ores “ “ee * ‘. ae: © © 4-6 4-6 48 8 . * . dtieetet . +999 ’ Peeve ‘7. we vie eee ¥-6. 4) THE JOURNAL OF COMPARATIVE NEUROLOGY AND PSYCHOLOGY EDITORIAL BOARD Henry H. DonALpsoN ApOLPH MEYER The Wistar Institute Pathological Institute, New York C. Jupson Herrick OuiveR 8. Strona University of Chicago Columbia University HERBERT S. JENNINGS JoHun B. Watson Johns Hopkins University Johns Hopkins University J. B. Jounston Rospert M. YERKES — University of Minnesota Harvard University VOLUME XIx 1909 PHILADELPHIA, PA. THE WISTAR INSTITUTE OF ANATOMY AND BIOLOGY i i. sea iG , : agpederate > 1 to verte bak * J tt ar rd ; etd rhb ae, 1) lag oe bs PO Ic) ee Tae eee i i S “ ( . 4 7 V ir < * o Coen ’ tA : * os nm . k - fh re -~ i ) i * ‘ A Peas f es a: ae * i \ , Se ‘ . a ' ; - i p 7 — \ £ ? ) ' af ¥ ; ‘ - = n ’ ' 4 , 4 : ¥ 7 x « , . & i ‘ " - A vi —_ a its m3 ; 5 ; : > ; / J 7 = 3 ; 2 , ss . ~ 4 f ) 4 { ‘ S t hl - oJ ‘ t iz ' ® - i i Pie Se - ae Sal as ay 7 ) ; n . / - = . \ i a - The Journal of Comparative Neurology and Psychology CONTENTS OF VOL. XIX, 1909. Number 1, April, 1909. PAG, Some Experiments Bearing upon Color Vision in Monkeys. By JouHn B. Watson. (From the Psychological Laboratory of the University of Ohicago.) With Five Pigures: ... 2.00... ee 5 oe clot cent nce rceene 1 The Expressions of Emotion in the Pigeons. I. The Blond Ring-dove. (Turtur Risorius). By Wattace Craic. (From the Department of Zoology of the University of Chicago.) With One Plate ............ 29 The Reaction to Tactile Stimuli and the Development of the Swimming Movement in Embryos of Diemyctylus Torosus, Hschscholtz. By G. EH. Cocuitn. (Studies from the Neurological Laboratory of Denison University. No. XXII.) With Six Figures .......:. 2.5.2... Keeueei ess 83 Sensations Following Nerve Division. By SHEPHERD Ivory FRANZ. (From the Laboratory of the Government Hospital for the Insane, Wash- ington, D. OC.) I. Pressure-like Sensations. With Five Figures .... 107 Alterations in the Spinal Ganglion Cells Following Neurotomy. By S. Wa.LreR Ranson. (From the Anatomical Laboratory of the Univer- Sion Cnicago:). Introdutions Wath Six Hg URS terse -)eleus ols teyehel'« 125 Number 2, May, 1909. On the Relation of the Body Length to the Body Weight and to the Weight of the Brain and of the Spinal Cord in the Albino Rat (Mus Norvegi- cus var. Albus). By Henry H. Donatpson. (Professor of Neurology at the Wistar Institute.) With Three Wigures!.::...-............0.. 155 Note on the Formulas Used for Calculating the Weight of the Brain in the Albino Rats. By SHINKISHI HaTaL. (From the Wistar Institute.) 169 The Nervus Terminalis (Nerve of Pinkus) in the Frog. By C. JupSON Herrick. (From the Anatomical Laboratory of the University of OECHIO,)) * WAM MAKES GoopdaaobooocaboogdooccuoMdogubduonOL 175 The Nervus Terminalis in the Carp. By R. E. SaHetpon. (From the Ana- tomical Laboratory of the University of Chicago.) With Seven Figures. 191 The Criteria of Homology in the Peripheral Nervous System. By C. Jup- son Herrick. (From the Anatomical Laboratory of the University Gj) CHRO) 35 Pro HD OA OOS WACO OGIO 2 con Co Gad olen Oo oD 203 LAiGEREy . IN(CTIGRS: As Oc OIRO Oo eaianiom 6 oidic Cao dod 05 6.0 oSelo ow Ga como Gar 211 lv Contents. Number 3, June, 1909. PAGH. On Sensations Following Nerve Division. By SHEPHERD Ivory FRANZ. (From the Laboratories of the Government Hospital for the Insane, Washington, D. C.) II. The Sensibility of the Hairs. With Seven [Dieaieae MAAN HOD Ooh AD coo Gