has attempted to leave the nest. They huddletogether in the middle of it for warmth. The epidermal scales, which haveincreased in number since the third day, are dropping off rapidly. Contactwith the vibrissae or with the surface of the body, frequently calls fortha motor reaction but neither light nor sound produces any visible changein behavior. The black and white regions of the skin are sufficientlydefinite now to enable one to distinguish the various individuals by theirmarkings. The mother was seen to dance repeatedly today, and she ate allthe food that was given to her. _Eighth day_. A fold is plainly visible where later the eyelids willseparate. The black pigment in the skin has increased markedly. _Ninth day_. The eyelids are taking form rapidly, but they I have notseparated. The body is covered with a thick coat of hair which is eitherpure white or black. Standing on the four legs is still a difficult task. _Tenth day_. To-day teats are plainly visible in the case of four of thefive individuals of the litter. Up to this time I had thought, fromstructural indications, that there were three males and two females; it isnow evident that there are four females and one male. The external ear, the pinna, is well formed, and has begun to stand out from the head, butno opening to the inner portion of the ear is present. The eyelids appearto be almost fully formed. _Eleventh day_. There are no very noticeable changes in appearance exceptin size, which continues to increase rapidly. They are able to regaintheir normal upright position almost immediately when they happen to rollover. The mother dances as usual. _Twelfth day_. It appears to-day as if the eyes were about to open. Theears are still closed, and there is no evidence of a sense of hearing. They squeaked considerably when in the nest, but not at all when I tookthem out to note their development. The mother stays outside of the nestbox much of the time now, probably to prevent the young ones from suckingcontinuously. _Thirteenth day_. One of the little mice came out of the nest box while Iwas watching the litter this morning, and was able to find his way backdirectly despite the lack of sight. The mice are still dependent upon themother for nourishment. I have not seen any of them attempt to eat thefood which is given to the mother. They are extremely neat and clean. Iwatched one of them wash himself this morning. Each foot was carefullylicked with the tongue. There seems to be special care taken to keep thetoes perfectly clean. _Fourteenth day_. An opening into the ear is visible to-day. When testedwith the Galton whistle, all five responded with quick, jerky movements ofthe head and legs. They evidently hear certain tones. During the past twodays the ears have changed rapidly. In one of the females, which seems tobe a little in advance of the others in development, certain peculiaritiesof behavior appeared to-day. She jumped and squeaked sharply when touchedand sprang out of my hand when I attempted to take her up. This is inmarked contrast with her behavior previously. _Fifteenth day_. The eyes are partly opened. All of the members of thelitter came out of the nest box this morning and ran around the cage, dancing frequently and trying to eat with the mother. Three out of thefive gave auditory reactions on first being stimulated; none of themresponded to repetitions of the stimulus. All appeared to be lesssensitive to sounds than yesterday. The quick, nervous, jerky movementsare very noticeable. _Sixteenth day_. The eyes of all five are fully opened. They dancevigorously and are outside the nest much of the time. _Seventeenth day_. No reactions to sound could be detected to-day. Thesense of sight gives evidence of being well developed. The nervous jumpingmovements persist. _Eighteenth day_. The young mice continue to suck, although they eat ofthe food which is given to the mother. They are now able to take care ofthemselves. _Nineteenth day_. There are no noteworthy changes except increase in sizeand strength. _Twentieth day_. No auditory reactions were obtained today, but otherforms of stimulation brought about unmistakable responses. _Twenty-first day_. They are now about half grown and there is no otherchange of special interest to be recorded. Growth continues for severalweeks. The statement made by Alexander and Kreidl to the effect that thedancer is almost full grown by the thirty-first day of life is false. Atthat age they may be sexually mature, but usually they are far from fullgrown. CHAPTER III BEHAVIOR: DANCE MOVEMENTS The peculiarities of behavior of the dancing mouse are responsible alikefor the widespread interest which it has aroused, and for its name. In alittle book on fancy varieties of mice, in which there is much valuableinformation concerning the care of the animals, one who styles himself "Anold fancier" writes thus of the behavior of the dancer: "I believe mostpeople have an idea that the waltzing is a stately dance executed on thehind feet; this is not so. The performer simply goes round and round onall fours, as fast as possible, the head pointing inwards. The giddywhirl, after continuing for about a dozen turns, is then reversed indirection, and each performance usually occupies from one to two minutes. Whether it is voluntary or not, is difficult to determine, but I aminclined to think the mouse can refrain if it wishes to do so, because Inever see them drop any food they may be eating, and begin to waltz in themidst of their meal. The dance, if such it can be called, generally seizesthe mouse when it first emerges from its darkened sleeping place, and thiswould lead one to suppose that the light conveys an impression of shock tothe brain, through the eyes, which disturbs the diseased centers andstarts the giddy gyrations. The mice can walk or run in a fairly straightline when they wish to do so. " Some of the old fancier's statements aretrue, others are mere guesses. Those who have studied the mice carefullywill doubtless agree that he has not adequately described the variousforms of behavior of which they are capable. I have quoted his descriptionas an illustration of the weakness which is characteristic of most popularaccounts of animal behavior. It proves that it is not sufficient to watchand then describe. The fact is that he who adequately describes thebehavior of any animal watches again and again under natural andexperimental conditions, and by prolonged and patient observation makeshimself so familiar with his subject that it comes to possess anindividuality as distinctive as that of his human companions. To thecasual observer the individuals of a strange race are almostindistinguishable. Similarly, the behavior of all the animals of aparticular species seems the same to all except the observer who hasdevoted himself whole-heartedly to the study of the subject and who hasthus become as familiar with their life of action as most of us are withthat of our fellow-men; for him each individual has its own unmistakablecharacteristics. I shall now describe the behavior of the dancing mouse in the light of theresults of the observation of scores of individuals for months at a time, and of a large number of experiments. From time to time I shall refer topoints in the accounts of the subject previously given by Rawitz (25 p. 236), Cyon (9 p. 214), Alexander and Kreidl (1 p. 542), Zoth (31 p. 147), and Kishi (21 p. 479). The most striking features of the ordinary behavior of the dancer arerestlessness and movements in circles. The true dancer seldom runs in astraight line for more than a few centimeters, although, contrary to thestatements of Rawitz and Cyon, it is able to do so on occasion for longerdistances. Even before it is old enough to escape from the nest it beginsto move in circles and to exhibit the quick, jerky head movements whichare characteristic of the race. At the age of three weeks it is able todance vigorously, and is incessantly active when not washing itself, eating, or sleeping. According to Zoth (31 p. 149) the sense of sight andespecially the sense of smell of the dancer "seem to be keenly developed;one can seldom remain for some time near the cage without one or anotherof the animals growing lively, looking out of the nest, and beginning tosniff around in the air (_windet_). They also seem to have stronglydeveloped cutaneous sensitiveness, and a considerable amount of curiosity, if one may call it such, in common with their cousin, the white mouse. " Ishall reserve what I have to say concerning the sense of sight for laterchapters. As for the sense of smell and the cutaneous sensitiveness, Zothis undoubtedly right in inferring from the behavior of the animal that itis sensitive to certain odors and to changes in temperature. One of themost noticeable and characteristic activities of the dancer is itssniffing. Frequently in the midst of its dancing it stops suddenly, raisesits head so that the nose is pointed upward, as in the case of one of themice of the frontispiece, and remains in that position for a second ortwo, as if sniffing the air. The restlessness, the varied and almost incessant movements, and thepeculiar excitability of the dancer have repeatedly suggested to casualobservers the question, why does it move about in that aimless, uselessfashion? To this query Rawitz has replied that the lack of certain sensescompels the animal to strive through varied movements to use to thegreatest advantage those senses which it does possess. In Rawitz's opinionthe lack of hearing and orientation is compensated for by the continuoususe of sight and smell. The mouse runs about rapidly, moves its head fromside to side, and sniffs the air, in order that it may see and smell asmuch as possible. In support of this interpretation of the restlessness ofthe dancer, Rawitz states that he once observed similar behavior in analbino dog which was deaf. This suggestion is not absurd, for it seemsquite probable that the dancer has to depend for the guidance of itsmovements upon sense data which are relatively unimportant in the commonmouse, and that by its varied and restless movements it does in part makeup for its deficiency in sense equipment. The dancing, waltzing, or circus course movement, as it is variouslyknown, varies in form from moment to moment. Now an individual moves itshead rapidly from side to side, perhaps backing a little at the same time, now it spins around like a top with such speed that head and tail arealmost indistinguishable, now it runs in circles of from 5 cm. To 30 cm. In diameter. If there are any objects in the cage about or through whichit may run, they are sure to direct the expression of activity. A tunnelor a hole in a box calls forth endless repetitions of the act of passingthrough. When two individuals are in the same cage, they frequently dancetogether, sometimes moving in the same direction, sometimes in oppositedirections. Often, as one spins rapidly about a vertical axis, the otherruns around the first in small circles; or again, both may run in a smallcircle in the same direction, so that their bodies form a living ring, which, because of the rapidity of their movements, appears perfectlycontinuous. The three most clearly distinguishable forms of dance are (1)movement in circles with all the feet close together under the body, (2)movement in circles, which vary in diameter from 5 cm. To 30 cm. , with thefeet spread widely, and (3) movement now to the right, now to the left, infigure eights ([Symbol: figure eight]). For convenience of referencethese types of dance may be called _whirling, circling_, and the _figureeight dance_. Zoth, in an excellent account of the behavior of the dancer(31 p. 156), describes "manège movements, " "solo dances, " and "centredances. " Of these the first is whirling, the second one form of circling, and the third the dancing of two individuals together in the mannerdescribed above. Both the whirling and the circling occur to the right (clockwise) and tothe left (anticlockwise). As certain observers have stated that it ischiefly to the left and others that it is as frequently to the right, Ihave attempted to get definite information concerning the matter byobserving a number of individuals systematically and at stated intervals. My study of this subject soon convinced me that a true conception of thefacts cannot be got simply by noting the direction of turning from time totime. I therefore planned and carried out a series of experimentalobservations with twenty dancers, ten of each sex. One at a time theseindividuals were placed in a glass jar, 26 cm. In diameter, and the numberof circle movements executed to the right and to the left during a periodof five minutes was determined as accurately as possible. This wasrepeated at six hours of the day: 9 and 11 o'clock A. M. , and 2, 4, 6, and8 o'clock P. M. In order that habituation to the conditions under which thecounts of turning were made might hot influence the results for the group, with ten individuals the morning counts were made first, and with theothers the afternoon counts. No attempt was made in the counting to keep aseparate record of the whirling and circling, although had it beenpracticable this would have been desirable, for, as soon became evident tothe observer, some individuals which whirl in only one direction, circlein both. In Table 2 the results of the counts for the males are recorded; in Table3 those for the females. Each number in the column headed "right" and"left" indicates the total number of circles executed by a certain dancerin a period of five minutes at the hour of the day named at the head ofthe column. I may point out briefly the curiously interesting and entirelyunexpected new facts which this method of observation revealed to me. First, there are three kinds of dancers: those which whirl almostuniformly toward the right, those which whirl just as uniformly toward theleft, and those which whirl about as frequently in one direction as in theother. To illustrate, No. 2 of Table 2 may be characterized as a "rightwhirler, " for he turned to the right almost uniformly. In the case of the6 P. M. Count, for example, he turned 285 times to the right, not once tothe left. No. 152, on the contrary, should be characterized as a "leftwhirler, " since he almost always turned to the left. From both of theseindividuals No. 210 is distinguished by the fact that he turned now to theleft, now to the right. For him the name "mixed whirler" seemsappropriate. Second, the amount of activity, as indicated by the number of times anindividual turns in a circle within five minutes, increases regularly andrapidly from 9 A. M. To 8 P. M. According to the general averages whichappear at the bottom of Table 2, the average number of circles executed bythe males at 9 A. M. Was 89. 8 as compared with 207. 1 at 8 P. M. In otherwords, the mice dance more in the evening than during the day. Third, as it appears in a comparison of the general averages of Tables 2and 3, the females dance more than the males, under the conditions ofobservation. At 9 A. M. The males circled 89. 8 times, the females 151. 0times; at 8 P. M. The males circled 207. 1 times, the females, 279. 0 times. Fourth, according to the averages for the six counts made with eachindividual, as they appear in Table 4, the males turn somewhat morefrequently to the left than to the right (the difference, however, is notsufficient to be considered significant); whereas, the females turn muchmore frequently to the right than to the left. I do not wish to emphasizethe importance of this difference, for it is not improbable that countsmade with a larger number of animals, or even with another group oftwenty, would yield different results. TABLE 2 NUMBER or WHIRLS TO THE RIGHT AND TO THE LEFT DURINGFIVE-MINUTE INTERVALS AS DETERMINED BY COUNTS MADE ATSIX DIFFERENT HOURS, FOR EACH OF TEN MALE DANCERS NUMBER 9 A. M 11 A. M. 2 P. M. OFANIMAL RIGHT LEFT RIGHT LEFT RIGHT LEFT 2 11 2 23 4 194 1 30 20 1 134 1 109 2 34 2 16 2 48 4 92 36 194 21 180 11 143 65152 7 48 3 171 6 79156 63 8 53 9 27 6210 3 9 7 41 225 21220 168 105 39 43 47 5410 2 67 10 27 8 103420 15 142 5 214 16 238 Averages 48. 5 41. 3 45. 6 56. 9 77. 9 61. 2 Gen. Av. 89. 8 102. 5 139. 1 NUMBER 4 P. M 6 P. M. 8 P. M. OFANIMAL RIGHT LEFT RIGHT LEFT RIGHT LEFT 2 70 3 285 0 237 10 30 154 0 107 6 134 5 34 7 158 5 118 6 147 36 173 14 170 11 325 19152 0 91 16 210 9 223156 85 2 72 26 139 26210 159 18 31 82 47 201220 45 38 78 17 69 33410 9 155 9 394 24 94420 18 243 16 291 3 320 Averages 72. 0 72. 2 78. 9 115. 5 99. 3 107. 8 Gen. Av. 144. 2 194. 4 207. 1 TABLE 3 NUMBER OF WHIRLS TO THE RIGHT AND TO THE LEFT DURING FIVE-MINUTE INTERVALS AS DETERMINED BY COUNTS MADE AT SIX DIFFERENT HOURS, FOR EACH OF TEN FEMALE DANCERS NUMBER 9 A. M. 11 A. M. 2 P. M. OFANIMAL RIGHT LEFT RIGHT LEFT RIGHT LEFT 29 9 18 17 30 7 22 33 287 0 329 1 352 3 35 48 15 198 46 208 14151 13 88 7 75 3 167157 57 6 50 45 53 12211 218 21 31 55 66 5215 67 216 33 105 37 226225 46 39 72 49 143 44415 23 0 156 0 34 3425 43 296 12 201 12 210 Averages 81. 1 69. 9 90. 5 60. 7 91. 5 70. 6 Gen. Av. 151. 0 151. 2 162. 1 NUMBER 4 P. M. 6 P. M. 8 P. M. OFANIMAL RIGHT LEFT RIGHT LEFT RIGHT LEFT 29 33 114 31 36 45 99 33 436 7 408 3 364 2 35 279 6 165 24 353 10151 3 8 2 285 2 217157 52 15 19 125 51 104211 190 7 86 31 67 250215 15 292 45 336 150 232225 133 86 48 39 177 81415 268 3 437 7 382 8425 12 242 19 210 4 192 Averages 142. 1 78. 0 126. 0 109. 6 159. 5 119. 5 Gen. Av. 220. 1 235. 6 279. 0 The most important results of this statistical study of turning are thedemonstration of the existence of individual tendencies to turn in aparticular direction, and of the fact that the whirling increases inamount from morning to evening. In order to discover whether the distribution of the dancers among thethree groups which have been designated as right, left, and mixed whirlersagrees in general with that indicated by Table 4 (approximately the samenumber in each group) I have observed the direction of turning in the caseof one hundred dancers, including those of the foregoing tables, and haveclassified them in accordance with their behavior as is indicated below. RIGHT LEFT MIXED WHIRLERS WHIRLERS WHIRLERS Males 19 19 12Females 12 23 15 Totals 31 42 27 The left whirlers occur in excess of both the right and the mixedwhirlers. This fact, together with the results which have already beenconsidered in connection with the counts of turning, suggests that atendency to whirl in a certain way may be inherited. I have examined mydata and conducted breeding experiments for the purpose of ascertainingwhether this is true. But as the results of this part of the investigationmore properly belong in a special chapter on the inheritance of behavior(XVIII), the discussion of the subject may be closed for the present withthe statement that the preponderance of left whirlers indicated above isdue to a strong tendency to turn to the left which was exhibited by theindividuals of one line of descent. TABLE 4 AVERAGE NUMBER OF WHIRLS TO THE RIGHT AND TO THE LEFT FORTHE SIX INTERVALS OF TABLES 2 AND 3, WITH A CHARACTERIZATIONOF THE ANIMALS AS RIGHT WHIRLERS, LEFT WHIRLERS, ORMIXED WHIRLERS. AVERAGE NO. AVERAGE NO. MALES AGE OF WHIRLS OF WHIRLS CHARACTERIZATION 2 12 mo. 136. 7 3. 3 Right whirler 30 2 mo. 109. 7 2. 5 Right whirler 34 2 mo. 4. 3 96. 5 Left whirler 36 2 mo. 197. 5 23. 5 Right whirler 152 6 mo. 6. 8 137. 0 Left whirler 156 1 mo. 73. 2 12. 8 Right whirler 210 3 mo. 78. 7 62. 0 Mixed whirler 220 4 mo. 74. 3 40. 2 Mixed whirler 410 3 mo. 10. 3 139. 0 Left whirler 420 3 mo. 12. 2 241. 3 Left whirler Average 70. 4 75. 8 4 Right whirlers 4 Left whirlers 2 Mixed whirlers FEMALES 29 2 mo. 23. 7 53. 2 Left whirler 33 2 mo. 362. 7 2. 7 Right whirler 35 2 mo. 208. 5 19. 2 Left whirler 151 6 mo. 5. 0 140. 0 Right whirler 157 1 mo. 47. 0 51. 2 Left whirler 211 3 mo. 109. 7 61. 5 Right whirler 215 3 mo. 57. 8 234. 5 Mixed whirler 225 4 mo. 103. 2 56. 3 Mixed whirler 415 3 mo. 216. 7 3. 5 Left whirler 425 3 mo. 17. 0 225. 2 Left whirler Average 115. 1 84. 7 3 Right whirlers 4 Left whirlers 3 Mixed whirlers The tendency of the dancer's activity to increase in amount towardevening, which the results of Tables 2, 3, and 4 exhibit, demands furtherconsideration. Haacke (7 p. 337) and Kishi (21 p. 458) agree that thedancing is most vigorous in the evening; but Alexander and Kreidl (i p. 544) assert, on the contrary, that the whirling of the individuals whichthey observed bore no definite relation to the time of day and apparentlywas not influenced in intensity thereby. Since the results of my ownobservations contradict many of the statements made by the latter authors, I suspect that they may not have watched their animals long enough todiscover the truth. The systematic records which I have kept indicate thatthe mice remain quietly in their nests during the greater part of the day, unless they are disturbed or come out to obtain food. Toward dusk theyemerge and dance with varying intensity for several hours. I have seldomdiscovered one of them outside the nest between midnight and daylight. Theperiod of greatest activity is between 5 and 10 o'clock P. M. Zoth states that he has observed the adult dancer whirl 79 times withoutan instant's interruption, and I have counted as many as 110 whirls. Itseems rather absurd to say that an animal which can do this is weak. Evidently the dancer is exceptionally strong in certain respects, althoughit may be weak in others. Such general statements as are usually made failto do justice to the facts. The supposition that light determines the periodicity of dancing is notborne out by my observations, for I have found that the animals continueto dance most vigorously toward evening, even when they are kept in a roomwhich is constantly illuminated. In all probability the periodicity ofactivity is an expression of the habits of the mouse race rather than ofthe immediate influence of any environmental condition. At some time inthe history of the dancer light probably did have an influence upon theperiod of activity; but at present, as a result of the persistence of awell-established racial tendency, the periodicity of dancing depends to agreater extent upon internal than upon external conditions. During itshours of quiescence it is possible to arouse the dancer and cause it towhirl more or less vigorously by stimulating it strongly with intenselight, a weak electric current, or by placing two individuals which arestrangers to one another in the same cage; but the dancing thus induced isseldom as rapid, varied, or as long-continued as that which ischaracteristic of the evening hours. One of the most interesting results of this study of the direction ofturning, from the observer's point of view, is the demonstration of thefact that the truth concerning even so simple a matter as this can bediscovered only by long and careful observation. The casual observer ofthe dancer gets an impression that it turns to the left more often than tothe right; he verifies his observation a few times and then asserts withconfidence that such is the truth about turning. That such a method ofgetting knowledge of the behavior of the animal is worse than valueless isclear in the light of the results of the systematic observations whichhave just been reported. But, however important the progress which we mayhave made by means of systematic observation of the phenomenon of turning, it must not for one moment be supposed that the whole truth has beendiscovered. Continued observation will undoubtedly reveal other importantfacts concerning circling, whirling, and the periodicity of dancing, notto mention the inheritance of peculiarities of dancing and thesignificance of the various forms of activity. CHAPTER IV BEHAVIOR: EQUILIBRATION AND DIZZINESS Quite as interesting and important as the general facts of behavior whichwe have been considering are the results of experimental tests of thedancer's ability to maintain its position under unusual spatialconditions--to climb, cross narrow bridges, balance itself on high places. Because of its tendency to circle and whirl, to dart hither and thitherrapidly and apparently without control of its movements, the study of themouse's ability to perform movements which demand accurate and delicatemuscular coördination, and to control its expressions of activity, are ofpeculiar scientific interest. That observers do not entirely agree as to the facts in this field isapparent from the following comparison of the statements made by Cyon andZoth (31 p. 174). Cyon states that the dancer Cannot run in a straight line, Cannot turn in a narrow space, Cannot run backward, Cannot run up an incline, Cannot move about safely when above the ground, because of fear and visual dizziness, Can hear certain tones. Zoth, on the contrary, maintains that the animal Can run in a straight line for at least 20 cm. , Can and repeatedly does turn in a narrow space, Can run backward, for he has observed it do so, Can run up an incline unless the surface is too smooth for it to gain a foothold, Can move about safely when above the ground, and gives no signs of fear or dizziness, Cannot hear, or at least gives no signs of sensitiveness to sounds. Such contradictory statements (and unfortunately they are exceedinglycommon) stimulated me to the repetition of many of the experiments whichhave been made by other investigators to test the dancer's behavior inunusual spatial relations. I shall state very briefly the generalconclusions to which these experiments have led me, with only sufficientreference to methods and details of results to enable any one who wishesto repeat the tests for himself to do so. For the sake of convenience ofpresentation and clearness, the facts have been arranged under threerubrics: equilibrational ability, dizziness, and behavior when blinded. Toour knowledge of each of these three groups of facts importantcontributions have come from the experiments of Cyon (9 p. 220), Alexanderand Kreidl (1 p. 545), Zoth (31 p. 157), and Kishi (21 p. 482), although, as has been stated, in many instances their results are so contradictoryas to demand reexamination. All in all, Zoth has given the mostsatisfactory account of the behavior and motor capacity of the dancer. If the surface upon which it is moving be sufficiently soft or rough tofurnish it a foothold, the dancer is able to run up or down inclines, eventhough they be very steep, to cross narrow bridges, to balance itself atheights of at least 30 cm. Above the ground, and even to climb up and downon rods, as is shown by certain of Zoth's photographs which are reproducedin Figure 4. Zoth himself says, and in this I am able fully to agree withhim on the basis of my own observations, "that the power of equilibrationin the dancing mouse, is, in general, very complete. The seeming reductionwhich appears under certain conditions should be attributed, not to visualdizziness, but in part to excitability and restlessness, and in part to areduced muscular power" (31 p. 161). The dancer certainly has far lessgrasping power than the common mouse, and is therefore at a disadvantagein moving about on sloping surfaces. One evidence of this fact is thecharacter of the tracks made by the animal. Instead of raising its feetfrom the substratum and placing them neatly, as does the common mouse(Figure 5), it tends to shuffle along, dragging its toes and thusproducing on smoked paper such tracks as are seen in Figure 6. From my ownobservations I am confident that these figures exaggerate the differences. My dancers, unless they were greatly excited or moving under conditions ofstress, never dragged their toes as much as is indicated in Figure 6. However, there can be no doubt that they possess less power of graspingwith their toes than do common mice. The animal is still furtherincapacitated for movement on inclined surfaces or narrow places by itstendency to move in circles and zigzags. The results of my own experimentsindicate that the timidity of the adult is greater than that of theimmature animal when it is placed on a bridge 1 or 2 cm. Wide at adistance of 20 cm. From the ground. Individuals three weeks old showedless hesitation about trying to creep along such a narrow pathway than didfull-grown dancers three or four months old; and these, in turn, were notso timid apparently as an individual one year old. But the younger animalsfell off more frequently than did the older ones. [Illustration: FIGURE 4. --Zoth's photographs of dancers crossing bridgesand climbing rods. Reproduced from _Pfluger's Archiv_, Bd. 86. ] [Illustration: FIGURE 5--Tracks of common mouse Reproduced from Alexanderand Kreidl's figure in _Pfluger's Archiv_, Bd 82] [Illustration: FIGURE 6--Tracks of dancing mouse Reproduced from Alexanderand Kreidl's figure in _Pfluger's Archiv_ Bd 82] Additional support for these statements concerning equilibrational abilityis furnished by the observations of Kishi (21 p. 482). He built a woodenbridge 60 cm. Long, 1 cm. Wide at one end, and 1/2 cm. At the other, andsupported it at a height of 30 cm. Above the ground by posts at the ends. On this bridge ten dancers were tested. Some attempted to move sidewise, others began to whirl and fell to the ground; only one of the tensucceeded in getting all the way across the bridge on the first trial. Thesecond time he was tested this individual crossed the bridge and found thepost; and the third time he crossed the bridge and climbed down the postdirectly. The others did not succeed in descending the post even afterhaving crossed the bridge safely, but, instead, finally fell to the floorfrom awkwardness or exhaustion. On the basis of these and other similarobservations, Kishi says that the dancer possesses a fair degree ofability to orient and balance itself. Inasmuch as equilibration occurs similarly in darkness and in daylight, Zoth thinks that there is neither visual dizziness nor fear of heights. But it is doubtful whether he is right concerning fear. There is no doubtin my mind, in view of the way the mice behave when placed on an elevatedsurface, that they are timid; but this is due probably to theuncomfortable and unusual position rather than to perception of theirdistance from the ground. That they lack visual dizziness seems fairlywell established. When rotated in a cyclostat[1] the dancer, unlike the common mouse, doesnot exhibit symptoms of dizziness. The following vivid description of thebehavior of both kinds of mice when rotated is given by Alexander andKreidl (1 p. 548). I have not verified their observations. [Footnote 1: An apparatus consisting of a glass cylinder with a mechanismfor turning it steadily and at different speeds about its vertical axis. ] The common mouse at first runs with increasing rapidity, as the speed ofrotation of the cyclostat cylinder is increased, in the direction oppositeto that of the cylinder itself. This continues until the speed of rotationhas increased to about 60 revolutions per minute. As the rotation becomesstill more rapid the mouse begins to crawl along the floor, its bodystretched out and clinging to the floor. At a speed of 250 revolutions perminute it lies flat on the floor with its limbs extended obliquely to themovement of rotation, and at times with its back bent against the axis ofthe cylinder; in this position it makes but few and feeble efforts tocrawl forward. When the rotation is suddenly stopped, the animal pullsitself together, remains for some seconds with extended limbs lying on thefloor, and then suddenly falls into convulsions and trembles violently. After several attacks of this kind, cramps appear and, despite itsresistance, the animal is thrown about, even into the air at times, as ifby an external force. This picture of the position assumed during rapidrotation, and of cramps after the cessation of rotation (the typicalpicture of rotation dizziness), is repeated with great uniformity in thecase of the common mouse. Within fifteen minutes after being returned toits cage the animal recovers from the effects of its experience. Thisdescription of the symptoms of rotation dizziness in the common mouseapplies equally well to the blinded and the seeing animal. In sharp contrast with the behavior of the common mouse in the cyclostatis that of the dancer. As the cylinder begins to rotate the dancer runsabout as usual in circles, zigzags, and figure-eights. As the speedbecomes greater it naturally becomes increasingly difficult for the mouseto do this, but it shows neither discomfort nor fear, as does the commonmouse. Finally the centrifugal force becomes so great that the animal isthrown against the wall of the cylinder, where it remains quietly withouttaking the oblique position. When the cyclostat is stopped suddenly, itresumes its dance movements as if nothing unusual had occurred. Itexhibits no signs of dizziness, and apparently lacks the exhaustion whichis manifest in the case of other kinds of mice after several repetitionsof the experiment. The behavior of the blinded dancer is very similar. If these statements are true, there is no reason to believe that thedancer is capable of turning or rotation dizziness. If it were, its dailylife would be rendered very uncomfortable thereby, for its whirling wouldconstantly bring about the condition of dizziness. Apparently, then, thedancer differs radically from most mammals in that it lacks visual androtational dizziness. In the next chapter we shall have to seek for thestructural causes for these facts. The behavior of the blinded animal is so important in its bearings uponthe facts of orientation and equilibration that it must be considered inconnection with them. Cyon insists that the sense of vision is of greatimportance to the dancer in orienting and equilibrating itself. When theeyes are covered with cotton wads fastened by collodion, this writerstates (9 p. 223) that the mice behave as do pigeons and frogs whosesemicircular canals have been destroyed. They perform violent forcedmovements, turn somersaults forward and backward, run up inclines and fallover the edges, and roll over and over. In a word, they show precisely thekind of disturbances of behavior which are characteristic of animals whosesemicircular canals are not functioning normally. Cyon, however, observedthat in certain dancers these peculiarities of behavior did not appearwhen they were blinded, but that, instead, the animals gave no otherindication of being inconvenienced by the lack of sight than do commonwhite mice. This matter of individual differences we shall have toconsider more fully later. No other observer agrees with Cyon in his conclusions concerning vision, or, for that matter, in his statements concerning the behavior of theblind dancer. Alexander and Kreidl (1 p. 550) contrast in the followingrespects the behavior of the white mouse and that of the dancer when theyare blinded. The white mouse runs less securely and avoids obstacles lesscertainly when deprived of vision. The dancer is much disturbed at firstby the shock caused by the removal of its eyes, or in case they arecovered, by the presence of the unusual obstruction. It soon recoverssufficiently to become active, but it staggers, swerves often from side toside, and frequently falls over. It moves clumsily and more slowly thanusual. Later these early indications of blindness may wholly disappear, and only a slightly impaired ability to avoid obstacles remains. It was noted by Kishi (21 p. 484), that the dancer when first blindedtrembles violently, jumps about wildly, and rolls over repeatedly, as Cyonhas stated; but Kishi believes that these disturbances of behavior aretemporary effects of the strong stimulation of certain reflex centers inthe nervous system. After having been blinded for only a few minutes thedancers observed by him became fairly normal in their behavior. They movedabout somewhat more slowly than usually, especially when in a positionwhich required accurately coordinated movements. He therefore fully agreeswith Alexander and Kreidl in their conclusion that vision is not soimportant for the guidance of the movements of the dancer as Cyonbelieves. In summing up the results of his investigation of this subject Zoth wellsays (31 p. 168), "the orientation of the positions of the body withrespect to the horizontal and vertical planes seems to take place withoutthe assistance of the sense of sight. " And, as I have already stated, thisexcellent observer insists that the ability of the dancer to place itsbody in a particular position (orientation), and its ability to maintainits normal relations to its surroundings (equilibration) are excellent indarkness and in daylight, provided only the substratum be not too smoothfor it to gain a foothold. It must be admitted that the contradictions which exist in the severalaccounts of the behavior of the dancer are too numerous and too serious tobe explained on the basis of careless observation. Only the assumption ofstriking individual differences among dancers or of the existence of twoor more varieties of the animal suffices to account for the discrepancies. That there are individual or variety differences is rendered practicallycertain by the fact that Cyon himself worked with two groups of dancerswhose peculiarities he has described in detail, both as to structure andbehavior. In the case of the first group, which consisted of three individuals, thesnout was more rounded than in the four individuals of the second group, and there were present on the head three large tufts of bristly black hairwhich gave the mice a very comical appearance. The animals of the secondgroup resembled more closely in appearance the common albino mouse. Theypossessed the same pointed snout and long body, and only the presence ofblack spots on the head and hips rendered them visibly different from thealbino mouse. In behavior the individuals of these two groups differed strikingly. Thoseof the first group danced frequently, violently, and in a variety of ways;they seldom climbed on a vertical surface and when forced to move on anincline they usually descended by sliding down backwards or sidewiseinstead of turning around and coming down head first; they gave no signswhatever of hearing sounds. Those of the second group, on the contrary, danced very moderately and in few ways; they climbed the vertical walls oftheir cage readily and willingly, and when descending from a height theyusually turned around and came down head first; two of the four evidentlyheard certain sounds very well. No wonder that Cyon suggests thepossibility of a different origin! It seems not improbable that theindividuals of the second group were of mixed blood, possibly the resultof crosses with common mice. As I shall hope to make clear in a subsequent discussion of the dancer'speculiarities of behavior, in a chapter on individual differences, thereis no sufficient reason for doubting the general truth of Cyon'sdescription, although there is abundant evidence of his inaccuracy indetails. If, for the present, we accept without further evidence thestatement that there is more than one variety of dancer, we shall be ableto account for many of the apparent inaccuracies of description which areto be found in the literature on the animal. As a result of the examination of the facts which this chapter presents wehave discovered at least six important peculiarities of behavior of thedancer which demand an explanation in terms of structure. These are: (1)the dance movements--whirling, circling, figure-eights, zigzags; (2)restlessness and the quick, jerky movements of the head; (3) lack ofresponsiveness to sounds; (4) more or less pronounced deficiency inorientational and equilibrational power; (5) lack of visual dizziness; (6)lack of rotational dizziness. Naturally enough, biologists from the first appearance of the dancingmouse in Europe have been deeply interested in what we usually speak of asthe causes of these peculiarities of behavior. As a result, the structureof those portions of the body which are supposed to have to do with thecontrol of movement, with the phenomena of dizziness, and with the abilityto respond to sounds, have been studied thoroughly. In the next chapter weshall examine such facts of structure as have been discovered and attemptto correlate them with the facts of behavior. CHAPTER V STRUCTURAL PECULIARITIES AND BEHAVIOR The activities of an animal are expressions of changes which occur in itsstructure, and they can be explained satisfactorily only when the facts ofstructure are known. Such peculiarities of activity as are exhibited bythe dancing mouse, as contrasted with the common mouse, suggest at oncethat this creature has a body which differs in important respects fromthat of the ordinary mouse. In this chapter I shall present what is knownconcerning the structural bases for the whirling, the lack ofequilibrational ability and of dizziness, the quick jerky head movements, the restlessness, and the partial or total deafness of the dancing mouse. Comparative physiologists have discovered that the ability of animals toregulate the position of the body with respect to external objects and torespond to sounds is dependent in large measure upon the groups of senseorgans which collectively are called the ear. Hence, with reason, investigators who sought structural facts with which to explain the formsof behavior characteristic of the dancer turned their attention first ofall to the study of the ear. But the ear of the animal is not, as might besupposed on superficial examination, a perfectly satisfactory naturalexperiment on the functions of this group of sensory structures, for it isextremely uncertain whether any one of the usual functions of the organ istotally lacking. Dizziness may be lacking, and in the adult hearing also, but in general the functional facts lead the investigator to expectmodifications of the sense organs rather than their absence. I shall now give an account of the results of studies concerning thestructure of the ear and brain of the dancer. Since the descriptions givenby different anatomists contradict one another in many important points, the several investigations which have been made may best be consideredchronologically. Bernhard Rawitz (25 p. 239) was the first investigator to describe thestructure of the ear of the Japanese or Chinese dancers, as he calls them. The definite problem which he proposed to himself at the beginning of hisstudy was, what is the structural basis of the whirling movement and ofthe deafness of the mice? In his first paper Rawitz described the form of the ears of five dancers. His method of work was to make microscopic preparations of the ears, andfrom the sections, by the use of the Born method, to reconstruct the earin wax. These wax models were then drawn for the illustration of theauthor's papers (Figures 8, 9, 10). The principal results of the early work of Rawitz are summed up in thefollowing quotation from his paper: "The Japanese dancing mice have onlyone normal canal and that is the anterior vertical. The horizontal andposterior vertical canals are crippled, and frequently they are growntogether. The utriculus is a warped, irregular bag, whose sections havebecome unrecognizable. The utriculus and sacculus are in wide-opencommunication with each other and have almost become one. The utriculusopens broadly into the scala tympani, and the nervous elements of thecochlea are degenerate. "The last-mentioned degeneration explains the deafness of the dancingmice; but in my opinion it is a change of secondary nature. The primarychange is the broad opening between the utriculus and the scala tympanifrom which results the streaming of the endolymph from the semicircularcanals into the cochlea. When, as a consequence of the rapid whirlingmovements, a great part of the endolymph is hurled into the scala tympani, the organ of Corti in the scala vestibuli is fixed and its parts arerendered incapable of vibration. The condition of atrophy which isobservable in the sense cells and in the nerve elements is probably due tothe impossibility of functional activity; it is an atrophy caused bydisuse "(25 p. 242). Ampulla externaAmpulla anteriorRamus utriculi Membrana basilaris Lagena Canalis utriculo-saccularis Membrana basilarisAmpulla posteriorMacula acustica sacculi [Illustration: FIGURE 7. --The inner ear of the rabbit. Reproduced fromSelenka after Retzius. ] To render the terms which occur in this and subsequent descriptions of theear of the dancer somewhat more intelligible to those who are not familiarwith the general anatomy of the vertebrate ear, a side view of the innerear of the rabbit is reproduced from a drawing by Retzius (Figure 7). Ihave chosen the ear of the rabbit for this purpose, not in preference tothat of the common mouse, but simply because I failed to find any reliabledescription of the latter with drawings which could be reproduced. Therabbit's ear, however, is sufficiently like that of the mouse to make itperfectly satisfactory for our present purpose. This drawing of the rabbit's ear represents the three semicircular canals, which occur in the ear of all mammals, and which are called, by reason oftheir positions, the anterior vertical, the posterior vertical, and thehorizontal. Each of these membranous canals possesses at one end, in anenlargement called the ampulla, a group of sense cells. In Figure 7 theampullae of the three canals are marked respectively, ampulla anterior, ampulla posterior, and ampulla externa. This figure shows also thecochlea, marked lagena, in which the organ of hearing of mammals (theorgan of Corti) is located. The ear sac, of which the chief divisions arethe utriculus and the sacculus, with which the canals communicate, is notshown well in this drawing. Within a few months after the publication of Rawitz's first paper on thestructure of the dancer's ear, another European investigator, Panse (23and 24) published a short paper in which he claimed that previous to theappearance of Rawitz's paper he had sectioned and mounted ears of thecommon white mouse and the dancing mouse side by side, and, as the resultof careful comparison, found such slight differences in structure that heconsidered them unworthy of mention. Panse, therefore, directlycontradicts the statements made by Rawitz. In fact, he goes so far as tosay that he found even greater differences between the ears of differentwhite mice than between them and the ears of the dancer (23 p. 140). In a somewhat later paper Panse (24 p. 498) expresses his belief that, since there are no peculiarities in the general form, sensory structures, or nerve supply of the ear of the dancer, which serve to explain thebehavior of the animal, it is probable that there are unusual structuralconditions in the brain, perhaps in the cerebellum, to which are due thedance movements and the deafness. The work of Panse is not veryconvincing, however, for his figures are poor and his descriptions meager;nevertheless, it casts a certain amount of doubt upon the reliability ofthe descriptions given by Rawitz. [Illustration: FIGURE 8. --The membranous labyrinth of the dancer's ear. Type I. This figure, as well as 9 and 10, are reproduced from Rawitz'sfigures in the _Archiv für Anatomie und Physiologie, PhysiologischeAbtheilung_, 1899. _C. S. _, anterior vertical canal; _C. P. _, posteriorvertical canal; _C. E. _, horizontal canal; _U. _, utriculus. ] The unfavorable light in which his report was placed by Panse's statementsled Rawitz to examine additional preparations of the ear of the dancer. Again he used the reconstruction method. The mice whose ears he studiedwere sent to him by the physiologist Cyon. As has been noted in Chapter IV, Cyon discovered certain differences inthe structure and in the behavior of these dancers (11 p. 431), which ledhim to classify them in two groups. The individuals of one group climbedreadily on the vertical walls of their cages and responded vigorously tosounds; those of the other group could not climb at all and gave noevidences of hearing. After he had completed his study of their behavior, Cyon killed the mice and sent their heads to Rawitz; but unfortunatelythose of the two groups became mixed, and Rawitz was unable to distinguishthem. When he examined the structure of the ears of these mice, Rawitz didfind, according to his accounts, two structural types between which verymarked differences existed. Were it not for the carelessness which isindicated by the confusion of the materials, and the influence of Cyon'ssuggestion that there should be different structures to account for thedifferences in behavior, Rawitz's statements might be accepted. As mattersstand there can be no doubt of individual differences in behavior, external appearance, and the structure of the ear; but until these havebeen correlated on the basis of thoroughgoing, careful observation, it isscarcely worth while to discuss their relations. [Illustration: FIGURE 9. --The membranous labyrinth of the dancer's ear. Type II. ] [Illustration: FIGURE 10. --The membranous labyrinth of the dancer's ear. Type III. ] To his previous description of the conditions of the ear sacs, senseorgans, and nerve elements of the dancer's ear, Rawitz adds nothing ofimportance in his second paper (26 p. 171). He merely reiterates hisprevious statements concerning the form of the canals, on the basis of hisfindings in the case of six additional dancers. Figures 8, 9, and 10 arereproduced from Rawitz to show the anatomical conditions which he claimsthat he found. As these figures indicate, the canals were found to beextremely variable, as well as unusual in form, and the sacs distorted. Inthe ears of some specimens there were only two canals, and in all casesthey were more or less reduced in size, distorted, or grown together. [Illustration: FIGURE 11. --Photograph of a wax model of the membranouslabyrinth of the ear of the dancer. Reproduced from Baginsky's figure inthe _Centralblatt für Physiologie_, Bd. 16. ] The work of Rawitz was unfavorably criticised by Alexander and Kreidl (2), Kishi (21), and Baginsky (4), as well as by Panse (23 and 24). To theircriticisms Rawitz replied by insisting that the other investigators couldnot with right attack his statements because they had not used thereconstruction method. In order to test the value of this contention, andif possible settle the question of fact, Baginsky had a model of the earof the dancer constructed by a skilled preparator (Herr Spitz) fromsections which had been prepared by the best neurological methods. Thismodel was made eighty times the size of the ear. It was then reduced inthe process of photographic reproduction to sixteen times the natural sizeof the ear in the mouse. Figure 11 is a photograph of Baginsky's model. Itshows beyond question the presence of three canals of the same generalform and relations as those of the common mouse and of other mammals. Baginsky's paper is brief and to the point. His criticisms of the work ofboth Cyon and Rawitz are severe, but they are justified in all probabilityby the carelessness of these investigators in the fixation of theirmaterials. Of the five skilled histologists who have examined the ear ofthe dancer, Rawitz alone found markedly abnormal canals. It is highlyprobable, therefore, that the canals in his preparations in some waybecame distorted before the ears were sectioned. He doubtless describedaccurately the conditions which he found, but the chances are that thoseconditions never existed in the living animals. The conflicting statements of Rawitz and Panse stimulated interest, and asa result two other investigators, without knowledge of one another's work, began careful researches on the dancer's ear. One, Alexander (2 and 3), worked in coöperation with the physiologist Kreidl; the other, Kishi (21), worked independently. The anatomical papers of Alexander and Kishiappeared at about the same time, and since neither contains a reference tothe other, it is evident that the investigations were carried on almostsimultaneously. Alexander's descriptions are more detailed than those ofRawitz and Panse, and in certain respects Kishi's are even morethoroughgoing. The first paper published by Alexander and Kreidl (1)contains the results of observations on the habits and behavior of thedancers. Having examined the chief facts of function, these investigatorsattempted to discover the structural conditions for the peculiarities ofbehavior which they had observed. As material for their anatomical work they made use of four dancers, onealbino mouse, and four common gray mice. The ears of these individualswere fixed, sectioned, and examined microscopically in connection withparts of the brain. In all, eight dancer ears and six common mouse earswere studied. Very extensive descriptions of these preparations, together withmeasurements of many important portions of the ear, are presented in theirpaper, the chief conclusions of which are the following:-- 1. The semicircular canals, the ampullae, the utriculus, and the cristaeacusticae of the canals are normal in their general form and relations toone another as well as in their histological conditions (2 p. 529). Thisis contradictory of the statements made by Rawitz. 2. There is destruction of the macula sacculi (2 p. 534). 3. There is destruction also of the papilla basilaris cochleae, withencroachment of the surrounding tissues in varying degrees. 4. There is diminution in the number of fibers of the branches and rootsof the ramus superior and ramus medius of the eighth nerve, and the fiberbundles are very loosely bound together. 5. Similarly the number of fibers in the inferior branch (the cochlearnerve) of the eighth nerve is very much reduced. 6. There is moderate reduction in the size of the two vestibular gangliaas a result of the unusually small number of nerve cells. 7. The ganglion spirale is extremely degenerate. There is therefore atrophy of the branches, ganglia, and roots of theentire eighth nerve, together with atrophy and degeneration of the parsinferior labyrinthii. The nerve endings are especially degenerate (2 p. 534). The above structural deviations of the ear of the dancer from that of thecommon mouse may be considered as primary or secondary according as theyare inherited or acquired. Since, according to Alexander and Kreidl, thedancers' peculiarities of behavior and deafness are directly and uniformlyinherited, it is obvious that certain primary structural deviations mustserve as a basis for these functional facts. But it is equally clear, inthe opinion of Alexander and Kreidl (2 p. 536), that other structuralpeculiarities of the dancer are the result of the primary changes, and inno way the conditions for either the dancing or the deafness. Theseauthors feel confident that the facts of behavior which are to beaccounted for are almost certainly due to the pathological changes whichthey have discovered in the nerves, ganglia, and especially in theperipheral nerve endings of the ear of the mouse (2 p. 537). It is further claimed by Alexander and Kreidl that there are very markedindividual differences among the dancers in the structure of the ear. Insome cases the otoliths and the sensory hairs are lacking; in others, theyare present in the state of development in which they are found in othervarieties of mouse. Sometimes the cochlea is much reduced in size; atother times it is found to be of normal size (2 p. 538). These variationsin structure, if they really exist, go far toward justifying the tendencyof Cyon and Alexander and Kreidl, as well as many other investigators, toregard the dancer as abnormal or even pathological. The functions of the ear as at present known to the comparativephysiologist are grouped as the acoustic and the non-acoustic. The cochleais supposed on very good grounds to have to do with the acousticfunctions, and the organs of the semicircular canals on equally goodevidence are thought to have to do with such of the non-acoustic functionsas equilibration and orientation. Just what the functions of the organs ofthe ear sacs are is not certainly known. These facts are of importancewhen we consider the attempts made by Alexander and Kreidl to correlatethe various peculiarities of behavior shown by the dancer with thestructural facts which their work has revealed. This correlation isindicated schematically below. The physiological facts to be accounted forin terms of structure are presented in the first column, and theanatomical facts which are thought to be explanatory, in the second (2 p. 539). FUNCTION 1 Lack of sensitiveness to auditory stimuli. {Structure 1, 2, 3 below} 2 Defective equilibrational ability. {Structure 4, 5, 6 below} 3 Lack of turning dizziness. {Structure 4, 5, 6 below} 4 Normal reactions to galvanic stimulation. (not related in table to anyStructure) STRUCTURE 1 Destruction of the papilla basilaris cochleae, etc. 2 Diminution of the inferior branch of the eighth nerve. 3 Marked degeneration of the ganglion spirale. 4 Destruction of the macula sacculi. 5 Diminution of the branches and roots of the superior and middle branchesof the eighth nerve. 6 Diminution of both ganglia vestibulii and of the nerve cells. Alexander and Kreidl themselves believe that the partial deafness of thedancers (for they admit that the total lack of hearing has not beensatisfactorily proved) is due to the defective condition of the cochlea. They account for the imperfect equilibrational ability of the animals bypointing out the structural peculiarities of the sacculus, the vestibularganglia, and the peripheral nerves. Similarly, the lack of dizziness theysuppose to be due to the diminution of the fibers of the nerves whichsupply the canal organs, the atrophied condition of the vestibularganglia, and a disturbance of the peripheral sense organs. Furthermore, there are no anatomical facts which would indicate a lack of galvanicdizziness (2 p. 552). Despite the fact that they seem to explain all the functionalpeculiarities of the dancer, the statements made by Alexander and Kreidlare neither satisfying nor convincing. Their statements concerning thestructure of the ear have not been verified by other investigators, andtheir correlation of structural with functional facts lacks anexperimental basis. In this connection it may be worth while to mention that a beautifultheory of space perception which Cyon (9) had constructed, largely on thebasis of the demonstration by Rawitz that the dancers have only one normalcanal, is totally destroyed by Panse, Baginsky, Alexander and Kreidl, andKishi, for all of these observers found in the dancer three canals ofnormal shape. Cyon had noted that the most abnormal of the voluntary aswell as of the forced movements of the dancer occur in the plane of thecanal which Rawitz found to be most strikingly defective. This fact heconnected with his observation that the fish Petromyzon, which possessesonly two canals, moves in only two spatial dimensions. The dancer withonly one functional canal in each ear moves in only one plane, and neitherit nor Petromyzon is able to move far in a straight line (11 p. 444). Fromthese and similar surmises, which his eagerness to construct an ingenioustheory led him to accept as facts quite uncritically, Cyon concluded thatthe perception of space depends upon the number and arrangement of thesemicircular canals, and that the dancer behaves as it does because itpossesses canals of unusual shape and relations to one another. Theabsurdity of Cyon's position becomes obvious when it is shown that thestructural conditions of which he was making use do not exist in thedancer. The results obtained by Kishi in his study of the ear of the dancer differin many important respects from those of all other investigators, butespecially from those of Rawitz and Alexander and Kreidl. Kishi's work was evidently done with admirable carefulness. His methods inthe preparation of his materials, so far as can be judged from his report, were safe and satisfactory, and his descriptions of results are minute andgive evidence of accuracy and conscientious thoughtfulness. The materialfor his histological work he obtained from three different animal dealers. It consisted of fifteen adult and nineteen young dancers, and, as materialfor comparison, ten common gray mice. The animals were studied firstbiologically, that their habits and behavior might be described accuratelyand so far as possible accounted for in the light of whatever histologicalresults might be obtained subsequently; then they were studiedphysiologically, that the functional importance of various organs whichwould naturally be supposed to have to do with the peculiarities of themouse might be understood; and, finally, they were killed and their earsand portions of their brains were studied microscopically, that structuralconditions for the biological and physiological facts might be discovered. The ear, which was studied by the use of several series of sections, aswell as in gross dissections, is described by Kishi under threeheadings:-- (1) The sound-receiving apparatus (auditory organs). (2) The static apparatus (equilibrational organs). (3) The sound-transmitting apparatus (ear drum, ear bones, etc. ). The chief results of his structural investigation may be stated brieflyunder these three headings. In the sound-receiving or auditory apparatus, Kishi failed to find the important deviations from the usual structure ofthe mammalian ear which had been described by Rawitz. The latterdistinctly says that although the organ of Corti is present in all of thewhirls of the cochlea, the auditory cells in it are noticeably degenerate. Kishi does not agree with Panse's statement (21 p. 476) that the auditoryorgan of the dancer differs in no important respects from that of thecommon mouse, for he found that in certain regions the hair cells of theorgan of Corti were fewer and smaller in the dancer. He thereforeconcludes that the auditory organ is not entirely normal, but at the sametime he emphasizes the serious discrepancy between his results and thoseof Rawitz. In not one of the ears of the twelve dancers which he studieddid Kishi find the direct communication between the utriculus and thescala tympani which Rawitz described, and such differences as appeared inthe organ of Corti were in the nature of slight deviations rather thanmarked degenerations. In the outer wall of the ductus cochlearis of the dancer the striavasculosa is almost or totally lacking, while in the gray mouse it isprominent. This condition of the stria vasculosa Kishi was the first tonotice in the dancer; Alexander and Kreidl had previously described asimilar condition in an albino cat. If, as has been supposed by somephysiologists, the stria vasculosa is really the source of the endolymph, this state of affairs must have a marked influence on the functions of theauditory apparatus and the static apparatus, for pressure differencesbetween the endolymph and the perilymph spaces must be present. And, asKishi points out, should such pressure differences be proved to exist, thefunctional disturbance in the organ of hearing which the lack of responsesto sounds seems to indicate might better be ascribed to them than to thestreaming of the endolymph from the canals into the cochlea as assumed byRawitz (21 p. 477). Kishi merely suggests that the condition of the striamay account for the deafness of the mouse; he does not feel at allconfident of the truth of his explanation, and he therefore promises inhis first paper a continuation of his work in an investigation of thefunctions of the stria. This, however, he seems not to have accomplishedthus far. [Illustration: FIGURE 12. --The inner ear of the dancer. Reproduced fromKishi's figure in the _Zeitschrift für wissenschaftliche Zoölogie_, Bd. 71. _c. C. _ crus simplex; o. B. Anterior vertical canal; _h. B. _ posteriorvertical canal; _a. B. _ horizontal canal. ] The static apparatus Kishi describes as closely similar in form to that ofthe gray mouse. In none of his twelve preparations of the ear of thedancer did he find such abnormalities of form and connections in thesemicircular canals as Rawitz's figures and descriptions represent. Rawitzstates that the anterior canal is normal except in its lack of connectionwith the posterior and that the posterior and horizontal are much reducedin size. Kishi, on the contrary, insists that all of the three canals arenormal in shape and that the usual connection between the anterior and theposterior canals, the crus simplex, exists. He justifies these statementsby presenting photographs of two dancer ears which he carefully removedfrom the head. Comparison of these photographs (Figures 12 and 13) withRawitz's drawings of the conditions of the canals and sacs as he foundthem (Figures 8, 9, and 10), and of both with the condition in the typicalmammalian ear as shown by Figure 7, will at once make clear the meaning ofKishi's statements. That Rawitz's descriptions of the canals are notcorrect is rendered almost certain by the fact that Panse, Baginsky, Alexander and Kreidl, and Kishi all agree in describing them as normal inform. The only important respects in which Kishi found the membranous labyrinth, that is, the canals and the ear sacs, of the dancer to differ from that ofthe gray mouse are the following. In the dancer the cupola of the cristaacustica is not so plainly marked and not so highly developed, and theraphae of the ampullae and canals, which frequently are clearly visible inthe gray mouse, are lacking (21 p. 478). [Illustration: FIGURE 13. --The inner ear of the dancer, showing the spiralform of the cochlea. After Kishi. ] The sound-transmitting apparatus of the dancer, according to Kishi, differs only very slightly from that of the gray mouse, and there is noreason to consider the differences which appear as important (21 p. 478). Almost as amusing as the way in which Cyon's theory of space perceptiondisappears in the light of critical research is Panse's explanation of thedeafness of the dancer. Failing to find any defects in the auditoryapparatus of the inner ear which seemed adequate to account for theobvious lack of responsiveness to sounds, this investigator concluded thatplugs of wax which he had noticed in the auditory meatus of the dancerexcluded sounds or in some way interfered with the functioning of thetympanic membrane. Kishi reports that he found such plugs of wax in theears of one gray mouse, but in none of the dancers which he examined didhe discover them (21 p. 479). Panse's explanation of the defective hearingof the dancer neither needs nor deserves further comment. As one result of his investigation, Kishi is convinced that the dancemovements are not due to peculiarities in the semicircular canals andtheir sense organs, as Rawitz claimed, for the general form and finerstructure of these organs in the dancer is practically the same as in thecommon mouse. Kishi is just as certain that the whirling is not due todefects in the canal organs, as Rawitz is that it is due to suchstructural conditions! It is rather surprising that any one should feelconfident of the power of the microscope to reveal all those structuralconditions which are important as conditions of function. Probably thereare histological differences between the ear of the dancer and that of thegray mouse, which, although undetectable by scientific means at present, furnish the structural basis for the marked differences in behavior. Ashas been set forth already (p. 9), Kishi accounts for the dance movementsby assuming the inheritance of an acquired character of behavior. Thisinherited tendency to dance, he thinks, has been accentuated by theconfinement of the mice in narrow cages and their long-continued movementin the wheels which are placed in the cages (21 p. 481). Rawitz, Cyon, and Alexander and Kreidl felt themselves under the necessityof finding peculiarities of behavior in the dancer which could be referredto the various abnormalities of structure which they had either seen oraccepted on faith; Kishi found himself in a very different predicament, for he had on his hands the commonly accepted statement that the animalsare deaf, without being able to find any structural basis for this defect. To avoid the difficulty he questions the existence of deafness! Ifperchance they are deaf, he thinks that it is possibly because of thedefect in the stria vasculosa. This suggestion Kishi makes despite thefact that our ignorance of the function of the stria renders it impossiblefor us to do otherwise than guess at its relation to hearing. We have now briefly reviewed the results of the various importantinvestigations of the behavior and structure of the dancer. The observations of Cyon, Zoth, and the writer establish beyond doubt theexistence of important individual differences in behavior if not ofdistinct divisions within the species of mouse, and the general results ofthe several anatomical investigations make it seem highly probable thatthe structure of the ear, as well as the externally visible structuralfeatures of the animals, vary widely. Unfortunately, the lack of agreementin the descriptions of the ear given by the different students of thesubject renders impossible any certain correlation of structural andfunctional facts. That the whirling and the lack of dizziness and ofhearing have their structural bases no one doubts, but whether it is inthe brain itself, in the sense organs, or in the labyrinth, our knowledgedoes not permit us to say. With this statement Rawitz, Cyon, and Alexanderand Kreidl would not agree, for they believe that they have discoveredstructural peculiarities which fully explain the behavior of the dancer. Panse and Kishi, on the other hand, contend that the ear gives nostructural signs of such peculiarities as the dancing and deafnesssuggest; they therefore look to the cerebellum for the seat of thedisturbance. With the same possibility in mind the author of "FancyVarieties of Mice" writes: "These quaint little creatures make amusingpets for any one who is not scientific, or very fond of knowing 'thereason why. ' In their case, the reason of the peculiarity which gives themtheir name is rather a sad one. It is now pretty conclusively establishedthat they are no more Japanese than they are of any other country inparticular, but that the originators of the breed were common fancy micewhich were suffering from a disease of the brain analogous to the 'gid' insheep. In the latter, the complaint is caused by a parasite in the brain;in the case of the Waltzing Mouse, it is probably due to an hereditarymalformation therein. Be this as it may, the breed is now a firmlyestablished one, and the children of waltzing mice waltz like theirparents" (32 p. 45). Although it is quite possible that peculiarities inthe central nervous system, rather than in the peripheral nervous system, may be responsible for the forms of behavior exhibited by the dancer, itmust be remembered that no such peculiarities have been revealed by theexamination of the central nervous system. The old fancier has neitherbetter nor worse grounds for his belief than have Panse and Kishi. So far as the reliability of the anatomical work which has been discussedis in question, it would seem that Rawitz's results are rendered somewhatunsatisfactory by the carelessness of Cyon in fixing the materials; thatPanse's descriptions and comparisons are neither careful nor detailedenough to be convincing; that the work of Alexander and Kreidl, as well asthat of Kishi, gives evidence of accuracy and trustworthiness. The factthat the statements of Alexander and Kreidl frequently do not agree withthose of Kishi proves that there are serious errors in the work of one oranother of these investigators. Cyon's discussion of the anatomy of thedancer is not to be taken too seriously, for by his theory of spaceperception and of a sixth sense he was unduly biased in favor of thestructural peculiarities described by Rawitz. Nevertheless, his discussionis not without interest, for the way in which he succeeded in making everystructural fact which Rawitz suggested fit into his theories and help toaccount for the functional peculiarities which he had himself observed, isextremely clever and indicates a splendid scientific imagination. To sum up: All the facts of behavior and physiology which have beenestablished lead us to expect certain marked structural differencesbetween the dancer and the common mouse. The bizarre movements, lack ofequilibrational ability, and the nervous shaking of the head suggest thepresence of peculiar conditions in the semicircular canals or their senseorgans; and the lack of sensitiveness to sounds indicates defects in thecochlea. Yet, strange as it may seem to those who are not familiar withthe difficulties of the study of the minute structure of these organs, nostructural conditions have been discovered which account satisfactorilyfor the dancer's peculiarities of behavior. That the ear is unusual inform is highly probable, since three of the four investigators who havestudied it carefully agree that it differs more or less markedly from thatof the common mouse. But, on the other hand, the serious lack of agreementin their several descriptions of the conditions which they observedrenders their results utterly inconclusive and extremely unsatisfactory. The status of our knowledge of the structure of the central nervous systemis even less satisfactory, if possible, than that of our knowledge ofthose portions of the peripheral nervous system which would naturally besupposed to have to do with such functional peculiarities as the dancerexhibits. So far as I have been able to learn, no investigator hascarefully examined the brain and spinal cord in comparison with those ofthe common mouse, and only those who have failed to find any structuralbasis for the facts of behavior in the organs of the ear have attempted toaccount for the dancer's whirling and deafness by assuming that thecerebellum is unusual in structure. We are, therefore, forced to concludethat our knowledge of the nervous system of the dancing mouse does not atpresent enable us to explain the behavior of the animal. It seems highly probable to me, in the light of my observation of thedancer and my study of the entire literature concerning the animal, thatno adequate explanation of its activities can be given in terms of thestructure of the peripheral or the central nervous system, or of both, butthat the structure of the entire organism will have to be taken intoaccount. The dancer's physiological characteristics, in fact, suggestmultitudinous structural peculiarities. I have confined my study to itsbehavior, not because the problems of structure seemed less interesting orless important, but simply because I found it necessary thus to limit thefield of research in order to accomplish what I wished within a limitedperiod. That there are structural bases for the forms of behavior which this bookdescribes is as certain as it could be were they definitely known; thatthey, or at least some of them, are discoverable by means of our present-day histological methods is almost as certain. It is, therefore, obviousthat this is an excellent field for further research. It is not anagreeable task to report inconclusive and contradictory results, and Ihave devoted this chapter to a brief account of the work that has beendone by others on the structure of the ear of the dancer rather for thesake of presenting a complete account of the animal as it is known to-daythan because of the value of the facts which could be stated. CHAPTER VI THE SENSE OF HEARING Repeatedly in the foregoing chapters mention has been made of the dancer'sirresponsiveness to sounds, but it has not been definitely stated whetherthis peculiarity of behavior is due to deafness or to the inhibition ofreaction. This chapter is concerned with the evidence which bears upon theproblem of the existence of a sense of hearing. Again I may be permittedto call attention to the observations of other investigators beforepresenting the results of my own experiments and stating the conclusionswhich I have reached through the consideration of all available facts. By the results of various simple tests which he made, Rawitz (25 p. 238)was convinced that the adult dancer is totally deaf. He did not experimentwith the young, but he says he thinks they may be able to hear, since thenecessary structural conditions are present. This guess which Rawitz madeon the basis of very indefinite and uncertain knowledge of the histologyof the ear of the young dancer is of special interest in the light offacts revealed by my own experiments. Unfortunately the study of hearingmade by Rawitz is casual rather than thorough, and although it may turnout that all of his statements are justified by his observations, thereader is not likely to get much satisfaction from his discussion of thesubject. Inasmuch as he could discover no structural basis for deafness, Panse (23p. 140) expressed himself as unwilling to believe that the mice are deaf, and this despite the fact that he observed no responses to the sounds madeby a series of tuning forks ranging from C5 to C8. He believes rather thatthey are strangely irresponsive to sounds and that their sensitiveness isdulled, possibly, by the presence of plugs of wax in the ears. Sinceanother investigator, Kishi, has observed the presence of similar plugs ofwax in the ears of common mice which could hear, there is but slightprobability that Panse is right in considering the plugs of wax as thecause of the dancer's irresponsiveness to sounds. Far more thoroughgoing tests than those of Rawitz or Panse were made byCyon (9 p. 218), who holds the unique position of being the only person onrecord who has observed the adult dancer give definite reactions tosounds. To a König Galton whistle so adjusted that it gave a tone of about7000 complete vibrations per second, which is said to be about the pitchof the voice of the dancer, some of the animals tested by Cyon respondedunmistakably, others not at all. In one group of four mice, two not onlyreacted markedly to the sound of the whistle but apparently listenedintently, for as soon as the whistle was blown they ran to the side of thecage and pressed their noses against the walls as if attempting toapproach the source of the stimulus. The remaining two mice gave not theslightest indication that the sound acted as a stimulus. By the repetitionof this sound from eight to twelve times Cyon states that he was able toarouse the mice from sleep. When thus disturbed, the female came out ofthe nest box before the male. Similarly when the mice were disturbed bythe whistle in the midst of their dancing, the female was first to retreatinto the nest box. There is thus, according to Cyon, some indication ofsex, as well as individual, differences in sensitiveness to the sound ofthe whistle. Cyon's statement that in order to evoke a response thewhistle must be held above the head of the dancer suggests at once thepossibility that currents of air or odors instead of sounds may have beenresponsible for the reactions which he observed. The work of thisinvestigator justifies caution in the acceptance of his statements. Neither the conditions under which the auditory tests were made nor thecondition of the animals is described with sufficient accuracy to makepossible the comparison of Cyon's work with that of other investigators. As will appear later, it is of the utmost importance that the influence ofother stimuli than sound be avoided during the tests and that the age ofthe mouse be known. The conclusion reached by Cyon is that some dancersare able to hear sounds of about the pitch of their own cries. The fact, emphasized by Cyon, that the mice respond to tones of about thepitch of their own voice is of peculiar interest in its relation to theadditional statements made by the same author to the effect that thefemale dancer is more sensitive to sounds than the male, and that themales either do not possess a voice or are much less sensitive todisagreeable stimuli than the females. In the case of the dancers which hefirst studied (9 p. 218), Cyon observed that certain strong stimuli evokedpain cries; but later in his investigation he noticed that fourindividuals, all of which were males, never responded thus to disagreeablestimulation (11 p. 431). He asks, therefore, does this mean that the maleslack a voice or that they are less sensitive than the females? The factthat he did not succeed in getting a definite answer to this simplequestion is indicative of the character of Cyon's work. My dancers haveprovided me with ample evidence concerning the matter. Both the males andthe females, among the dancers which I have studied, possess a voice. The females, especially during periods of sexual excitement, are much morelikely to squeak than the males. At such times they give their shrill crywhenever they are touched by another mouse or by the human hand. A slightpinching of the tail will frequently cause the female to squeak, but themale seldom responds to the same stimulus by crying out. The mostsatisfactory way to demonstrate the existence of a voice in the male is tosubject him to the stimulating effect of an induced current, so weak thatit is barely appreciable to the human hand. To this unexpected stimuluseven the male usually responds by a sudden squeak. There can be no doubt, then, of the possession of a voice by both males and females. The malesmay be either less sensitive or less given to vocal expression, but theyare quite able to squeak when favorable conditions are presented. Thepossession of a voice by an animal is presumptive evidence in favor of asense of hearing, but it would scarcely be safe to say that the mice mustbe able to hear their own voices. Cyon, however, thinks that some dancerscan. What further evidence is to be had? Although they obtained no visible motor reactions to such noises as theclapping of the hands, the snapping of the fingers, or to the tones oftuning forks of different pitches and the shrill tones of the Galtonwhistle, Alexander and Kreidl (1 p. 547) are not convinced of the totaldeafness of the dancer, for, as they remark, common mice which undoubtedlyhear do not invariably respond visibly to sounds. Furthermore, theanatomical conditions revealed by their investigation of the ear of thedancer are not such as to render sensitiveness to sounds impossible. Theyrecognize also that the existence of the ability to produce sounds is anindication of hearing. They have no confidence in the results reported byCyon, for they feel that he did not take adequate precautions to guardagainst the action of other than auditory stimuli. Zoth (31 p. 170) has pointed out with reason and force that testing thesensitiveness of the mice is especially difficult because of theirrestlessness. They are almost constantly executing quick, jerky movements, starting, stopping, or changing the direction of movement, and it istherefore extremely difficult to tell with even a fair degree of certaintywhether a given movement which occurs simultaneously with a sound is aresponse to the sound or merely coincident with it. With great care in theexclusion of the influence of extraneous stimuli, Zoth tried a largenumber of experiments to test the hearing of both young and adult dancers. Not once did he observe an indubitable auditory reaction. As he says, "Ihave performed numerous experiments with the Galton whistle, with asqueaking glass stopper, with caps and cartridges, without being able tocome to any certain conclusion. With reference to the Galton whistle andparticularly to the tone which was said to have been heard extremely wellby Cyon's mice, I believe I am rather safe in asserting that my mice, young (12-13 days) as well as old, do not react to the König Galtonwhistle (7210 Vs. ). They could not be awakened out of sleep by repetitionsof the sound, nor enticed out of their nests, and their dancing could notbe interrupted" (31 p. 170). Zoth's experiments appear to be the mostcareful and critical of those thus far considered. Last to be mentioned, but in many respects of greatest interest and value, is the work of Kishi (21 p. 482) on the problem of hearing. To this acuteobserver belongs the credit of calling attention emphatically to the earmovements which are exhibited by the dancer. Frequently, as he remarks, the ears move as if the animal were listening or trying to determine thedirection whence comes a sound, yet usually the mouse gives no other signof hearing. That the absence of ordinary reactions to sounds is due todeafness, Kishi, like Panse, is led to doubt because his anatomicalstudies have not revealed any defects in the organs of hearing which wouldseem to indicate the lack of this sense. This historical survey of the problem of hearing has brought out a fewimportant facts. No one of the several investigators of the subject, withthe exception of Cyon, is certain that the dancer can hear, and no one ofthem, with the exception of Rawitz, is certain that it cannot hear! Cyonalmost certainly observed two kinds of dancing mice. Those of his dancerswhich exhibited exceptional ability to climb in the vertical direction andwhich also gave good evidence of hearing certain sounds may have beenhybrids resulting from the crossing of the dancer with a common mouse, orthey may have been exceptional specimens of the true dancer variety. Athird possibility is suggested by Rawitz's belief in the ability of theyoung dancer to hear. Cyon's positive results may have been obtained withimmature individuals. I am strongly inclined to believe that Cyon didobserve two types of dancer, and to accept his statement that some of themice could hear, whereas others could not. It is evident, in the light ofour examination of the experimental results thus far obtained by otherinvestigators, that neither the total lack of sensitiveness to sounds inthe adult nor the presence of such sensitiveness in the young dancer hasbeen satisfactorily proved. I shall now report in detail the results of my own study of the sense ofhearing in the dancer. As the behavior of the young differs greatly fromthat of the adult, by which is meant the sexually mature animal, I shallpresent first the results of my experiments with adults and later, incontrast, the results obtained with mice from one to twenty-eight daysold. My preliminary tests were made with noises. While carefully guardingagainst the interference of visual, tactual, temperature, and olfactorystimuli, I produced noises of varying degrees of loudness by clapping thehands together suddenly, by shouting, whistling, exploding pistol caps, striking steel bars, ringing an electric bell, and causing another mouseto squeak. To these sounds a common mouse usually responds either bystarting violently, or by trembling and remaining perfectly quiet for afew seconds, as if frightened. The adult dancers which I have tested, andI have repeated the experiment scores of times during the last three yearswith more than a hundred different individuals, have never givenunmistakable evidence of hearing. Either they are totally deaf or there isa most surprising lack of motor reactions. Precisely the same results were obtained in tests made with the Galtonwhistle throughout its range of pitches, and with Appuun whistles which, according to their markings, ranged from 2000 Vs. (C_4) to 48, 000 (G_9), but which undoubtedly did not correspond at all exactly to this range, andwith a series of König tuning forks which gave tones varying in pitch from1024 to 16, 382 complete vibrations. I am willing to trust these experimental results the more fully becauseduring all the time I have had adult dancers under observation I havenever once seen a reaction which could with any fair degree of certaintybe referred to an auditory stimulus. Never once, although I have triedrepeatedly, have I succeeded in arousing a dancer from sleep by producingnoises or tones, nor have I ever been able to observe any influence ofsounds on the dance movements. All of Cyon's signs have failed with mymice. Occasionally what looked like a response to some sound appeared, butcritical observation invariably proved it to be due to some other causethan the auditory stimulus. A sound produced above the animal is verylikely to bring about a motor reaction, as Cyon claims; but I have alwaysfound it to be the result of the currents of air or odors, which usuallyinfluence the animal when the experimenter is holding any object above it. I do not wish to maintain that Cyon's conclusions are false; I merelyemphasize the necessity for care in the exclusion of other stimuli. Themice are extremely sensitive to changes in temperature, such, for example, as are produced by the breath of the experimenter, and one must constantlyguard against the misinterpretation of behavior. In a single experiment with mice over a month old, I observed what mightpossibly indicate sensitiveness to sound. While holding a mouse, thirty-five days old, in my hand I pursed my lips and made a very shrill sound bydrawing in air; the mouse seemed to start perceptibly according to theindications given by my sense of touch. I repeated the stimulus severaltimes and each time I could see and feel the animal start slightly. Withtwo other individuals which I tested the reaction was less certain, andwith several others I failed to get any indication of response. This wouldseem to prove that the three individuals which responded happened to besensitive to that particular tone at the age of five weeks. The test isunsatisfactory because the vibrations from my own body may have broughtabout the reaction instead of the air vibrations produced by my lips, andI therefore merely mention it in the enumeration of the variousexperimental tests which I have made. If we should conclude from all the negative evidence that is available, orthat could be obtained, that the dancer is totally deaf, it might fairlybe objected that the conclusion is unsafe, since an animal does notnecessarily respond to stimuli by a visible change in the position orrelations of its body. Death feigning may fairly be considered a responseto a stimulus or stimulus complex, yet there may be no sign of movement. The green frog when observed in the laboratory usually gives no indicationwhatever, by movements that are readily observable, that it hears soundswhich occur about it, but I have been able to show by means of indirectmethods of study that it is stimulated by these same sounds. [1] Its rateof respiration is changed by the sounds, and although a sound does notbring about a bodily movement, it does very noticeably influence movementsin response to other stimuli which occur simultaneously with the sound. Idiscovered that under certain rather simple experimental conditions thegreen frog would regularly respond to a touch on the back by drawing itshind leg up toward the body. Under the same conditions the sound of anelectric bell caused no visible movement of the leg, but if at the instantthe back was touched the bell was rung, the leg movement was much greaterthan that brought about by the touch alone. This suggests at once thedesirability of studying the sense of hearing in the dancer by someindirect method. The animal may be stimulated, and yet it may not give anyvisible sign of the influence of the auditory stimulus. [Footnote 1: "The Sense of Hearing in Frogs. " _Journal of ComparativeNeurology and Psychology_, Vol. XV, p. 288, 1905. ] Were not the dancing so extremely variable in rapidity and duration, itmight be used as an index of the influence of auditory stimuli. Cyon'sstatements would indicate that sounds interfere with the dancing, but as Iobtained no evidence of this, I worked instead with the following indirectmethod, which may be called the method of auditory choice. The apparatus which was used is described in detail in Chapter VII. Figures 14 and 15 will greatly aid the reader in understanding itsessential features. Two small wooden boxes, identical in form and asclosely similar as possible in general appearance, were placed in a largerbox in such positions that a mouse was forced to enter and pass throughone of them in order to get to the nest-box. On the bottom of each ofthese small boxes was a series of wires through which an electric currentcould be made to pass at the will of the experimenter. The boxes couldreadily be interchanged in position. At one side of the large wooden boxand beyond the range of vision of the mouse was an electric bell whichcould be caused to ring whenever the mouse approached the entrance to oneof the small boxes. The point of the experiment was to determine whetherthe dancer could learn to avoid the box-which-rang when it was approached. The method of conducting the tests was as follows. Each day at a certainhour the mouse was placed in that part of the large box whence it couldescape to the nest-box only by passing through one of the small boxes. Ifit approached the wrong box (whether it happened to be the one on theright or the one on the left depended upon the experimenter's decision), the bell began to ring as a warning against entering; if it approached theother box, all was silent. As motives for the choice of the box-which-did-not-ring both reward and punishment were employed. The reward consisted offreedom to return to the nest-box _via_ the passage which led from thebox-which-did-not-ring; the punishment, which consisted of a disagreeableelectric shock, was given whenever the mouse entered the wrong box, thatis, the one which had the sound as a warning. Entering the wrong boxresulted in a disagreeable stimulus and in the necessity of returning tothe large box, for the exit to the nest-box by way of the passage fromthis box was closed. My assumption, on the basis of extended study of theability of the dancer to profit by experience, was that if it could hearthe sound of the bell it would soon learn to avoid the box-that-rang andenter instead the one which had no sound associated with it. Systematic tests were made with No. 4 from the 3d to the 12th of February, inclusive, 1906. Each day the mouse was permitted to find his way to thenest-box through one of the small boxes ten times in succession. Usuallythe experimenter rang the bell alternately for the box on the left and thebox on the right. The time required for such a series of experimentsvaried, according to the rapidity with which the mouse made his choice, from ten to thirty minutes. If in these experiments the animal approachedand entered the right, or soundless box, directly, the choice indicatednothing so far as ability to hear is concerned; if it entered the wrong, or sounding box, despite the ringing of the bell, it indicated either thelack of the influence of experience or inability to hear the sound; but ifit regularly avoided the box-which-sounded it thus gave evidence ofability to hear the sound of the bell. The purpose of the test was todetermine, not whether the mouse could learn, but whether it could hear. For ten successive days this experiment was carried on with No. 4 withoutthe least indication of increasing ability to avoid the wrong box by theassociation of the sound of the bell with the disagreeable electric shockand failure to escape to the nest-box. In fact, the experiment wasdiscontinued because it became evident that an impossible task had beenset for the mouse. Day by day as the tests were in progress I noticed thatthe animal became increasingly afraid of the entrances to the small boxes;it seemed absolutely helpless in the face of the situation. Partly becauseof the definiteness of the negative results obtained with No. 4 and partlybecause of the cruelty of subjecting an animal to disagreeable conditionswhich it is unable to avoid, the experiment was not repeated with otherindividuals. I have never conducted an experiment which gave me as muchdiscomfort as this; it was like being set to whip a deaf child because itdid not learn to respond to stimuli which it could not feel. By a very similar method No. 18 was tested for his sensitiveness to thenoise and jar from the induction apparatus which was used in connectionwith many of my experiments on vision and the modifiability of behavior. In this experiment the wrong box was indicated by the buzzing sound of theapparatus and the slight vibrations which resulted from it. Although No. 18 was tested, as was No. 4, for ten successive days, ten trials each day, it gave no evidence of ability to avoid the box-which-buzzed. Since both direct and indirect methods of testing the hearing of thedancer have uniformly given negative results, in the case of mice morethan five weeks old, I feel justified in concluding that they are totallydeaf and not merely irresponsive to sounds. Rawitz's statements, and the fact that what may have been auditoryreactions were obtained with a few individuals of five weeks of age, suggest that the mice may be able to hear at certain periods of life. Todiscover whether this is true I have tested the young of twenty differentlitters from the first day to the twenty-eighth, either daily or atintervals of two or three days. In these tests König forks, steel bars, and a Galton whistle were used. The results obtained are curiouslyinteresting. During the first two weeks of life none of the mice which I tested gaveany visible motor response to the various sounds used. During the thirdweek certain of the individuals responded vigorously to sudden high tonesand loud noises. After the third week I have seen only doubtful signs ofhearing. I shall now describe in detail the method of experimentation, thecondition of the animals, and the nature of the auditory reactions. Between the twelfth and the eighteenth day the auditory canal becomes opento the exterior. The time is very variable in different litters, for theirrate of growth depends upon the amount of nourishment which the mother isable to supply. Without exception, in my experience, the opening to theear appears before the eyes are open. Consequently visual stimuli usuallyare not disturbing factors in the auditory tests with mice less thansixteen days old. There is also a sudden and marked change in the behaviorof the mice during the third week. Whereas, for the first fourteen oreighteen days they are rather quiet and deliberate in their movements whenremoved from the nest, some time in the third week their behavior suddenlychanges and they act as if frightened when taken up by the experimenter. They jump out of his hand, squeak, and sometimes show fight. This is sopronounced that it has attracted my attention many times and I havestudied it carefully to determine, if possible, whether it is due to someprofound change in the nervous system which thus suddenly increases thesensitiveness of the animal or to the development of the sexual organs. Iam inclined to think that it is a nervous phenomenon which is intimatelyconnected with the sexual condition. Within a day or two after it appearsthe mice usually begin to show auditory reactions and continue to do sofor three to five days. I shall now describe the results obtained with a few typical litters. Alitter born of Nos. 151 and 152 gave uniformly negative results in allauditory tests up to the fourteenth day. On that day the ears were open, and the following observations were recorded. The five individuals of thelitter, four females and one male, were taken from the nest one at a timeat 7 A. M. And placed on a piece of paper in the bright sunlight. Thewarmth of the sun soon quieted them so that auditory tests could be madeto advantage. As soon as an individual had become perfectly still, theGalton whistle was held at a distance of about four inches from its headin such a position that it could not be seen nor the currents of aircaused by it felt, and suddenly blown. Each of the five mice responded tothe first few repetitions of this stimulus by movements of the ears, twitchings of the body, and jerky movements of the legs. The most violentreactions resulted when the individual was lying on its back with its legsextended free in the air. Under such circumstances the four legs wereoften drawn together suddenly when the whistle was sounded. Similarresponses were obtained with the lip sound already mentioned. Two otherobservers saw these experiments, and they agreed that there can be nodoubt that the mice responded to the sound. The sounds which wereeffective lay between 5000 and 10, 000 complete vibrations. On the fifteenth day the eyes were just beginning to open. Three of themice responded definitely to the sounds, but the other two slightly, if atall. On the sixteenth day they were all too persistently active forsatisfactory auditory tests, and on the seventeenth, although they weretested repeatedly under what appeared to be favorable conditions, no signsof sensitiveness were noted. Although I continued to test this litter, atintervals of three or four days, for two weeks longer, I did not onceobserve a response to sound. This was the first litter with which I obtained perfectly definite, clear-cut responses to sounds. That the reactive ability had not been presentearlier than the fourteenth day I am confident, for I had conducted thetests in precisely the same manner daily up to the time of the appearanceof the reactions. To argue that the mice heard before the fourteenth day, but were unable to react because the proper motor mechanism had notdeveloped sufficiently would be short-sighted, for if the responsedepended upon the development of such a mechanism, it is not likely thatit would disappear so quickly. I am therefore satisfied that thesereactions indicate hearing. With another litter the following results were obtained. On the thirteenthday each of the eight members of the litter responded definitely anduniformly to the Galton whistle, set at 5 (probably about 8000 completevibrations), and to a König steel bar of a vibration rate of 4096 Vs. Thelargest individuals, for almost always there are noticeable differences insize among the members of a litter, appeared to be most sensitive tosounds. On the fifteenth day and again on the seventeenth unmistakable responsesto sound were observed; on the eighteenth the responses were indefinite, and on the nineteenth none were obtained. I continued the tests up to thetwenty-eighth day without further indications of hearing. Certain individuals in this litter reacted so vigorously to the loud soundproduced by striking the steel bar a sharp blow and also to the Galtonwhistle, during a period of five days, that I have no hesitation in sayingthat they evidently heard during that period of their lives. Other membersof the litter seemed to be less sensitive; their reactions were sometimesso indefinite as to leave the experimenter in doubt about the presence ofhearing. A third litter, which developed very slowly because of lack of sufficientfood, first showed unmistakable reactions to sound on the twenty-firstday. On this day only two of the five individuals reacted. The reactionswere much more obvious on the twenty-second day, but thereafter theybecame indefinite. Still another litter, which consisted of one female and four males, beganto exhibit the quick, jerky movements, already mentioned, on thefourteenth day. On the morning of the fifteenth day three members of thelitter definitely reacted to the tone of the steel bar, and also to thehammer blow when the bar was held tightly in the hand of the experimenter. My observations were verified by another experimenter. Two individualswhich appeared to be very sensitive were selected for special tests. Theirreactions were obvious on the sixteenth, seventeenth, and eighteenth days;on the nineteenth day they were indefinite, and on the twentieth nonecould be detected. Some individuals of this litter certainly had theability to hear for at least five days. A sixth litter of four females and two males first gave indications of thechange in behavior which by this time I had come to interpret as a sign ofthe approach of the period of auditory sensitiveness, on the seventeenthday. I had tested them almost every day previous to this time withoutobtaining evidence of hearing. The tests with the steel bar and the Galtonwhistle were continued each day until the end of the fourth week withoutpositive results. To all appearances the individuals of this litter wereunable to hear at any time during the first month of life. Practically the same results were obtained with another litter of fourfemales. The change in their behavior was obvious on the eighteenth day, but at no time during the first month did they give any satisfactoryindications of hearing. In the accompanying table, I have presented in condensed form the resultsof my auditory tests in the case of twelve litters of young dancers. TABLE 5 PERIOD OF AUDITORY REACTION IN YOUNG DANCERS PARENTS No. In Change in Ears Auditory Reactions Litter Behavior Open Appear Disappear 152+151 5 13th day 14th day 14th day 16th day152+15l 8 (?) 13th day 13th day 17th day152+151 5 13th day 13th day 13th day 17th day152+151 4 10th day 12th day 13th day 15th day410+415 5 14th day 15th day 15th day 19th day410+415 6 13th day 14th day 14th day 18th day420+425 2 12th day 14th day 14th day 17th day210+215 5 17th day 13th day 17th day 19th day210+215 6 11th day 14th day No reactions220+225 6 16th day 14th day No reactions220+225 6 17th day 13th day No reactions212+211 4 15th day 14th day No reactions Certain of the litters tested responded definitely to sounds, others gaveno sign of hearing at any time during the first four weeks of life. Of thetwelve litters for which the results of auditory tests are presented inTable 5, eight evidently passed through an auditory period. It isimportant to note that all except one of these were the offspring of Nos. 151 and 152, or of their descendants Nos. 410 and 415 and Nos. 420 and425. In fact every one of the litters in this line of descent which I havetested, and they now number fifteen, has given indications of auditorysensitiveness. And, on the other hand, only in a single instance have thelitters born of Nos. 210 and 215, or of their descendants, given evidenceof ability to hear. These two distinct lines of descent may be referred to hereafter as the400 and the 200 lines. I have observed several important differencesbetween the individuals of these groups in addition to the one alreadymentioned. The 200 mice were sometimes gray and white instead of black andwhite; they climbed much more readily and danced less vigorously thanthose of the 400 group. These facts are particularly interesting inconnection with Cyon's descriptions of the two types of dancer which heobserved. In criticism of my conclusion that the young dancers are able to hearcertain sounds for a few days early in life, and then become deaf, it hasbeen suggested that they cease to react because they rapidly becomeaccustomed to the sounds. That this is not the case, is evident from thefact that the reactions often increase in definiteness during the firsttwo or three days and then suddenly disappear entirely. But even if thiswere not true, it would seem extremely improbable that the mouse shouldbecome accustomed to a sudden and startlingly loud sound with so fewrepetitions as occurred in these tests. On any one day the sounds were notmade more than five to ten times. Moreover, under the same externalcondition, the common mouse reacts unmistakably to these sounds day afterday when they are first produced, although with repetition of the stimulusat short intervals, the reactions soon become indefinite or disappear. The chief results of my study of hearing in the dancer may be summed up ina very few words. The young dancer, in some instances, hears sounds for afew days during the third week of life. The adult is totally deaf. Shortlybefore the period of auditory sensitiveness, the young dancer becomesextremely excitable and pugnacious. CHAPTER VII THE SENSE OF SIGHT: BRIGHTNESS VISION The sense of sight in the dancer has received little attention hitherto. In the literature there are a few casual statements to the effect that itis of importance. Zoth, for example (31 p. 149), remarks that it seems tobe keenly developed; and other writers, on the basis of their observationof the animal's behavior, hazard similar statements. The descriptions ofthe behavior of blinded mice, as given by Cyon, Alexander and Kreidl, andKishi (p. 47), apparently indicate that the sense is of some value; they donot, however, furnish definite information concerning its nature and itsrole in the daily life of the animal. The experimental study of this subject which is now to be described wasundertaken, after careful and long-continued observation of the generalbehavior of the dancer, in order that our knowledge of the nature andvalue of the sense of sight in this representative of the Mammalia mightbe increased in scope and definiteness. The results of this studynaturally fall into three groups: (1) those which concern brightnessvision, (2) those which concern color vision, and (3) those which indicatethe role of sight in the life of the dancer. Too frequently investigators, in their work on vision in animals, haveassumed that brightness vision and color vision are inseparable; or, ifnot making this assumption, they have failed to realize that the samewave-length probably has markedly different effects upon the retinalelements of the eyes of unlike organisms. In a study of the sense of sightit is extremely important to discover whether difference in the quality, as well as in the intensity, of a visual stimulus influences the organism;in other words, whether color sensitiveness, as well as brightnesssensitiveness, is present. If the dancer perceives only brightness orluminosity, and not color, it is evident that its visual world isstrikingly different from that of the normal human being. The experimentsnow to be described were planned to show what the facts really are. [Illustration: Figure 14. --Discrimination box. _W_, electric-box withwhite cardboards; B, electric-box with black cardboards. Drawn by Mr. C. H. Toll. ] As a means of testing the ability of the dancer to distinguish differencesin brightness, the experiment box represented by Figures 14 and 15 wasdevised. Figure 14 is the box as seen from the position of theexperimenter during the tests. Figure 15 is its ground plan. This box, which was made of wood, was 98 cm. Long, 38 cm. Wide, and 17 cm. Deep, asmeasured on the outside. The plan of construction and its significance inconnection with these experiments on vision will be clear from thefollowing account of the experimental procedure. A mouse whose brightnessvision was to be tested was placed in the nest-box, A (Figure 15). Thenceby pushing open the swinging door at _I_, it could pass into the entrancechamber, _B_. Having entered _B_ it could return to _A_ only by passingthrough one of the electric-boxes, marked _W_, and following the alley to_O_, where by pushing open the swing door it could enter the nest-box. Thedoor at _I_ swung inward, toward _B_, only; those at _O_, right and left, swung outward, toward _A_, only. It was therefore impossible for the mouseto follow any other course than _A-I-B-L-W-E-O_ or _A-I-B-R-W-E-O_. Thedoors at _I_ and _O_ were pieces of wire netting of 1/2 cm. Mesh, hingedat the top so that a mouse could readily open them, in one direction, bypushing with its nose at any point along the bottom. On the floor of eachof the electric-boxes, _W_, was an oak board 1 cm. In thickness, whichcarried electric wires by means of which the mouse could be shocked in _W_when the tests demanded it. The interrupted circuit constituted by thewires in the two electric-boxes, in connection with the inductionapparatus, _IC_, the dry battery, _C_, and the hand key, _K_, was made bytaking two pieces of No. 20 American standard gauge copper wire andwinding them around the oak board which was to be placed on the floor ofeach electric-box. The wires, which ran parallel with one another, 1/2 cm. Apart, fitted into shallow grooves in the edges of the board, and thus, aswell as by being drawn taut, they were held firmly in position. The coilsof the two pieces of wire alternated, forming an interrupted circuitwhich, when the key _K_ was closed, was completed if the feet of a mouserested on points of both pieces of wire. Since copper wire stretcheseasily and becomes loose on the wooden base, it is better to use phosphorbronze wire of about the same size, if the surface covered by theinterrupted circuit is more than three or four inches in width. Thephosphor bronze wire is more difficult to wind satisfactorily, for it isharder to bend than the copper wire, and it has the further disadvantageof being more brittle. But when once placed properly, it forms a far morelasting and satisfactory interrupted circuit for such experiments as thoseto be described than does copper wire. In the case of the electric-boxesunder consideration, the oak boards which carried the interrupted circuitswere separate, and the two circuits were joined by the union of the wiresbetween the boxes. The free ends of the two pieces of wire whichconstituted the interrupted circuit were connected with the secondary coilof a Porter inductorium whose primary coil was in circuit with a No. 6Columbia dry battery. In the light of preliminary experiments, made inpreparation for the tests of vision, the strength of the induced currentreceived by the mouse was so regulated, by changing the position of thesecondary coil with reference to the primary, that it was disagreeable butnot injurious to the animal. What part the disagreeable shock played inthe test of brightness vision will now be explained. [Illustration: FIGURE 15. --Ground plan of discrimination box. _A_, nest-box; _B_, entrance chamber; _W, W_, electric-boxes; _L_, doorway of leftelectric-box; _R_, doorway of right electric-box; _E_, exit from electric-box to alley; _I_, swinging door between _A_ and _B_; _O_, swinging doorbetween alley and _A_; _IC_, induction apparatus; _C_, electric cell; _K_, key in circuit. ] An opportunity for visual discrimination by brightness difference wasprovided by placing dead black cardboard at the entrance and on the insideof one of the electric-boxes, as shown in Figure 14, _B_, and whitecardboard similarly in the other box. These cardboards were movable andcould be changed from one box to the other at the will of theexperimenter. The test consisted in requiring the mouse to choose acertain brightness, for example, the white cardboard side, in order toreturn to the nest-box without receiving an electric shock. The questionwhich the experimenter asked in connection with this test really is, Can adancer learn to go to the white box and thus avoid discomfort? If weassume its ability to profit by experience within the limits of the numberof experiences which it was given, such a modification of behavior wouldindicate discrimination of brightness. Can the dancer distinguish whitefrom black; light gray from dark gray; two grays which are almost of thesame brightness? The results which make up the remainder of this and thefollowing chapter furnish a definite answer to these questions. To return to the experimental procedure, the mouse which is being testedis placed by the experimenter in the nest-box, where frequently in theearly tests food and a comfortable nest were attractions. If it does notof its own accord, as a result of its abundant random activity, passthrough _I_ into _B_ within a few seconds, it is directed to the doorwayand urged through. A choice is now demanded of the animal; to return tothe nest-box it must enter either the white electric-box or the black one. Should it choose the white box, it is permitted to return directly to _A_by way of the doorway _E_, the alley, and the swinging door at _O_, and itthus gets the satisfaction of unobstructed activity, freedom to whirl, tofeed, and to retreat for a time to the nest. Should it choose to attemptto enter the black box, as it touches the wires of the interrupted circuitit receives a shock as a result of the closing of the key in the circuitby the experimenter, and further, if it continues its forward courseinstead of retreating from the "stinging" black box, its passage through_E_ is blocked by a barrier of glass temporarily placed there by theexperimenter, and the only way of escape to the nest-box is an indirectroute by way of _B_ and the white box. Ordinarily the shock was given onlywhen the mouse entered the wrong box, not when it retreated from it; itwas never given when the right box was chosen. The box to be chosen, whether it was white, gray, or black, will be called the right box. Theelectric shock served as a means of forcing the animal to use itsdiscriminating ability. But the question of motives in the tests is not sosimple as might appear from this statement. The reader will wonder why the mouse should have any tendency to enter_B_, and why after so doing, it should trouble to go further, knowing, asit does from previous experiences, that entering one of the electric-boxesmay result in discomfort. The fact is, a dancer has no very constanttendency to go from _A_ to _B_ at the beginning of the tests, but after ithas become accustomed to the box and has learned what the situationdemands, it shows eagerness to make the trip from _A_ to _B_, and thenceby way of either the right or the left route to _A_. That the mouse shouldbe willing to enter either of the electric-boxes, after it has experiencedthe shock, is even more surprising than its eagerness to run from _A_ to_B_. When first tested for brightness discrimination in this apparatus, adancer usually hesitated at the entrance to the electric-boxes, and thishesitation increased rapidly unless it were able to discriminate the boxesby their difference in brightness and thus to choose the right one. Duringthe period of increasing hesitancy in making the choice, the experimenter, by carefully moving from _I_ toward the entrances to the electric-boxes apiece of cardboard which extended all the way across _B_, greatlyincreased the mouse's desire to enter one of the boxes by depriving it ofdancing space in _B_. If an individual which did not know which entranceto choose were permitted to run about in _B_, it would often do so forminutes at a time without approaching the entrance to the boxes; but thesame individual, when confined to a dancing space 4 or 5 cm. Wide in frontof the entrances, would enter one of the electric-boxes almostimmediately. This facilitation of choice by decrease in the amount ofspace for whirling was not to any considerable extent the result of fear, for all the dancers experimented with were tame, and instead of forcingthem to rush into one of the boxes blindly and without attempt atdiscrimination, the narrowing of the space simply increased their effortsto discriminate. The common mouse when subjected to similar experimentalconditions is likely to be frightened by being forced to approach theentrances to the boxes, and fails to choose; it rushes into one boxdirectly, and in consequence it is as often wrong as right. The danceralways chooses, but its eagerness to choose is markedly increased by therestriction of its movements to a narrow space in front of the entrancesbetween which it is required to discriminate. It is evident that theanimal is uncomfortable in a space which is too narrow for it to whirl infreely. It must have room to dance. This furnished a sufficiently strongmotive for the entering of the electric-boxes. It must avoid disagreeableand unfavorable stimuli. This is a basis for attempts to choose, by visualdiscrimination, the electric-box in which the shock is not given. It maysafely be said that the success of the majority of the experiments of thisbook depended upon three facts: (1) the dancer's tendency to avoiddisagreeable external conditions, (2) its escape-from-confinement-impelling need of space in which to dance freely, and (3) its abundant andincessant activity. Of these three conditions of success in the experiments, the second andthird made possible the advantageous use of the first. For the avoidanceof a disagreeable stimulus could be made use of effectively in the testsjust because the mice are so restless and so active. In fact theireagerness to do things is so great that the experimenter, instead ofhaving to wait for them to perform the desired act, often is forced tomake them wait while he completes his observation and record. In thisrespect they are unlike most other animals. My experiments with the dancer differ from those which have been made bymost students of mammalian behavior in one important respect. I have usedpunishment instead of reward as the chief motive for the properperformance of the required act. Usually in experiments with mammalshunger has been the motive depended upon. The animals have been requiredto follow a certain devious path, to escape from a box by working abutton, a bolt, a lever, or to gain entrance to a box by the use of teeth, claws, hands, or body weight and thus obtain food as a reward. There aretwo very serious objections to the use of the desire for food as a motivein animal behavior experiments--objections which in my opinion render italmost worthless in the case of many mammals. These are the discomfort ofthe animal and the impossibility of keeping the motive even fairlyconstant. However prevalent the experience of starvation may be in thelife of an animal, it is not pleasant to think of subjecting it to extremehunger in the laboratory for the sake of finding out what it can do toobtain food. Satisfactory results can be obtained in an experiment whosesuccess depends chiefly upon hunger only when the animal is so hungry thatit constantly does its best to obtain food, and when the desire for foodis equally strong and equally effective as a spur to action in therepetitions of the experiment day after day. It is easy enough to getalmost any mammal into a condition of utter hunger, but it is practicallyimpossible to have the desire for food of the same strength day after day. In short, the desire for food is unsatisfactory as a motive in animalbehavior work, first, because a condition of utter hunger, as has beendemonstrated with certain mammals, is unfavorable for the performance ofcomplex acts, second, because it is impossible to control the strength ofthe motive, and finally, because it is an inhumane method ofexperimentation. In general, the method of punishment is more satisfactory than the methodof reward, because it can be controlled to a greater extent. Theexperimenter cannot force his subject to desire food; he can, however, force it to discriminate between conditions to the best of its knowledgeand ability by giving it a disagreeable stimulus every time it makes amistake. In other words, the conditions upon which the avoidance of adisagreeable factor in the environment depends are far simpler and muchmore constant than those upon which the seeking of an agreeable factordepends. Situations which are potentially beneficial to the animal attractit in varying degrees according to its internal condition; situationswhich are potentially disagreeable or injurious repel it with a constancywhich is remarkable. The favorable stimulus solicits a positive response;the unfavorable stimulus demands a negative response. Finally, in connection with the discussion of motives, it is an importantfact that forms of reward are far harder to find than forms of punishment. Many animals feed only at long intervals, are inactive, do not try toescape from confinement, cannot be induced to seek a particular spot, in aword, do not react positively to any of the situations or conditions whichare employed usually in behavior experiments. It is, however, almostalways possible to find some disagreeable stimulus which such an animalwill attempt to avoid. As it happens, the dancer is an animal which does not stand the lack offood well enough to make hunger a possible motive. I was driven to makeuse of the avoiding reaction, and it has proved so satisfactory that I amnow using it widely in connection with experiments on other animals. Theuse of the induction shock, upon which I depended almost wholly in thediscrimination experiments with the dancer, requires care; but I amconfident that no reasonable objection to the conduct of the experimentscould be made on the ground of cruelty, for the strength of the currentwas carefully regulated and the shocks were given only for an instant atintervals. The best proof of the humaneness of the method is the fact thatthe animals continued in perfect health during months of experimentation. The brightness discrimination tests demanded, in addition to motives forchoice, adequate precautions against discrimination by other than visualfactors, and, for that matter, by other visual factors than that ofbrightness. The mouse might choose, for example, not the white or theblack box, but the box which was to the right or to the left, inaccordance with its experience in the previous test. This would bediscrimination by position. As a matter of fact, the animals have a strongtendency at first to go uniformly either to the right or to the leftentrance. This tendency will be exhibited in the results of the tests. Again, discrimination might depend upon the odors of the cardboards orupon slight differences in their shape, texture, or position. Beforeconclusive evidence of brightness discrimination could be obtained, all ofthese and other possibilities of discrimination had to be eliminated bycheck tests. I shall describe the various precautions taken in theexperiments to guard against errors in interpretation, in order to showthe lengths to which an experimenter may be driven in his search forsafely interpretable results. To exclude choice by position, the cardboards were moved from oneelectric-box to the other. When the change was made regularly, so thatwhite was alternately on the right and the left, the mouse soon learned togo alternately to the right box and the left without stopping to noticethe visual factor. This was prevented by changing the position of thecardboards irregularly. Discrimination by the odor, texture, shape, and position of the cardboardswas excluded by the use of different kinds of cardboards, by changing theform and position of them in check tests, and by coating them withshellac. The brightness vision tests described in this chapter were made in a roomwhich is lighted from the south only, with the experiment box directedaway from the windows. The light from the windows shone upon thecardboards at the entrances to the electric-boxes, not into the eyes ofthe mouse as it approached them. Each mouse used in the experiments wasgiven a series of ten tests in succession daily. The experiment wasconducted as follows. A dancer was placed in _A_, where it usually ranabout restlessly until it happened to find its way into _B_. Havingdiscovered that the swing door at _I_ could be pushed open, the animalseemed to take satisfaction in passing through into _B_ as soon as it hadbeen placed in or had returned to _A_. In _B_, choice of two entrances, one of which was brighter than the other, was forced by the animal's needof space for free movement. If the right box happened to be chosen, themouse returned to _A_ and was ready for another test; if it entered thewrong box, the electric shock was given, and it was compelled to retreatfrom the box and enter the other one instead. In the early tests with anindividual, a series sometimes covered from twenty to thirty minutes; inlater tests, provided the condition of discrimination was favorable, itdid not occupy more than ten minutes. To exhibit the methods of keeping the records of these experiments andcertain features of the results, two sample record sheets are reproducedbelow. The first of these sheets, Table 6, represents the results given byNo. 5, a female, [1] in her first series of white-black tests. Table 7presents the results of the eleventh series of tests given to the sameindividual. [Footnote 1: It is to be remembered that the even numbers always designatemales; the odd numbers, females. ] In the descriptions of the various visual experiments of this and thefollowing chapters, the first word of the couplet which describes thecondition of the experiment, for example, white-black, always designatesthe visual condition which the animal was to choose, the second that whichit was to avoid on penalty of an electric shock. In the case of Tables 6and 7, for example, white cardboard was placed in one box, black in theother, and the animal was required to enter the white box. In the tablesthe first column at the left gives the number of the test, the second thepositions of the cardboards, and the third and fourth the result of thechoice. The first test of Table 6 was made with the white cardboard on thebox which stood at the left of the mouse as it approached from _A_, and, consequently, with the black cardboard on the right. As is indicated bythe record in the "wrong" column, the mouse chose the black instead of thewhite. The result of this first series was choice of the white box fourtimes as compared with choice of the black box six times. On the eleventhday, that is, after No. 5 had been given 100 tests in this brightnessvision experiment, she made no mistakes in a series of ten trials (Table7). TABLE 6 BRIGHTNESS DISCRIMINATION White-Black, Series 1 Experimented on No. 5 January 15, 1906 POSITION OFTEST CARDBOARDS RIGHT WRONG 1 White left -- Wrong2 White right -- Wrong3 White left -- Wrong4 White right -- Wrong5 White left Right --6 White right Right --7 White left -- Wrong8 White right Right --9 White left -- Wrong10 White right Right -- Totals 4 6 Before tests, such as have been described, can be presented as conclusiveproof of discrimination, it must be shown that the mouse has no preferencefor the particular brightness which the arrangement of the test requiresit to select. That any preference which the mouse to be tested might havefor white, rather than black, or for a light gray rather than a dark gray, might be discovered, what may be called preference test series were givenbefore the discrimination tests were begun. These series, two of whichwere given usually, consisted of ten tests each, with the whitealternately on the left and on the right. The mouse was permitted to entereither the white or the black box, as it chose, and to pass through to thenest-box without receiving a shock and without having its way blocked bythe glass plate. The conditions of these preference tests may be referredto hereafter briefly as "No shock, open passages. " The preference tests, which of course would be valueless as such unless they preceded thetraining tests, were given as preliminary experiments, in order that theexperimenter might know how to plan his discrimination tests, and how tointerpret his results. TABLE 7 BRIGHTNESS DISCRIMINATION White-Black, Series II Experimented on No. 5 February 2, 1906 POSITIONTEST OF CARDBOARDS RIGHT WRONG 1 White left Right -- 2 White left Right -- 3 White right Right -- 4 White right Right -- 5 White right Right -- 6 White left Right -- 7 White left Right -- 8 White left Right -- 9 White right Right --10 White right Right -- Totals 10 0 The results given in the white-black preference tests by ten males and tenfemales are presented in Table 8. Three facts which bear upon thebrightness discrimination tests appear from this table: (1) black ispreferred by both males and females, (2) this preference is more marked inthe first series of tests than in the second, and (3) it is slightlystronger for the first series in the case of females than in the case ofmales. TABLE 8 WHITE-BLACK PREFERENCE TESTS MALES FIRST SERIES SECOND SERIES WHITE BLACK WHITE BLACK No. 10 3 7 3 7 18 5 5 5 5 20 2 8 4 6 152 4 6 6 4 210 4 6 4 6 214 6 4 3 7 220 5 5 3 7 230 4 6 2 8 410 4 6 5 5 420 4 6 9 1 Averages 4. 1 5. 9 4. 4 5. 6 FEMALES FIRST SERIES SECOND SERIES WHITE BLACK WHITE BLACK No. 11 5 5 4 6 151 6 4 5 5 215 2 8 2 8 213 2 8 5 5 225 4 6 2 8 227 4 6 6 4 235 6 4 4 6 415 2 8 4 6 425 5 5 8 2 229 2 8 5 5 Averages 3. 8 6. 2 4. 5 5. 5 That the dancers should prefer to enter the dark rather than the light boxis not surprising in view of the fact that the nests in which they werekept were ordinarily rather dark. But whatever the basis of thepreference, it is clear that it must be taken account of in the visualdiscrimination experiments, for an individual which strongly preferredblack might choose correctly, to all appearances, in its first black-whiteseries. Such a result would demonstrate preference, and therefore one kindof discrimination, but not the formation of a habit of choice bydiscrimination. The preference for black is less marked in the secondseries of tests because the mouse as it becomes more accustomed to theexperiment box tends more and more to be influenced by other conditionsthan those of brightness. The record sheets for both series almostinvariably indicate a strong tendency to continue to go to the left or theright entrance according to the way by which the animal escaped the firsttime. This cannot properly be described as visual choice, for the mouseapparently followed the previous course without regard to the conditionsof illumination. We have here an expression of the tendency to therepetition of an act. It is only after an animal acquires considerablefamiliarity with a situation that it begins to vary its behavior inaccordance with relatively unimportant factors in the situation. It isthis fact, illustrations of which may be seen in human life, as well asthroughout the realm of animal behavior, that renders it imperative thatan animal be thoroughly acquainted with the apparatus for experimentationand with the experimenter before regular experiments are begun. Any animalwill do things under most experimental conditions, but to discover thenature and scope of its ability it is necessary to make it thoroughly athome in the experimental situation. As the dancer began to feel at home inthe visual discrimination apparatus it began to exercise itsdiscriminating ability, the first form of which was choice according toposition. Since there appears to be a slight preference on the part of most dancers'for the black box in comparison with the white box, white-black trainingtests were given to fifty mice, and black-white to only four. The testswith each individual were continued until it had chosen correctly in allof the tests of three successive series (thirty tests). As thereproduction of all the record sheets of these experiments would fillhundreds of pages and would provide most readers with little moreinformation than is obtainable from a simple statement of the number ofright and wrong choices, only the brightness discrimination records ofTables 6 and 7 are given in full. As a basis for the comparison of the results of the white-black tests withthose of the black-white tests, two representative sets of data for eachof these conditions of brightness discrimination are presented (Tables 9and 10). In these tables only the number of right and wrong choices foreach series of ten tests appears. Tables 9 and 10 indicate--if we grant that the precautionary tests to bedescribed later exclude the possibility of other forms of discrimination--that the dancer is able to tell white from black; that it is somewhateasier, as the preference tests might lead us to expect, for it to learnto go to the black than to the white, and that the male forms the habit ofchoosing on the basis of brightness discrimination more quickly than thefemale. TABLE 9WHITE-BLACK TESTS No. 210 No. 215 AGE, 28 DAYS AGE, 28 DAYSSERIES DATE RIGHT WRONG RIGHT WRONG (WHITE) (BLACK) (WHITE) (BLACK) A June 22 4 6 2 8 B 23 4 6 2 8 1 24 4 6 3 7 2 25 6 4 5 5 3 26 7 3 7 3 4 27 5 5 8 2 5 28 7 3 9 1 6 29 8 2 8 2 7 30 9 1 9 1 8 July 1 10 0 10 0 9 2 10 0 9 1 10 3 10 0 10 0 11 4 -- -- 10 0 12 5 -- -- 10 0 TABLE 10WHITE-BLACK TESTS No. 14 No. 13 AGE, 32 DAYS AGE, 32 DAYSSERIES DATE RIGHT WRONG RIGHT WRONG (WHITE) (BLACK) (WHITE) (BLACK) 1 May 13[1] 5 5 7 3 2 14 8 2 6 4 3 15 7 3 9 1 4 16 9 1 9 1 5 17 10 0 10 0 6 18 10 0 9 1 7 19 10 0 10 0 8 20 -- -- 10 0 9 21 -- -- 10 0 [Footnote 1: No preference tests were given. ] It is now necessary to justify the interpretation of these results asevidence of brightness discrimination by proving that all other conditionsfor choice except brightness difference may be excluded withoutinterfering with the animal's ability to select the right box. We shallconsider in order the possibility of discrimination by position, by odor, and by texture and form of the cardboards. The tendency which the dancer has in common with many, if not all, animalsto perform the same movement or follow the same path under uniformconditions is an important source of error in many habit-formationexperiments. This tendency is evident even from casual observation of thebehavior of the dancer. The ease with which the habit of choosing the boxon the left or the box on the right is formed in comparison with that ofchoosing the white box or the black box is strikingly shown by thefollowing experiment. Five mice were given one series of ten trials eachin the discrimination box of Figure 14 without the presence of cardboardsor of other means of visual discrimination. The electric shock was givenwhenever the box on the left was entered. Thus without other guidance thanthat of direction, for the boxes themselves were interchanged in position, and, as was proved by additional tests, the animals were utterly unable totell one from the other, the mouse was required to choose the box on itsright. Only one of the five animals went to the box on the left after onceexperiencing the electric shock. The results of the series are given inTable 11. TABLE 11 CHOICE BY POSITION Choices of Choices of Box on Right Box on LeftFirst mouse 9 1Second mouse 8 2Third mouse 9 1Fourth mouse 9 1Fifth mouse 9 1 This conclusively proves that the habit of turning in a certain directionor of choosing by position can be formed more readily than a habit whichdepends upon visual discrimination. A rough comparison justifies thestatement that it takes from six to ten times as long for the dancer tolearn to choose the white box as it does to learn to choose the box on theright. Since this is true, it is exceedingly important that thepossibility of choice by position or direction of movement be excluded inthe case of tests of brightness discrimination. To indicate how this waseffectively accomplished in the experiments, the changes in the positionof the cardboards made in the case of a standard set of white-black seriesare shown in Table 12. The number of the series, beginning at the top ofthe table with the two lettered preference series, is given in the firstcolumn at the left, the number of the tests at the top of the table, andthe position of the white cardboard, left or right, is indicated below bythe letters l (left) and r (right). TABLE 12 POSITION OF WHITE CARDBOARDS FOR A SET OF 150 TESTS SERIES 1 2 3 4 5 6 7 8 9 10 Preference A l r l r l r l r l r B r l r l r l r l r l 1 r l r l r l r l r l 2 l l r r l r l l r r 3 r r l r l l r l r l 4 l l l r r r l r r l 5 r l r l r l r l r l 6 l l r l r r l r l r 7 r l l l r r r l r l 8 r r l l r l r l r l 9 r r r l l l r l r l 10 l l l l r r r r l r 11 r l r r r l l l r l 12 r l r l r r l l r l 13 r l r l l l r r r l 14 l l l l r r r r l r 15 r l r r r l l l r l It is to be noted that in the case of each series of ten tests the whitecardboard was on the left five times and on the right five times. Thus theestablishment of a tendency in favor of one side was avoided. Theirregularity of the changes in position rendered it impossible for themouse to depend upon position in its choice. It is an interesting factthat the dancer quickly learns to choose correctly by position if thecardboards are alternately on the left box and on the right. The prevalent, although ill-founded, impression that mice have anexceedingly keen sense of smell might lead a critic of these experimentsto claim that discrimination in all probability was olfactory rather thanvisual. As precautions against this possibility the cardboards wererenewed frequently, so that no odor from the body of the mouse itselfshould serve as a guiding condition, different kinds of cardboard wereused from time to time, and, as a final test, the cardboards were coatedwith shellac so that whatever characteristic odor they may have had forthe dancer was disguised if not totally destroyed. Despite all theseprecautions the discrimination of the boxes continued. A still moreconclusive proof that we have to do with brightness discrimination, andthat alone, in these experiments is furnished by the results of white-black tests made with an apparatus which was so arranged that light wastransmitted into the two electric-boxes through a ground glass plate inthe end of each box. No cardboards were used. The illumination of each boxwas controlled by changes in the position of the sources of light. Underthese conditions, so far as could be ascertained by critical examinationof the results, in addition to careful observation of the behavior of theanimals as they made their choices, there was no other guiding factor thanbrightness difference. Nevertheless the mice discriminated the white fromthe black perfectly. This renders unnecessary any discussion of thepossibility of discrimination by the texture or form of the cardboards. Since a variety of precautionary tests failed to reveal the presence, inthese experiments, of any condition other than brightness difference bywhich the mice were enabled to choose correctly, and since evidence ofability to discriminate brightness differences was obtained by the use ofboth reflected light (cardboards) and transmitted light (lamps behindground glass), it is necessary to conclude that the dancer possessesbrightness vision. CHAPTER VIII THE SENSE OF SIGHT: BRIGHTNESS VISION (Continued) Since the ability of the dancer to perceive brightness has beendemonstrated by the experiments of the previous chapter, the next step inthis investigation of the nature of vision is a study of the delicacy ofbrightness discrimination, and of the relation of the just perceivabledifference to brightness value. Expressed in another way, the problems ofthis portion of the investigation are to determine how slight a differencein brightness enables the dancer to discriminate one light from another, and what is the relation between the absolute brightnesses of two lightsand that amount of difference which is just sufficient to render thelights distinguishable. It has been discovered in the case of the humanbeing that a stimulus must be increased by a certain definite fraction ofits own value if it is to seem different. For brightness, within certainintensity limits, this increase must be about one one-hundredth; abrightness of 100 units, for example, is just perceivably different fromone of 101 units. The formulation of this relation between the amount of astimulus and the amount of change which is necessary that a difference benoted is known as Weber's law. Does this law, in any form, hold for thebrightness vision of the dancing mouse? Two methods were used in the study of these problems. For the firstproblem, that of the delicacy of brightness discrimination, I first usedlight which was reflected from gray papers, according to the method ofChapter VII. For the second, the Weber's law test, transmitted light wasused, in an apparatus which will be described later. Either of thesemethods might have been used for the solution of both problems. Which ofthem is the more satisfactory is definitely decided by the results whichmake up the material of this chapter, Under natural conditions the dancerprobably sees objects which reflect light more frequently than it doesthose which transmit it; it would seem fairer, therefore, to require it todiscriminate surfaces which differ in brightness. This the use of graypapers does. But, on the other hand, gray papers are open to theobjections that they may not be entirely colorless (neutral), and thattheir brightness values cannot be changed readily by the experimenter. Aswill be made clear in the subsequent description of the experiments withtransmitted light, neither of these objections can be raised in connectionwith the second method of experimentation. To determine the delicacy of discrimination with reflected light it isnecessary to have a series of neutral grays (colorless) whose adjacentmembers differ from one another in brightness by less than the thresholdof discrimination of the animal to be tested. A series which promised tofulfill these conditions was that of Richard Nendel of Berlin. It consistsof fifty papers, beginning with pure white, numbered 1, and passing byalmost imperceptible steps of decrease in brightness through the grays toblack, which is numbered 50. For the present we may assume that thesepapers are so nearly neutral that whatever discrimination occurs is due tobrightness. The differences between successive papers of the series areperceptible to man. The question is, can they, under favorable conditionsof illumination, be perceived by the dancer? On the basis of the fact that the dancer can discriminate between whiteand black, two grays which differed from one another in brightness by aconsiderable amount were chosen from the Nendel series; these were numbers10 and 20. It seemed certain, from the results of previous experiments, that the discrimination of these papers by brightness difference would bepossible, and that therefore the use of papers between these two extremeswould suffice to demonstrate the delicacy of discrimination. In Figure 16we have a fairly accurate representation of the relative brightness of theNendel papers Nos. 10, 15, and 20. [Illustration: FIGURE 16. Three of Nendel's gray papers: Nos. 10, 15, and20. To exhibit differences in brightness. ] Pieces of the gray papers were pasted upon cardboard carriers so that theymight be placed in the discrimination box as were the white and blackcardboards in the tests of brightness vision previously described. Micewhich had been trained to choose the white box by series of white-blacktests were now tested with light gray (No. 10) and dark gray (No. 20), myassumption being that they would immediately choose the brighter of thetwo if they were able to detect the difference. As a matter of fact thisdid not always occur; some individuals had to be trained to discriminategray No. 10 from gray No. 20. As soon as an individual had been so trainedthat the ability to choose the lighter of these grays was perfect, it wastested with No. 10 in combination with No. 15. If these in turn proved tobe discriminable, No. 10 could be used with No. 14, with No. 13, and so onuntil either the limit of discrimination or that of the series had beenreached. That it was not necessary to use other combinations than 10 with 20, and10 with 15 is demonstrated by the results of Table 13. Mouse No. 420, whose behavior was not essentially different from that of three otherindividuals which were tested for gray discrimination, learned withdifficulty to choose gray No. 10 even when it was used with No. 20. Twoseries of ten tests each were given to this mouse daily, and not until thetwentieth of these series (200 tests) did he succeed in making ten correctchoices in succession. Immediately after this series of correct choices, tests with grays No. 10 and No. 15 were begun. In the case of this amountof brightness difference twenty series failed to reveal discrimination. The average number of right choices in the series is slightly in excess ofthe mistakes, 5. 8 as compared with 4. 2. From the experiments with gray papers we may conclude that under theconditions of the tests the amount by which Nendel's gray No. 10 differsin brightness from No. 20 is near the threshold of discrimination, or, inother words, that the difference in the brightness of the adjacent graysof Figure 16 is scarcely sufficient to enable the dancer to distinguishthem. TABLE 13 GRAY DISCRIMINATION The Delicacy of Brightness Discrimination No. 420 GRAYS NOS. 10 GRAYS NOS. 20 AND 20 AND 15SERIES DATE DATE NO. 10 NO. 2 NO. 10 NO. 15 (RIGHT) (WRONG) (RIGHT) (WRONG) 1 Jan. 26 5 5 Feb. 6 8 2 2 27 8 2 6 5 5 3 28 6 4 7 9 1 4 28 2 8 7 7 3 5 29 1 9 8 5 5 6 29 6 4 8 6 4 7 30 9 1 9 5 5 8 30 7 3 9 6 4 9 31 6 4 10 8 2 10 31 4 6 10 3 7 11 Feb. 1 7 3 11 4 6 12 1 8 2 11 4 6 13 2 7 3 12 7 3 14 2 8 2 12 7 3 15 3 9 1 13 6 4 16 3 9 1 13 4 6 17 4 6 4 14 4 6 18 4 9 1 14 5 5 19 5 6 4 15 5 5 20 5 10 0 15 8 2 Averages 6. 6 3. 4 5. 8 4. 2 This result of the tests with gray papers surprised me very much at thetime of the experiments, for all my previous observation of the dancer hadled me to believe that it is very sensitive to light. It was only after along series of tests with transmitted light, in what is now to bedescribed as the Weber's law apparatus, that I was able to account for themeager power of discrimination which the mice exhibited in the gray tests. As it happened, the Weber's law experiment contributed quite asimportantly to the solution of our first problem as to that of the second, for which it was especially planned. For the Weber's law experiment a box similar to that used in the previousbrightness discrimination experiments (Figure 14) was so arranged that itstwo electric-boxes could be illuminated independently by the light fromincandescent lamps directly above them. The arrangements of the light-boxand the lamps, as well as their relations to the other important parts ofthe apparatus, are shown in Figure 17. The light-box consisted of twocompartments, of which one may be considered as the upward extension ofthe left electric-box and the other of the right electric-box. The light-box was pivoted at A and could be turned through an angle of 180° by theexperimenter. Thus, by the turning of the light-box, the lamp which in thecase of one test illuminated the left electric-box could be brought intosuch a position that in the case of the next test it illuminated the rightelectric-box. The practical convenience of this will be appreciated whenthe number of times that the brightnesses of the two boxes had to bereversed is considered. The light-box was left open at the top forventilation and the prevention of any considerable increase in thetemperature of the experiment box. In one side of each of the compartmentsof the light-box a slit (B, B of the figure) was cut out for anincandescent lamp holder. A strip of leatherette, fitted closely into inchgrooves at the edges of the slit, prevented light from escaping throughthese openings in the sides of the light-box. By moving the strips ofleatherette, one of which appears in the figure, C, the lamps could bechanged in position with reference to the bottom of the electric-box. Ascale, S, at the edge of each slit enabled the experimenter to determinethe distance of the lamp from the floor of the electric-box. The front ofthe light-box was closed, instead of being open as it appears in thefigure. [Illustration: Figure l7. --Weber's law apparatus for testing brightnessdiscrimination. Lower part, discrimination box similar to that of Figure14. Upper part, rotatory light-box, pivoted at A, and divided into twocompartments by a partition P in the middle. L, L, incandescent lampsmovable in slits, B, B, in which a narrow strip of leatherette, C, servesto prevent the escape of light. S, scale. ] This apparatus has the following advantages. First, the electric-boxes, between which the mouse is expected to discriminate by means of theirdifference in brightness, are illuminated from above and the lighttherefore does not shine directly from the lamps into the eyes of theanimal, as it approaches the entrances to the boxes. Choice is required, therefore, between illuminated spaces instead of between two directlyilluminated surfaces. Second, the amount of illumination of each electric-box can be accurately measured by the use of a photometer. Third, sincethe same kind of lamp is used in each box, and further, since the lampsmay be interchanged at any time, discrimination by qualitative instead ofquantitative difference in illumination is excluded. And finally, fourth, the tests can be made expeditiously, conveniently, and under such simpleconditions that there should be no considerable error of measurement or ofobservation within the range of brightness which must be used. It was my purpose in the experiment with this apparatus to ascertain howgreat the difference in the illumination of the two electric-boxes must bein order that the mouse should be able to choose the brighter of them. This I attempted to do by fixing an incandescent lamp of a certain knownilluminating power at such a position in one compartment of the light-boxthat the electric-box below it was illuminated by what I call a standardvalue, and by moving the incandescent lamp in the other compartment of thelight-box until the illumination of the electric-box below it was justsufficiently less than that of the standard to enable the dancer todistinguish them, and thereby to choose the brighter one. The light whichwas changed from series to series I shall call the _variable_, in contrastwith the _standard_, which was unchanged. The tests, which were made in a dark-room under uniform conditions, weregiven in series of fifty each; usually only one such series was given perday, but sometimes one was given in the morning and another in theafternoon of the same day. To prevent choice by position the lights werereversed in position irregularly, first one, then the other, illuminatingthe right electric-box. For the fifty tests of each initial series theorder of the changes in position was as follows: standard (brighter light)on the _l_ (left), _l, r_ (right), _r, l, l, r, r, l, r, l, r, l, l, r, r, l, l, r, r, l, l, l, r, r, r, l, r, l, r, r, r, l, l, l, r, r, r, l, l, r, l, r, l, r, l, r, l, r, l_. Twenty-five times in fifty the standard lightilluminated the right electric-box, and the same number of times itilluminated the left electric-box. When a second series was given underthe same conditions of illumination, a different order of change wasfollowed. In order to discover whether Weber's law holds in the case of thebrightness vision of the dancer it was necessary for me to determine thejust perceivable difference between the standard and the variable lightsfor two or more standard values. I chose to work with three values, 5, 20, and 80 hefners, and I was able to discover with a fair degree of accuracyhow much less than 5, 20, or 80 hefners, as the case might be, thevariable light had to be in order that it should be discriminable from theother. For the work with the 5 hefner standard I used 2-candle-powerlamps, [1] for the 20, 4-candle-power, and for the 80, 16-candle-power. [Footnote 1: I give merely the commercial markings of the lamps. They hadbeen photometered carefully by two observers by means of a Lummer-Brodhunphotometer and a Hefner amyl acetate lamp previous to their use in theexperiment. For the photometric measurements in connection with theWeber's law tests I made use of the Hefner lamp with the hope of attaininggreater accuracy than had been possible with a standard paraffine candle, in the case of measurements which I had made in connection with theexperiments on color vision that are reported in Chapters IX and X. TheHefner unit is the amount of light produced by an amyl acetate lamp at aflame height of 40 mm. (See Stine's "Photometrical Measurements. ") Aparaffine candle at a flame height of 50 mm. Is equal to 1. 2 Hefnerunits. ] For reasons which will soon appear, Weber's law tests were made with onlyone dancer. This individual, No. 51, had been thoroughly trained in white-black discrimination previous to the experiments in the apparatus which isrepresented in Figure 17. Having given No. 51 more than two hundredpreliminary tests in the Weber's law apparatus with the electric-boxessufficiently different in brightness to enable her to discriminatereadily, I began my experiments by trying to ascertain how much less thevalue of the illumination of one electric-box must be in order that itshould be discriminable from a value of 20 hefners in the other electric-box. In recording the several series of tests and their results hereafter, I shall state in Hefner units the value of the fixed or standard light andthe value of the variable light, the difference between the two in termsof the former, and the average number of wrong choices in per cent. With the lamps so placed that the difference in the illumination of thetwo electric-boxes was . 53 of the value of the standard, that is about onehalf, No. 51 made twenty wrong choices in one hundred, or 20 per cent. When the difference was reduced to . 36 (one third) the number of errorsincreased to 36 per cent, and with an intermediate difference of . 48 therewere 26 per cent of errors (see Table 14). Are these results indicative of discrimination, or are the errors inchoice too numerous to justify the statement that the dancer was able todistinguish the boxes by their difference in brightness? Evidently thisquestion cannot be answered satisfactorily until we have decided what thepercentage of correct choices should be in order that it be accepted asevidence of ability to discriminate, or, to put it in terms of errors, what percentage of wrong choices is indicative of the point of justperceivable difference in brightness. Theoretically, there should be asmany mistakes as right choices, 50 per cent of each, when the twoelectric-boxes are equally illuminated (indiscriminable), but in practicethis does not prove to be the case because the dancer tends to return tothat electric-box through which in the previous test it passed safely, whereas it does not tend in similar fashion to reënter the box in which ithas just received an electric shock. The result is that the percentage ofright choices, especially in the case of series which have the right boxin the same position two, three, or four times in succession, rises ashigh as 60 or 70, even when the visual conditions are indiscriminable. Abundant evidence in support of this statement is presented in ChaptersVII and IX, but at this point I may further call attention to the resultsof an experiment in the Weber's law apparatus which was made especially totest the matter. The results appear under the date of May 27 in Table 14. In this experiment, despite the fact that both boxes were illuminated by80 hefners, the mouse chose the standard (the illumination in which it wasnot shocked) 59 times in 100. In other words the percentage of error was41 instead of 50. It is evident, therefore, that as low a percentage oferrors as 40 is not necessarily indicative of discrimination. Anythingbelow 40 per cent is likely, however, to be the result of ability todistinguish the brighter from the darker box. To be on the safe side wemay agree to consider 25 wrong choices per 100 as indicative of a justperceivable difference in illumination. Fewer mistakes we shall considerindicative of a difference in illumination which is readily perceivable, and more as indicative of a difference which the mouse cannot detect. Thereader will bear in mind as he examines Table 14 that 25 per cent of wrongchoices indicates the point of just perceivable difference in brightness. TABLE 14 RESULTS OF WEBER'S LAW EXPERIMENTSBrightness vision DATE NUMBER STANDARD VARIABLE DIFFERENCE % OF ERRORS OF TESTS LIGHT LIGHT May 13 100 20 9. 4 . 53 20 15 100 20 12. 8 . 36 36 16 100 20 10. 8 . 46 26 20 50 80 37. 6 . 53 6 21 50 80 51. 3 . 36 10 22 100 80 71. 1 . 11 35 24 100 80 60. 0 . 25 21 25 100 80 65. 0 . 19 25 27 100 80 80 0 41 28 50 5 2. 5 . 50 18 29 50 5 4. 0 . 20 14 29 100 5 4. 5 . 10 25 31 50 5 4. 25 . 15 20June 1 50 5 4. 85 . 03 48 2 50 20 15. 0 . 25 16 3 50 20 17. 4 . 13 22 3 100 20 18. 0 . 10 22 4 100 80 72. 0 . 10 18 5 100 5 4. 5 . 10 12 7 100 5 4. 67 . 067 46 8 50 80 74. 67 . 067 56 9 50 20 18. 67 . 067 44 If we apply this rule to the results of the first tests, reported above, it appears that a standard of 20 hefners was distinguished from a variableof 9. 4 hefners (. 53 difference), for the percentage of errors was only 20. But in the case of a difference of . 36 in the illuminations lack ofdiscrimination is indicated by 36 per cent of errors. A difference of . 46gave a frequency of error so close to the required 25 (26 per cent) that Iaccepted the result as a satisfactory determination of the justperceivable difference for the 20 hefner standard and proceeded toexperiment with another standard value. The results which were obtained in the case of this second standard, thevalue of which was 80 hefners, are strikingly different from those for the20 hefner standard. Naturally I began the tests with this new standard bymaking the differences the same as those for which determinations had beenmade in the case of the 20 standard. Much to my surprise only 6 per centof errors resulted when the difference in illumination was . 53. I finallydiscovered that about . 19 difference (about one fifth) could bediscriminated with that degree of accuracy which is indicated by 25 percent of mistakes. So far as I could judge from the results of determinations for the 20 andthe 80 hefner standards, Weber's law does not hold for the dancer. Withthe former a difference of almost one half was necessary fordiscrimination; with the latter a difference of about one fifth could beperceived. But before presenting additional results I should explain theconstruction of Table 14, and comment upon the number of experiments whichconstitutes a set. The table contains the condensed results of several weeks of difficultexperimentation. From left to right the columns give the date of theinitial series of a given set of experiments, the number of experiments inthe set, the value of the standard light in hefners, the value of thevariable light, the difference between the lights in terms of the standard(the variable was always less than the standard), and the percentage oferrors or wrong choices. Very early in the investigation I discovered thatone hundred tests with any given values of the lights sufficed to revealwhatever discriminating ability the mouse possessed at the time. In someinstances either the presence or the lack of discrimination was so clear, as the result of 50 tests (first series), that the second series of 50 wasnot given. Consequently in the table the number of tests for the variousvalues of the lights is sometimes 100, sometimes 50. After finishing the experiments with the 80 standard on May 27 (see table)I decided, in spite of the evidence against Weber's law, to make testswith 5 as the standard, for it seemed not impossible that the lights weretoo bright for the dancer to discriminate readily. I even suspected that Imight have been working outside of the brightness limits in which Weber'slaw holds, if it holds at all. The tests soon showed that a difference ofone tenth made discrimination possible in the case of this standard. Ifthe reader will examine the data of the table, he will note that adifference of . 20 gave 14 per cent of mistakes; a difference of . 03, 48per cent. Evidently the former difference is above the threshold, thelatter below it. But what of the interpretation of the results in terms ofWeber's law? The difference instead of being one half or one fifth, as itwas in the cases of the 20 and 80 standards respectively, has now becomeone tenth. Another surprise and another contradiction! Had these three differences either increased or decreased regularly withthe value of the standard I should have suspected that they indicated aprinciple or relationship which is different from but no less interestingthan that which Weber's law expresses. But instead of reading 5 standard, difference one tenth; 20 standard, difference one fifth; 80 standard, difference one half: or 5 standard, difference one half; 20 standard, difference one fifth; 80 standard, difference one tenth: they read 5standard, difference one tenth; 20 standard, difference one half; 80standard, difference one fifth. What does this mean? I could think of noother explanation than that of the influence of training. It seemed notimpossible, although not probable, that the mouse had been improving inability to discriminate day by day. It is true that this in itself wouldbe quite as interesting a fact as any which the experiment might reveal. To test the value of my supposition, I made additional experiments withthe 20 standard, the results of which appear under the dates June 2 and 3of the table. These results indicate quite definitely that the animal hadbeen, and still was, improving in her ability to discriminate. For insteadof requiring a difference of about one half in order that she mightdistinguish the 20 standard from the variable light she was now able todiscriminate with only 22 per cent of errors when the difference was onetenth. As it seemed most improbable that improvement by training should continuemuch longer, I next gave additional tests with the 80 standard. Again adifference of one tenth was sufficient for accurate discrimination (18 percent of errors). These series were followed immediately by further testswith the 5 standard. As the results indicated greater ease ofdiscrimination with a difference of one tenth in the case of this standardthan in the case of either of the others I was at first uncertain whetherthe results which I have tabulated under the dates June 3, 4, and 5 of thetable should be interpreted in terms of Weber's law. Up to this point the experiments had definitely established two facts:that the dancer's ability to discriminate by means of brightnessdifferences improves with training for a much longer period and to a fargreater extent than I had supposed it would; and that a difference of onetenth is sufficient to enable the animal to distinguish two lights in thecase of the three standard values, 5, 20, and 80 hefners. The questionremains, is this satisfactory evidence that Weber's law holds with respectto the brightness vision of the dancer, or do the results indicate rather, that this difference is more readily detected in the case of 5 as astandard (12 per cent error) than in the case of 20 as a standard (22 percent error)? For the purpose of settling this point I made tests for each of the threestandards with a difference of only one fifteenth. In no instance did Iobtain the least evidence of ability to discriminate. These final tests, in addition to establishing the fact that the limit of discrimination forNo. 51, after she had been subjected to about two thousand tests, laybetween one tenth and one fifteenth, proved to my satisfaction, when takenin connection with the results already discussed, that Weber's law doeshold for the brightness vision of the dancer. In concluding this discussion of the Weber's law experiment I wish to callattention to the chief facts which have been revealed, and to make acritical comment. In my opinion it is extremely important that the studentof animal behavior should note the fact that the dancer with which Iworked week after week in the Weber's law investigation gradually improvedin her ability to discriminate on the basis of brightness differencesuntil she was able to distinguish from one another two boxes whosedifference in illumination was less than one tenth[1] that of the brighterbox. At the beginning of the experiments a difference of one half did notenable her to choose as certainly as did a difference of one tenth aftershe had chosen several hundred times. Evidently we are prone tounderestimate the educability of our animal subjects. [Footnote 1: Under the conditions of the experiment I was unable todistinguish the electric-boxes when they differed by less than onetwentieth. ] The reason that the experiments were carried out with only one mouse mustnow be apparent. It was a matter of time. The reader must not suppose thatmy study of this subject is completed. It is merely well begun, and Ireport it here in its unfinished state for the sake of the value of themethod which I have worked out, rather than for the purpose of presentingthe definite results which I obtained with No. 51. The critical comment which I wish to make for the benefit of those who areworking on similar problems is this. The phosphor bronze wires, on thebottom of the electric-boxes, by means of which an electric shock could begiven to the mouse when it chose the wrong box, are needless sources ofvariability in the illumination of the boxes. They reflect the light intothe eyes of the mouse too strongly, and unless they are kept perfectlyclean and bright, serious inequalities of illumination appear. To avoidthese undesirable conditions I propose hereafter to use a box within abox, so that the wires shall be hidden from the view of the animal as itattempts to discriminate. A brief description of the behavior of the dancer in the brightnessdiscrimination experiments which have been described may veryappropriately form the closing section of this chapter. For theexperimenter, the incessant activity and inexhaustible energy of theanimal are a never-failing source of interest and surprise. When a danceris inactive in the experiment box, it is a good indication either ofindisposition or of too low a temperature in the room. In no animal withwhich I am familiar is activity so much an end in itself as in this oddspecies of mouse. With striking facility most of the mice learn to openthe wire swing doors from either side. They push them open with theirnoses in the direction in which they were intended by the experimenter towork, and with almost equal ease they pull them open with their teeth inthe direction in which they were not intended to work. In the rapiditywith which this trick is learned, there are very noticeable individualdifferences. The pulling of these doors furnished an excellent opportunityfor the study of the imitative tendency. When confronted with the two entrances of the electric-boxes, the dancermanifested at first only the hesitation caused by being in a strangeplace. It did not seem much afraid, and usually did not hesitate longbefore entering one of the boxes. The first choice often determined themajority of the choices of the preference series. If the mouse happened toenter the left box, it kept on doing so until, having become so accustomedto its surroundings that it could take time from its strenuous runningfrom _A_ by way of the left box to the alley and thence to _A_, to examinethings in _B_ a little, it observed the other entrance and in a seeminglyhalf-curious, half-venturesome way entered it. In the case of otherindividuals, the cardboards themselves seemed to determine the choicesfrom the first. The electric shock, as punishment for entering the wrong box, came as asurprise. At times an individual would persistently attempt to enter, oreven enter and retreat from the wrong box repeatedly, in spite of theshock. This may have been due in some instances to the effects of fright, but in others it certainly was due to the strength of the tendency tofollow the course which had been taken most often previously. The nexteffect of the shock was to cause the animal to hesitate before theentrances to the boxes, to run from one to the other, poking its head intoeach and peering about cautiously, touching the cardboards at theentrances, apparently smelling of them, and in every way attempting todetermine which box could be entered safely. I have at times seen a mouserun from one entrance to the other twenty times before making its choice;now and then it would start to enter one and, when halfway in, draw backas if it had been shocked. Possibly merely touching the wires with itsfore paws was responsible for this simulation of a reaction to the shock. The gradual waning of this inhibition of the forward movement was one ofthe most interesting features of the experiment. Could we but discoverwhat the psychical states and the physiological conditions of the animalwere during this period of choosing, comparative psychology and physiologywould advance by a bound. If the conditions at the entrances of the two boxes were discriminable, the mouse usually learned within one hundred experiences to choose theright box without much hesitation. Three distinct methods of choice wereexhibited by different individuals, and to a certain extent by the sameindividual from time to time. These methods, which I have designated_choice by affirmation_, _choice by negation_, and _choice by comparison_, are of peculiar interest to the psychologist and logician. When an individual runs directly to the entrance of the right box, and, after stopping for an instant to examine it, enters, the choice may bedescribed as recognition of the right box. I call it choice by affirmationbecause the act of the animal is equivalent to the judgment--"this is it. "If instead it runs directly to the wrong box, and, after examining it, turns to the other box and enters without pause for examination, itsbehavior may be described as recognition of the wrong box. This I callchoice by negation because the act seems equivalent to the judgment--"thisis not it. " Further, it seems to imply the judgment--"therefore the otheris it. " In the light of this fact, this type of choice might appropriatelybe called choice by exclusion. Finally, when the mouse runs first to onebox and then to the other, and repeats this anywhere from one to fiftytimes, the choice may be described as comparison of the boxes; therefore, I call it choice by comparison. Certain individuals choose first bycomparison, and later almost uniformly by affirmation and negation. Whenever the conditions are difficult to discriminate, choice bycomparison occurs most frequently and persistently. If, however, theconditions happen to be absolutely indiscriminable, as was true, forexample, in the case of the sound tests, in certain of the Weber's lawtests, and in the plain electric-box tests, the period of hesitationrapidly increases during the first three or four series of tests, then themouse seems to lessen its efforts to discriminate and more and more tendsto rush into one of the boxes without hesitation or examination, andapparently with the expectation of a shock, but with the intention ofgetting it over as soon as possible. Now and then under such conditionsthere is a marked tendency to enter the same box each time. Indiscriminable conditions are likely to render the animals fearful of theexperiment; instead of going from _A_ to _A_ willingly, they fight againstmaking the trip. They refuse to pass from _A_ to _B_; and when in _B_, they fight against being driven toward the entrances to the electric-boxes. In marked contrast with this behavior on the part of the mouse underconditions which do not permit it to choose correctly is that of theanimal which has learned what is expected of it. The latter, far fromholding back or fighting against the conditions which urge it forward, isso eager to make the trip that it sometimes has to be forced to wait whilethe experimenter records the results of the tests. There is evidence ofdelight in the freedom of movement and in the variety of activity whichthe experiment furnishes. The thoroughly trained dancer runs into _B_almost as soon as it has been placed in _A_ by the experimenter; itchooses the right entrance by one of the three methods described above, immediately, or after whirling about a few times in _B_; it runs through_E_ and back to _A_ as quickly as it can, and almost before theexperimenter has had time to record the result of the choice it is againin _B_ ready for another choice. CHAPTER IX THE SENSE OF SIGHT: COLOR VISION Is the dancing mouse able to discriminate colors as we do? Does it possessanything which may properly be called color vision? If so, what is thenature of its ability in this sense field? Early in my study of the mice Iattempted to answer these and similar questions, for the fact that theyare completely deaf during the whole or the greater part of their livessuggested to me the query, are they otherwise defective in senseequipment? In the following account of my study of color vision, I shalldescribe the evolution of my methods in addition to stating the resultswhich were obtained and the conclusions to which they led me. For in thiscase the development of a method of research is quite as interesting asthe facts which the method in its various stages of evolution revealed. Observation of the behavior of the dancer under natural conditions causedme to suspect that it is either defective in color vision or possesses asense which is very different from human vision. I therefore devised thefollowing extremely simple method of testing the animal's ability todistinguish one color from another. In opposite corners of a wooden box 26cm. Long, 23 cm. Wide, and 11 cm. Deep, two tin boxes 5 cm. In diameterand 1. 5 cm. Deep were placed, as is shown in part I of Figure 18. One ofthese boxes was covered on the outside with blue paper (_B_ of Figure 18), and the other with orange[1] (_O_ of Figure 18). A small quantity of"force" was placed in the orange box. As the purpose of the test was todiscover whether the animals could learn to go directly to the box whichcontained the food, the experiments were made each morning before the micehad been fed. The experimental procedure consisted in placing theindividual to be tested in the end of the large wooden box opposite thecolor boxes, and then permitting it to run about exploring the box untilit found the food in the orange box. While it was busily engaged in eatinga piece of "force" which it had taken from the box and was holding in itsfore paws, squirrel fashion, the color boxes were quickly and withoutdisturbance shifted in the directions indicated by the arrows of Figure18, I. Consequently, when the animal was ready for another piece of"force, " the food-box was in the corresponding corner of the opposite endof the experiment box (position 2, 18, II). After the mouse had againsucceeded in finding it, the orange box was shifted in position as isindicated by the arrows in Figure 18, II. Thus the tests were continued, the boxes being shifted after each success on the part of the animal insuch a way that for no two successive tests was the position of the food-box the same; it occupied successively the positions 1, 2, 3, and 4 of thefigure, and then returned to 1. Each series consisted of 20 tests. [Footnote 1: These were the Milton Bradley blue and orange papers. ] [Illustration: FIGURE 18. --Food-box apparatus for color discriminationexperiments. _O_, orange food-box; _B_, blue food-box; 1, 2, 3, 4, different positions of the food-boxes, _O_ and _B_; I, II, III, IV, figures in which the arrows indicate the direction in which the food-boxeswere moved. ] [Illustration: FIGURE 19. --Food-box apparatus with movable partitions. _O_, orange food-box; _B_, blue food-box; _X_, starting point for mouse;_A_, point at which both food-boxes become visible to the mouse as itapproaches them; 1, 2, two different positions of the food-boxes; _T_, _T_, movable partitions. (After Doctor Waugh. )] An improvement on this method, which was suggested by Doctor Karl Waugh, has been used by him in a study of the sense of vision in the commonmouse. It consisted in the introduction, at the middle of the experimentbox, of two wooden partitions which were pivoted on their mid-verticalaxes so that they could be placed in either of the positions indicated inFigure 19. Let us suppose that a mouse to be tested for color vision inthis apparatus has been placed at _X_. In order to obtain food it mustpass through _A_ and choose either the orange or the blue box. If itchooses the former, the test is recorded as correct; if it goes to theblue box first, and then to the orange, it is counted an error. While theanimal is eating, the experimenter shifts the boxes to position 1 ofFigure 19, and at the same time moves the partitions so that they occupythe position indicated by the dotted lines. The chief advantage of thisimprovement in method is that the animal is forced to approach the colorboxes from a point midway between them, instead of following the sides ofthe experiment box, as it is inclined to do, until it happens to come tothe food-box. This renders the test fairer, for presumably the animal hasan opportunity to see both boxes from _A_ and can make its choice at thatpoint of vantage. Two males, A and B, of whose age I am ignorant, were each given seventeenseries of tests in the apparatus of Figure 18. A single series, consistingof twenty choices, was given daily. Whether the animal chose correctly ornot, it was allowed to get food; that is, if it went first to the bluebox, thus furnishing the condition for a record of error, it was permittedto pass on to the orange box and take a piece of "force. " No attempt wasmade to increase the animal's desire for food by starving it. Usually itsought the food-box eagerly; when it would not do so, the series wasabandoned and work postponed. "Force" proved a very convenient form offood in these tests. The mice are fond of it, and they quickly learned totake a flake out of the box instead of trying to get into the box and sitthere eating, for when they attempted the latter they were promptly pushedto one side by the experimenter and the box, as well as the food, wasremoved to a new position. The results of the tests appear in Table 15. No record of the choices inthe first two of the 17 series was kept. The totals therefore include 15series, or 300 tests, with each individual. Neither the daily records northe totals of this table demonstrate choice on the basis of colordiscrimination. Either the dancers were not able to tell one box from theother, or they did not learn to go directly to the orange box. It might beurged with reason that there is no sufficiently strong motive for theavoidance of an incorrect choice. A mistake simply means a moment's delayin finding food, and this is not so serious a matter as stopping todiscriminate. I am inclined, in the light of result of other experiments, to believe that there is a great deal in this objection to the method. Reward for a correct choice should be supplemented by some form ofpunishment for a mistake. This conclusion was forced upon me by theresults of these preliminary experiments on color vision and by myobservation of the behavior of the animals in the apparatus. At the timethe above tests were made I believed that I had demonstrated the inabilityof the dancer to distinguish orange from blue, but now, after two years'additional work on the subject, I believe instead that the method wasdefective. The next step in the evolution of a method of testing the dancer's colorvision was the construction of the apparatus (Figures 14 and 15) which wasdescribed in Chapter VII. In connection with this experiment box the basisfor a new motive was introduced, namely, the punishment of mistakes by anelectric shock. Colored cardboards, instead of the white, black, or graysof the brightness tests, were placed in the electric-boxes. TABLE 15 ORANGE-BLUE TESTS, WITH FOOD-BOX MOUSE A MOUSE BSERIES DATE 1904 RIGHT WRONG RIGHT WRONG (ORANGE) (BLUE) (ORANGE) (BLUE) 1 Dec. 6 -- -- -- -- 2 7 -- -- -- -- 3 8 12 8 12 8 4 9 10 10 9 11 5 10 15 5 10 10 6 11 10 10 12 8 7 12 9 11 9 11 8 13 10 10 9 11 9 14 12 8 12 8 10 15 13 7 12 8 11 16 13 7 10 10 12 17 12 8 10 10 13 18 11 9 10 10 14 19 13 7 8 12 15 20 13 7 9 11 16 22 14 6 12 8 17 23 10 10 9 11 TOTALS 177 123 153 147 In preliminary tests, at the rate of four per day, the colored cardboardswere placed only at the entrances to the boxes, not inside, and as wastrue also in the case of brightness tests under like conditions, noevidence of discrimination was obtained from ten days' training. Thisseemed to indicate that a considerable area of the colored surface shouldbe exposed to the mouse's view, if discrimination were to be madereasonably easy. This conclusion was supported by the results of other preliminaryexperiments in which rectangular pieces of colored papers[1] 6 by 3 cm. , were placed on the floor at the entrances to the electric-boxes, insteadof on the walls of the boxes. Mouse No. 2 was given five series of tentests each with a yellow card to indicate the right box and a red card atthe entrance to the wrong box. At first he chose the red almost uniformly, and at no time during these fifty tests did he exhibit ability to choosethe right box by color discrimination. I present the results of theseseries in Table 16, because they indicate a fact to which I shall have torefer repeatedly later, namely, that the brightness values of differentportions of the spectrum are not the same for the dancer as for us. Previous to this yellow-red training, No. 2, as a result of ten days ofwhite-black training (two tests per day), had partially learned to go tothe brighter of the two electric-boxes. It is possible therefore that thechoice of the box in the case of these color experiments was in realitythe choice of what appeared to the mouse to be the brighter box. If thiswere not true, how are the results of Table 16 to be accounted for? [Footnote 1: These were the only Hering papers used in my experiments. ] TABLE 16 YELLOW-RED TESTS In Color Discrimination Box with 6 by 3 cm. Pieces of HeringPapers at Entrances to Boxes No. 2 SERIES DATE RIGHT WRONG 1906 (Yellow) (Red) 1 Jan. 16 1 9 2 17 3 7 3 18 4 6 4 19 5 5 5 20 5 5 Without further mention of the many experiments which were necessary forthe perfecting of this method of testing color vision, I may at oncepresent the final results of the tests which were made with reflectedlight. These tests were made with the discrimination apparatus inessentially the same way as were the brightness discrimination tests ofChapter VII. In all of the color experiments, unless otherwise stated, a series of tentests each day was given, until satisfactory evidence of discrimination orproof of the lack of the ability to discriminate had been obtained. Thedifficulties of getting conclusive evidence in either direction will beconsidered in connection with the results themselves. For all of thesetests with reflected light the Milton Bradley colored papers were used. These colored papers were pasted on white cardboard carriers. I shalldesignate, in the Bradley nomenclature, the papers used in eachexperiment. With colored cardboards inside the electric-boxes as well as at theirentrances (see Figure 14 for position of cardboards) blue-orange testswere given to Nos. 2 and 3 until they discriminated perfectly. The paperswere Bradley's blue tint No. 1 and orange. Number 2 was perfect in thetwelfth series (Table 17), No 3 in the fourteenth and again in thesixteenth. They were then tested with a special brightness check serieswhich was intended by the experimenter to reveal any dependence upon apossible brightness difference rather than upon the color difference ofthe boxes. TABLE 17 LIGHT BLUE-ORANGE TESTS IN COLOR DISCRIMINATION BOX SERIES DATE NO. 2 NO. 3 1906 RIGHT WRONG RIGHT WRONG (LIGHT (ORANGE) (LIGHT (ORANGE) BLUE) BLUE) 1 Jan. 26 7 3 1 9 2 27 7 3 5 5 3 28 7 3 6 4 4 29 7 3 7 3 5 30 7 3 4 6 6 31 10 0 7 3 7 Feb. 1 9 1 7 3 8 2 8 2 6 4 9 3 9 1 9 1 10 5 7 3 5 5 11 6 8 2 5 5 12 7 10 0 5 5Special brightness check series (see Table 18) 13 8 10 0 7 3Special light blue-dark blue series 14 9 8 2 10 0 15 10 9 1 9 1Special light blue-dark blue series 16 11 9 1 10 0 Special brightness check series 17 12 10 0 9 1 TABLE 18 LIGHT BLUE-ORANGE Brightness check series Mouse No. 2, Series 13 Feb. 8, 1906 TEST CONDITION RIGHT WRONG 1 Light blue on right Orange on left Right ____ 2 Light blue on left Orange on right Right ____ 3 Light blue on right Red substituted for orange Right ____ 4 Light blur on left Red substituted for orange Right ____ 5 Dark blue on right Orange on left Right ____ 6 Dark blue on left Orange on right Right ____ 7 Dark blue on left Orange on right Right ____ 8 Dark blue on right Red substituted for orange Right ____ 9 Dark blue on left Red substituted for orange Right ____ 10 Dark blue on left Red substituted for orange Right ____ Totals 10 0 The nature of this brightness check series, as well as the results whichNo. 2 gave when tested by it, may be appreciated readily by reference toTable 18. Tint No. 1 of the blue, which is considerably brighter, in myjudgment, than the Bradley blue, was replaced at intervals in this seriesby the latter. For it was thought that in case the mouse were choosing theblue of the series because it seemed brighter than the orange, thissubstitution might mislead it into choosing the orange. These blues arereferred to in the table as light blue (tint No. 1) and dark blue(standard blue). Again a change in the opposite direction was made bysubstituting Bradley red for orange. As this was for the human eye thesubstitution of a color whose brightness was considerably less than thatof the orange, it seemed possible that the mouse, if it had formed thehabit of choosing the box which seemed the darker, might by this change bemisled into choosing the red instead of the light blue. In a word, changesin the conditions of the experiments were made in such a way that now onecolor, now the other, appeared to be the brighter. But I did not attemptto exclude brightness discrimination on the part of the mouse bydependence upon the human judgment of brightness equality, for it ismanifestly unsafe to assume that two colors which are of the samebrightness for the human eye have a like relation for the eye of thedancer or of any other animal. My tests of color vision have beenconducted without other reference to human standards of judgment orcomparisons than was necessary for the description of the experimentalconditions. In planning the experiments I assumed neither likeness nordifference between the human retinal processes and those of the dancer. Itwas my purpose to discover the nature of the mouse's visual discriminativeability. As is indicated in the tables, neither the substitution of dark blue forlight blue, nor the replacement of the orange by red or dark blue renderedcorrect choice impossible, although certain of the combinations did renderchoice extremely difficult. In other words, despite all of the changeswhich were made in the brightness of the cardboards in connection with thelight blue-orange tests, the mice continued to make almost perfectrecords. What are we to conclude from this? Either that the ability todiscriminate certain colors is possessed by the dancer, or that for somereason the tests are unsatisfactory. If it be granted that the possibilityof brightness discrimination was excluded in the check series, the firstof these alternatives apparently is forced upon us. That such apossibility was not excluded, later experiments make perfectly clear. Thefact was that not even in the check series was the brightness value of theorange as great as that of the blue. Consequently the mice may have chosenthe brighter box each time while apparently choosing the blue. Although conclusive proof of the truth of this statement is furnished onlyby later experiments, the results of the light blue-orange series, asgiven in Table 17, strongly suggest such a possibility. Mouse No. 3 hadnot been experimented with previous to these color discrimination tests. Her preference for the orange, which in the case of the first series was 9to 1, consequently demands an explanation. If she had been trainedpreviously to choose the white instead of the black, as was true of No. 2, it might be inferred that she went to the orange box because it appearedbrighter than the blue. As this explanation is not available, we aredriven back upon the results of the white-black preference tests inChapter VII, which proved that many dancers prefer the black to the white. This may mean that they prefer the lower degree of brightness orillumination, and if so it might be argued, in turn, that the orange waschosen by No. 3 because it appeared darker than the blue. Since, as hasalready been stated, the orange was far brighter for me than the blue, this would also mean that the brightness values of different colors arenot the same for man and mouse. Practically the same kind of color tests as those described for Nos. 2 and3 were given to Nos. 1000 and 5. The results appear in Table 19. Thesetests followed upon the formation of a habit to choose white instead ofblack (that is, the greater brightness). From the first both No. 1000 andNo. 5 chose the light blue in preference to the orange or the red. Ittherefore seems probable that the former was considerably brighter thanthe latter. Number 1000, to be sure, was led into three erroneous choicesby the brightness check series (series 7), but, on the other hand, No. 5was not at all disturbed in her choices by similar check tests. It seemsnatural to conclude from these facts that both of these mice chose theblue at first because of its relatively greater brightness, and that theycontinued to do so for the same reason. In other words, their behaviorindicates that the brightness check tests were valueless because notenough allowance had been made for the possible differences between thevision of mouse and man. TABLE 19LIGHT BLUE-ORANGE AND DARK BLUE-RED TESTS No. 1000 No. 5SERIES DATE Condition Right Wrong Right Wrong (Light (Orange (Light (Orange Blue or or Blue or Dark Red) Dark Red) Blue) Blue) 1 Jan. 25 Blue-red 8 2 10 0 2 26 Blue-red or Light blue-orange 10 0 10 0 3 27 Light blue-orange 10 0 5 5 4 29 Light blue-orange 9 1 8 2 5 30 Light blue-orange 10 0 8 2 6 31 Light blue-orange 10 0 10 0 7 Feb. 1 Light blue-orange or Dark blue-red 7 3 10 0 If only the final results of my experiments with the dancer and theconclusions to which they lead were of interest, all of this descriptionof experiments which served merely to clear the ground and thus makepossible crucial tests might be omitted. It has seemed to me, however, that the history of the investigation is valuable, and I am thereforepresenting the evolution of my methods step by step. To be sure, not everydetail of this process can be mentioned, and only a few of the individualresults can be stated, but my purpose will have been fulfilled if Isucceed in showing how one method of experimentation pointed the way toanother, and how one set of results made possible the interpretation ofothers. As the results of my color vision experiments seemed to indicate that thered end of the spectrum appears much darker to the dancer than to us, tests were now arranged with colors from adjacent regions of the spectrum, green and blue. The papers used were the Bradley green and tint No. 1 ofthe blue. They were not noticeably different in brightness for the humaneye. Green marked the box to be chosen. Three of the individuals which hadpreviously been used in the light blue-orange series, and which thereforehad perfect habits of going to the light blue, were used for the green-light blue tests. Of these individuals, No. 1000 became inactive on thefifth day of the experiment, and the tests with him were discontinued. Twenty series were given to each of the other mice, with the results whichappear in Table 20. To begin with, both No. 4 and No. 5 exhibited apreference for the light blue, as a result of the previous light blue-orange training. As this preference was gradually destroyed by theelectric shock which was received each time the light blue box wasentered, they seemed utterly at a loss to know which box to enter. Occasionally a record of six, seven, or even eight right choices would bemade in a series, but in no case was this unquestionably due to colordiscrimination; usually it could be explained in the light of the order ofthe changes in the positions of the cardboards. For example, series 9, inwhich No. 5 made a record of 8 right and 2 wrong, had green on the rightfor the first three tests. The animal happened to choose correctly in thefirst test, and continued to do so three times in succession simplybecause there was no change in the position of the cardboards. I haveoccasionally observed a record of seven right choices result when it wasperfectly evident to the observer that the mouse could not discriminatevisually. It was to avoid unsafe conclusions and unfair comparisons, asthe result of such misleading series, that three perfect series insuccession were required as evidence of a perfectly formed habit ofdiscrimination. TABLE 20 GREEN-LIGHT BLUE TESTS Date No. 1000 No. 4 No. 5SERIES 1906 RIGHT WRONG RIGHT WRONG RIGHT WRONG (GREEN) (BLUE) (GREEN) (BLUE) (GREEN) (BLUE) 1 Feb. 3 2 8 3 7 3 7 2 5 7 3 5 5 5 5 3 6 5 5 6 4 5 5 4 7 5 5 5 5 5 5 5 8 2 8 5 5 4 6 6 9 7 3 7 3 7 10 4 6 3 7 8 10 6 4 4 6 9 12 6 4 8 2 10 13 6 4 6 4 11 14 5 5 3 7 12 15 6 4 7 3 13 16 5 5 7 3 14 17 3 7 6 4 15 19 6 4 6 4 16 20 7 3 5 5 17 21 4 6 8 2 18 22 3 7 4 6 19 23 6 4 4 6 20 24 6 4 5 5 Twenty series, 200 tests for each of the individuals in the experiment, yielded no evidence whatever of the dancer's ability to tell green fromblue. As it has already been proved that they readily learn to choose theright box under discriminable conditions, it seems reasonable to concludeeither that they lack green-blue vision, or that they have it in arelatively undeveloped state. If it be objected that the number of training tests given was too small, and that the dancer probably would exhibit discrimination if it were given1000 instead of 200 tests in such an experiment, I must reply that thebehavior of the animal in the tests is even more satisfactory evidence ofits inability to choose than are the results of Table 20. Had there beenthe least indication of improvement as the result of 200 tests, I shouldhave continued the experiment; as a matter of fact, the mice each dayhesitated more and more before choosing, and fought against being driventoward the entrance to the experiment box. That they were helpless was soevident that it would have been manifestly cruel to continue theexperiment. TABLE 21 VIOLET-RED TESTS With Odor of All Cardboards the Same SERIES DATE NO. 7 NO. 998 RIGHT WRONG RIGHT WRONG (VIOLET) (RED) (VIOLET) (RED) A MAR. 7 8 2 5 5 B 7 3 7 2 8 1 14 3 7 6 4 2 15 4 6 4 6 3 16 5 5 5 5 4 19 4 6 4 6 5 20 5 5 6 4 6 21 4 6 8 2 7 22 8 2 4 6 8 23 4 6 6 4 9 24 6 4 4 6 10 25 4 6 6 4 Further color tests with reflected light were made with violet and red. Two dancers, Nos. 998 and 7, neither of which had been in any experimentpreviously, were subjected to the ten series of tests whose results are tobe found in Table 21. In this experiment the cardboards used had beencoated with shellac to obviate discrimination by means of odor. It istherefore impossible to give a precise description of the color orbrightness by referring to the Bradley papers. [1] Both the violet and thered were rendered darker, and apparently less saturated, by the coating. [Footnote 1: The violet was darker than Bradley's shade No. 2, and the redwas lighter than Bradley's red. ] These violet-red tests were preceded by two series of preference tests(_A_ and _B_), in which no shock was given and escape was possible througheither electric-box. Although the results of these preference tests asthey appear in Table 21 seem to indicate a preference for the red on thepart of No. 998, examination of the record sheets reveals the fact thatneither animal exhibited color preference, but that instead both chose byposition. Number 998 chose the box on the right 15 times in 20, and No. 7chose the box on the left 15 times in 20. Ten series of tests with the violet-red cardboards failed to furnish theleast indication of discrimination. The experiment was discontinuedbecause the mice had ceased to try to discriminate and dashed into one orthe other of the boxes on the chance of guessing correctly. When wrongthey whirled about, rushed out of the red box and into the violetimmediately. They had learned perfectly as much as they were able to learnof what the experiment required of them. Although we are not justified inconcluding from this experiment that dancers cannot be taught todistinguish violet from red, there certainly is good ground for thestatement that they do not readily discriminate between these colors. The experiments on color vision which have been described and the recordswhich have been presented will suffice to give the reader an accurateknowledge of the nature of the results, only a few of which could beprinted, and of the methods by which they were obtained. In brief, these results show that the dancer, under the conditions of theexperiments, is not able to tell green from blue, or violet from red. Theevidence of discrimination furnished by the light blue-orange tests is notsatisfactory because the conditions of the experiment did not permit theuse of a sufficiently wide range of brightnesses. It is obvious, therefore, that a method of experimentation should be devised in which theexperimenter can more fully control the brightness of the colors which heis using. I shall now describe a method in which this was possible. CHAPTER X THE SENSE OF SIGHT: COLOR VISION (_Continued_) There are three well-known ways in which colors may be used as stimuli inexperiments on animals: by the use of colored papers (reflected light); bythe use of a prism (the spectrum which is obtained may be used as directlytransmitted or as reflected light); and by the use of light filters(transmitted light). In the experiments on the color vision of the dancerwhich have thus far been described only the first of these three methodshas been employed. Its advantages are that it enables the experimenter towork in a sunlit room, with relatively simple, cheap, and easilymanipulated apparatus. Its chief disadvantages are that the brightness ofthe light can neither be regulated nor measured with ease and accuracy. The use of the second method, which in many respects is the most desirableof the three, is impracticable for experiments which require as large anilluminated region as do those with the mouse; I was therefore limited tothe employment of light filters in my further tests of colordiscrimination. The form of filter which is most conveniently handled is the coloredglass, but unfortunately few glasses which are monochromatic aremanufactured. Almost all of our so-called colored glasses transmit thelight of two or more regions of the spectrum. After making spectroscopicexaminations of all the colored glasses which were available, I decidedthat only the ruby glass could be satisfactorily used in my experiments. With this it was possible to get a pure red. Each of the other colors wasobtained by means of a filter, which consisted of a glass box filled witha chemical solution which transmitted light of a certain wave length. For the tests with transmitted light the apparatus of Figures 20 and 21was constructed. It consisted of a reaction-box essentially the same asthat used in the brightness vision tests, except that holes were cut inthe ends of the electric-boxes, at the positions _G and R_ of Figure 20, to permit the light to enter the boxes. Beyond the reaction-box was a longlight-box which was divided lengthwise into two compartments by apartition in the middle. A slit in the cover of each of these compartmentscarried an incandescent lamp _L_ (Figure 20). Between the two lamps, _L, L_, and directly over the partition in the light-box was fastened amillimeter scale, _S_, by means of which the experimenter could determinethe position of the lights with reference to the reaction-box. The light-box was separated from the reaction-box by a space 6 cm. Wide in whichmoved a narrow wooden carrier for the filter boxes. This carrier, as shownin Figure 20, could be moved readily from side to side through a distanceof 20 cm. The filter boxes, which are represented in place in Figures 20and 21, consisted of three parallel-sided glass boxes 15 cm. Long, 5 cm. Wide, and 15 cm. Deep. Each box contained a substance which acted as a rayfilter. Tightly fitted glass covers prevented the entrance of dust and theevaporation of the solutions in the boxes. Figures 20 and 21 represent thetwo end boxes, _R, R_, as red light filters and the middle one, _G_, as agreen light filter. Three filters were used thus side by side in orderthat the position of a given color with reference to the electric-boxesmight be changed readily. As the apparatus was arranged, all theexperimenter had to do when he wished to change from green-left, red-rightto green-right, red-left was to push the carrier towards the right untilthe green filter covered the hole on the right at the end of the electric-box. When this had been accomplished the red filter at the left end of thecarrier covered the hole on the left at the end of the electric-box. Thusquickly, noiselessly, easily, and without introducing any other change inconditions than that of the interchange of lights, the experimenter wasable to shift the positions of his colored lights at will. [Illustration: FIGURE 20. --Color discrimination apparatus. _A, _ nest-box;_B, _ entrance chamber; _R, R, _ red filters; _G, _ green filter; _L, L, _incandescent lamps in light-box; _S, _ millimeter scale on light-box; _I, _door between _A_ and _B; O, O, _ doors between alleys and _A_. ] [Illustration: FIGURE 21--Ground plan of color discrimination apparatus. _E, E_, exits from electric-boxes. _LB_, light-box; _R, G, R_, filterboxes on carrier; _L_, left electric-box; _R_, right electric-box; _IC_induction apparatus; _C_, electric cell; _K_, key; _S_, millimeter scale. ] In the tests which are now to be reported, three portions of the spectrumwere used: the red end, the blue-violet end, and a middle region, chieflygreen. The red light was obtained by the use of a filter which was made byplacing two plates of ruby glass in one of the glass boxes, filling thebox with filtered water and then sealing it to prevent evaporation. Theblue-violet was obtained by the use of a filter box which contained a 5per cent solution of copper ammonium sulphate. The green, which, however, was not monochromatic, was obtained by the use of a filter box whichcontained a saturated solution of nickel nitrate. These three sets offilters were examined spectroscopically both before the experiments hadbeen made and after their completion. [1] The red filters, of which I hadtwo for shifting the lights, transmitted only red light. The blue-violetfilters, two also, at first appeared to transmit only portions of the blueand violet of the spectrum, but my later examination revealed a trace ofgreen. It is important to note, however, that the red and the blue-violetfilters were mutually exclusive in the portions of the spectrum which theytransmitted. Of all the filters used the green finally proved the leastsatisfactory. I detected some yellow and blue in addition to green in myfirst examination, and later I discovered a trace of red. Apparently thetransmitting power of the solutions changed slightly during the course ofthe experiments. On this account certain solutions are undesirable forexperiments on color vision, for one must be certain of the constancy ofthe condition of stimulation. It is to be understood, of course, that eachof the three filters transmitted, so far as the eye is concerned, only thecolor named. I consider the red filter perfectly satisfactory, the blue-violet very good, and the green poor. Henceforth, in testing color visionin animals, I shall make use of colored glasses as filters, if it is inany way possible to obtain or have manufactured blue, green, and yellowglasses which are as satisfactory as the ruby. [Footnote 1: A Janssen-Hoffman spectroscope was used. ] The apparatus needs no further description, as its other importantfeatures were identical with those of the reflected light experiment box. The use of artificial light for the illumination of the electric-boxesmade it necessary to conduct all of the following tests in a dark-room. The method of experimentation was practically the same as that alreadydescribed. A mouse which had been placed in _A_ by the experimenter waspermitted to enter _B_ and thence to return to _A_ by entering one of theelectric-boxes, the red or blue or green one, as the case might be. Mistakes in choice were punished by an electric shock. One further pointin the method demands description and discussion before the results of thetests are considered, namely, the manner of regulating and measuring thebrightness of the lights. Regulating brightness with this apparatus was easy enough; measuring itaccurately was extremely difficult. The experimenter was able to controlthe brightness of each of the two colored lights which he was using bychanging the position or the power of the incandescent lamps in the light-box. The position of a lamp could be changed easily between tests simplyby moving it along toward or away from the electric-box in the slit whichserved as a lamp carrier. As the distance from the entrances of theelectric-boxes to the further end of the light-box was 120 cm. , aconsiderable range or variation in brightness was possible without changeof lamps. Ordinarily it was not necessary to change the power of thelamps, by replacing one of a given candle power by a higher or lower, during a series of tests. Both the candle power of the lamps and theirdistance from the filters were recorded in the case of each test, but forthe convenience of the reader I have reduced these measurements to candlemeters[1] and report them thus in the descriptions of the experiments. [Footnote 1: The illuminating power of a standard candle at a distance ofone meter. ] But measuring the actual brightness of the red light or the green lightwhich was used for a particular series of tests, and the variations intheir brightnesses, was not so simple a matter as might appear from thestatements which have just been made. The influence of the light filtersthemselves upon the brightness must be taken into account. The two redfilters were alike in their influence upon the light which entered them, for they were precisely alike in construction, and the same was true ofthe two blue-violet filters. The same kind of ruby glass was placed ineach of the former, and a portion of the same solution of copper ammoniumsulphate was put into each of the filter boxes for the latter. But it isdifficult to say what relation the diminution in brightness caused by ared filter bore to that caused by a blue-violet or a green filter. My onlymeans of comparison was my eye, and as subjective measurement wasunsatisfactory for the purposes of the experiment, no attempt was made toequalize the amounts of brightness reduction caused by the severalfilters. So far as the value of the tests themselves, as indications ofthe condition of color vision in the dancer is concerned, I have noapology for this lack of measurement, but I do regret my inability to givethat accurate objective statement of brightness values which would enableanother experimenter with ease and certainty to repeat my tests. Thenearest approach that it is possible for me to make to such an objectivemeasurement is a statement of the composition and thickness of the filtersand of the candle-meter value of the light when it entered the filter. Thedistance from this point to the entrance to the electric-box was 20 cm. To sum up and state clearly the method of defining the brightness of thelight in the following experiments: the candle-meter value of each lightby which an electric-box was illuminated, as determined by the use of aLummer-Brodhun photometer and measurements of the distance of the sourceof light from the filter, is given in connection with each of theexperiments. This brightness value less the diminution caused by thepassage of the light through a filter, which has been defined as tocomposition and thickness of the layer of solution, gives that degree ofbrightness by which the electric-box was illuminated. Tests of the dancer's ability to discriminate green and blue[1] in thetransmitted light apparatus were made with four animals. An incandescentlamp marked 16-candle-power was set in each of the light-boxes. Theselamps were then so placed that the green and the blue seemed to be ofequal brightness to three persons who were asked to compare themcarefully. Their candle-meter values in the positions selected wererespectively 18 and 64, as appears from the statement of conditions at thetop of Table 22. [Footnote 1: Hereafter the light transmitted by the blue-violet filterwill be referred to for convenience as blue. ] TABLE 22 GREEN-BLUE TESTS Brightnesses Equal for Human Eye Green 18 candle meters Blue 64 candle meters SERIES DATE NO. 10 NO. 11 1906 RIGHT WRONG RIGHT WRONG (GREEN) (BLUE) (GREEN) (BLUE)A and B[1] April 2 10 10 12 8 1 3 6 4 5 5 2 4 5 5 6 4 3 5 5 5 5 5 4 6 5 5 5 5 5 7 7 3 5 5 6 8 7 3 3 7 7 9 7 3 5 5 8 10 3 7 7 3 9 11 5 5 4 6 10 12 5 5 6 4[Footnote: A single preference series of twenty tests. ] Numbers 10 and 11 exhibited no preference for either of these colors inthe series of 20 tests which preceded the training tests, and neither ofthem gave evidence of ability to discriminate as the result of ten seriesof training tests. In this case, again, the behavior of the animals was asstrongly against the inference that they can tell green from blue as arethe records of choices which appear in the table. Granted, that they areunable to discriminate green from blue when these colors are of about thesame brightness for the human eye, what results when they differ markedlyin brightness? Table 23 furnishes a definite answer to this question. Numbers 5 and 12 were given eight series of green-blue tests with eachlight at 18 candle meters. Little, if any, evidence of discriminationappeared. Then, on the supposition that the difference was not greatenough for easy discrimination, the blue light was reduced almost to 0, the green being left at 18. The tests (series 9) immediately indicateddiscrimination. For series 10 the green was made 64 candle meters, theblue 18, and again there was discrimination. These results were soconclusively indicative of the lack of color vision and the presence ofbrightness vision, that there appeared to be no need of continuing theexperiment further. Accepting provisionally the conclusion that the dancers cannot tell greenfrom blue except by brightness differences, we may proceed to inquirewhether they can discriminate other colors. Are green and reddistinguishable? Green-red discrimination now was tested by a method which it was hopedmight from the first prevent dependence upon brightness. The light in thelight-box on the left was so placed that it had a value of 18 candlemeters, that in the light-box on the right so that it had a value of 1800candle meters. Neither light was moved during the first four series of thegreen-red tests which were given to Nos. 151 and 152. TABLE 23 GREEN-BLUE TESTS Brightnesses Different for Human Eye Green 18 candle meters Blue 18 candle meters No. 5 No. 12 DATESERIES 1906 RIGHT WRONG RIGHT WRONG (GREEN) (BLUE) (GREEN) (BLUE) 1 April 10 6 4 5 5 2 11 5 5 7 3 3 12 6 4 7 3 4 13 4 6 7 3 5 14 7 3 5 5 6 15 4 6 6 4 7 16 6 4 8 2 8 17 5 5 4 6 As it was now evident that the intensity difference was not sufficient torender discrimination easy, the blue was reduced to 0 and the green leftat 18. 9 17 7 3 8 2 Now the brightnesses were made, green 64, blue 18, just the reverse ofthose of series of Table 22. 10 17 8 2 8 2 Each of these series consisted of 20 tests instead of 10. As a result ofthe arrangement of the lights just mentioned, the green appeared to mevery much brighter than the red when it was on the right and very muchdarker when it was on the left. If this were true for the mouse also, itis difficult to see how it could successfully depend upon brightness forguidance in its choices. Such dependence would cause it to choose now thegreen, now the red. The first four series of green-red tests so clearly demonstrateddiscrimination, of some sort, that it was at once necessary to alter theconditions of the experiment. The only criticism of the above method ofexcluding brightness discrimination, of which I could think, was that thered at no time had been brighter than the green. In other words, thatdespite a value of 1800 candle meters for the red and only 18 candlemeters for the green, the latter still appeared the brighter to the mouse. To meet this objection, I made the extreme brightness values 1 and 1800candle meters in some of the later series, of which the results appear inTable 24. From day to day different degrees of brightness were used, as isindicated in the second column of the table. Instead of having first onecolor and then the other the brighter, after the fourth series I changedthe position of the lights each time the position of the filters waschanged; hence, the table states a certain brightness value for each colorinstead of for each electric-box. Series 5 to 14 so clearly indicated discrimination, that it seemednecessary to devise some other means than that of changing thebrightnesses of the colored lights themselves to test the assumption thatthe animals were choosing the brighter light. I therefore removed thelight filters so that the colors which had been present as conditions ofdiscrimination were lacking, and arranged the apparatus so that first onebox, then the other, was illuminated the more brightly. The purpose ofthis was to discover whether as the result of their green-red training themice had acquired the habit of choosing uniformly either the lighter orthe darker box. One series was given under the conditions of illuminationspecified in Table 24 with the result that the brighter box was choseneight times in ten by No. 151 and every time by No. 152. Since neither ofthese individuals had previously been trained by white-black tests to goto the white, and since, furthermore, the dancers usually manifest aslight preference for the lower instead of the higher illumination, thisresult may be interpreted as indicative of dependence upon brightness inthe previous color tests. It looks very much indeed as if the green hadbeen chosen, not because of its greenness, but on account of itsrelatively greater brightness. This test of brightness preference was followed by two series, 16 and 17, under conditions similar to those of the first four series of the table. For series 16 the value of the light in the left box was 1 candle meter, that of the light in the right box 1800 candle meters. Discrimination wasperfect. For series 17 the value for the left remained at 1 candle meter, but that of the right box was decreased to 0. In this series No. 152 wasentirely at a loss to know which box to choose. Of course this was anentirely new set of conditions for choice, namely, a colored box, thegreen or the red as the case might be, beside a dark box, the one whichwas not illuminated. If the mice really had been choosing correctlybecause of a habit of avoiding the red or of seeking the green, thismethod should bring out the fact, for the red box, since with it thedisagreeable electric shock had always been associated, should be a box tobe avoided. For No. 151 this seemed to be the case. Series 23 to 27 of Table 24 were given as final and crucial tests of therelation of brightness discrimination to color discrimination. As it isnot possible to express in a simple formula the conditions of the tests, asample series which indicates the brightness of the colors in each of thetwenty tests of a series, and in addition the results given by No. 151 inthe first of these final series, is reproduced in Table 25. For an animalwhich had presumably learned perfectly to choose green in preference tored, the record of 8 mistakes in 20 choices as a result of changes inrelative brightness is rather bad, and it renders doubtful the existenceof color discrimination in any of these experiments. No. 152 showed noability whatever to choose the green in the first of the series (series 23of Table 24) of which that of Table 25 is a sample. His record, 10mistakes in 20 choices, was even poorer than that of No. 151. That both ofthese mice learned to choose fairly accurately in these final tests isshown by the results of series 24, 25, 26, and 27. I must admit, however, that these records indicate little ability on the part of the animals todiscriminate colors. TABLE 24 GREEN-RED TESTS Brightnesses Extremely Different for Human EyeIntensities are given in candle meters (c. M. ) NO. 151 NO. 152SERIES DATE CONDITIONS RIGHT WRONG RIGHT WRONG (GREEN) (RED) (GREEN) (RED) 1 April 26 18 c. M. On left 1800 c. M. On right 11 9 7 13 2 27 Same 16 4 16 4 3 28 Same 20 0 17 3 4 29 Same 19 1 19 1 5 30 Green 18 c. M. Red 18 c. M. 9 1 10 0 6 30 Green 64 c. M. Red 18 c. M. 9 1 8 2 7 May 1 Green 6 c. M. Red 1500 c. M. 7 3 9 1 8 1 Green 4 c. M. Red 1500 c. M. 8 2 7 3 9 2 Both varied from 4 to 1500 c. M. 18 2 18 2 10 3 Green 2 c. M. Red 1800 c. M. 6 4 7 3 11 3 Same 10 0 10 0 12 4 Same 7 3 8 2 13 4 Same 8 2 6 4 14 5 Green 1 c. M. Red 1800 c. M. 19 1 19 1 Filters were now removed. An illumination of 15 c. M. Was established onone side and an illumination of 0 on the other side, in order to ascertainwhether the mice would choose the brighter box. This was done to test theassumption that the green in the previous tests had always appearedbrighter to the mice than did the red, and that in consequence they hadchosen the brighter box instead of the green box. TABLE 24--CONTINUED No. 151 No. 152 SERIES DATE CONDITIONS RIGHT WRONG RIGHT WRONG (GREEN) (RED) (GREEN) (RED) 15 May 5 Brighter 15 c. M. 8[1] 2[2] 10[1] 0[2] Darker 0 c. M. 16 5 1 c. M. On left 1800 c. M. On right 10 0 10 0 17 5 1 c. M. On left 0 c. M. On right 9 1 4 6 18 5 Green 18 c. M. Red 18 c. M. 19 1 17 3 19 9 Same 9 1 9 1 20 9 Same 10 0 10 0 21 10 Same 10 0 10 0 22 11 Same 10 0 10 0 23 June 1 Both varied from 1 to 1800 c. M. 12 8 10 10 24 2 Same 18 2 14 6 25 June 3 Both varied from 2 to 1800 c. M. 19 1 17 3 26 4 Same 17 3 17 3 27 5 Same 18 2 18 2 [Footnote 1: Brighter. ][Footnote 2: Darker. ] These long-continued and varied tests with Nos. 151 and 152 revealed threefacts: that the mice depend chiefly upon brightness differences in visualdiscrimination; that they probably have something which corresponds to ourred-green vision, although their color experience may be totally unlikeours; and that the red end of the spectrum seems much darker to them thanto us, or, in other words, that the least refrangible rays are of lowerstimulating value for them than for us. TABLE 25 GREEN-RED TESTS June 1, 1906 No. 151 BRIGHTNESS VALUE IN CANDLE RIGHT WRONGTEST POSITION METERS (GREEN) (RED) 1 Green on left Green 4, Red 448 Right -- 2 Green on right Green 448, Red 4 Right -- 3 Green on right Green 4, Red 448 Right -- 4 Green on left Green 448, Red 4 Right -- 5 Green on left Green 3, Red 1800 -- Wrong 6 Green on right Green 1800, Red 3 -- Wrong 7 Green on right Green 3, Red 1800 -- Wrong 8 Green on left Green 1800, Red 3 Right -- 9 Green on right Green 5, Red 34 Right -- 10 Green on left Green 34, Red 5 Right -- 11 Green on right Green 6, Red 74 Right -- 12 Green on left Green 74, Red 6 Right -- 13 Green on left Green 4, Red 448 -- Wrong 14 Green on right Green 448, Red 4 Right -- 15 Green on right Green 4, Red 448 -- Wrong 16 Green on left Green 448, Red 4 Right -- 17 Green on right Green 3, Red 1800 -- Wrong 18 Green on left Green 1800, Red 3 -- Wrong 19 Green on right Green 1800, Red 3 -- Wrong 20 Green on left Green 3, Red 1800 Right -- Totals 12 8 So many of the results of my color experiments have indicated the all-important role of brightness vision that I have hesitated to interpret anyof them as indicative of true color discrimination. But after I had madeall the variations in brightness by which it seemed reasonable to supposethat the mouse would be influenced under ordinary conditions, and after Ihad introduced all the check tests which seemed worth while, there stillremained so large a proportion of correct choices that I was forced toadmit the influence of the quality as well as of the intensity of thevisual stimulus. The first of the facts mentioned above, that brightness discrimination ismore important in the life of the mouse than color discrimination, isattested by almost all of the experiments whose results have beenreported. The second fact, namely, that the dancer possesses somethingwhich for the present we may call red-green vision, also has been provedin a fairly satisfactory manner by both the reflected and the transmittedlight experiments. I wish now to present, in Table 26, results whichstrikingly prove the truth of the statement that red appears darker to thedancer than to us. The brightness conditions which appeared to make the discriminationbetween green and red most difficult were, so far as my experiments permitthe measurement thereof, green from 1 to 4 candle meters with red from1200 to 1600. Under these conditions the red appeared extremely bright, the green very dark, to the human subject. According to the description of conditions in Table 26, Nos. 2 and 5 wererequired to distinguish green from red with the former about 3 candlemeters in brightness and the latter about 1800 candle meters. In theeighth series of 20 tests, each of these animals made a perfect record. Asit seemed possible that they had learned to go to the darker of the twoboxes instead of to the green box, I arranged the following check test. The filters were removed, the illumination of one electric-box was made 74candle meters, that of the other 3, and the changes of the lighter boxfrom left to right were made at irregular intervals. In February, No. 2had been trained to go to the black in black-white tests, and at the sametime No. 5 had been trained to go to the white in white-black tests. Theresults of these brightness check tests, as they appear in the table, series 8 _a_, are indeed striking. Number 2 chose the darker box eachtime; No. 5 chose it eight times out of ten. Were it not for the fact thatmemory tests four weeks after his black-white training had proved that No. 2 had entirely lost the influence of his previous experience (he chosewhite nine times out of ten in the memory series), it might reasonably beurged that this individual chose the darker box because of his experiencein the black-white experiment. And what can be said in explanation of thechoices of No. 5? I can think of no more reasonable way of accounting forthis most unexpected result of the brightness tests than the assumptionthat both of these animals had learned to discriminate by brightnessdifference instead of by color. TABLE 26 GREEN-RED TESTS Brightnesses Different for Human Eye No. 2 No. 5 SERIES DATE BRIGHTNESS RIGHT WRONG RIGHT WRONG VALUES (GREEN) (RED) (GREEN) (RED) 1 May 7 Green Red 1800 c. M. 10 10 12 8 2 8 Same 12 8 11 9 3 9 Same 15 5 14 6 4 10 Same 18 2 12 8 5 11 Same 18 2 14 6 6 12 Same 19 1 16 4 7 13 Same 19 1 18 2 8 14 Same 20 0 20 0 Brightness tests without colors were now given to determine whether themice had been choosing the brighter or the darker instead of the green. TABLE 26--CONTINUED NO. 2 NO. 5 SERIES DATE BRIGHTNESS VALUES RIGHT WRONG RIGHT WRONG (GREEN) (RED) (GREEN) (RED) 8a 14 Brighter 74 c. M. 0[1] 10[2] 2[1] 8[2] Darker 3 c. M. 9 15 3 c. M. On left 1800 c. M. On right 8 12 16 4 10 16 4 c. M. On left 36 c. M. On right 5 5 7 3 11 16 Green 4 c. M. Red 36 c. M. 9 1 8 2 12 17 11 c. M. On left 1800 c. M. On right 7 3 6 4 13 17 Green 11 c. M. Red 1800 c. M. 9 1 8 2 14 18 Mixed values 3 to 1800 c. M. 7 3 8 2 15 19 Same 7 3 7 3 16 20 Same 7 3 7 3 17 21 Same 7 3 9 1 18 22 Same 9 1 8 2 19 23 Same 7 3 9 1 20 24 Same 10 0 8 2 21 25 Same 10 0 9 1 22 26 Same 9 1 10 0 [Footnote 1: Brighter][Footnote 2: Darker] Immediately after the brightness series, the influence of making first onecolor, then the other, the brighter was studied. Throughout series 9 thebrightness value of the left box remained 3 candle meters, that of theright side 1800 candle meters. Number 2 was so badly confused by thischange that his mistakes in this series numbered 12; No. 5 made only 4incorrect choices. Then series after series was given under widelydiffering conditions of illumination. The expression "mixed values, " whichoccurs in Table 26 in connection with series 14 to 22 inclusive, meansthat the brightnesses of the green and the red boxes were changed fromtest to test in much the way indicated by the sample series of Table 25. In view of the results of these 22 series, 320 tests for each of two mice, it is evident that the dancer is able to discriminate visually by someother factor than brightness. What this factor is I am not prepared tosay. It may be something akin to our color experience, it may be distanceeffect. No other possibilities occur to me. Table 26 shows that discrimination was relatively easy for Nos. 2 and 5with green at 3 candle meters and red at 1800. That their discriminationwas made on the basis of the greater brightness of the red, instead of onthe basis of color, is indicated by the results of the brightness checkseries 8a. Increase in the brightness of the green rendered discriminationdifficult for a time, but it soon improved, and by no changes in therelative brightness of the two colors was it possible to prevent correctchoice. In addition to giving point to the statement that red appears darker tothe dancer than to us, the above experiment shows that the animals dependupon brightness when they can, and that their ability to discriminatecolor differences is extremely poor, so poor indeed that it is doubtfulwhether their records are better than those of a totally color blindperson would be under similar conditions. Surely in view of such resultsit is unsafe to claim that the dancer possesses color vision similar toours. Perfectly trained as they were, by their prolonged green-red tests, tochoose the green, or what in mouse experience corresponds to our green, Nos. 2 and 5 offered an excellent opportunity for further tests of blue-green discrimination. For in view of their previous training there shouldbe no question of preference for the blue or of a tendency to depend uponbrightness in the series whose results constitute Table 27. TABLE 27 BLUE-GREEN TESTS NO. 2 NO. 5 SERIES DATE BRIGHTNESS VALUES RIGHT WRONG RIGHT WRONG (BLUE) (GREEN) (BLUE) (GREEN) 1 June 1 Blue 74 c. M. Green 36 c. M. 3 7 3 7 2 2 Same 5 5 4 6 3 3 Same 5 5 6 4 4 4 Same 6 4 3 7 5 5 Same 6 4 5 5 6 6 Blue 21 c. M. Green 21 c. M. 6 4 7 3 7 7 Same 2 8 3 7 8 8 Same 5 5 4 6 9 9 Same 3 7 6 4 10 10 Same 2 8 4 6 11 12 Same 6 4 3 7 12 13 Blue 36 c. M. Green 21 c. M. 3 7 4 6 13 14 Same 5 5 14 15 Blue 62 c. M. Green 21 c. M. 4 6 15 16 Same 5 5 16 17 Same 5 5 17 18 Same 6 4 Now, as a final test, blue and green glasses were placed over theelectric-boxes, the brightness of the two was equalized for the human eye, and the tests of series 18 and 19 were given to No. 2:-- TABLE 27--CONTINUED NO. 2SERIES DATE BRIGHTNESS VALUES RIGHT WRONG (Blue) (Green) 18 18 Blue 62 c. M. Green 21 c. M 4 6 19 19 Same 6 4 20 20 Blue 21 c. M. Green 88 c. M. 2 8 The green was now made much the brighter. 21 21 Blue 21 c. M. Green 18 c. M. 7 3 22 23 Same 8 2 To begin with, the blue and the green were made quite bright for the humansubject, blue 74 candle meters, green 36. Later the brightness of both wasfirst decreased, then increased, in order to ascertain whetherdiscrimination was conditioned by the absolute strength of illumination. No evidence of discrimination was obtained with any of the severalconditions of illumination in seventeen series of ten tests each. On the supposition that the animals were blinded by the brightness of thelight which had been used in some of the tests, similar tests were madewith weaker light. The results were the same. I am therefore convincedthat the animals did justice to their visual ability in these experiments. Finally, it seemed possible that looking directly at the source of lightmight be an unfavorable condition for color discrimination, and that achamber flooded with colored light from above and from one end would provemore satisfactory. To test this conjecture two thicknesses of blue glasswere placed over one electric-box, two plates of green glass over theother; the incandescent lamps were then fixed in such positions that theblue and the green within the two boxes appeared to the experimenter, ashe viewed them from the position at which the mouse made its choice, ofthe same brightness. Mouse No. 2 was given two series of tests, series 18 and 19, under theseconditions, with the result that he showed absolutely no ability to tellthe blue box from the green box. The opportunity was now taken todetermine how quickly No. 2 would avail himself of any possibility ofdiscriminating by means of brightness. With the blue at 21 candle meters, the green was increased to about 1800. Immediately discriminationappeared, and in the second series (22 of Table 27) there were only twomistakes. The results of the blue-green experiments with light transmitted from infront of the animal and from above it are in entire agreement with thoseof the experiments in which reflected light was used. Since the range ofintensities of illumination was sufficiently great to exclude thepossibility of blinding and of under illumination, it is necessary toconclude that the dancer does not possess blue-green vision. Again I must call attention to the fact that the behavior of the mice inthese experiments is even more significant of their lack of discriminatingability than are the numerical results of the tables. After almost everyseries of tests, whether or not it came out numerically in favor ofdiscrimination, I was forced to add the comment, "No satisfactory evidenceof discrimination. " We have now examined the results of green-red, green-blue, and blue-greentests. One other important combination of the colors which were used inthese experiments is possible, namely, blue-red. This is the mostimportant of all the combinations in view of the results alreadydescribed, for these colors represent the extremes of the visiblespectrum, and might therefore be discriminable, even though those whichare nearer together in the spectral series were not. TABLE 28BLUE-RED TESTS No. 2 No. 205SERIES DATE BRIGHTNESS VALUES RIGHT WRONG RIGHT WRONG (BLUE) (RED) (BLUE) (RED) 1 July 31 1800 c. M. On left 24 c. M. On right 5 5 6 42 Aug. 1 21 c. M. On left 1800 c. M. On right 6 4 6 43 2 1800 c. M. On left 21 c. M. On right 8 2 6 44 3 19 c. M. On left 1800 c. M. On right 9 1 6 45 4 1800 c. M. On left 7 c. M. On right 7 3 5 56 5 6 c. M. On left 1800 c. M. On right 10 0 7 37 6 18 c. M. On left 74 c. M. On right 10 0 9 18 7 1800 c. M. On left 7 c. M. On right 8 2 8 29 8 7 c. M. On left 1800 c. M. On right 7 3 8 210 9 Mixed values 6 to 1800 c. M. 8 2 9 111 10 Blue 3 c. M. Red 1800 c. M. 7 3 6 4 Brightness tests were now made, without the use of colors. 11a 10 4 6 5 5 12 10 Blue 3 c. M. Red 8 c. M. 4 6 6 413 11 Blue 3 c. M. Red 7200 c. M. 8 2 5 514 13 Mixed values 3 to 7200 c. M. 7 3 7 315 13 Same 7 3 9 116 14 Blue 3 to 6 c. M. Red 112 to 3650 c. M. 10 0 10 0 Series were now given to test the assumption that red appears dark to thedancer. 17 14 Darkness on one side Red 3 c. M. 5 5 7 318 14 Blue 3 to 3650 c. M. Red 3 to 3650 c. M. 10 0 10 019 15 Darkness on one side Red 3 c. M. 5 5 4 620 15 Blue 3 to 3650 c. M. Red 3 to 3650 c. M. 10 0 9 121 16 Darkess on one side Red 72 c. M. 5 5 7 322 16 Darkness on one side Red 1800 c. M. 6 4 10 0 As is shown by the results in Table 28, no combination of brightnessesrendered correct choice impossible in the case of the blue-red tests whichare now to be described. Choice was extremely difficult at times, evenmore so perhaps than the table would lead one to suppose, and it is quitepossible that color played no part in the discrimination. But thatbrightness difference in the colors was not responsible for whateversuccess these mice attained in selecting the right box is proved by thebrightness-without-color series which follows series II of the table. Neither No. 2 nor No. 205 showed preference for the lighter or the darkerbox. At the end of the sixteenth blue-red series, I was convinced that oneof two conclusions must be drawn from the experiment: either the dancerspossess a kind of blue-red vision, or red is of such a value for them thatno brightness of visible green or blue precisely matches it. The latter possibility was further tested by an experiment whose resultsappear in series 17 to 22 inclusive, of Table 28. The conditions of series17 were a brightness value of 0 in one box (darkness) and in the other redof a brightness of 3 candle meters. Despite the fact that they had beenperfectly trained in _blue-red tests_ to avoid the red, neither of themice seemed able to discriminate the red from the darkness and to avoidit. This was followed by a series in which the brightness of both the blueand the red was varied between 3 and 3650 candle meters, with the strikingresult that neither mouse made any mistakes. In series 19 red was usedwith darkness as in series 17, and again there was a total lack ofdiscrimination. Series 20 was a repetition of series 18, with practicallythe same result. I then attempted to find out, by increasing thebrightness of the red, how great must be its value in order that thedancers should distinguish it readily from darkness. For the tests ofseries 21 it was made 72 candle meters, but discrimination did not clearlyappear. At 1800 candle meters, as is shown in series 22, the red wassufficiently different in appearance from total darkness to enable No. 205to discriminate perfectly between the two electric-boxes. For No. 2discrimination was more difficult, but there was no doubt about hisability. It would appear from these tests that the dancers had not learnedto avoid red. Therefore we are still confronted with the question, canthey see colors? TABLE 29 VISUAL CHECK TESTSWith the Electric-boxes Precisely Alike Visually No. 151 No. 152SERIES DATE RIGHT WRONG RIGHT WRONG 1 Sept. 29 6 4 4 62 30 5 5 6 63 Oct. 1 3 7 4 64 2 5 5 3 75 3 3 7 5 56 4 6 4 5 57 5 5 5 5 58 6 -- -- 3 79 7 -- -- 6 410 8 -- -- 4 6 Averages 4. 7 5. 3 4. 5 5. 5 The account of my color vision experiments is finished. If it be objectedthat other than visual conditions may account for whatever measure ofdiscriminating ability, apart from brightness discrimination, appears insome of the series, the results of the series of Table 29, in which allconceivable visual means of discrimination were purposely excluded, andthose of the several check tests which have been described from time totime in the foregoing account, should furnish a satisfactory and definiteanswer. I am satisfied that whatever discrimination occurred was due tovision; whether we are justified in calling it color vision is quiteanother question. I conclude from my experimental study of vision that although the dancerdoes not possess a color sense like ours, it probably discriminates thecolors of the red end of the spectrum from those of other regions bydifference in the stimulating value of light of different wave lengths, that such specific stimulating value is radically different in nature fromthe value of different wave lengths for the human eye, and that the red ofthe spectrum has a very low stimulating value for the dancer. In the lightof these experiments we may safely conclude that many, if not most, of thetests of color vision in animals which have been made heretofore by otherinvestigators have failed to touch the real problem because thepossibility of brightness discrimination was not excluded. Under the direction of Professor G. H. Parker, Doctor Karl Waugh hasexamined the structure of the retina of the dancing mouse for me, with theresult that only a single type of retinal element was discovered. Apparently the animals possess rod-like cells, but nothing closely similarto the cones of the typical mammalian retina. This is of peculiar interestand importance in connection with the results which I have reported in theforegoing pages, because the rods are supposed to have to do withbrightness or luminosity vision and the cones with color vision. In fact, it is usually supposed that the absence of cones in the mammalian retinaindicates the lack of color vision. That this inference of functionalfacts from structural conditions is correct I am by no means certain, butat any rate all of the experiments which I have made to determine thevisual ability of the dancer go to show that color vision, if it exists atall, is extremely poor. It is gratifying indeed to learn, after such astudy of behavior as has just been described, that the structuralconditions, so far as we are able to judge at present, justify theconclusions which have been drawn. CHAPTER XI THE ROLE OF SIGHT IN THE DAILY LIFE OF THE DANCER Darting hither and thither in its cage, whirling rapidly, now to the left, now to the right, running in circles, passing through holes in the nestbox quickly and neatly, the dancer, it would seem, must have excellentsight. But careful observation of its behavior modifies this inference. For it appears that a pair of mice dancing together, or near one another, sometimes collide, and that it is only those holes with which the animalis familiar that are entered skillfully. In fact, the longer one observesthe behavior of the dancer under natural conditions, the more he comes tobelieve in the importance of touch, and motor tendencies. Sight, which atfirst appears to be the chief guiding sense, comes to take a secondaryplace. In this chapter it is my purpose to show by means of simpleexperiments what part sight plays in the dancer's life of habit formation. The evidence on this subject has been obtained from four sources: (1)observation of the behavior of dancers in their cages; (2) observation oftheir behavior when blinded; (3) observation of their behavior in a greatvariety of discrimination experiments, many of which have already beendescribed; and (4) observation of their behavior in labyrinth experimentswhich were especially planned to exhibit the importance of the severalkinds of vision which the dancer might be supposed to possess. Theevidence from the first three of these sources may be presented summarily, for much of it has already appeared in earlier chapters. That from thefourth source will constitute the bulk of the material of this chapter. My observation of the behavior of the mice has furnished conclusiveevidence of their ability to see moving objects. But that they do not seevery distinctly, and that they do not have accurate perception of the formof objects, are conclusions which are supported by observations that Ihave made under both natural and experimental conditions. In Chapters VII, VIII, IX, and X, I have presented an abundance of evidence of brightnessvision and, in addition, indications of a specific sensitiveness to wavelength which may be said to correspond to our color vision. It isnoteworthy, however, that all of the experimental proofs of visual abilitywere obtained as the result of long periods of training. Seldom, indeed, in my experience with them, have the dancers under natural conditionsexhibited forms of activity which were unquestionably guided by vision. It is claimed by those who have experimented with blinded dancers that theloss of sight decreases the amount and rapidity of movement, and theability of the animals to avoid obstacles. By means of the discrimination method previously used in the preliminaryexperiments on color vision, a full description of which may be found inChapter IX, p. 133, the dancers' ability to perceive form was tested. Immediately after the two males _A_ and _B_ had been given the "food-box"tests, whose results appear in Table 15, they were tested in the sameapparatus and by the same method for their ability to discriminate arectangular food-box from a round one. In the case of the colordiscrimination tests, it will be remembered that the circular tin boxes 5cm. In diameter by 1. 5 cm. In depth, one of which was covered with bluepaper, the other with orange, were used. For the form discrimination testsI used instead one of the circular boxes of the dimensions given above anda rectangular box 8. 5 cm. Long, 5. 5 cm. Wide and 2. 5 cm. Deep. "Force" wasplaced in the circular box. The tests were given, in series of 20, daily. TABLE 30 VISUAL FORM TESTS SERIES DATE MOUSE A MOUSE B RIGHT WRONG RIGHT WRONG (CIRCULAR (RECTANGU- (CIRCULAR (RECTANGU- BOX) LAR BOX) BOX) LAR BOX) 1 Jan. 5 10 10 9 11 2 7 12 8 13 7 3 10 6 14 10 10 4 11 7 13 10 10 5 12 9 11 10 10 6 13 11 9 11 9 7 14 13 7 9 11 8 15 10 10 11 9 9 16 10 10 11 9 10 17 11 9 9 11 11 18 11 9 12 8 12 19 12 8 10 10 13 20 10 10 12 8 14 21 10 10 8 12 15 22 10 10 10 10 Totals 152 148 155 145 The results of 15 series of these tests, as may be seen by the examinationof Table 30, are about as definitely negative, so far as formdiscrimination is in question, as they possibly could be. From the firstseries to the last there is not one which justifies the inference thateither of the dancers depended upon the form of the boxes in making itschoice. In view of the general criticisms I have made concerning the useof hunger as a motive in experiments on animal behavior, and in view ofthe particular criticisms of this very method of testing thediscriminating powers of the mouse, it may seem strange that space shouldbe given to a report of these tests. I sympathize with the feeling, if anyone has it, but, at the same time, I wish to call attention to the factthat almost any mammal which is capable of profiting by experience, andwhich, under the same conditions, could distinguish the rectangular boxfrom the circular one, would have chosen the right box with increasingaccuracy as the result of such experience. The results are important in myopinion, not because they either prove or disprove the ability of thedancer to discriminate these particular forms, the discrimination of whichmight fairly be expected of any animal with an image-forming eye, butbecause they demonstrate an important characteristic of the dancing mouse, namely, its indifference to the straightforward or direct way of doingthings. Most mammals which have been experimentally studied have proved theireagerness and ability to learn the shortest, quickest, and simplest routeto food without the additional spur of punishment for wandering. With thedancer it is different. It is content to be moving; whether the movementcarries it directly towards the food is of secondary importance. On itsway to the food-box, no matter whether the box be slightly or strikinglydifferent from its companion box, the dancer may go by way of the wrongbox, may take a few turns, cut some figure-eights, or even spin like a topfor seconds almost within vibrissa-reach of the food-box, and all thiseven though it be very hungry. Activity is pre-eminently important in thedancer's life. In passing I may emphasize the importance of the fact that at no time didthe brightness or color discrimination tests furnish evidence of attemptson the part of the dancers to choose by means of slight differences in theform of the cardboards or the cardboard carriers. Several times formdifferences, which were easily perceivable by the human subject, wereintroduced in order to discover whether the mice would detect them andlearn to discriminate thereby instead of by the visual conditions ofbrightness or color. As these experiments failed to furnish evidences ofform discrimination, the following special test in the discrimination boxwas devised. [Illustration: FIGURE 22. --Cards used for tests of form discrimination. ] The color discrimination box of Chapter X was arranged so that the lightat the entrance to each electric-box had a value of 20 candle meters, lessthe diminution caused by a piece of ground glass which was placed over theend of the electric-boxes to diffuse the light. The windows through whichthe light entered the electric-boxes were covered with pieces of blackcardboard; in one of these cardboards I had cut a circular opening 4 cm. In diameter, and in the other an opening of the same area but markedlydifferent shape. These openings are shown in Figure 22. As the mouseapproached the entrance to the electric-boxes, it was confronted by thesetwo equally illuminated areas, whose chief difference was one of form. Difference in the amount of light within the boxes was excluded so far aspossible. The question which I asked was, can the dancer discriminate bymeans of this difference in visual form? For the purpose of settling this point and of gaining additional knowledgeof the role of vision, two individuals were tested in the discriminationbox under the conditions which have just been described. During the firstten days of the experiment each of these mice, Nos. 420 and 425, was givena series of ten tests daily. At the end of this period experimentationwith No. 425 had to be discontinued, and the number of daily tests givento No. 420 was increased to twenty. Instead of taking space for the presentation of the daily records, I maystate the general results of the tests. Neither of the mice learned tochoose the right box by means of form discrimination. In fact, there wasabsolutely no sign of discrimination at any time during the tests. Thisresult is as surprising as it is interesting. I could not at first believethat the mice were unable to perceive the difference in the lighted areas, but assumed that they were prevented from getting the outlines of theareas by the blinding effect of the light. However, decreasing theintensity of the illumination did not alter the result. According to theindications of this experiment, the dancer's ability to perceive visualform is extremely poor. Thus far the purpose of our experiments has been to ascertain what thedancer is enabled to do by sight. Suppose we now approach the problem ofthe role of this sense by trying to find out what it can do without sight. [Illustration: FIGURE 23. --Labyrinth B. _I_, entrance; _O_, exit; 1, 2, 3, doorways between alleys. ] For the investigation of this matter the labyrinth method seemed eminentlysuitable. The first form of labyrinth which was used in these visual testsappears in ground plan in Figure 23. It was made of 1-1/2 cm. Boards. Thelength was 52 cm. , the width 17 cm. , the depth 10 cm. Each of thedoorways, _I_ (the entrance), 1, 2, 3, and _O_ (the exit), was 5 by 5 cm. The alleys were 2-1/2 cm. Wide. For this width the necessity is obviousfrom what has already been said of the animal's propensity to whirl on alloccasions. As the mice almost never tried to climb up the walls, no coverfor the labyrinth was needed. The direct route is indicated by the symbols_I_-1-2-3-_O_. If an error be defined as a choice of the wrong path as theanimal progressed toward the exit, five mistakes were possible in theforward course: the first by turning to the left at the entrance; thesecond by failing to pass through doorway 1; the third by turning to theright after passing through doorway 1; the fourth by failing to passthrough doorway 3, and the fifth by turning to the left after passingthrough 3. In case the mouse retraced its course, any mistakes made as itagain progressed towards _O_ were counted, as at first, no matter how manytimes it went over the same ground. Thus an individual might make the samemistake several times in the course of a single test in the labyrinth. With this labyrinth Nos. 7, 998, 15, 16, 151, and 152 were tested. Atfirst a record was kept of the time which elapsed from the instant theanimal entered _I_ to the instant it emerged at _O_, of the path which itfollowed, and of the number of errors which it made; but later only thenumber of errors was recorded. TABLE 31 THE ROLE OF SIGHT Labyrinth-B Experiments NO. 7 NO. 998 TEST DATE TIME ERRORS TIME ERRORS 1 June 16 66" 8 127" 19 2 16 11 0 94 12 3 16 15 2 18 3 4 16 7 0 13 2 5 16 5 0 10 1 6 18 61 15 12 3 7 18 13 3 14 4 8 18 14 5 8 1 9 18 24 9 16 2 10 18 10 1 9 1 11 19 36 13 80 17 12 19 8 3 10 1 13 19 6 1 7 1 14 19 9 1 8 0 15 19 12 2 7 0 16 20 14 1 25 0 17 20 28 3 18 20 No efforts No efforts to escape to escape TABLE 32 LABYRINTH-B EXPERIMENTS with Electric Shock given as Punishment for Mistakes No. 7 No. 998TEST DATE CONDITION ERRORS CONDITION ERRORS 1 June 29 Light 4 Light 92 29 Light 1 Light 33 29 Light 1 Light 24 29 Light 0 Light 05 29 Light 0 Light 06 29 Light 0 Light 07 29 Light 1 Light 08 29 Light 0 Light 09 29 Light 1 Darkness 010 29 Light 1 Light 011 29 Light 1 Darkness 012 29 Light 0 Light 013 29 Light 0 Light 014 29 Light 0 Light 015 29 Light 0 Light 016 29 Light 0 Light 017 29 Darkness 2 Darkness 018 29 Light 2 Light 0 with paper19 29 Light 0 Light 020 29 Darkness 0 Light 0 with paper21 29 Light 0 Light 022 29 Light 0 Darkness 023 29 Light 0 Odorless 024 June 29 Light 0 Darkness 025 29 Light 026 29 Darkness 427 29 Light with paper 128 29 Light 029 29 Light with paper 130 29 Darkness 031 29 Odorless 232 29 Darkness 4 As the results in Table 31 show, the time and number of errors rapidlydiminished. Number 7, for example, made no errors in the second test. Thechiefly significant fact which appeared in these preliminary experiments, however, was that the mice soon ceased to care whether they got out of thelabyrinth or not. After they knew the path perfectly, they would enter thewrong passages repeatedly and wander about indefinitely. It was obvious, therefore, that the labyrinth could not be used to reveal the role ofsight unless some sufficiently strong motive for continuous effort toescape from it could be discovered. Naturally I looked to the electricshock for aid. The labyrinth of Figure 23, which for convenience in distinguishing itfrom several other forms to be described later I have designated aslabyrinth B, was placed upon a board 90 cm. Long and 30 cm. Wide aboutwhich had been wound two pieces of phosphor bronze wire after the mannerdescribed on p. 94. At _O_, Figure 24, there was an opening closed by aswinging door which led into a box 40 by 24 cm. In one corner of this boxwas a small nest-box. The significance of this rearrangement of thelabyrinth is apparent. As in the preliminary tests, the dancer was startedat I, but instead of being allowed to wander about without any otherresult than delay in escape, it was given a shock each time it made anerror. The satisfaction of escaping from the narrow bounds of thelabyrinth's passages, which alone was not strong enough to impel a dancerconstantly to do its best to escape, was thus supplemented by the powerfuland all-controlling tendency to avoid the disagreeable stimulus whichresulted from entering certain of the passages. The result of thismodification of method is strikingly exhibited by the data of Table 32. [Illustration: Figure 24. --Labyrinth B on an interrupted circuit board. _I_-1-2-3-_O_, labyrinth path; _B_, nest-box; _N_, nest; _EW_, board woundwith phosphor bronze wire; _IC_, induction apparatus; _C_ electric cell;_K_, key. ] This table was constructed for the purpose of exhibiting the principalfeatures of the results obtained with labyrinth _B_ in certain preliminaryexperiments in which the conditions were changed in various ways. Chiefamong the important facts which appear in the illustrative data (for Nos. 7 and 998) which are presented, are the following. The dancers readilylearn the path of labyrinth B so that they can follow it quickly and withperfect accuracy. After familiarity with the direct path from entrance toexit has been gained, they become indifferent about escaping and tend towander aimlessly. The introduction of the electric shock as punishment forthe choice of the wrong passage impels them to do their best to avoiderrors. The path once learned can be followed in total darkness with fewor no errors. Table 32 indicates marked differences in the behavior of No. 7 and No. 998. The latter learned the path readily and was littledisturbed by any of the changes in conditions. In total darkness hefollowed the path rapidly and accurately, as was indicated by the time ofthe trip and the path that he left on a sheet of smoked paper that hadbeen placed on the floor of the labyrinth as a means of obtaining a recordof the errors made. The presence of the smoked paper did not seem tointerfere at all with his behavior, nor did the thorough washing of thelabyrinth and the resultant removal of its odors. In the case of No. 7 theopposite was true. She did not learn the path readily, was confused by anychange in conditions, had great difficulty in finding her way in darkness, made errors when the smoked paper was placed on the floor and after theodors of the labyrinth had been removed by washing. Of the six dancerswhich were observed in these preliminary tests, No. 7 alone gaveconvincing evidence of the importance of sight. I think we may say in the light of the results of the table that sucherrors as appear in the darkness tests are due rather to the disturbinginfluence of a change in the conditions of the experiment than to theexclusion of visual data, for as many or more errors were sometimes causedsimply by changing the position of the labyrinth, placing smoked paper onthe floor, or by introducing a new odor at some point. The exclusion ofthe possibility of guidance by smell and touch did not seriously interferewith the animal's ability to follow the path. The results which have just been considered seemed to be of sufficientinterest and importance to justify the further use of the labyrinth methodin the investigation of the role of vision. A series of experiments withlabyrinth B was therefore planned so that the importance of sight, touch, and smell in connection with this form of habit should be moresatisfactorily exhibited. Does the dancer follow the path by sight, touch, smell, by all, or by no one of them? This series of tests with labyrinth B, whose several purposes may best beexplained in connection with the various kinds of tests enumerated below, consisted of: I. A preliminary test in which the dancer was permitted to wander about inthe labyrinth, without being shocked, until it finally escaped to thenest-box by way of the exit. Thus the animal was given an opportunity todiscover that escape from the maze was possible. II. This was immediately followed by a series of tests at the rate ofabout one per minute, with an electric shock as punishment for everymistake. This was continued without interruption until the path had beenfollowed without error five times in succession. III. The labyrinth was now moved about 3 cm. To one side so that itcovered a new floor area, and a test was given for the purpose ofascertaining whether the mouse had been following a trail on the floor. IV. Tests with smoked paper on the floor were now alternated with tests inwhich the floor was plain. The alternation was rendered necessary by thefact that the paper was laid over the electric wires and thereforeprevented the punishment of mistakes. The purpose of these tests was todiscover whether the smoked paper, which was an essential condition forthe next test, was itself a disturbing condition. These tests werecontinued until the animal had followed the path correctly, despite thesmoked paper, twice in succession. V. The electric lights were now turned out and tests were given in totaldarkness, with smoked paper on the floor as a means of obtaining a recordof the number of errors. These tests were continued until the path hadbeen followed once correctly. VI. The labyrinth was now thoroughly washed with warm water, to which alittle kerosene had been added, and quickly dried over a steam radiator. This usually necessitated a delay of about five minutes. As soon as thelabyrinth was dry, tests were given to discover whether the odors of thevarious passages had been serving as important guiding conditions. Thesetests were continued until the path had been followed once without error. VII. A final test in darkness completed the series. As it was not possible for the observer to watch the animal and thus tocount the number of mistakes which it made in total darkness, the simplemethod of placing a piece of smoked paper on the floor of the labyrinthwas used. The mouse left a graphic record of its path on the paper andfrom this the number of errors could be ascertained. In the tests now tobe described the smoked paper was placed upon the electric wires, butlater a form of electric labyrinth was devised in which it was underneathand therefore did not interfere with the electric shock. The above series of tests was given under the same external conditions ina dark-room to six pairs of dancers. In all cases, two individuals, a maleand a female, which had been kept in the same cage, were experimented withat the same time, _i. E. _ one was permitted to rest in the nest-box whilethe other was being put through a test. This was done in order that thecomparison of the results for males and females should be perfectly fair. The detailed results of this long series of tests may be presented foronly two individuals, Nos. 210 and 215, Table 33. In this table linesseparate the results of the seven different kinds of tests. TABLE 33 THE ROLE OF SIGHT, TOUCH, AND SMELL IN LABYRINTH EXPERIMENTS No. 210 No. 215 TEST CONDITION ERRORS CONDITION ERRORS I. 1 No shock 9 I. No shock 2 II. 2 Shock 5 II. Shock 3 3 Shock 4 Shock 1 4 Shock 2 Shock 0 5 Shock 3 Shock 0 6 Shock 0 Shock 0 7 Shock 0 Shock 0 8 Shock 0 Shock 0 9 Shock 0 III. Labyrinth 0 moved 10 Shock 0 IV. Paper on floor 4 III. 11 Labyrinth 0 No paper (shock) 0 moved IV. 12 Paper on 0 0 floor 13 No paper 0 No paper 0 (shock) 14 Paper 1 Paper 1 15 No paper 0 No paper 0 16 Paper 7 Paper 4 17 No paper 0 No paper 0 18 Paper 0 Paper 0 19 No paper 0 No paper 0 20 Paper 4 Paper 0 21 No paper 0 No paper 0 22 Paper 2 V. Darkness 0 23 No paper 2 VI. Labyrinth 2 24 Paper 1 washed 0 25 No paper 0 VII. Darkness 2 26 Paper 0 27 No paper 0 28 Paper 0 29 No paper 0 V. 30 Darkness 0 VI. 31 Labyrinth 2 washed 32 0VII. 33 Darkness 0 The average results for the twelve individuals (six of each sex) whichwere subjected to the tests, I have arranged in Table 34. The Romannumerals at the top of the table designate the seven groups of tests, andthe figures under each, the numerical results of the tests. I may explainand comment upon the averages of the several columns of this table inturn. Column I gives the number of errors made in the preliminary test. Curiously enough, the males made many more errors than the females. For the second group of tests (II) two results have been tabulated: thenumber of the first correct test, and the total number of tests before thepath was followed correctly five times in succession. The first correcttrip came usually after not more than five or six tests, but fivesuccessive correct trips demanded on the average at least fourteentraining tests. Destruction of the floor path by movement of the labyrinth to one side, without changing its relations to the points of the compass, disturbed themice very little. Only four of the twelve individuals made any mistakes asa result of the change in the tactual conditions, and the average error asit appears in Column III is only . 3. TABLE 34 ROLE OF SIGHT, TOUCH, AND SMELL IN LABYRINTH EXPERIMENTS II. IV. TRAINING TESTS SMOKED I. NO OF TESTS BEFORE III. PAPER ONMALES PRELIMINARY CORRECT LABYRINTH FLOOR TEST. _____________________ MOVED. NO OF TIMES ERRORS FIRST TIME FIVE TIMES ERRORS BEFORE COR- RECT TWICE 210 9 5 9 0 9 212 2 3 8 1 3 214 6 10 28 0 22 220 25 4 8 0 14 410 11 6 20 0 10 420 14 6 14 1 7 AVERAGES 11. 2 5. 7 14. 5 . 3 10. 8 FEMALES 211 16 6 10 1 5 213 7 5 14 1 21 215 2 3 7 0 6 225 14 6 18 0 14 415 6 6 13 0 3 425 10 7 13 0 8 AVERAGES 9. 2 5. 5 12. 5 . 3 9. 5 V. DARKNESS VI. MALES LABYRINTH VII. ERRORS IN NO. OF TESTS WASHED. DARKNESS. FIRST TEST BEFORE COR'CT ERRORS ERRORS 210 0 1 2 0 212 2 2 0 0 214 0 1 -- 0 220 2 4 2 0 410 1 3 2 1 420 2 4 1 4 Averages 1. 2 2. 5 1. 2 0. 8 FEMALES 211 2 2 0 0 213 2 2 -- 3 215 0 1 2 2 225 3 2 0 0 415 1 3 2 1 425 1 7 0 0 Averages 1. 5 2. 8 0. 7 1. 0 That covering the floor with smoked paper forced the mice to relearn thepath, in large measure, is evident from the results of Column IV. Anaverage of ten tests was necessary to enable the mice to follow the pathcorrectly. It is almost certain, however, that the interference with theperfectly formed labyrinth habit which this change in the condition of thefloor caused was not due to the removal of important tactual sense data. As Column V shows, the number of errors in total darkness is very small. Some individuals gave no sign of being disturbed by the absence of visualguidance, others at first seemed confused. I have given in the table thenumber of errors in the first darkness test and the number of the firsttest in which no mistakes occurred. No more disturbance of the dancer's ability to follow the path which ithad learned resulted from washing the labyrinth thoroughly than fromdarkening the room. Indeed it is clear from Column VI that the path wasnot followed by the use of smell. However, the test in darkness, after theodor of the box had been removed, proved conclusively that in most casesthe mice could follow the path correctly without visual or olfactoryguidance. The behavior of 18 individuals as it was observed in labyrinth B makesperfectly evident three important facts, (1) In following the path whichit has learned, the dancer in most instances is not guided to anyconsiderable extent by a trail (odor or touch) which has been formed byits previous journeys over the route; (2) sight is quite unnecessary forthe easy and perfect execution of the labyrinth habit, for even thoseindividuals which are at first confused by the darkening of the experimentroom are able after a few tests to follow the path correctly; (3) and, finally, smell, which according to current opinion is the chieflyimportant sense of mice and rats, is not needful for the performance ofthis habitual act. At this point we may very fittingly ask, what sense data are necessary forthe guidance of the series of acts which constitutes the labyrinth habit?I answer, probably none. A habit once formed, the senses have done theirpart; henceforth it is a motor process, whose initiation is conditioned bythe activity of a receptive organ (at times a sense receptor), but whoseform is not necessarily dependent upon immediate impressions from eye, nose, vibrissae, or even from internal receptors. These are statements ofmy opinion; whether they express the truth, either wholly or in part, onlyfurther experimentation can decide. In considering the results of these labyrinth tests it is important thatwe distinguish clearly those which have to do with the conditions of habitformation from those which instead have to do with the conditions of habitperformance. Sense data which are absolutely necessary for the learning ofa labyrinth path may be of little or no importance for the execution ofthe act of following the path after the learning process has beencompleted. Thus far in connection with the labyrinth tests we havediscussed only the relations of sight, touch, and smell to what I havecalled habit performance. We may now ask what part these senses play inthe formation of a labyrinth habit. A very definite answer to this question is furnished by observation of thebehavior of the dancers in the tests. Most of them continuously made useof their eyes, their noses, and their vibrissae. Some individuals used oneform of receptive organ almost exclusively. I frequently noticed thatthose individuals which touched and smelled of the labyrinth passages mostcarefully gave least evidence of the use of sight. It is safe to say, then, that under ordinary conditions habit formation in the dancer isconditioned by the use of sight, touch, and smell, but that these sensesare of extremely different degrees of importance in different individuals. And further, that, although in the case of some individuals the loss ofsight would not noticeably delay habit formation, in the case of others itwould seriously interfere with the process. When deprived of one sense, the dancer depends upon its remaining channels of communication withenvironment. Indeed there are many reasons for inferring that if deprivedof sight, touch, and smell it would still be able to learn a labyrinthpath; and there are reasonable grounds for the belief that a habit onceformed can be executed in the absence of all special sense data. Apparently the various receptive organs of the body furnish the dancerwith impressions which serve as guides to action and facilitate habitformation, although they are not necessary for habit performance. The reader may wonder why I have not carried out systematic experiments todetermine accurately and quantitatively the part which each sense plays inthe formation of a labyrinth habit instead of basing my inferences uponincidental observation of the behavior of the dancers. The reason issimply this: the number and variety of experiments which were suggested bythe several directions in which this investigation developed rendered theperformance of all of them impossible. I have chosen to devote my time toother lines of experimentation because a very thorough study of theconditions of habit formation has recently been made by Doctor Watson. [1] [Footnote 1: Watson, J. B. , _Psychological Review_, Monograph Supplement, Vol. 8, No. 2, 1907. ] What is the role of sight in the dancing mouse? How shall we answer thequestion? The evidence which has been obtained in the course of my studyof the animal indicates that brightness vision is fairly acute, that colorvision is poor, that although form is not clearly perceived, movement isreadily perceived. My observations under natural conditions justify theconclusion that sight is not of very great importance in the daily life ofthe dancer, and my observations under experimental conditions stronglysuggest the further conclusion that movement and changes in brightness arethe only visual conditions which to any considerable extent control theactivity of the animal. CHAPTER XII EDUCABILITY: METHODS OF LEARNING Nearly all of the experiments described in earlier chapters have revealedfacts concerning the educability of the dancer. In order to supplement theknowledge of this subject thus incidentally gained and to discover theprinciples of educability, the specially devised experiments whose resultsappear in this and succeeding chapters were arranged and carried out witha large number of mice. In the work on the modifiability of behavior Ihave attempted to determine (1) by what methods the dancer is capable ofprofiting by experience, (2) the degree of rapidity of learning, (3) thepermanency of changes wrought in behavior, (4) the effect of one kind oftraining upon others, (5) the relation of re-training to training, and (6)the relation of all these matters to age, sex, and individuality. As it is obvious that knowledge of these subjects is a necessary conditionfor the intelligent appreciation of the capacities of an animal, as wellas of the choice of methods by which it may be trained advantageously, perhaps it is not too much to expect that this investigation of the natureand conditions of educability in the dancing mouse may give us some newinsight into the significance of certain aspects of human education andmay serve to suggest ways in which we may measure and increase theefficiency of our educational methods. Merely for the sake of convenience of description I shall classify themethods which have been employed as problem methods, labyrinth methods, and discrimination methods. That these names are not wholly appropriate issuggested by the fact that discrimination necessarily occurs in connectionwith each of them. As problem methods we may designate those tests ofinitiative and modifiability which involve the opening of doors by pushingor pulling them, and the climbing of an inclined ladder. An example of thelabyrinth method has been presented in Chapter XI. The name discriminationmethod I have applied to those tests which involve the choice of one oftwo visual, tactual, or olfactory conditions. The white-blackdiscrimination tests, for example, served to reveal the rapidity andpermanency of learning as well as the presence of brightness vision. In the case of most mammals whose educability has been studiedexperimentally, problem methods have proved to be excellent tests ofdocility and initiative. The cat, the raccoon, the monkey, in theirattempts to obtain food, learn to pull strings, turn buttons, presslatches, slide bolts, pull plugs, step on levers. The dancer does none ofthese things readily. Are we therefore to infer that it is lessintelligent, that it is less docile, than the cat, the raccoon, or themonkey? Not necessarily, for it is possible that these methods do not suitthe capacity of the animal. As a matter of fact, all of the tests whichare now to be described in their relation to the educability of the dancerbear witness to the importance of the selection of methods in the light ofthe motor equipment and the habits of the animal which is to be tested. Judged by ordinary standards, on the basis of results which it yields inproblem and labyrinth tests, the dancer is extremely stupid. But that thisconclusion is not justified is apparent when it is judged in the light oftests which are especially adapted to its peculiarities. Problems which are easy for other mammals because of their energetic andpersistent efforts to secure food in any way which their motor capacitymakes possible are useless as tests of the dancer's abilities, because itis not accustomed to obtain its food as the result of strenuous and variedactivities. There are problems and problems; a condition or situationwhich presents a problem to one organism may utterly lack interest for anorganism of different structure and behavior. What is a problem test inthe case of the cat or even of the common mouse, is not necessarily aproblem for the dancer. Similarly, in connection with the labyrinthmethod, it is clear that the value of the test depends upon the desire ofthe organism to escape from the maze. The cat, the rat, the tortoise dotheir best to escape; the dancer is indifferent. Clearly, then, methods oftraining should be chosen on the basis of a knowledge of thecharacteristics of the animal whose educability is to be investigated. The simplest possible test of the intelligence of the dancer which I coulddevise was the following. Beside the cage in which the mice were kept Iplaced a wooden box 26 cm. Long, 23 cm. Wide, and 12 cm. Deep. Neitherthis box nor the cage was covered, for the animals did not attempt toclimb out. As a way of passing from one of these boxes to the other Iarranged a ladder made of wire fly-screen netting. This ladder was about 8cm. Broad and it extended from the middle of one side of the wooden boxupward at an angle of about 30° to the edge of the box and then descendedat the same angle into the cage. A dancer when taken from the nest-box and placed in the wooden box couldreturn to its cage and thus find warmth, food, and company by climbing theladder. It was my aim to determine, by means of this apparatus, whetherthe dancers can learn such a simple way of escape and whether they learnby watching one another. As it turned out, a third value belonged to thetests, in that they were used also to test the influence of putting themice through the act. In the first experiment three dancers, Nos. 1000, 2, and 6, were togetherplaced in the wooden box. At the end of 15 minutes not one of them hadsucceeded in returning to the cage. They were then driven to the bottom ofthe ladder and started upward by the experimenter; with this assistanceall escaped to the nest-box. At the expiration of 5 minutes they wereagain placed in the wooden box, whence the chilly temperature (about 60°F. ) and the lack of food made them eager to return to their cage. Noattempt to climb up the ladder was made by any of them within 15 minutes, so the experimenter directed them to the ladder and started them upward asin the first test. This completed the experiment for the day. Thefollowing day two tests were given in the same way. In the second of thesetests, that is, on its fourth trial, No. 1000 climbed over of his owninitiative in 5 minutes. The others had to be assisted as formerly. On thethird day No. 1000 found his way back to the nest-box quickly and fairlydirectly, but neither No. 2 or No. 6 climbed of its own initiative in thefirst test. When their movements were restricted to the region of the boxabout the base of the ladder, both of them returned to the cage quickly. And on the second test of the third day all the mice climbed the ladderdirectly. In Table 35 I have given the time required for escape in the case of 40tests which were given to these 3 individuals at the rate of 2 tests perday. When the time exceeded 15 minutes the mice were helped out by theexperimenter; a record of 15 minutes, therefore, indicates failure. Naturally enough the motives for escape were not sufficiently strong orconstant to bring about the most rapid learning of which the dancer iscapable. Sometimes they would remain in the wooden box washing themselvesfor several minutes before attempting to find a way of escape. On thisaccount I made it a rule to begin the time record with the appearance ofactive running about. The daily average time of escape as indicated in thetable does not decrease regularly and rapidly. On the fourth day, whichwas the first on which all three of the dancers returned to the cage byway of the ladder of their own initiative in both tests, the average is214 seconds. In contrast with this, on the twentieth day the time was only5 seconds. It is quite evident that the dancers had learned to climb theladder. At the end of the twentieth day the experiment was discontinued with Nos. 2 and 6, and after two weeks they were given memory tests, which showedthat they remembered perfectly the ladder-climbing act, for when placed inthe wooden box, with Nos. 4 and 5 as controls, they returned to the cageby way of the ladder immediately and directly. TABLE 35 LADDER CLIMBING TEST Time in Minutes and Seconds No. Of Date No. 1000 No. 2 No. 6 Average Daily Av. Exp. 1905 For All For All 1 Nov. 14 15' 15' 15' -- -- 2 15' 15' 15' -- -- 3 15 15' 15' 15' -- -- 4 300" 15' 15' -- -- 5 16 480" 15' 15' -- -- 6 180" 300" 420" 300" 300" 7 17 450" 240" 540" 410" 8 20" 15" 18" 18" 214" 9 18 90" 180" 135" 135" 10 135" 105" 165" 135" 135" 11 19 480" 240" 330" 350" 12 30" 120" 90" 80" 143" 13 20 360" 75" 120" 185" 14 5" 6" 8" 6" 95" 15 21 105" 450" 120" 192" 16 8" 80" 20" 54" 123" 17 22 255" 300" 180" 245" 18 10" 30" 270" 103" 174" 19 23 300" 660" 450" 470" 20 90" 120" 150" 120" 295" 21 24 240" 125" 225" 197" 22 4" 6" 168" 59" 128" 23 Nov. 25 305" 85" 130" 173" 24 5" 6" 118" 43" 108" 25 26 3" 8" 44" 18" 26 19" 1" 176" 98" 58" 27 27 150" 79" 269" 166" 28 26" 3" 31" 20" 93" 29 28 214" 18" 267" 166" 30 40" 3" 4" 16" 91" 31 29 130" 45" 250" 142" 32 12" 3" 25" 13" 77" 33 Dec. 2 61" 35" 44" 47" 34 50" 5" 24" 26" 36" 35 3 66" 18" 2" 29" 36 8" 5" 10" 8" 19" 37 4 9" 4" 3" 5" 38 10" 5" 6" 7" 6" 39 5 5" 3" 5" 4" 40 10" 4" 3" 6" 5" One of the most interesting and important features of the behavior of thedancer in the ladder experiment was a halt at a certain point on theladder. It occurred just at the edge of the wooden box at the point wherethe ladder took a horizontal position, and led over into the cage. Everyindividual from the first test to the last made this halt. Although fromthe point of view of the experimenter the act was valueless, it may haveoriginated as an attempt to find a way to escape from the uncomfortableposition in which the animal found itself on reaching the top of theladder. Its persistence after a way of escape had been found is anindication of the nature of habit. Day after day the halt became shorteruntil finally it was little more than a pause and a turn of the headtoward one side of the ladder. I think we may say that in this act we haveevidence of the persistence of a particular resolution of physiologicalstates which is neither advantageous nor disadvantageous to the organism. Had the act resulted in any gain, it would have become more marked andelaborate; had it resulted in injury or discomfort, it would havedisappeared entirely. I have observed the same kind of behavior in thefrog and in other animals. What the animal begins to do it persists inunless the act is positively harmful or conflicts with some beneficialactivity. The only explanation of certain features of behavior is to befound in the conditions of their original occurrence. They persist bysheer force of conservatism. They have value only in the light of thecircumstances under which they first appeared. Although this is merely afact of habit formation, it suggests that many of the problems which havepuzzled students of behavior for ages may be solved by a study of thehistory of activity. That there are marked individual differences in intelligence in thedancing mice is apparent from the results of the ladder-climbingexperiment. No. 1000 learned to climb quickly, and largely by his owninitiative; Nos. 2 and 6, on the contrary, learned only by reason oftuition (being put through the required act by the experimenter). Itoccurred to me that this experiment, since it was difficult for someindividuals and easy for others, might be used to advantage as a test ofimitation. If a dancer which knows how to escape to the cage by way of theladder be placed in the wooden box with one which, despite abundantopportunity, has proved unable to form the habit on his own initiative, will the latter profit by the activity of the former and thus learn themethod of escape? On November 20, Nos. 4 and 5 were placed in the wooden box and left therefor half an hour. As they had failed to escape at the end of thisinterval, they were taken out of the box by the experimenter and returnedto the nest-box. November 21 and 22 this test of their ability to learn toclimb the ladder was repeated with the same result. On November 23 theywere placed in the box with the three mice which had previously beentrained to climb the ladder. The latter escaped at once. Apparently theattention of Nos. 4 and 5 was drawn to the ladder by the disappearance oftheir companions, for they approached its foot and No. 5 climbed up ashort distance. Neither succeeded in escaping, however, and they made nofurther efforts that day. On the 24th, and daily thereafter until the29th, these two dancers were placed in the box for half an hour, withnegative results. At the end of the half hour on the 29th, Nos. 2 and 6were placed in the box and permitted to go back and forth from one box tothe other repeatedly within sight of Nos. 4 and 5. The latter made noattempts to follow them, although at times they seemed to be watchingtheir movements as they ascended the ladder. To render the results of this test of imitation still more conclusive No. 5 was given further opportunity to learn from No. 1000. Beginning December2, the following method of experimentation was employed with these twoindividuals. They were placed in the wooden box together. No. 1000 usuallyclimbed out almost immediately. Sometimes No. 5 apparently saw himdisappear up the ladder; sometimes she paid no attention whatever eitherto the presence or absence of her companion. After he had been in thenest-box for a few seconds, No. 1000 was returned to the wooden box by theexperimenter and again permitted to climb out for the benefit of No. 5. This mode of procedure was kept up until No. 1000 had made from three toten trips. No. 5 was left in the box for half an hour each day. This testwas repeated on 18 days within a period of 3 weeks. No. 5 showed no signsof an imitative tendency, and she did not learn to climb the ladder. To this evidence of a lack of an imitative tendency in the dancer I mayhere add the results of my observations in other experiments. In thediscrimination tests and in the labyrinth tests I purposely so arrangedconditions, in certain instances, that one individual should have anopportunity to imitate another. In no case did this occur. Seldom indeeddid the animals so much as follow one another with any considerable degreeof persistence. They did not profit by one another's acts. Excellent evidence in support of this conclusion was furnished by thebehavior of the mice in the discrimination experiments. Some individualslearned to pull as well as to push the swinging wire doors of theapparatus and were thus enabled to pass through the doorways in eitherdirection; other individuals learned only to pass through in the directionin which the doors could be pushed open. Naturally I was interested todiscover whether those which knew only the trick of opening the doors bypushing would learn to pull the doors or would be stimulated to try byseeing other individuals do so. At first I arranged special tests ofimitation in the discrimination box; later I observed the influence of thebehavior of one mouse upon that of its companion in connection with visualdiscrimination experiments. This was made possible by the fact thatusually a pair of individuals was placed in the discrimination box and thetests given alternately to the male and to the female. Both individualshad the freedom of the nest-box and each frequently saw the other passthrough the doorway between the nest-box, _A_, and the entrance chamber, _B_ (Figure 14), either from _A_ to _B_ by pushing the swing door or from_B_ to _A_ by pulling the door. Although abundant opportunity for imitation in connection with the openingof the doors in the discrimination box was given to twenty-fiveindividuals, I obtained no evidence of ability to learn by imitation. Thedancers did not watch the acts which were performed by their companions, and in most instances they did not attempt to follow a mate from nest-boxto entrance chamber. These problem tests, simple as they are, have revealed two important factsconcerning the educability of the dancer. First, that it does not learn byimitation to any considerable extent, and, second, that it is aided bybeing put through an act. Our general conclusion from the results of theexperiments which have been described in this chapter, if any generalconclusion is to be drawn thus prematurely, must be that the dancing mousein its methods of learning differs markedly from other mice and from rats. CHAPTER XIII HABIT FORMATION: THE LABYRINTH HABIT The problem method, of which the ladder and door-opening tests of thepreceding chapter are examples, has yielded interesting results concerningthe individual initiative, ingenuity, motor ability, and ways of learningof the dancer; but it has not furnished us with accurate measurements ofthe rapidity of learning or of the permanency of the effects of training. In this chapter I shall therefore present the results of labyrinthexperiments which were planned as means of measuring the intelligence ofthe dancer. The four labyrinths which have been used in the investigation may bedesignated as _A, B, C, _ and _D_. They differ from one another in thecharacter of their errors, as well as in the number of wrong choices of apath which the animal might make on its way from entrance to exit. In theuse of the labyrinth method, as in the case of the discrimination methodof earlier chapters, the steps by which a satisfactory form of labyrinthfor testing the dancer was discovered are quite as interesting andimportant for those who have an intelligent appreciation of the problemsand methods of animal psychology as are the particular results which wereobtained. For this reason, I shall describe the various forms of labyrinthin the order in which they were used, whether they proved satisfactory ornot. At the outset of this part of my investigation, it was my purpose tocompare directly the capacity for habit formation in the dancer with thatof the common mouse. This proved impracticable because the same labyrinthis not suited to the motor tendencies of both kinds of mice. [Illustration: FIGURE 25. --Labyrinth A. _I_. Entrance; _O_, exit; 1, 2, 3, 4, blind alleys. ] The first of the four labyrinths, A, appears in ground plan in Figure 25. It was constructed of wood, as were the other labyrinths also, andmeasured 60 cm. In length and width, and 10 cm. In depth. The outsidealleys were 5 cm. Wide. In the figure, _I_ marks the starting point orentrance to the maze, and _O_ the exit through which the mouse waspermitted to pass into its nest-box. Any turn in the wrong direction whichthe animal made in its progress from entrance to exit was recorded as anerror. The four errors, exclusive of the mistake of turning back, whichwere possible in this labyrinth, are indicated in the figure by thenumerals 1, 2, 3, and 4. By retracing its steps a mouse might repeat anyone or all of these errors, and add to them the error of turning back. In the experiments a mouse was permitted to enter the maze from a smallbox which had been placed by the experimenter at _I_, and an accuraterecord was kept of the number of errors which it made in finding its wayfrom entrance to exit, and of the time occupied. Each of five dancers wasgiven 31 tests in this labyrinth. The number of tests per day varied, asis indicated in Table 36, from 1 to 4. The results of the tests, so far aserrors and times are in question, appear in the table. _T_ at the head ofa column is an abbreviation for time, _E_ for errors. The dancers did not learn to escape from this labyrinth easily andquickly. In fact, the average time of the thirty-first test (198") isconsiderably longer than that of the first (130"). The number of errorsdecreased, it is true, but even for the last test it was 6. 6 as comparedwith only a little more than twice that number for the first test. Thelast column of the table furnishes convincing proof of the truth of thestatement that the animals did not acquire a perfect labyrinth-A habit. Was this due to inability to learn so complex a path, or to the fact thatthe method is not adapted to their nature? Observation of the behavior ofthe mice in the experiments enables me to say with certainty that therewas no motive for escape sufficiently strong to establish a habit offollowing the direct path. Often, especially after a few experiences inthe maze, a dancer would wander back and forth in the alleys and centralcourts, dancing much of the time and apparently exploring its surroundingsinstead of persistently trying to escape. This behavior, and the time anderror results of the accompanying table, lead me to conclude that thelabyrinth method, as it has been employed in the study of the intelligenceof several other mammals, is not a satisfactory test of the ability of thedancer to profit by experience. That the fault is not in the labyrinthitself is proved by the results which I obtained with common mice. TABLE 36 RESULTS OF LABYRINTH A TESTS WITH DANCERS AVERAGETEST DATE No. 1000 No. 2 No. 6 No. 4 No. 5 FOR ALL 1905 T E T E T E T E T E T E 1 Nov 23 130" 14 100" 8 170" 13 60" 6 190" 26 130" 13. 4 2 24 140 19 78 7 60 8 149 6 211 25 128 13. 0 3 25 392 31 87 1 98 5 185 13 120 9 176 11. 8 4 26 448 38 38 3 47 2 50 3 121 12 141 11. 3 5 27 142 8 21 2 27 3 27 2 17 1 47 3. 2 6 28 45 2 61 7 63 5 102 8 33 4 61 5. 2 7 29 303 17 64 7 36 3 42 2 57 4 100 6. 6 8 30 222 15 26 2 37 5 42 3 7 0 67 5. 0 9 Dec 1 185 9 36 5 48 3 63 3 94 8 85 5. 6 10 2 52 2 71 4 19 0 196 5 95 11 87 4. 4 11 3 180 8 32 2 107 4 52 3 38 4 82 4. 2 12 4 310 10 133 11 65 3 242 6 125 6 175 7. 2 13 4 153 9 335 55 130 10 195 15 154 18 193 21. 4 14 5 330 7 69 2 42 2 201 6 130 10 154 5. 4 15 5 287 7 34 4 61 4 136 7 25 2 109 4. 8 16 5 455 15 65 4 25 0 110 8 160 15 183 8. 4 17 6 120 15 280 9 33 0 168 4 39 2 128 6. 0 18 6 120 4 164 10 81 4 101 5 85 4 110 5. 4 19 6 132 12 78 7 110 6 40 2 151 12 102 7. 8 20 7 258 10 223 16 33 1 92 5 37 1 129 6. 6 21 7 110 7 23 3 44 4 20 4 305 23 100 8. 2 22 7 100 4 60 8 167 15 44 7 58 4 86 7. 6 23 8 43 1 179 7 356 6 34 3 65 3 135 4. 0 24 8 92 5 56 5 42 3 17 1 23 1 46 3. 0 25 9 85 5 114 3 62 3 129 8 31 0 84 3. 8 26 9 30 2 36 4 109 15 12 1 34 2 44 4. 8 27 9 69 5 40 4 85 6 36 3 16 1 49 3. 8 28 10 169 7 80 3 28 0 142 5 35 2 89 3. 4 29 10 155 5 266 8 91 5 27 0 37 2 115 4. 0 30 10 29 1 25 2 124 14 83 6 111 12 74 7. 0 31 10 465 6 208 8 95 3 65 3 159 13 198 6. 6 On the basis of two tests per day, two common mice, a white one and a grayone, quickly learned to escape from labyrinth _A_ by the shortest path. The time of escape for the gray individual (Table 37) decreased from 180"in the first test to 21" in the tenth, and the number of errors from 6 to1. Similarly in the case of the white individual, the time decreased from122" to 8", and the errors from 5 to 1. A fraction of the number of teststo which the dancer had been subjected sufficed to establish a habit ofescape in the common mouse. It is evident, therefore, that the dancerdiffers radically from the common mouse in its behavior in a maze, and itis also clear that the labyrinth method, if it is to be used to advantage, must be adapted to the motor tendencies of the animal which is to betested. TABLE 37 RESULTS OF LABYRINTH A TESTS WITH COMMON MICE GREY MOUSE WHITE MOUSETEST T E T E 1 180" 6 122" 5 2 26 2 80 6 3 37 1 56 4 4 18 0 27 1 5 68 2 33 2 6 10 1 19 1 7 11 1 17 1 8 13 1 17 1 9 10 0 8 1 10 21 1 8 1 The behavior of the dancer made obvious two defects in labyrinth A. Itspassages are so large that the mouse is constantly tempted to dance, andit lacks the basis for a strong and constant motive of escape by thedirect path. To obviate these shortcomings labyrinth B was constructed, asis shown in Figures 23 and 24, with very narrow passages, and a floorwhich was covered with the wires of an interrupted electric circuit sothat errors might be punished. The length of this labyrinth was 52 cm. Andthe passages were 2. 5 cm. Wide and 10 cm. Deep. Dancing in these narrowalleys was practically impossible, for the mice could barely turn aroundin them. In the case of all except the common mice and two dancers, adepth of 10 cm. Was sufficient to keep the animals in the maze without theuse of a cover. As an account of the behavior of the dancer in labyrinth B has alreadybeen given in Chapter XI, I may now state the general results of theexperiments. In all, thirty individuals were trained in this labyrinth. Each individual was given tests at the rate of one per minute until it hadsucceeded in following the correct path five times in succession. The weakelectric shock, which was given as a punishment for mistakes, provided anactivity-impelling motive for escape to the nest-box. An idea of the extreme individual difference in the rapidity with whichthe labyrinth-B path was learned by these dancers may be obtained by anexamination of Table 38, from which it appears that the smallest number oftraining tests necessary for a successful or errorless trip through themaze was one and the largest number fourteen. It is to be remembered thateach mouse was given an opportunity to pass through the labyrinth oncewithout punishment for errors, and thus to discover, before the trainingtests were begun, that a way of escape existed. This first test we maydesignate as the preliminary trial. Table 38 further indicates that thefemales acquired the labyrinth habit more quickly than did the males. TABLE 38 RESULTS OF LABYRINTH-B EXPERIMENTS, WITH TWENTY DANCERS MALES FEMALES NO. OF NO. OF FIRST NO. OF LAST OF NO. OF NO. OF FIRST NO. OF LAST OF MOUSE CORRECT FIVE CORRECT MOUSE CORRECT FIVE CORRECT TEST TESTS TEST TESTS 76 8 14 75 4 15 78 5 20 77 7 11 86 13 22 87 12 22 58 2 14 49 1 5 50 6 23 57 3 20 60 13 37 59 14 28 410 6 20 415 4 13 220 4 8 225 6 18 212 3 7 211 6 10 214 10 28 213 5 14 AV. 7. 0 19. 3 AV. 6. 2 15. 6 A graphic representation of certain of the important features of theprocess of formation of the labyrinth-B habit is furnished by Figure 26 inwhich the solid line is the curve of learning for the ten males of Table38, and the broken line for the ten females. These two curves were plottedfrom the number of errors made in the preliminary trial (P in the figure)and in each of the subsequent tests up to the sixteenth. In the case ofboth the males and the females, for example, the average number of errorsin the preliminary trial was 11. 3, as is indicated by the fact that thecurves start at a point whose value is given in the left margin as 11. 3. In the second training test the number of errors fell to 3. 3 for the malesand 2. 7 for the females. The number of the test is to be found on the baseline; the number of errors in the left margin. If these two curves oflearning were carried to their completion, that for the males would endwith the thirty-seventh test, and that for the females with the twenty-eighth. [Illustration: FIGURE 26. --Curves of habit formation, plotted from thedata of labyrinth-B tests with ten males and ten females. The figures inthe left margin indicate the number of errors; those below the base linethe number of the test. _P_ designates the preliminary test. Males____[solid line]; Females ----[broken line]. ] Time records are not reported for these and subsequent labyrinth testsbecause they proved to be almost valueless as measures of the rapidity ofhabit formation. At any point in its progress through a labyrinth, thedancer may suddenly stop to wash its face, look about or otherwise examineits surroundings; if a shock be given to hurry it along it may besurprised into an error. It is my experience, and this is true of otheranimals as well as of the dancing mouse, that a long trip, as measured intime units, does not necessarily indicate the lack of ability to followthe labyrinth path correctly and rapidly. Hence, whenever it is possible(and the experimenter can always plan his tests so that it shall bepossible), the number of errors should be given first importance and thetime of the test second place. I have presented in Table 38 the number ofthe first correct test, and the number of the last of five successivecorrect tests. Space cannot be spared for records of the errors made inthe several tests by each individual. In general, labyrinth B proved very satisfactory as a means of testing theability of the dancer to learn a simple path. The narrow passageseffectively prevented dancing, and the introduction of the electric shockas a punishment for mistakes developed a motive for escape which wasuniform, constant, and so strong that the animals clearly did their bestto escape from the labyrinth quickly and without errors. This maze was sosimple that it did not tend to discourage them as did the one which isnext to be described. It must be admitted, however, that, though labyrinthB is perfectly satisfactory as a test of the dancer's ability to learn tofollow a simple path, it is not an ideal means of measuring the rapidityof habit formation. This is due to the fact that the preliminary trial andthe first training test play extremely different roles in the case ofdifferent individuals. A dancer which happens to follow the correct pathfrom entrance to exit in the preliminary trial may continue to do so, withonly an occasional error, during several of the early training tests, andit may therefore fail for a considerable time to discover that there areerrors which should be avoided. The learning process is delayed by itsaccidental success. On the other hand, an individual which happens to makemany mistakes to begin with immediately attempts to avoid the points inthe maze at which it receives the electric shock. I was led to conclude, as a result of the labyrinth-B experiments, that the path was too easy, and that a more complex labyrinth would, in all probability, furnish amore satisfactory means of measuring the rapidity of habit formation. [Illustration: FIGURE 27--A record sheet, showing the plan of labyrinth C(as made on the sheet by means of a rubber stamp) on which theexperimenter recorded the path followed by the mouse. This sample sheetpresents the path records for the first, fifth, tenth, and eleventh testsof No. 2 in labyrinth C. 1, 2, 3, 4, 5 designate the several errors of thelabyrinth. ] On the basis of the supposition that a maze whose path was so complex thatthe animal would not be likely to follow it correctly in the early trialswould be more to the purpose than either A or B, labyrinth C was devised. As is shown in the plan of this maze, Figure 27, five mistakes in choiceof path were possible on the forward trip. These errors, as a rule, weremore difficult for the dancers to avoid than those of labyrinths A and B. Those which are designated by the numerals 2, 3, and 4 were especiallydifficult. Error 4 was much more troublesome for left whirlers than forright whirlers because, after turning around abruptly at the entrance tothe blind alley, the former type of dancer almost always followed the sidewall of the maze so far that it missed the correct path. Undoubtedly thevarious errors are not of the same value for different individuals; but itwould be extremely difficult, if not impossible, to devise a maze whichshould be equally difficult for several normal individuals. In order that records of the path followed by a mouse in test after testmight be kept with ease and accuracy by the experimenter, the plan of thislabyrinth, and also that of labyrinth D, were cast in rubber. The outlinesof labyrinths C and D which appear in Figures 27 and 28 respectively weremade with the rubber stamps which were thus obtained. Figure 27 is thereproduction of a record sheet which presents the results of the first, the fifth, the tenth, and the eleventh tests of No. 2 in labyrinth C. Thepath followed by this individual in the first test was far too complex tobe traced accurately on the record sheet. The record therefore representsmerely the number of errors which was made in each region of the maze. Forthe fifth test, and again for the tenth and the eleventh, the path wasrecorded accurately. This simple device for making record blanks which canreadily be filled in at the time of the experiment should recommend itselfto all students of animal behavior. In labyrinth C ten pairs of dancers were given continuous training testsat the rate of one test per minute until they were able to follow thedirect path correctly. Because of the difficulty in learning this mazeperfectly, it was not demanded of the mice that they should follow thepath correctly several times in succession, but instead the training wasterminated after the first successful trip. TABLE 39 RESULTS OF LABYRINTH-C EXPERIMENTS, WITH TWENTY DANCERS MALES FEMALES NO. OF NO. OF FIRST NO. OF NO. OF FIRST MOUSE CORRECT TEST MOUSE CORRECT TEST 2 11 29 15 30 33 49 34 50 49 57 15 52 22 59 15 58 16 215 10 60 17 415 10 76 3 75 8 78 6 77 11 86 5 87 9 88 25 85 11 AV. 18. 7 AV. 13. 8 The results of the experiments with this labyrinth as they are presentedin Table 39 indicate that its path is considerably more difficult for thedancer to learn than that of labyrinth B, that the females learn morequickly than the males, and finally, that individual differences are justas marked as they were in the case of the simpler forms of labyrinth. Ittherefore appears that increasing the complexity of a labyrinth does not, as I had supposed it might, diminish the variability of the results. Certain of the individual differences which appear in Table 39 are due, however, to the fact that in some cases training in labyrinth B hadpreceded training in labyrinth C, whereas in the other cases C was thefirst labyrinth in which the animals were tested. But even this does notserve to account for the wide divergence of the results given by No. 2 andNo. 50, for the latter had been trained in B previous to his training inC, and the former had not been so trained. Yet, despite the advantagewhich previous labyrinth experience gave No. 50, he did not learn the pathof C as well in fifty tests as No. 2 did in eleven. The facts concerningthe value of training in one form of labyrinth for the learning ofanother, as they were revealed by these experiments, may more fittingly bediscussed in a later chapter in connection with the facts of memory andre-learning. [Illustration: FIGURE 28. --Plan of Labyrinth _D_, as reproduced from aprint made with a rubber stamp. _I_, entrance; _O_, exit; numerals 1 to13, errors. ] Labyrinth C is a type of maze which might properly be described asirregular, since the several possible errors are extremely different innature. In view of the results which this labyrinth yielded, it seemedimportant that the dancer be tested in a perfectly regular maze of thelabyrinth-D type. The plan which I designed as a regular labyrinth hasbeen reproduced, from a rubber stamp print, in Figure 28. As is true alsoof the mazes previously described, it provides four kinds of possiblemistakes: namely, by turning to the left (errors 1, 5, 9, and 13), byturning to the right (errors 3, 7, and 11), by moving straight ahead(errors 2, 4, 6, 8, 10, and 12), and by turning back and retracing thepath just followed. The formula for the correct path of _D_ is simple inthe extreme, in spite of the large number of mistakes which are possible, for it is merely "a turn to the right at the entrance, to the left at thefirst doorway, and thereafter alternately to the right and to the leftuntil the exit is reached. " This concise description would enable a man tofind his way out of such a maze with ease. Labyrinth D had beenconstructed with an exit at 10 so that it might be used as a nine-errormaze if the experimenter saw fit, or as a thirteen-error maze by theclosing of the opening at 10. In the experiments which are now to bedescribed only the latter form was used. Can the dancer learn a regular labyrinth path more quickly than anirregular one? Again, I may give only a brief statement of results. Eachof the twenty dancers, of Table 40, which were trained in labyrinth D hadpreviously been given opportunity to learn the path of C, and most of themhad been trained also in labyrinth B. All of them learned this regularpath with surprising rapidity. The numerical results of the tests withlabyrinths B, C, and D, as well as the behavior of the mice in theseseveral mazes, prove conclusively that the nature of the errors is farmore important than their number. Labyrinth D with its thirteen chances oferror on the forward trip was not nearly as difficult for the dancer tolearn to escape from as labyrinth C with its five errors. That thefacility with which the twenty individuals whose records are given inTable 40 learned the path of D was not due to their previous labyrinthexperience rather than to the regularity of the maze is proved by theresults which I obtained by testing in D individuals which were new tolabyrinth experiments. Even in this case, the number of tests necessaryfor a successful trip was seldom greater than ten. If further evidence ofthe ease with which a regular labyrinth path may be followed by the dancerwere desired, it might be obtained by observation of the behavior of anindividual in labyrinths C and D. In the former, even after it has learnedthe path perfectly, the mouse hesitates at the doorways from time to timeas if uncertain whether to turn to one side or go forward; in the latterthere is seldom any hesitation at the turning points. The irregularlabyrinth is followed carefully, as by choice of the path from point topoint; the regular labyrinth is followed in machine fashion, --oncestarted, the animal dashes through it. TABLE 40 RESULTS OF LABYRINTH-D EXPERIMENTS, WITH TWENTY DANCERS MALES FEMALES NO. OF NO. OF FIRST NO. OF LAST OF NO. OF NO. OF FIRST NO. OF LAST OF MOUSE CORRECT TWO CORRECT MOUSE CORRECT TWO CORRECT TEST TESTS TEST TESTS 2 3 7 29 10 11 58 7 10 49 7 8 30 9 10 57 3 6 60 10 14 215 6 10 402 10 11 415 7 8 76 4 7 75 4 13 78 4 5 77 11 12 86 3 9 87 4 9 88 4 8 85 3 4 90 7 8 83 4 7 Av. 6. 1 8. 9 Av. 5. 9 8. 8 From the results of these labyrinth experiments with dancers I am led toconclude that a standard maze for testing the modifiability of behavior ofdifferent kinds of animals should be constructed in conformity with thefollowing suggestions. Errors by turning to the right, to the left, and bymoving forward should occur with equal frequency, and in such order thatno particular kind of error occurs repeatedly in succession. If we shoulddesignate these three types of mistake by the letters _r, l_, and _s_respectively, the error series of labyrinth C would read _l-l-r-s-l_. Ittherefore violates the rule of construction which I have just formulated. In the case of labyrinth D the series would read _l-s-r-s-l-s-r-s-l-s-r-s-l_. This also fails to conform with the requirement, for there are threeerrors of the first type, four of the second, and six of the third. Again, in a standard maze, the blind alleys should all be of the same length, andcare should be taken to provide a sufficiently strong and uniform motivefor escape. In the case of one animal the desire to escape fromconfinement may prove a satisfactory motive; in the case of another, thedesire for food may conveniently supplement the dislike of confinement;and in still other cases it may appear that some form of punishment forerrors is the only satisfactory basis of a motive for escape. Readers ofthis account of the behavior of the dancing mouse must not infer from myexperimental results that the electric shock as a means of forcingdiscrimination will prove satisfactory in work with other animals or evenwith all other mammals. As a matter of fact it has already been proved byDoctor G. Van T. Hamilton that the use of an electric shock may sointimidate a dog that experimentation is rendered difficult and of littlevalue. And finally, in connection with this discussion of a standardLabyrinth, I wish to emphasize the importance of so recording the resultsof experiments that they may be interpreted in terms of an animal'stendency to turn to the right or to the left. My work with the dancer hasclearly shown that the avoidance of a particular error may be extremelydifficult for left whirlers and very easy for right whirlers. I hope I have succeeded in making clear by the foregoing account of myexperiments that the labyrinth method is more satisfactory in general thanthe problem method as a means of measuring the rapidity of habit formationin the dancer, and I hope that I have made equally clear the fact that itis very valuable as a means of discovering the roles of the various sensesin the acquirement of a habit (Chapter XI). From my own experience in theuse of the labyrinth with the dancer and with other animals, I am forcedto conclude that its chief value lies in the fact that it enables theexperimenter so to control the factors of a complex situation that he mayreadily determine the importance of a given kind of sense data for theformation or the execution of a particular habit. As a means of measuringthe intelligence of an animal, of determining the facility with which itis capable of adjusting itself to new environmental conditions, and ofmeasuring the permanency of modifications which are wrought in itsbehavior by experimental conditions, I value the labyrinth method muchless highly now than I did previous to my study of the dancer. It isnecessarily too complex for the convenient and reasonably certaininterpretation of results. Precisely what is meant by this statement willbe evident in the light of the results of the application of thediscrimination method to the dancer, which are to be presented in the nextchapter. The labyrinth method is an admirable means of getting certainkinds of qualitative results; it is almost ideal as a revealer of the roleof the senses, and it may be used to advantage in certain instances forthe quantitative study of habit formation and memory. Nevertheless, Ithink it may safely be said that the problem method and the discriminationmethod are likely to do more to advance our knowledge of animal behaviorthan the labyrinth method. CHAPTER XIV HABIT FORMATION: THE DISCRIMINATION METHOD Discrimination is demanded of an animal in almost all forms of the problemand labyrinth methods, as well as in what I have chosen to call thediscrimination method. In the latter, however, discrimination as the basisof a correct choice of an electric-box is so obviously important that ithas seemed appropriate to distinguish this particular method of measuringthe intelligence of the dancer from the others which have been used, bynaming it the discrimination method. It has been shown that neither the problem nor the labyrinth method proveswholly satisfactory as a means of measuring the rapidity of learning, orthe duration of the effects of training, in the case of the dancer. Theformer type of test serves to reveal to the experimenter the generalnature of the animal's capacity for profiting by experience; the latterserves equally well to indicate the parts which various receptors (some ofwhich are sense organs) play in the formation and execution of habits. Butneither of them is sufficiently simple, easy of control, uniform as toconditions which constitute bases for activity, and productive ofinterpretable quantitative results to render it satisfactory. The problemmethod is distinctly a qualitative method, and, in the case of the dancingmouse, my experiments have proved that the labyrinth method also yieldsresults which are more valuable qualitatively than quantitatively. I hadanticipated that various forms of the labyrinth method would enable me tomeasure the modifiability of behavior in the dancer with great accuracy, but, as will now be made apparent, the discrimination method proved to bea far more accurate method for this purpose. Once more I should emphasize the fact that my statements concerning thevalue of methods apply especially to the dancing mouse. Certain of thetests which have proved to be almost ideal in my study of this peculiarlittle rodent would be useless in the study of many other mammals. Anexperimenter must work out his methods step by step in the light of thedaily results of patient and intelligent observation of the motorcapacity, habits, instincts, temperament, imitative tendency, intelligence, hardihood, and life-span of the animal which he is studying. The fact that punishment has proved to be more satisfactory than reward inexperiments with the dancer does not justify the inference that it is moresatisfactory in the case of the rat, cat, dog, or monkey. Methods whichyielded me only qualitative results, if applied to other mammals mightgive accurate quantitative results; and, on the other hand, thediscrimination method, which has proved invaluable for my quantitativework, might yield only qualitative results when applied to another kind ofanimal. The form of the discrimination method whose results are to be presented inthis chapter has already been described as white-black discrimination. Inthe discrimination box (Figures 14 and 15, p. 92) the two electric-boxeswhich were otherwise exactly alike in appearance were rendereddiscriminable for the mouse by the presence of white cardboards in one andblack cardboards in the other. In order to escape from the narrow spacebefore the entrances to the two electric-boxes, the dancer was required toenter the white box. If it entered the black box a weak electric shock wasexperienced. After two series of ten tests each, during which the animalwas permitted to choose either the white or the black box without shock orhindrance, the training was begun. These two preliminary series serve toindicate the natural preference of the animal for white or black previousto the training. An individual which very strongly preferred the whitemight enter, from the first, the box thus distinguished, whereas anotherindividual whose preference was for the black might persistently enter theblack box in spite of the disagreeable shocks. First of all, therefore, the preliminary tests furnish a basis for the evaluation of the results ofthe subsequent training tests. On the day succeeding the last series ofpreliminary tests, and daily thereafter until the animal had acquired aperfect habit of choosing the white box, a series of training tests wasgiven. These experiments were usually made in the morning between nine andtwelve o'clock, in a room with south-east windows. The entrances to theelectric-boxes faced the windows, consequently the mouse did not have tolook toward the light when it was trying to discriminate white from black. All the conditions of the experiment, including the strength of thecurrent for the shock, were kept as constant as possible. Choice by position was effectively prevented, as a rule, by shifting thecardboards so that now the left now the right box was white. The order ofthese shifts for the white-black series whose results are quantitativelyvaluable appear in Table 12 (p. III). That the order of these changes inposition may be criticised in the light of the results which the testsgave, I propose to show hereafter in connection with certain other facts. The significant point is that the defects which are indicated by theaverages of thousands of tests could not have been predicted withcertainty even by the most experienced investigator in this field. In Table 41 are to be found the average number of errors in each series often white-black discrimination tests for five males and for five femaleswhich were trained by being given ten tests per day, and similarly for thesame number of individuals of each sex, trained by being given twentytests per day. Since the results for these two conditions of training arevery similar, the averages for the twenty individuals are presented in thelast column of the table. For the present we may neglect the interestingindividual, sex, and age differences which these experiments revealed andexamine the significant features of the general averages, and of thewhite-black discrimination curve (Figure 29). TABLE 41 WHITE BLACK DISCRIMINATION TESTS. NUMBER OF ERRORS IN THE VARIOUS SERIES MALES FEMALES AVERAGES AVERAGES GENERAL AVERAGES AVERAGES GENERAL AVERAGESSERIES FOR 5, FOR 5, AVERAGES FOR 5, FOR 5, AVERAGES FOR ALL 10 TESTS 20 TESTS FOR 10 10 TESTS 20 TESTS FOR 10 (20) MALES PER DAY PER DAY PER DAY PER DAY AND FEMALES A 5. 8 6. 0 5. 9 5. 8 5. 8 5. 8 5. 85 B 5. 6 6. 2 5. 9 5. 8 5. 6 5. 7 5. 8 1 5. 0 5. 0 5. 0 5. 6 4. 6 5. 1 5. 05 2 2. 6 4. 6 3. 6 4. 4 5. 0 4. 7 4. 15 3 3. 0 3. 4 3. 2 3. 4 3. 4 3. 4 3. 3 4 2. 6 3. 8 3. 2 2. 4 2. 2 2. 3 2. 75 5 2. 4 2. 0 2. 2 2. 6 1. 8 2. 2 2. 2 6 1. 6 1. 6 1. 6 1. 0 2. 2 1. 6 1. 6 7 1. 0 1. 4 1. 2 2. 0 0. 4 1. 2 1. 2 8 0. 2 0. 6 . 4 1. 4 1. 6 1. 5 . 95 9 0. 2 1. 0 . 6 0. 6 0. 8 . 7 . 65 10 0 . 8 . 4 1. 0 0. 8 . 9 . 65 11 0 . 8 . 4 0. 8 0 . 4 . 40 12 0 . 6 . 3 0. 4 0 . 2 . 25 13 0 0 0 0 0 0 0 14 0 0 0 0 0 0 15 0 0 0 0 0 0 [Illustration: FIGURE 29. --Error curve plotted from the data given bytwenty dancers in white-black discrimination tests. The figures in theleft margin indicate the number of errors; those below the base line, thenumber of the series. _A_ and _B_ designate the preference series. ] The preference series, _A_ and _B_, reveal a constant tendency to choosethe black box, whose strength as compared with the tendency to choose thewhite box is as 5. 8 is to 4. 2. In other words, the dancer on the averagechooses the black box almost six times in ten. The first series oftraining tests reduced this preference for black to zero, and succeedingseries brought about a rapid and fairly regular decrease in the number oferrors, until, in the thirteenth series, the white was chosen every time. Since I arbitrarily define a perfect habit of discrimination as theability to choose the right box in three successive series of ten testseach, the tests ended with the fifteenth series. The discrimination curve, Figure 29, is a graphic representation of thegeneral averages of Table 41. --It is an error curve, therefore. Startingat 5. 85 for the first preliminary series, it descends to 5. 8 for thesecond series, and thence abruptly to 5. 05 for the first training series. This series of ten tests therefore served to reduce the black preferencevery considerably. The curve continues to descend constantly until thetenth series, for which the number of errors was the same as for thepreceding series, . 65. This irregularity in the curve, indicative, as itwould appear, of a sudden cessation in the learning process, demands anexplanation. My first thought was that an error in computation on my partmight account for the shape of the curve. The error did not exist, but inmy search for it I discovered what I now believe to be the cause of theinterruption in the fall of the error curve. In all of the training seriesup to the tenth the white cardboard had been on the right and the leftalternately or on one side two or three times in succession, whereas inthe tenth series, as may be seen by referring to Table 12 (p. 111), it wason the left for the first four tests, then on the right four times, and, finally, on the left for the ninth test and on the right for the tenth. This series was therefore a decidedly more severe test of the animal'sability to discriminate white from black and to choose the white boxwithout error than were any that had preceded it. If my interpretation ofthe results is correct, it was so much more severe than the ninth seriesthat the process of habit formation was obscured. It would not be fair tosay that the mouse temporarily ceased to profit by its experience; insteadit profited even more than usually, in all probability, but theunavoidably abrupt increase in the difficultness of the tests was justsufficient to hide the improvement. As I have suggested, the plan of experimentation may be criticisedadversely in the light of this irregularity in the error curve. Had theconditions been perfectly satisfactory the curve would not have taken thisform. I admit this, but at the same time I am glad that I chose thatseries of shifts in the position of the cardboards which, as it happens, served to exhibit an important aspect of quantitative measures of themodifiability of behavior that otherwise would not have been revealed. Ourmistakes in method often teach us more than our successes. I have takenpains, therefore, to describe the unsatisfactory as well as thesatisfactory steps in my study of the dancer. [Illustration: FIGURE 30. --Error curve plotted from the data given bythirty dancers, of different ages and under different conditions oftraining, in white-black discrimination tests. ] The form of the white-black discrimination curve of Figure 29 is moresurprising than disappointing to me, for I had anticipated many moreirregularities than appear. What I had expected, as the result of trainingfive or even ten pairs of mice, was the kind of curve which is presented, for contrast with the one already discussed, in Figure 30. This also is anerror curve, but, unlike the previous one, it is based upon results whichwere got from individuals of different ages which were trained accordingto the following different methods. Ten of these individuals were giventwo or five tests daily, ten were given ten tests daily, and ten weregiven twenty tests daily. The form of the curve serves to call attentionto the importance of uniform conditions of training, in case the resultsare to be used as accurate measures of the rapidity of learning. Examination of the detailed results of the white-black discriminationtests as they appear in the tables of Chapter VII will reveal the factthat some individuals succeeded in choosing correctly in a series of tentests after not more than five series, whereas others required at leasttwice as many tests as the basis of a perfect series. In very fewinstances, however, was a perfect habit of discrimination established byfewer than one hundred tests. As the averages just presented in Table 41indicate, fifteen series, or one hundred and fifty tests, were requiredfor the completion of the experiment. One might search a long time, possibly, for another mammal whose curve of error in a simplediscrimination test would fall as gradually as that of the dancer. It isfair to say that this animal learns very slowly as compared with mostmammals which have been carefully studied. It is to be remembered, however, that quantitative results such as are here presented for thedancer are available for few if any other animals except the white rat. Neither in the form of the curve of learning nor in the behavior of theanimal as it makes its choice of an electric-box is there evidence ofanything which might be described as a sudden understanding of thesituation. The dancer apparently learns by rote. It exhibits neitherintelligent insight into an experimental situation nor ability to profitby the experience of its companions. That the selection of the white boxoccurs in various ways in different individuals, and even in the sameindividual at different periods in the training process, is the onlyindication of anything suggestive of implicit reasoning. Naturally enoughcomparison of the two boxes is the first method of selection. It takes thedancers a surprisingly long time to reach the point of making thiscomparison as soon as they are confronted by the entrances to the twoelectric-boxes. The habit of running from entrance to entrance repeatedlybefore either is entered, once having been acquired, is retained oftenthroughout the training experiments. But in other cases, an individualfinally comes to the point of choosing by what appears to be the immediaterecognition of the right or the wrong box. In the former case the mouseenters the white box immediately; in the latter, it rushes from the blackbox into the white one without hesitation. So much evidence thediscrimination tests furnish of forms of behavior which in our fellow-menwe should interpret as rational. [Illustration: FIGURE 31. --Curve of habit formation, plotted from the dataof labyrinth-D tests with ten males and ten females. ] Comparison of the error curves for the labyrinth tests (Figures 26 and 31)with those for the discrimination tests (Figures 29 and 30) revealsseveral interesting points of difference. The former fall very abruptly atfirst, then with decreasing rapidity, to the base line; the latter, on thecontrary, fall gradually throughout their course. Evidently the labyrinthhabit is more readily acquired by the dancer than is the visualdiscrimination habit. Certain motor tendencies can be established quickly, it would seem, whereas others, and especially those which depend for theirguidance upon visual stimuli, are acquired with extreme slowness. Fromthis it might be inferred that the labyrinth method is naturally farbetter suited to the nature of the dancer than is any form of thediscrimination method. I believe that this inference is correct, but atthe same time I am of the opinion that the discrimination method is ofeven greater value than the labyrinth method as a means of discovering thecapacity of the animal for modification of behavior. Inasmuch as my first purpose in the repetition of white-blackdiscrimination tests with a number of individuals was to obtainquantitative results which should accurately indicate individual, age, andsex differences in the rapidity of learning, it is important to considerthe reliability of the averages with which we have been dealing. Possiblytwo groups of five male dancers each, chosen at random, would yield verydifferent results in discrimination tests. This would almost certainly betrue if the animals were selected from different lots, or were kept beforeand during the tests under different environmental conditions. But from myexperiments it has become apparent that the average of the results givenby five individuals of the same sex, age, and condition of health, whenkept in the same environment and subjected to the same experimental tests, is sufficiently constant from group to group to warrant its use as anindex of modifiability for the race. This expression, index ofmodifiability, is a convenient mode of designating the average number oftests necessary for the establishment of a perfect habit of white-blackdiscrimination. Hereafter I shall use it instead of a more lengthydescriptive phrase. As an indication of the degree of accuracy of measurements of the rapidityof learning which are obtained by the use of 5 individuals I may offer thefollowing figures. For one of two directly comparable groups of 5 maledancers which were chosen from 16 individuals which had been trained, thenumber of tests which resulted in a perfect habit of white-blackdiscrimination was 92; for the other group it was 96. These indices forstrictly comparable groups of 5 individuals each differ from one anotherby less than 5 per cent. Similarly, in the case of two groups of females, the indices of modifiability were 94 and 104. These figures designate thenumber of tests up to the point at which errors ceased for at least threesuccessive series (30 tests). The determination of the probable error of the index of modifiabilityfurther aids us in judging of the reliability of the measure of therapidity of learning which is obtained by averaging the results for 5individuals. For a group of 5 males (Table 43, p. 243) the index was 72 ±3. 5; and for a group of 5 females of the same age as the males andstrictly comparable with respect to conditions of white-black training, itwas 104 ± 2. 9. A probable error of ± 3. 5 indicates the reliability of thefirst of these indices of modifiability; one of ± 2. 9, that of the second. I do not doubt that 10 individuals would furnish a more reliable averagethan 5, but I do doubt whether the purposes of my experiments would havejustified the great increase in work which the use of averages based uponso large a group would have necessitated. Further discussion of the index of modifiability may be postponed untilthe several indices which serve as measures of the efficiency of differentmethods of training have been presented in the next chapter. From the data which constitute the materials of the present chapter it isapparent that the results of the discrimination method are amenable tomuch more accurate quantitative treatment than are those of the problemmethod or the labyrinth method. But I have done little more as yet thandescribe the method by which it is possible to measure certain dimensionsof the intelligence of the dancer, and to state some general results ofits application. In the remaining chapters it will be our task to discoverthe value of this method and of the results which it has yielded. CHAPTER XV THE EFFICIENCY OF TRAINING METHODS The nature of the modifications which are wrought in the behavior of anorganism varies with the method of training. This fact is recognized byhuman educators, as well as by students of animal behavior (makers of thescience of comparative pedagogy), but unfortunately accurate measurementsof the efficiency of our educational methods are rare. Whatever the subject of investigation, there are two preeminentlyimportant aspects of the educative process which may be taken asindications of the value of the method of training by which it wasinitiated and stimulated. I refer to the rapidity of the learning processand its degree of permanency, or, in terms of habit formation, to therapidity with which a habit is acquired, and to its duration. Of these twoeasily measurable aspects of the modifications in which training results, I have chosen the first as a means to the special study of the efficiencyof the training to which the dancing mouse has been subjected in myexperiments. The reader who has followed my account of the behavior of the dancer up tothis point will recall that in practically all of the discriminationexperiments the number of tests in a series was ten. Some readersdoubtless have wondered why ten rather than five or twenty tests wasselected as the number in each continuous series. I shall now attempt toanswer the question. It was simply because the efficiency of that numberof tests, given daily, when taken in connection with the amount of timewhich the conduct of the experiments required, rendered it the mostsatisfactory number. But this statement demands elaboration andexplanation. Very early in my study of the dancer, I learned that a single experiencein a given experiment day after day had so little effect upon the animalthat a perfect habit could not be established short of several weeks ormonths. Similarly, experiments in which two tests per day were givenproved that even a simple discrimination habit cannot be acquired by theanimal under this condition of training with sufficient rapidity to enablethe experimenter to study the formation of the habit advantageously. Next, ten tests in succession each day were given. The results provedsatisfactory, consequently I proceeded to carry out my investigation onthe basis of a ten-test series. After this method had been thoroughlytried, I decided to investigate the efficiency of other methods for thepurpose of instituting comparisons of efficiency and discovering thenumber of tests per day whose efficiency, as measured by the rapidity ofthe formation of a white-black discrimination habit, is highest. For this purpose I carefully selected five pairs of dancers of the sameage, descent, and previous experience, and gave them white-black tests inseries of two tests per day (after the twentieth day the number wasincreased to five) until they had acquired a perfect habit ofdiscriminating. Similarly other dancers were trained by means of series often tests, twenty tests, or one hundred tests per day. Since it was my aimto make the results of these various tests strictly comparable, I sparedno pains in selecting the individuals, and in maintaining constancy ofexperimental conditions. The order of the changes in the position of thecardboards which was adhered to in these efficiency tests was that givenin Table 12. At the beginning of the two-test training I thought it possible that theanimals might acquire a perfect habit with only a few more days' trainingthan is required by the ten-test method. This did not prove to be thecase, for at the end of the twentieth day (after forty tests in all) theaverage number of mistakes, as Table 42 shows, was 3. 2 for the males and3. 0 for the females. Up to this time there had been clear evidence of theformation of a habit of discriminating white from black, but, on the otherhand, the method had proved very unsatisfactory because the first testeach day usually appeared to be of very different value from the second. On account of the imminent danger of the interruption of the experiment bythe rapid spread of an epidemic among my mice, I decided to increase thenumber of tests in each series to five in order to complete the experimentif possible before the disease could destroy the animals. On the twenty-first day and thereafter, five-test series were given instead of two-test. Unfortunately I was able to complete the experiment up to the point ofthirty successive correct tests with only six of the ten individuals whosenumbers appear at the top of Table 42. That the results of this table arereliable, despite the fact that some of the individuals had to be takenout of the experiment on account of bad condition, is indicated by thefact that all the mice continued to do their best to discriminate so longas they were used. Possibly the habit would have been acquired a littlemore quickly by some of the individuals had they been stronger and moreactive. It should be explained at this point that the results in all theefficiency-of-training tables of this chapter are arranged, as in theprevious white-black discrimination tables, in tens, that is, each figurein the tables indicates the number of errors in a series of ten tests. Inall cases _A_ and _B_ mark preliminary series of tests which were given atthe rate of ten tests per series. The numbers in the first column of thesetables designate groups of ten tests each, and not necessarily dailyseries. In Table 42, for example, 1 includes the results of the first fivedays of training, 2, of the next five days, and so on. The table showsthat No. 80 made seven wrong choices in the first five series of two testseach. This method of grouping results serves to make the data for thedifferent methods directly comparable, and at the same time it saves spaceat the sacrifice of very little valuable information concerning the natureof the daily results. It is to be noted, with emphasis, that the two-fivetests per day training established a perfect habit after four weeks oftraining. This method is therefore costly of the experimenter's time. TABLE 42 EFFICIENCY OF TRAINING. WHITE-BLACK TESTS AT THE RATEOF 2 OR 5 PER DAY MALES FEMALESSETS 80 82 84 86 88 AV. 73 79 83 85 89 AV. OF 10 A 5 5 4 8 5 5. 4 5 6 7 7 6 6. 2 B 5 3 6 5 6 5. 0 7 5 7 6 7 6. 4 1 7 7 6 6 6 6. 4 7 6 9 4 6 6. 4 2 2 1 0 6 6 3. 0 6 5 6 5 5 5. 4 3 4 5 5 1 2 3. 2 6 5 2 4 1 3. 6 4 3 4 7 2 0 3. 2 4 3 1 4 3 3. 0 5 2 3 3 2 4 2. 8 - 3 4 3 1 2. 7 6 2 2 - 2 2 2. 0 - 0 2 2 0 1. 0 7 - 1 - 0 1 0. 7 - 1 0 2 1 1. 0 8 - - - 1 1 1. 0 - 1 1 0 0 0. 5 9 - - - 0 1 0. 5 0 1 1 0 0 0. 510 - - - 0 0 0 - 0 0 0 0 011 - - - 0 0 0 - 0 0 012 - - - 0 0 - 0 0 0 TABLE 43 EFFICIENCY OF TRAINING. WHITE-BLACK TESTS AT THE RATE OF10 PER DAY MALES FEMALESSETS 210 220 230 410 420 AV. 215 225 235 415 425 AV. OF 10 A 6 5 6 6 6 5. 8 8 4 4 8 5 5. 8 B 6 8 8 5 1 5. 6 8 7 6 6 2 5. 8 1 6 7 6 2 4 5. 0 7 6 5 6 4 5. 6 2 4 3 1 2 3 2. 6 5 6 4 2 5 4. 4 3 3 1 4 3 4 3. 0 3 3 4 2 5 4. 4 4 5 0 3 3 2 3. 2 2 1 3 3 3 2. 4 5 3 0 4 1 4 2. 4 1 3 3 3 3 2. 6 6 2 1 4 0 1 1. 6 2 1 1 1 0 1. 0 7 1 0 3 1 0 1. 0 1 1 2 3 3 2. 0 8 0 0 1 0 0 0. 2 0 0 2 2 3 2. 0 9 0 0 0 1 0 0. 2 1 0 0 1 1 0. 610 0 0 0 0 0 2 1 0 2 1. 011 0 0 0 0 3 0 1 0 0. 812 0 0 0 0 0 2 0 0. 413 0 0 0 0 014 0 0 015 0 0 TABLE 44 EFFICIENCY OF TRAINING. WHITE-BLACK TESTS AT THE RATE OF20 PER DAY MALES FEMALESSETS 72 74 208 240 402 AV. 217 230 245 403 407 AV. OF 10 A 4 6 7 7 6 6. 0 5 4 7 7 6 5. 8 B 6 4 6 8 7 6. 2 7 3 5 8 5 5. 6 1 3 5 7 5 5 5. 0 3 6 4 4 6 4. 6 2 4 3 7 5 4 4. 6 7 3 5 4 6 5. 0 3 3 3 3 5 3 3. 4 4 3 3 2 5 3. 4 4 6 3 1 4 5 3. 8 5 0 1 2 3 2. 2 5 4 1 0 2 3 2. 0 6 0 0 1 2 1. 8 6 3 1 0 2 2 1. 6 4 1 1 0 6 2. 2 7 3 2 0 1 1 1. 4 1 0 0 0 1 0. 4 8 2 0 1 1 0. 6 0 3 3 0 2 1. 6 9 2 1 1 1 1. 0 1 0 0 3 0. 810 1 2 1 0 0. 8 0 1 1 2 0. 811 3 1 0 0 0. 8 0 0 0 0 012 1 2 0 0 0. 6 0 0 0 0 013 0 0 0 0 0 0 0 014 0 0 015 0 0 0 The results of the ten-test training as they appear in Table 43 need nospecial comment, for quite similar data have already been examined inother connections. In the case of this table it is to be remembered thateach figure represents the number of errors for a single day as well asfor a series of ten successive tests. The results of Table 44, on theother hand, appear as subdivided series, since each daily series wasconstituted by two series of ten tests, or in all twenty tests. Finally, in Table 45 I have arranged the results of what may fairly becalled the continuous training method. In connection with several of thelabyrinth experiments of Chapter XIII continuous training proved verysatisfactory. It therefore seemed worth while to ascertain whether thesame method would not be more efficient than any other for theestablishment of a white-black discrimination habit. That this method wasnot applied to ten individuals as were the two-five-test, the ten-test, and the twenty-test methods is due to the fact that it proved practicallyinadvisable to continue the tests long enough to complete the experiment. I have usually designated the method as one hundred or more tests daily. Iapplied this training method first to individuals Nos. 51 and 60. At theend of one hundred and twenty tests with each of these individuals I wasforced to discontinue the experiment for the day because of the approachof darkness. In the table the end of a series for the day is indicated bya heavy line. The following day Nos. 51 and 60 succeeded in acquiring aperfect habit after a few more tests. TABLE 45 EFFICIENCY OF TRAINING. WHITE-BLACK TESTS AT THE RATE OF100 OR MORE PER DAY SETS 51[1] 60 87 Av. OF 10 A 5 5 6 5. 3B 5 3 7 5. 0 1 6 6 5 5. 72 3 2 5 3. 33 5 4 7 5. 34 7 4 5 5. 35 6 2 3 3. 76 1 1 3 1. 77 4 2 3 3. 08 3 3 0 2. 09 2 2 3 2. 310 5 0 2 2. 311 1 2 2 1. 712 2 1 1 1. 3 13 4 1 2 2. 314 1 2 1 1. 315 3 1 5 3. 016 3 3 2 2. 717 1 0 1 0. 718 2 0 1 1. 019 0 0 2 0. 720 0 0 021 0 1 0. 322 -23 -24 - [Footnote 1: Age of No. 51, 22 weeks. Age of No. 60, 17 weeks. Age of No. 87, 8 weeks. ] The results of the continuous training method for these two mice were sostrikingly different from those yielded by the other methods that I atonce suspected the influence of some factor other than that of the numberof tests per day. The ages of Nos. 51 and 60 at the time of their testswere twenty-two and seventeen weeks, respectively, whereas all theindividuals used in connection with the other efficiency tests were fourweeks of age. It seemed possible that the slow habit formation exhibitedin the continuous training experiments might be due to the greater age ofthe mice. I therefore selected a healthy active female which was onlyeight weeks old, and tried to train her by the continuous training method. With this individual, No. 87, the results were even more discouraging thanthose previously obtained, for she was still imperfect in herdiscrimination at the end of two hundred and ten tests. At that point theexperiment was interrupted, and it seemed scarcely worth while to continueit further at a later date. The evidence of the extremely low efficiencyof the continuous method in comparison with the other methods which wehave been considering is so conclusive that further comment seemssuperfluous. We are now in a position to compare the results of the several methods oftraining which have been applied to the dancer, and to attempt to getsatisfactory quantitative expressions of the efficiency of each method. Ihave arranged in Table 46 the general averages yielded by the fourmethods. Although these general results hide certain important facts whichwill be exhibited later, they clearly indicate that an increase in thenumber of tests per day does not necessarily result in an increase in therapidity of habit formation. Should we attempt, on superficialexamination, to interpret the figures of this table, we would doubtlesssay that in efficiency the two-five-test method stands first, thecontinuous-test method last, while the ten-test and twenty-test methodsoccupy intermediate positions. TABLE 46 EFFICIENCY OF TRAINING Number of Errors in White-Black Series for Different Methods ofTraining SETS OF 10 2 OR 5 TESTS 10 TESTS 20 TESTS 100 OR MORE PER DAY PER DAY PER DAY TESTS PER DAY A 5. 8 5. 8 5. 9 5. 3B 5. 7 5. 7 5. 9 5. 0 1 6. 4 5. 3 4. 8 5. 72 4. 2 3. 5 4. 8 3. 33 3. 4 3. 2 3. 4 5. 34 3. 1 2. 5 3. 0 5. 35 2. 7 2. 5 1. 9 3. 76 1. 5 1. 3 1. 9 1. 77 0. 9 1. 5 0. 9 3. 08 0. 7 0. 8 1. 1 2. 09 0. 5 0. 4 0. 9 2. 310 0 0. 5 0. 8 2. 311 0 0. 4 0. 4 1. 712 0 0. 2 0. 3 1. 313 0 0 2. 314 0 0 1. 315 0 0 3. 016 2. 717 0. 718 1. 019 0. 720 0 We may now apply to the results of our efficiency-of-training tables themethod of measuring efficiency which was mentioned at the end of thepreceding chapter as the _index of modifiability (that number of testsafter which no errors occur for at least thirty tests)_. By taking theaverage number of tests for the several individuals in each of the Tables42, 43, 44, and 45 we obtain the following expressions of efficiency:-- METHOD INDEX OF MODIFIABILITY (EFFICIENCY) Two-five-test 81. 7 ± 2. 7Ten-test 88. 0 ± 4. 1Twenty-test 91. 0 ± 5. 3Continuous-test 170. 0 ± 4. 8 Since the difference between the indices for the ten-test and the twenty-test methods lies within the limits of their probable errors (±4. 1 and±5. 3) it is evident that it is not significant. Except for this, I thinkthese indices may be accepted as indications of real differences in thevalue of the several methods of training. A somewhat different interpretation of our results is suggested by thegrouping of individuals according to sex. In Table 47 appear the generalaverages for the males and the females which were tested by the severalmethods. The most striking fact exhibited by this table is that of thehigh efficiency of the twenty-test method for the females. Apparently theyprofited much more quickly by this method than by the ten-test method, whereas just the reverse is true of the males. I present the data of thistable merely to show that general averages may hide important facts. TABLE 47 EFFICIENCY OF TRAINING CONDITION MALES FEMALES INDEX OF MODIFIABILITY INDEX OF MODIFIABILITY2 or 5 tests per day 85. 0 80. 010 tests per day 72. 0 104. 020 tests per day 94. 0 88. 0100 or more tests per day 160. 0 180. 0 From all considerations that have been mentioned thus far the reader wouldbe justified in concluding that I made a mistake in selecting the ten-testmethod for my study of the modifiability of the behavior of the dancer. That this conclusion is not correct is due to the time factor in theexperiments. If the dancer could acquire a perfect habit as a result oftwelve days' training, no matter whether two, five, ten, or twenty testswere given daily, it would, of course, be economical of time for theexperimenter to employ the two-test method. But if, on the contrary, thetwo-test method required twice as many days' training as the five-testmethod, it would be economical for him to use the five-test method despitethe fact that he would have to give a larger number of tests than the two-test method would have demanded. In a word, the time which the workrequires depends upon the number of series which have to be given, as wellas upon the number of tests in each series. As it happens, the ten-testmethod demands less of the experimenter's time than do methods with fewertests per day. The twenty-test method is even more economical of time, butit has a fatal defect. It is at times too tiresome for both mouse and man. These facts indicate that a balance should be struck between number oftests and number of series. The fewer the tests per day, within the limitsof two and one hundred, the higher the efficiency of the method oftraining, as measured in terms of the total number of tests necessary forthe establishment of a perfect habit, and the lower its efficiency asmeasured in terms of the number of series given. The greater the number oftests per day, on the other hand, the higher the efficiency of the methodin terms of the number of series, and the lower its efficiency in terms ofthe total number of tests. By taking into account these facts, togetherwith the fact of fatigue, we are led to the conclusion that ten tests perday is the most satisfactory number. If my time and attention had not been fully occupied with other problems, I should have determined the efficiency of various methods of training interms of the duration of habit, as well as in terms of the rapidity of itsformation. As these two measures of efficiency might give contradictoryresults, it is obvious that a training method cannot be fairly evaluatedwithout consideration of both the rapidity of habit formation and thepermanency of the habit. A _priori_ it seems not improbable that slownessof learning should be directly correlated with a high degree ofpermanency. By the further application of the method which I have used inthis study of the efficiency of training we may hope to get a definiteanswer to this and many other questions concerning the nature of theeducative process and the conditions which influence it. CHAPTER XVI THE DURATION OF HABITS: MEMORY AND RE-LEARNING The effects of training gradually disappear. Habits wane with disuse. Inthe dancer, it is not possible to establish with certainty the existenceof memory in the introspective psychological sense; but it is possible tomeasure the efficiency of the training to which the animal is subjected, and the degree of permanency of habits. The materials which constitutethis chapter concern the persistence of unused habits, and the influenceof previous training on the re-acquisition of a habit which has been lostor on the acquisition of a new habit. For convenience of description, Ishall refer to certain of the facts which are to be discussed as facts ofmemory, with the clear understanding that consciousness is not necessarilyimplied. By memory, wherever it occurs in this book, I mean the ability ofthe dancer to retain the power of adaptive action which it has acquiredthrough training. I first discovered memory in the dancer, although there was previously noreason for doubting its existence, in connection with the ladder-climbingtests of Chapter XII. In this experiment two individuals which hadperfectly learned to escape from the experiment box to the nest-box by wayof the wire ladder, when tested after an interval of two weeks, duringwhich they had remained in the nest-box without opportunity to exercisetheir newly acquired habit, demonstrated their memory of the method ofescape by returning to the nest-box by way of the ladder as soon as theywere given opportunity to do so. As it did not lend itself readily toquantitative study, no attempts were made to measure the duration of thisparticular habit. At best the climbing of a wire ladder is of veryuncertain value as an indication of the influence of training. Similarly, the persistence of habits has been forced upon my attention dayafter day in my various experiments with the mice. It is obvious, then, that the simple fact of memory is well established, and that we may turnat once to an examination of the facts revealed by special memory and re-learning experiments. The visual discrimination method, which proved invaluable as a means ofmeasuring the rapidity of habit formation, proved equally serviceable inthe measurement of the permanency or duration of habits. Memory tests fordiscrimination habits were made as follows. After a dancer had beentrained in the discrimination box so that it could choose the correctelectric-box, white, red, blue, or green as it might be, in threesuccessive daily series of ten tests each, it was permitted to remain fora certain length of time without training and without opportunity toexercise its habit of visual discrimination and choice. At the expirationof the rest interval, as we may designate the period during which thehabit was not in use, the mouse was placed in the discrimination box underprecisely the same conditions in which it had been trained and was given aseries of ten memory tests with the box to be chosen alternately on theright and on the left. In order that the entire series of ten tests, andsometimes two such series given on consecutive days, might be available asindications of the duration of a habit, the mouse was permitted to enterand pass through either of the electric-boxes without receiving a shock. Had the shock been given as punishment for a wrong choice, it is obviousthat only the first test of the memory series would be of value as anindication of the existence of a previously acquired habit. Even under theconditions of no shock and no stop or hindrance the first test of eachmemory series is of preeminent importance, for the mouse tends to persistin choosing either the side or the visual condition (sometimes one, sometimes the other) which it chooses in the first test. If the wrong boxis chosen to begin with, mistakes are likely to continue because of thelack of punishment; in this case the animal discriminates, but there is noevidence that it remembers the right box. Likewise, if the right electric-box is chosen in the first test, correct choices may continue simplybecause the animal has discovered that it can safely enter that particularbox; again, the animal discriminates without depending necessarily uponits earlier experience. I have occasionally observed a series of tencorrect choices, made on the basis of an accidental right start, followedby another series in which almost every choice was wrong, because theanimal happened to start wrong. As the results of my tests of memory are of such a nature that they cannotadvantageously be averaged, I have arranged in Table 48 a number oftypical measurements of the duration of visual discrimination habits. Inthis table I have indicated the number and age of the individual tested, the habit of discrimination which had been acquired, the length of therest interval, the result of the first test (right or wrong), and thenumber of errors made in each series of ten memory tests. TABLE 48 MEASUREMENTS OF THE DURATION OF A HABIT Memory ERRORSNO. AGE NAME OF TEST REST FIRST FIRST SECOND INTERVAL CHOICE SERIES SERIES 1000 25 weeks White-black 4 weeks Right 0 5 27 White-black 4 Right 5 7 210 15 White-black 8 Right 5 220 15 White-black 8 Right 4 230 15 White-black 8 Wrong 5 215 15 White-black 8 Right 5 225 15 White-black 8 Right 2 235 15 White-black 8 Right 7 410 15 White-black 8 Wrong 4 415 15 White-black 8 Wrong 6 420 15 White-black 8 Wrong 3 425 15 White-black 8 Right 3 2 28 Black-white 4 Wrong 9 7 17 Black-white 2 Wrong 1 7 21 Black-white 6 Right 1 7 27 Black-white 10 Right 1 6 998 18 Black-white 2 Wrong 3 998 22 Black-white 4 Right 0 998 28 Black-white 10 Right 5 5 13 10 Black-white 4 Right 3 14 10 Black-white 4 Right 3 15 10 Black-white 4 Right 2 16 10 Black-white 4 Right 41000 25 Light blue-orange 4 Right 4 2 28 Light blue-orange 2 Wrong 5 5 28 Light blue-orange 6 Wrong 4 6 3 25 Light blue-orange 4 Wrong 8 10 24 Light blue-orange 2 Right 8 10 26 Light blue-orange 2 Right 5 11 25 Light blue-orange 2 Right 6 11 27 Light blue-orange 2 Wrong 5 151 13 Green-red 2 Right 1 0 152 13 Green-red 2 Right 5 1 This quantitative study of the duration of simple habits of choice showedthat in the majority of cases a perfectly acquired habit persists for atleast two weeks. To be perfectly fair to the animal I must restrict thisstatement to visual conditions other than colors, for the dancer exhibitedlittle ability either to acquire or to retain a habit of distinguishingspectral colors. Altogether, I made a large number of white-black andblack-white memory tests after rest intervals of four, six, eight, or tenweeks. The results for the four-week interval show extreme individualdifferences in memory. Number 1000, for example, was able to choosecorrectly every time in a series of white-black tests after a restinterval of four weeks, whereas No. 5 was wrong as often as she was rightafter the same interval. I have placed the results for these twoindividuals at the head of the table because they suggest the variationswhich render averages undesirable. Number 1000 had a perfect habit at theend of four weeks of disuse; No. 5 had no habit whatever. I shall reservefurther discussion of age, sex, and individual differences in thepermanency of habits for the next chapter. With Nos. 7 and 998 memory tests were made after three different restintervals. At the end of two weeks the black-white habit was present inboth individuals, although it was not perfect. After six and four weeks, respectively (see Table 48), it still persisted; in fact, it apparentlyhad improved as the result of additional training after the earlier memorytests. At the expiration of ten weeks it had wholly disappeared. In herfirst series of memory tests after the ten-week interval No. 7 made onlyone error, but a chance choice of the black (right) in the first test andthe subsequent choice of the box in which no shock had been received serveto account for results which at first appear to be indicative of memory. That this explanation is correct is proved by the fact that a secondmemory series, in which the first choice happened to be wrong, resulted insix mistakes. Evidently she had lost the habit. In no instance have memory tests definitely indicated the presence of ahabit after a rest interval of more than eight weeks. It is safe, therefore, to conclude from the results which have been obtained that awhite-black or black-white discrimination habit may persist during aninterval of from two to eight weeks of disuse, but that such a habit isseldom perfect after more than four weeks. The measurements of memory which were made in connection with colordiscrimination experiments are markedly different from those which wereobtained in the brightness tests. As might have been anticipated (?), inview of the extreme difficulty with which the dancer learns todiscriminate colors, the habit of discriminating between qualitativelydifferent visual conditions does not persist very long. I have neverobtained evidence of a perfect habit after an interval of more than twoweeks, and usually, as is apparent from Table 48, the tests indicated veryimperfect memory at the end of that interval. It seems probable that evenin these so-called color tests discrimination is partly by brightnessdifference, and that the imperfection of the habit and its short durationare due to the fact that the basis of discrimination is inadequate. Thisis the only explanation which I have to offer for the difference which hasbeen demonstrated to exist between the duration of brightnessdiscrimination habits and color discrimination habits. The duration of a discrimination habit having been measured with a fairdegree of accuracy, I undertook the task of ascertaining whether trainingwhose results have wholly disappeared, so far as memory tests are inquestion, influences the re-acquisition of the same habit. Can a habit bere-acquired with greater facility than it was originally acquired? Is re-learning easier than learning? To obtain an answer to the question whichmay be asked in these different forms, ten individuals were experimentedwith in accordance with a method whose chief features are now to bestated. In each of these ten individuals a perfect white-black habit wasestablished by the use of the standard series of tests the order of whichis given in Table 12. At the expiration of a rest interval of eight weeksprecisely the same series of tests were repeated as memory and re-trainingtests. In this repetition, the preliminary series, _A_ and _B_, served asmemory tests, and the subsequent training series, as re-training series. [Illustration: FIGURE 32. --Error curves plotted from the data given by tendancers in white-black discrimination tests. The solid line ([Symbol:solid line]) is the error curve of the original learning process; thebroken line (------) is that of the re-learning process, after an intervalof eight weeks. ] The striking results of this investigation of re-learning are exhibited inthe curves of learning and re-learning of Figure 32. These curves make itappear that the mice re-acquired the white-black discrimination habit muchmore readily than they had originally acquired it. But in addition tofurnishing the basis for some such statement as the foregoing, the curvessuggest a serious criticism of the experiment. In the original tests, the preliminary series indicated a strongpreference for black. In series _A_ it was chosen on the average 5. 8 timesin 10, and in series _B_, 5. 7 times. This preference was rapidly overcomeby the training series, and at the end of 130 tests discrimination wasperfect. All this appears in the curve of learning (solid line of figure). On the other hand, these preliminary series when repeated as memory tests, after a rest-interval of eight weeks, gave markedly different results. Series _A_ indicated preference for white (5. 6 times in 10) instead ofblack, and series _B_ indicated only a slight preference for black. Inbrief, series _A_ and _B_ show that the preference for black wasconsiderably stronger at the beginning of the training than at thebeginning of the re-training. In the light of these facts it is fair to claim that the effects of thewhite-black training had not wholly disappeared as the result of eightweeks of rest, and that the experiment therefore fails to furnishsatisfactory grounds for the statement that re-learning occurs morerapidly than learning. I accept this criticism as pertinent, although notnecessarily valid, and at the same time I freely admit that the resultshave a significance which I had not anticipated. But they are not lessinteresting or valuable on that account. Granting, then, that at leastsome of the ten individuals which took part in the experiment had notcompletely lost the memory of their white-black training at the end ofeight weeks, it is still possible that an examination of the individualresults may justify some conclusion concerning the question which wasproposed at the outset of the investigation. Such an examination is madepossible by Tables 49 and 50, in which I have arranged separately theresults for the males and the females. TABLE 49 WHITE-BLACK TRAINING. TEN TESTS PER DAY Males TRAINING RETRAINING 210 220 230 410 420 AV. 210 220 230 410 420 AV. A 6 5 6 6 6 5. 8 5 4 5 4 3 4. 2 B 6 8 8 5 1 5. 6 8 4 5 4 6 5. 4 1 6 7 6 2 4 5. 0 3 3 4 7 3 4. 0 2 4 3 1 2 3 2. 6 2 4 2 5 3 3. 2 3 3 1 4 3 4 3. 0 1 4 1 4 1 2. 2 4 5 0 3 3 2 2. 6 0 1 0 1 2 0. 8 5 3 0 4 1 4 2. 4 0 2 0 2 0 0. 8 6 2 1 4 0 1 1. 6 0 1 0 0 2 0. 6 7 1 0 3 1 0 1. 0 0 0 0 0 8 0 0 1 0 0 0. 2 0 0 1 0. 2 9 0 0 0 1 0 0. 2 0 0 010 0 0 0 0 1 0. 211 0 0 0 0 012 0 0 0 013 0 01415 Only three of the ten individuals failed to re-acquire the habit of white-black discrimination more quickly than it had originally been acquired, and, in the case of these exceptions, No. 220 required exactly the samenumber of tests in each case, and No. 420 was placed at a slightdisadvantage in the re-learning series by an interruption of the trainingbetween the seventh and the eighth series. Had his training been completedby the sixth series he too would have had the same number of tests intraining and re-training. Moreover, and this is of preëminent importancefor a fair interpretation of the results, in several instances even thoseindividuals which exhibited as strong a preference for the black in thememory series as in the preliminary series re-learned more quickly thanthey had learned. Number 210, for example, although he gave no evidence ofmemory, and, in fact, chose the black more frequently in the memory seriesthan he did in the preliminary series, re-acquired the discriminationhabit in less than half the number of tests which had been necessary forthe establishment of the habit originally. TABLE 50 WHITE-BLACK TRAINING. TEN TESTS PER DAY Females TRAINING RE-TRAINING 215 225 235 415 425 Av. 215 225 235 415 425 Av. A 8 4 4 8 5 5. 8 5 2 7 6 3 4. 6 B 8 7 6 6 2 5. 8 8 5 6 4 3 5. 2 1 7 6 5 6 4 5. 6 4 1 5 4 3 3. 4 2 5 6 4 2 5 4. 4 1 1 1 2 3 1. 6 3 3 3 4 3 4 3. 4 1 0 3 6 0 2. 0 4 2 1 3 3 3 2. 4 0 0 3 3 1 1. 4 5 1 3 3 3 3 2. 6 0 0 died 2 0 0. 5 6 2 1 1 1 0 1. 0 0 1 0 0. 2 7 1 1 2 3 3 2. 0 0 0 0 8 0 0 2 2 3 1. 4 1 0. 2 9 1 0 0 1 1 0. 6 0 010 0 2 1 0 2 1. 0 0 011 0 3 0 1 0 0. 8 0 012 0 0 0 2 0 0. 413 0 0 0 0 014 0 0 015 0 0 The facts which have been presented thus far become more significant whenthe indices of modifiability for the learning and the re-learningprocesses are compared. INDICES OF MODIFIABILITY LEARNING RE-LEARNING Females . . . . . . . 104 42. 5Males . . . . . . . . 72 54 The behavior of the mice in the experiments, the detailed results ofTables 49 and 50, and the indices of modifiability together justify thefollowing conclusions. Most of the ten dancers, at the end of a restinterval of eight weeks, had so far lost the habit of white-blackdiscrimination that memory tests furnished no conclusive evidence of theinfluence of previous training; a few individuals seemed to possess tracesof the habit after such an interval. In the case of each group ofindividuals re-training brought about the establishment of a perfect habitfar more quickly than did the original training. This suggests theexistence of two kinds or aspects of organic modification in connectionwith training; those which constitute the basis of a definite form ofmotor activity, and those which constitute the bases or dispositions forthe acquirement of certain types of behavior. There are severalindications that further study of the modifiability of behavior willfurnish the facts which are necessary to render this suggestionmeaningful. Closely related to the facts which have been revealed by the re-trainingexperiments are certain results of the labyrinth experiments. For thestudent of animal behavior, as for the human educator, it is of importanceto learn whether one kind of training increases the efficiency of similarforms of training. Can a dancer learn a given labyrinth path the morereadily because it has previously had experience in another form oflabyrinth? The answer to this question, which my experimental results furnish, isgiven in Table 51. In the upper half of the table have been arranged theresults for six individuals which were trained first in labyrinth B, thenin labyrinth C, and finally in labyrinth D. Below, in similar fashion, aregiven the results for six individuals which were trained in the same threelabyrinths in the order C, B, D, instead of B, C, D. My purpose in givingthe training in these two orders was to ascertain whether labyrinth C, which had proved to be rather difficult for most individuals, would bemore easily learned if the training in it were preceded by training inlabyrinth C. TABLE 51 THE INFLUENCE OF ONE LABYRINTH HABIT UPON THE FORMATIONOF ANOTHER LABYRINTH B LABYRINTH C LABYRINTH D NO. OF NO. OF NO. OF NO. OF NO. OF NO. OFNO. FIRST COR- LAST OF FIRST COR- LAST OF FIRST COR- LAST OF RECT TEST FIVE COR- RECT TEST FIVE COR- RECT TEST FIVE COR- RECT TESTS RECT TESTS RECT TESTS 76 8 14 3 19 4 778 5 20 6 14 4 586 13 22 5 12 3 975 4 15 8 19 4 1377 7 11 11 29 11 1287 12 22 9 20 4 9 AV. 8. 2 17. 3 7. 0 18. 8 5. 0 9. 2 LABYRINTH C LABYRINTH B LABYRINTH D 58 16 -- 2 14 7 1060 17 -- 13 37 10 1488 25 35 9 22 4 849 34 -- 1 5 7 857 15 -- 3 20 3 685 11 18 2 11 3 4 AV. 19. 7 26. 5 5. 0 18. 2 5. 7 8. 3 The results are sufficiently definite to warrant the conclusion thatexperience in B rendered the learning of C easier than it would have beenhad there been no previous labyrinth training. Those individuals whosefirst labyrinth training was in C made their first correct trip as theresult of 19. 7 trials, whereas those which had previously been trained inlabyrinth B were able to make a correct trip as the result of only 7. 0trials. Similarly the table shows that training in C rendered thesubsequent learning of B easier. To master B when it was the firstlabyrinth required 8. 2 trials; to master it after C had been learnedrequired only 5 trials. In addition to proving that the acquisition of oneform of labyrinth habit may facilitate the acquisition of others, comparison of the averages of Table 51 furnishes evidence of the truth ofthe statement that no results of training can be properly interpreted inthe absence of knowledge of the previous experience of the organism. CHAPTER XVII INDIVIDUAL, AGE, AND SEX DIFFERENCES IN BEHAVIOR All dancers are alike in certain important respects, but to the trainedobserver of animal behavior their individual peculiarities are quite asevident, and even more interesting than their points of resemblance. Omitting consideration of the structural marks of individuality, we shallexamine the individual, age, and sex differences in general behavior, rapidity of learning, memory, and discrimination, which have been revealedby my experiments. Observations which bear on the subject of differencesare scattered through the preceding chapters, but in no case have theybeen given sufficient prominence to force them upon the attention of thosewho are not especially interested in individual peculiarities. It hasseemed worth while, therefore, to assemble all the available material inthis chapter for systematic examination and interpretation. In the pages which follow, individual, age, and sex peculiarities arediscussed in turn. Within each of these three groups of differences I havearranged in order what Royce has appropriately named the facts ofdiscriminating sensitiveness, docility, and initiative. Individuals of thesame age and sex no less than those which differ in sex or age exhibitimportant differences in ability to discriminate among sense impressions("discriminative sensitiveness"), in ability to profit by experience("docility"), and in ability to try new kinds of behavior ("initiative"). Individual differences in sensitiveness to visual, auditory, tactual, andolfactory stimuli have been revealed by many of my experiments. Thebrightness discrimination tests conclusively proved that a degree ofdifference in illumination which is easily detectable by one dancer may bebeyond the discriminating sensitiveness of another. Both the tests withgray papers and those with the Weber's law apparatus furnished strikingevidence of individual differences in the kind of visual sensitivenesswhich throughout this book has been called brightness vision. I suspectthat certain of the differences which were observed should be referred tothe experience of the individuals rather than to the capacity of thevisual organs, for training improves visual discrimination to a muchgreater extent than would ordinarily be thought possible. To the truth ofthis statement the results of the Weber's law experiments with No. 51 bearwitness. Likewise in color discrimination there are individualdifferences, examples of which may be discovered by the examination of theresults given in Chapters IX and X. No differences in auditory sensitiveness appeared in my adult dancers, forin none of them was there definite response to sounds, but among the youngindividuals differences were prominent. I may call attention to the dataon this subject which Table 5, p. 89, contains. The mice in four out oftwelve litters gave no indications of hearing any sounds that I was ableto produce; the remaining individuals responded with varying degrees ofsensitiveness. I made no attempt to measure this sensitiveness, but itobviously differed from mouse to mouse. I feel justified, therefore, instating that the young dancers exhibit extreme individual difference insensitiveness to sounds. My observations of differences in sensitiveness to other forms ofstimulation were made in connection with training tests, and although theyare not quantitative, I venture to call attention to them. Indeed, I amled by the results of my study of various aspects of the dancer's behaviorto conclude that the race exhibits individual differences indiscriminating sensitiveness to a far greater extent than do most mammals, not excepting man. The importance of this fact (for I am confident thatany one who carefully examines the detailed results of the variousexperiments which are described in this book will agree that it is anestablished fact) cannot be overlooked. It alters our interpretation ofthe results of training, memory, heredity, and discrimination experiments, and it leads us to suspect that the dancing race is exceedingly unstable. I do not venture to make comparison of my own observations of the dancer'ssense equipment with those of Cyon, Rawitz, Zoth, and Kishi, for thedifferences are too great in many instances to be thought of as other thanspecies or variety peculiarities. It has seemed fairer to compare onlyindividuals of the same breed, or, as I have done and shall continue to dothroughout this chapter, of two lines of descent. With respect to docility individual differences are prominent. We needonly turn to the various tables of results to discover that inmodifiability of behavior, in memory, in re-learning, not to mention otheraspects of docility, dancers of the same sex and age differed strikingly. Let me by way of illustration cite a few cases of difference in docility. Number 1000 learned to discriminate white from black more quickly andretained his habit longer than any other dancer with which I haveexperimented. I should characterize him as an exceptionally docileindividual. Table 44 offers several examples. Numbers 403 and 407, thoughthey were born in the same litter and were alike in appearance and inconditions of life, acquired the white-black habit with a difference inrapidity which is expressed by the indices of modifiability 50 and 100. Inother words, it took No. 407 twice as long to acquire this habit as ittook No. 403. Similarly the ladder-climbing tests revealed importantindividual differences in ability to profit by experience. In the tablesof labyrinth tests (38, 39, 40) individual differences are too numerous tomention. It required forty-nine tests to establish in No. 50 a labyrinth-Chabit which was approximately equal in degree of perfection to that whichresulted from twenty-two tests in the case of No. 52. The figures in thisand other instances do not exaggerate the facts, for repeatedly I havetested individuals of the same litter, the same sex, and, so far as Icould judge, of the same stage of development, and obtained results whichdiffer as markedly as do those just cited. If space limits permitted, Icould present scores of similar differences in docility which the problem, labyrinth, and discrimination methods have revealed. In examining the detailed individual results of the various tables fordifferences of this sort, it is important to bear in mind that sex, age, and descent should be taken into account, for with each of them, as willbe shown clearly later in this chapter, sensitiveness, docility, andinitiative vary. I have therefore based my statements concerningindividual differences in docility upon the results of comparison of miceof the same litter, sex, and age. It is safe to say that human beingssimilarly selected for comparison do not exhibit greater differences inability to profit by experience than did these dancing mice. The facts concerning individual differences in initiative which I havediscovered are not less definite than those of the preceding paragraphs. From the beginning of my study of the dancer I observed that what oneindividual would readily learn of his own initiative another neverlearned. For example, in the ladder-climbing experiment No. 1000distinguished himself for his initiative, whereas Nos. 4 and 5 neveracquired the habit of escaping from confinement by using the ladder. Inoticed, in this test of the animal's ability to learn, that while oneindividual would be scurrying about trying all ways of escape, investigating its surroundings, looking, sniffing, and dancing by turns, another would devote all its time to whirling, circling, or washingitself. One in the course of its activity would happen upon the way ofescape, the other by reason of the limited scope of its activity, not thelack of it, would fail hour after hour to discover even the simplest wayof getting back to its nest, to food, and to its companions. Hundreds oftimes during the past three years I have noticed important individualdifferences in initiative in connection with the discriminationexperiments. The swinging wire doors which one dancer learned to push openbefore he had been in the box five minutes, another might not becomefamiliar with through his own initiative for hours or days. In fact, itwas not seldom that I had to teach an individual to pass from onecompartment to the other by gently pushing him against the door until itopened sufficiently to allow him to squeeze through. Occasionally a mouselearned to pull the doors open so that he could pass through the openingsin either direction with facility. This was a form of individualinitiative which I had not anticipated and did not especially desire, so Idid not encourage its development, but, nevertheless, at least one fourthof the mice which I experimented with in the discrimination box learnedthe trick. The other three fourths, although they were used in the box dayafter day sometimes for weeks, never discovered that they might return tothe nest-box by pulling the swing-door through which they had just passedas well as by entering one of the electric-boxes. Another indication of individual initiative in action appeared in thetendency of certain mice to climb out of the experiment boxes orlabyrinths. It would have been extremely easy for any of the mice toescape from the labyrinths by scaling the walls of the alleys, for theywere only 10 cm. In height, and when a dancer stood on its hind legs itcould easily reach the top with its nose. But, strange though it will seemto any one who has not worked with the dancer, not more than one in ten ofthe animals which I observed made any attempt to escape in this manner. They lacked initiative. That it was not due to a lack of the power toclimb, I abundantly demonstrated by teaching a few individuals that ascramble in one corner meant easy escape from the maze of paths. I do notthink any one of the mice was physically incapable of climbing, but I amconfident that they differed markedly, not only in the willingness to trynew modes of action, but in the readiness with which they could climb. Ihave already said that individuals differ noticeably in the scope of theiractivity. By this statement I mean that they try a varying number of kindsof activity. As in the case of men, so in mice, one individual will do agreater number of things in a few hours than another will in weeks ormonths. The dancers differ in versatility, in individual initiative, as dowe, albeit not so markedly. Important differences which may with certainty be described as agedifferences are not so obvious as are such marks of individuality as havebeen set forth in the preceding pages. I have noted few changes indiscriminative sensitiveness, other than those with regard to auditorysensitiveness, which could be correlated with age. In certain instancesadults appeared to be able to discriminate more accurately and more easilythan young mice, but it is difficult to say whether this change belongsunder sensitiveness or docility. I have not made an ontogenetic study ofthe senses, and I am therefore unable to describe in detail the course oftheir development and decline. Of one important fact I am certain, thatdiscriminative sensitiveness increases up to a certain point with age andwith training. Differences in docility which are obviously to be correlated with ageabound. In the prime of its life (from the second to the tenth month) thedancer is active, full of energy, quick to learn; in its senility (duringthe second year) it is inactive, but at times even more docile than duringthe period of greatest physical development. Frequently I have noticed inconnection with labyrinth tests that individuals of the age of a year ormore learn much more quickly than do individuals of the age of two orthree months. But, on the other hand, I have contradictory observations, for now and then I obtained just the opposite result in experiments totest docility. Evidently this is a matter which demands systematic, quantitative investigation. Casual observation may suggest conclusions, but it will not justify them. Early in my investigation of the behavior of the dancer I conceived theidea of determining the relation of modifiability of behavior (docility)to age. The question which was foremost in my mind and for which I firstsought an answer may be stated thus: can the dancer acquire a given habitwith the same facility at different ages? Since the visual discriminationexperiment seemed to be well suited for the investigation of this problemI planned to train, in the white-black discrimination experiment, fivepairs of dancers at the age of one month, and the same number for each ofthe ages four, seven, ten, thirteen, sixteen, and nineteen months. [1] [Footnote 1: I have not been able thus far to determine the average lengthof the dancer's life. The greatest age to which any of my individuals hasattained is nineteen months. ] To test the same individuals month after month would be the ideal way ofobtaining an answer to our question, but I could devise no satisfactoryway of doing this. The effects of training last so long, as the results ofthe previous chapter proved, and the uncertainty of their entiredisappearance is so serious, that the same training process cannot be usedat successive ages. The use of different methods of training is even moreunsatisfactory because it is extremely difficult to make accuratequantitative comparison of their results. It was these considerations thatforced me to attempt to discover the relation of docility to age bycarrying out the same experiments with groups of individuals of differentages. As my plan involved the execution of precisely the same set of tests withat least seventy individuals whose age, history, and past experience wereaccurately known, and of which some had to be kept for nineteen monthsbefore they could be trained, the amount of labor and the risk of mishapwhich it entailed were great. To make possible the completion of theinvestigation within two years, I accumulated healthy individuals forseveral months without training any of them. In March, 1907, I hadsucceeded in completing the tests for the age of one month, and I had onhand for the remaining tests almost a hundred individuals, whose agesranged from a few days to eighteen months. Had everything gone well, thework would have been finished within six months. Suddenly, and withoutdiscoverable external cause, my mice began to die of an intestinaltrouble, and despite all my efforts to check the disease by changing foodsupply and environment, all except a single pair died within a few weeks. Thus ended a number of experiments whose final results I had expected tobe able to present in this volume. However, the work which I have done isstill of value, for the single pair of survivors have made possible thecontinuance of my tests with other individuals of the same line of descentas those which perished, and I have to regret only the loss of time andlabor. As I have on hand results for ten individuals of the age of one month, andfor four individuals of the age of four months, it has seemed desirable tostate the problem, method, and incomplete results of this study of therelation of modifiability to age. The indices of modifiability for thesetwo groups of dancers differ so strikingly that I feel justified inpersisting in my efforts to obtain comparable data for the seven ageswhich have been mentioned. TABLE 52 PLASTICITY (RELATION or MODIFIABILITY TO AGE) Number of Errors in Successive Daily Series of Ten White-Black Tests, with Dancers Four Months Old SERIES MALES FEMALES NO. 76 NO. 78 AV. NO. 75 NO. 77 AV. GENERAL AV. A 7 7 7. 0 4 8 6. 0 6. 50 B 8 6 7. 0 6 5 5. 5 6. 25 1 5 5 5. 0 5 5 5. 0 5. 00 2 5 4 4. 5 2 2 2. 0 3. 25 3 4 5 4. 5 2 5 3. 5 4. 00 4 3 4 3. 5 1 1 1. 0 2. 25 5 5 2 3. 5 0 1 0. 5 2. 00 6 3 2 2. 5 1 0 0. 5 1. 50 7 2 1 1. 5 1 2 1. 5 1. 50 8 5 1 3. 0 0 0 0 1. 50 9 1 3 2. 0 0 0 0 1. 00 10 1 2 1. 5 1 0 0. 5 1. 00 11 1 1 1. 0 0 0 0. 50 12 1 1 1. 0 0 0 0. 50 13 0 0 0 0 0 0 14 0 0 0 0 15 0 0 0 0 [Illustration: FIGURE 33. --Plasticity curves. In the left margin are giventhe indices of modifiability (the number of tests necessary for theestablishment of a perfect habit). Below the base line the age of theindividuals is given in months. Curve for males, --•--•--•--; curve forfemales, - - - -; curve for both males and females, ----. When these threeplasticity curves are completed, they will represent the indices ofmodifiability as determined for ten individuals at the age of 1 month, andsimilarly for the same number of individuals at each of the ages, 4, 7, 10, 13, 16, and 19 months. ] The detailed results for the one-month old individuals appear in Table 43;those for the four-month individuals in Table 52. The general averages forthe former are to be found in the third column of Table 46, under theheading "10 tests per day"; those for the latter in the last column ofTable 52. Mere inspection of these tables reveals the curious sexdifference which goes far towards justifying the presentation of thisuncompleted work. The index of modifiability for the ten one-monthindividuals is 88 (that is, 88 tests were necessary for the establishmentof a habit); for the four-month individuals it is 102. 5. The heavy solidline of Figure 33 joins the points on the ordinates at which these valuesare located. Apparently, then, the dancer acquires the white-blackdiscrimination habit less readily at the age of four months than at theage of one month. Further analysis of the results proves that this statement is not true. When the averages for the two sexes are compared, it appears that themales learned much less quickly at four months than at one month, whereasjust the reverse is true of the females. The dash and dot line of thefigure extends from the index of modifiability of the one-month males (72)to that of the four-month males (120); and the regularly interrupted linesimilarly joins the indices of the one-month (104) and the four-month (85)females. In seeking to discover age differences in docility or ability toprofit by experience we have stumbled upon what appears to be an importantsex difference. Perhaps I should add to this presentation of partialresults the following statement. Since there are only four individuals inthe four-month group, two of each sex, the indices are not very reliable, and consequently too much stress should not be laid upon the age and sexdifferences which are indicated. In view of this impressive instance of the way in which averages mayconceal facts and lead the observer to false inferences, I wish to remarkthat my study of the dancer has convinced me of the profound truth of thestatement that the biologist, whether he be psychologist, anthropologist, physiologist, or morphologist, should work with the organic individual andshould first of all deal with his results as individual results. Averageshave their place and value, but to mass data before their individualsignificance has been carefully sought out is to conceal or distort theirmeaning. Too many of us, in our eagerness for quantitative results and inour desire to obtain averages which shall justify general statements, getthe cart before the horse. Figure 33 presents the beginning of what I propose to call plasticitycurves. When these three curves are completed on the basis of experimentswith five dancers of each sex for each of the ages indicated on the baseline of the figure, they will indicate what general changes in plasticity, modifiability of behavior, or ability to learn (for all of theseexpressions have been used to designate much the same capacity of theorganism) occur from the first month to the nineteenth in the male and thefemale dancer, and in the race without respect to sex. So far as I know, data for the construction of plasticity curves such as I hope in the nearfuture to be able to present for the dancing mouse have not been obtainedfor any mammal. At present it would be hazardous for me to attempt to state any generalconclusion concerning the relation of docility to age. The initiative of the dancer certainly varies with its age. In scope theaction system rapidly increases during the first few months of life, andif the animal be subjected to training tests, this increase may continuewell into old age. The appearance of noticeable quiescence does notnecessarily indicate diminished initiative. Frequently my oldest mice haveshown themselves preëminent in their ability to adjust their behavior tonew conditions. However, I have not studied individuals of more thaneighteen months in age. One would naturally expect initiative to decreasein senility. All that I can say is that I have seen no indications of it. We may now briefly consider the principal sex differences which have beenrevealed by the experiments. In sensitiveness I have discovered nodifference, but it should be stated that no special attention has beengiven to the matter. In docility the males usually appeared to be superiorto the females. This was especially noticeable early in my visualdiscrimination tests. The males almost invariably acquired a perfect habitquicker than the females. I may cite the following typical instances. Number 14 acquired the black-white habit with 40 tests; No. 13, with 60(Table 10, p. 109). Of the five pairs of individuals whose records inwhite-black training appear in Table 43, not one contradicts the statementwhich has just been made. It is to be noted, however, that under certainconditions of training, for example, 20 tests per day, the female is at anadvantage. Recently I have with increasing frequency obtained measures ofdocility which apparently favor the female. That this difference in theresults is due to a difference in age is probable. In labyrinth tests the female is as much superior to the male as the maleis to the female in discrimination tests. From the tables of Chapter XIIII may take a few averages to indicate the quantitative nature of thisdifference. A degree of proficiency in labyrinth B attained by the malesafter 7. 0 trials was equaled by the females after 6. 2 trials. In labyrinthC the males acquired a habit as a result of 18. 7 trials; the females, as aresult of 13. 8. And similarly in labyrinth D, 6. 1 trials did no more forthe males than 5. 9 did for the females. That at the age of about one month the male dancer should be able toacquire a visual discrimination habit more rapidly than the female, whereas the female can acquire a labyrinth habit more readily than themale, suggests an important difference in the nature of their equipmentfor habit formation. One might hazard the suggestion that the male dependsmore largely upon discrimination of external conditions, whereas thefemale depends to a greater extent than does the male upon the internal, organic changes which are wrought by acts. At any rate the female seems tofollow a labyrinth path more mechanically, more accurately, more easily, and with less evidence of sense discrimination than does the male. Finally, in concluding this chapter, I may add that in those aspects ofbehavior which received attention in the early chapters of this volume thedancers differ very markedly. Some climb readily on vertical or inclinedsurfaces to which they can cling; others seldom venture from theirhorizontally placed dance floor. Some balance themselves skillfully onnarrow bridges; others fall off almost immediately. My own observations, as well as a comparison of the accounts of the behavior of the dancerwhich have been given by Cyon, Zoth, and other investigators, lead me toconclude that there are different kinds of dancing mice. This may be theresult of crosses with other species of mice, or it may be merely anexpression of the variability of an exceptionally unstable race. I can see no satisfactory grounds for considering the dancer eitherabnormal or pathological. It is a well-established race, with certainpeculiarities to which it breeds true; and no pathological structuralconditions, so far as I have been able to learn, have been discovered. I have presented in this chapter on differences a program rather than acompleted study. To carry out fully the lines of work which have beensuggested by my observations and by the presentation of results wouldoccupy a skilled observer many months. I have not as yet succeeded inaccomplishing this, but my failure is not due to lack of interest or ofeffort. CHAPTER XVIII THE INHERITANCE OF FORMS OF BEHAVIOR In a general way those peculiarities of behavior which suggested the namedancing mouse are inherited. Generation after generation of the mice runin circles, whirl, and move the head restlessly and jerkily from side toside. But these forms of behavior vary greatly. Some individuals whirlinfrequently and sporadically; others whirl frequently and persistently, at certain hours of the day. Some are unable to climb a vertical surface;others do so readily. Some respond to sounds; others give no indicationsof ability to hear. I propose in this chapter to present certain factsconcerning the inheritance of individual peculiarities of behavior, and tostate the results of a series of experiments by which I had hoped to testthe inheritance of individually acquired forms of behavior. My study of the nature of the whirling tendency of the dancer has revealedthe fact that certain individuals whirl to the right almost uniformly, others just as regularly to the left, and still others now in onedirection, now in the other. On the basis of this observation, the animalshave been classified as right, left, or mixed whirlers. Does the dancertransmit to its offspring the tendency to whirl in a definite manner? Records of the direction of whirling of one hundred individuals have beenobtained. For twenty of these mice the determination was made by countingthe number of complete turns in five-minute intervals at six differenthours of the day. For the remaining eighty individuals the direction wasdiscovered by observation of the activity of the animals for a briefinterval at five different times. Naturally, the former results are themore exact; in fact, they alone have any considerable quantitative value. But for the problem under consideration all of the determinations aresufficiently accurate to be satisfactory. The distribution of the individuals which were examined as to direction ofwhirling is as follows. RIGHT WHIRLERS LEFT WHIRLERS MIXED WHIRLERS TOTAL Males 19 19 12 50Females 12 23 15 50 The frequency of occurrence of left whirlers among the females isunexpectedly high. Is this to be accounted for in terms of inheritance? Inmy search for an answer to this question I followed the whirling tendencyfrom generation to generation in two lines of descent. These two groups ofmice have already been referred to as the 200 line and the 400 line. Theformer were descended from Nos. 200 and 205, and the latter from Nos. 152and 151. Individuals which resulted from the crossing of these lines willbe referred to hereafter as of mixed descent. There were some strikingdifferences in the behavior of the mice of the two lines of descent. As arule the individuals of the 200 line climbed more readily, were moreactive, danced less vigorously, whirled less rapidly and lesspersistently, and were in several other respects much more like commonmice than were the individuals of the 400 line. It is also to be noted(see Table 5) that few of the litters of the 200 line exhibited auditoryreactions, whereas almost all of the litters of the 400 line which weretested gave unmistakable evidence of sensitiveness to certain sounds. These differences at once suggest the importance of an examination of thewhirling tendency of each line of descent. The results for the several generations of each line which I hadopportunity to examine are unexpectedly decisive so far as the question inpoint is concerned. INDIVIDUALS OF THE 200 LINE MALES FEMALES First generation No. 200, ? No. 205, ?Second generation No. 210, Mixed whirler No. 215, Left whirlerThird generation No. 220, Mixed whirler No. 225, Mixed whirlerFourth generation No. 230, Right whirler No. 235, Mixed whirlerFifth generation No. 240, Right whirler No. 245, Left whirler INDIVIDUALS OF THE 400 LINE MALES FEMALES First generation No. 152, Left whirler No. 151, Left whirlerSecond generation No. 410, Left whirler No. 415, Right whirlerThird generation No. 420, Left whirler No. 425, Left whirler One line of descent exhibited no pronounced whirling tendency; the otherexhibited a strong tendency to whirl to the left. Are these statementstrue for the group of one hundred individuals whose distribution among thethree classes of whirlers has been given? In order to obtain an answer tothis question I have reclassified these individuals according to descentand direction of whirling. INDIVIDUALS OF THE 200 LINE RIGHT WHIRLERS LEFT WHIRLERS MIXED WHIRLERS TOTAL Males 7 6 8 21Females 5 8 8 21 12 14 16 42 INDIVIDUALS OF THE 400 LINE RIGHT WHIRLERS LEFT WHIRLERS MIXED WHIRLERS TOTAL Males 4 9 1 14Females 6 9 4 19 10 18 5 33 INDIVIDUALS OF MIXED DESCENT 9 10 6 25 Three interesting facts are indicated by these results: first, theinheritance of a tendency to whirl to the left in the 400 line of descent;second, the lack of any definite whirling tendency in the 200 line; andthird, the occurrence of right and left whirlers with equal frequency as aresult of the crossing of these two lines of descent. It is quite possible, and I am inclined to consider it probable, that thepure dancer regularly inherits a tendency to whirl to the left, and thatthis is obscured in the case of the 200 line by the influences of a crosswith another variety of mouse. It is to be noted that the individuals ofthe 200 line were predominantly mixed whirlers, and I may add that many ofthem whirled so seldom that they might more appropriately be classed ascirclers. THE INHERITANCE OF INDIVIDUALLY ACQUIRED FORMS OFBEHAVIOR The white-black discrimination experiments which were made in connectionwith the study of vision and the modifiability of behavior were so plannedthat they should furnish evidence of any possible tendency towards theinheritance of modifications in behavior. The problem may be stated thus. If a dancing mouse be thoroughly trained to avoid black, by beingsubjected to a disagreeable experience every time it enters a black box, will it transmit to its offspring a tendency to avoid black? Systematic training experiments were carried on with individuals of boththe 200 and 400 lines of descent. For each of these lines a male and afemale were trained at the age of four weeks to discriminate between thewhite and the black electric-boxes and to choose the former. After theyhad been thoroughly trained these individuals were mated, and in course oftime a male and female, chosen at random from their first litter, weresimilarly trained. All the individuals were trained in the same way andunder as nearly the same conditions as could be maintained, and accuraterecords were kept of the behavior of each animal and of the number oferrors of choice which it made in series after series of tests. What dothese records indicate concerning the influence of individually acquiredforms of behavior upon the behavior of the race? TABLE 53 THE INHERITANCE OF THE HABIT OF WHITE-BLACK DISCRIMINATION Number of Errors in Daily Series of Ten Tests MALES FEMALES SERIES FIRST SECOND THIRD FOURTH FIRST SECOND THIRD FOURTH GENERA- GENERA- GENERA- GENERA- GENERA- GENERA- GENERA- GENERA- TION TION TION TION TION TION TION TION No. 210 No. 220 No. 230 No. 240 No. 215 No. 225 No. 235 No. 245 A 6 5 6 7 8 4 4 7 B 6 8 8 8 8 7 6 5 1 6 7 6 5 7 6 5 4 2 4 3 1 5 5 6 4 5 3 3 1 4 5 3 4 4 3 4 5 0 3 4 2 1 3 1 5 3 0 4 2 1 3 3 0 6 2 1 4 2 2 1 1 1 7 1 0 3 1 1 1 2 0 8 0 0 1 0 0 0 2 3 9 0 0 0 1 1 0 0 0 10 0 0 1 0 2 1 1 11 0 0 0 3 0 0 12 0 0 0 0 0 13 0 0 0 0 14 0 I have records for four generations in the 200 line and for threegenerations in the 400 line. [1] As the results are practically the samefor each, I shall present the detailed records for the former group alone. In Table 53 are to be found the number of errors made in successive seriesof ten tests each by the various individuals of the 200 line which weretrained in this experiment. The most careful examination fails to revealany indication of the inheritance of a tendency to avoid the black box. No. 240, in fact, chose the black box more frequently in the preferenceseries than did No. 210, and he required thirty more tests for theestablishment of a perfect habit than did No. 210. Apparently descent fromindividuals which had thoroughly learned to avoid the black box gives thedancer no advantage in the formation of a white-black discriminationhabit. There is absolutely no evidence of the inheritance of thisparticular individually acquired form of behavior in the dancer. [Footnote 1: This experiment was interrupted by the death of the animalsof both lines of descent. ] INDEX Abnormal dancers. Acquired forms of behavior. Act, useless, repeated. Activity, periods of. Affirmation, choice by. Age, peculiarities; maximum age; and intelligence. Albino cat; dog. Alexander and Kreidl, young dancer; behavior; tracks of mice; behavior in cyclostat; behavior of white mouse and dancer; structure of ear; deafness. Allen, G. M. , drawing of dancer; heredity in mice. Alleys, width of, in labyrinths. Amyl acetate for photometry. Anatomy of dancer. Animals, education of. Appuun whistles. Audition. _See_ Hearing. Averages, dangers in. Baginsky, B. , model of ear of dancer. Bateson, W. , breeding experiments. Behavior, of dancer; inheritance of; when blinded; equilibration; dizziness; structural bases of; of young; changes in; useless acts; under experimental conditions; in indiscriminable conditions; value of sight; in labyrinth experiments; modifiability of; history of; explanations of; individual differences in. Blinded dancers, behavior of. Blue-orange tests; blue-red tests; blue-green tests; blue-green blindness. Bradley papers. Brain, structure of. Breeding of dancers. Brehm, A. E. , "Tierleben". Brightness vision; preference; check experiments; relation to color vision. Cages for dancers. Candle meter. Candle power. Cardboards, for tests of vision; positions of. Care of dancer. Castle, W. E. , drawing of mouse; cages. Cat, albino; training of. Cerebellum of dancer. Characters, acquired. Check experiments. China, dancers of. Choice, exhibition of; by affirmation; by negation; by comparison; methods of. Circling, a form of dance. Circus course mice. Cleghorn, A. G. Climbing of dancer. Cochlea, functions of. Color blindness. Color discrimination apparatus. Colored glasses. Colored papers. Color patterns of dancers. Color vision, problem; methods of testing; tests with colored papers, tests with ray filters, orange-blue tests, yellow-red tests, light blue-orange tests, dark blue-red tests, green-light blue tests, violet-red tests, green-blue tests, green-red tests, blue-green tests, blue-red tests, structure of the retina, conclusions, of different animals, Comparative pedagogy, Comparison, choice by, Cones, lacking in eye of dancer, Corti, organ of, in dancer, Cotton mouse, Curves, of habit formation, irregularities of, of labyrinth habit, of discrimination habit, of learning and re-learning, of plasticity, Cyclostat, behavior of dancer in, Cyon, E. De, dancer pathological, behavior, behavior of blinded dancers, varieties of dancer, space perception, individual differences, anatomy of dancer, hearing of dancer, pain cries. Dancers, occurrence among common mice, varieties of, hybrid, Dancing, forms of dance movement, whirling, circling, figure-eights, manège movements, solo dance, centre dance, direction of, periods of, amount of, causes of, sex differences in, individual differences in, Darbishire, A. D. , breeding experiments with dancers, Deafness of dancer, causes of, Descent, lines of, Development of young dancer, Differences, individual, sex, Direction of movement, choice by, Direction of whirling, Discrimination, visual, box, of brightness, white-black and black-white, of grays, habits, by odor, by form, method, habit defined, Diseases of dancer, Dizziness, visual, rotational, Docility, Dog, albino, training of, fear of electric shock. Ear, structure of, structural types, model of, of rabbit, functions of, movements of, Educability of dancer, Education, human, methods of, of vision, Efficiency of training, Electric-box for visual tests, Electric-labyrinth for habit experiments, Electric-shock as punishment for mistakes, Epidemic among dancers, Equilibration in dancer, Error curves, form of, Error records versus time records, Errors, in labyrinths, nature of, types of, value of, number of, Even numbers to designate males, Excitability of dancer, Experience, value of, influence of, Eyes, of dancer opening of retina of Fear, in dancerFemales, designated by odd numbers dancing of voice of _See_ SexFighting of dancersFigure eight danceFilters for obtaining colored lightFood of dancerForm discriminationFrog, reactions of repetition of act byFunctions of eye Galton whistleGestation, period of, in dancerGray papersGreen blue testsGreen-red testsGrouping for averagesGuaita, G von, breeding experiments with dancers Haacke, W, description of dancer, origin of dancer breeding experimentsHabit, of dancing, discrimination, useless labyrinth, duration of, reacquisition of, relations of, Habit formation, and the senses, _versus_ habit performance, in the dancer and in the common mouse, curves of, speed ofHabituation to soundsHacker, dancing shrewsHair, appearance ofHamilton, G V, experiments with dogHatai, S, the dancerHead, shape of, in dancerHearing, in dancer in young in adult methods of testing, in frogHefner unit of lightHeredity _See_ InheritanceHering, E, colored papersHistory, of dancer of actsHunger as motive in experimentsHybrid dancers Imitation in dancerIndex of modifiabilityIndividualityInheritanceInhibition of an actInitiative of dancerInsight of dancerIntelligence, measures of, comparisonsInterrupted circuit for experimental useIrregular labyrinths Janssen-Hoffman spectroscopeJapan, dancers inJudgment in dancer Kammerer, P, dancing wood mice, Kishi, K, dancer in Japan origin of race, equilibration, blinded dancer, structure of ear, wax in ears, tests of hearingKönig tuning forks, steel barsKreidl, A _See_ Alexander Labyrinth, forms of, labyrinth A, errors in, tests, labyrinth B, tests, labyrinth C, labyrinth D, a standard labyrinth, regular and irregular labyrinthsLabyrinth errors and individual tendenciesLabyrinth habits, Labyrinth method, Labyrinth path, formula, method of recording, Ladder climbing tests, Landois, H, account of dancer, Lathrop, A, dancers, Learning, process, methods of in dancer, by being put through act, by imitation, by rote, rapidity of, permanency of, learning and relearning, curves of, Left whirlers, Life span of dancer, Light, reflected, transmitted, unit of measurement, control of, Litter, size of, in dancer, Lummer-Brodhun photometer. Males, dancing of, fighting and killing young, designation of, voice of, _See_ SexManège movements, Mark, E L, cages, Maze _See_ LabyrinthMeasurements, of light, of rapidity of habit formation, of intelligence, of efficiency of training, Memory, defined, for ladder climbing, tests of, measurements of, span of, for brightness, for color, Method, of studying dance, for testing hearing, indirect, for testing vision, motives, for brightness vision, for color vision, of shifting filters, of testing form discrimination, of testing Weber's law, development of methods, of choice, food box, labyrinth, of recording errors, of training problem method, labyrinth method, discrimination method, of recording labyrinth path, qualitative versus quantitative, of studying senses, values of methods, of measuring intelligence, quantitative, comparisons of, Milne-Edwards, origin of dancer, Mitsukun, K, the dancer in Japan, Mixed whirlers, Modifiability, of behavior, of useless acts, index of, Motives, for activity, for choice, avoidance of discomfort, in labyrinths, desire to escape, to get food, to avoid pain, Motor, tendencies, ability, capacityMovements, of ears, _Mus musculus L_, _Mus spiciosus L_, _Mus sylvaticus L_. _Nankin nesumi_, name for dancer, Negation, choice by, Nendel, R, gray papers, Nerve, eighth, Nervous system, Nest materials, Noises, effects of, Numbers, odd for females, even for males, reference, _See_ Bibliographic List. Odors, discrimination by, Old Fancier's description of dancer, Olfactory sense _See_ SmellOrange-blue tests, Orientation of dancer, Origin of dancer; by selectional breeding; by inheritance of an acquired character; by mutation; by pathological changes; by natural selection. Panse, R. , structure of ear; explanation of deafness. Papers, Nendel's grays; Bradley's colored; Hering's colored. Parker, G. H. Path in labyrinth, record of. Pathological condition of dancer. Pedagogy, comparative. Perception, of brightness; of color; of movement; of form. Peru, dancers in. Petromyzon, semicircular canals of. Photometer, Lummer-Brodhun. Plasticity of dancer; curves of. Position choice by, of cardboards. Preference for brightness, tests of. Preliminary tests. Probable error. Problems, of structure; of method. Punishment versus reward. Putting-through, training by. Qualitative methods. Quantitative methods. Rabbit, ear of. Rawitz, B. , behavior of dancer; structure of ear; deafness of dancer; hearing in young. Ray filters. Reactions, to sounds; to disagreeable stimuli; valueless. Reasoning, implicit. Reconstruction method. Records, of markings of dancers; of time; of errors; of path. Red, stimulating value of; vision. Reference numbers to literature. _See_ Literature on Dancer. Reflected light. Refrangibility and vision of dancer. Regular labyrinth. Re-learning, relation to learning; curves of. Reliability of averages. Repetition of useless acts. Rest-interval. Restlessness, of dancer; cause of. Retina of dancer. Retzius, ear of rabbit. Reward, for performance of act; versus punishment. Right whirlers; behavior in labyrinth; occurrence of; inheritance of tendency. Rods of retina. Rotational dizziness. Rubber stamps of labyrinths. Saint-Loup, R. Schlumberger, C. ; wood carving with dancers. Selenka, ear of rabbit. Semicircular canals. Sense organs. Senses, and habit formation; differences in. SensitivenessSex, recognition of, designation of, peculiaritiesShellac to coat cardsShrews, dancingSight, role of, _See_ Vision, Brightness Vision, and Color VisionSmell sense of, in labyrinth habitsSniffing by dancerSolutions as ray filters_Sorex vulgaris L_Sound, reactions toSpace perceptionSpectroscopeSpectrum, stimulating value ofStandard, candle, light, labyrinthStine, W M, photometrical measurementsStrength of dancerStructure, of brain, of ear, of eyeSwinhoe, mice in China Temperament of animalTemperature senseTests, visual, number of, per day, Threshold of discriminationTime recordsTouch, and labyrinth habitsTraining conditions of, Weber's law, methods of, and retraining, in labyrinths, efficiency of, two test, ten-test, twenty-test, continuous, relation to methods, spread ofTransmitted light. Variability of dancer, Variable light. Varieties of dancerViolet red testsVision, brightness vision, color vision, training of, importance of, conclusions concerningVisual dizzinessVoice of dancer Watson, J B, habit formationWaugh, K, color vision apparatus, retina of mouseWax, plugs of, in ear of mouseWeber's law, tests of, apparatusWeldon, W F R, breeding experiments, Whirling of dancer Yellow Red testsYoung dancers, killing of, by male, description of, development of, hearing of, intelligence of, size of Zoth, O, origin of dancer, size of young mice, the senses of dancer, behavior, dancing, equilibration, climbing dancers, individual differences, tests of hearing, vision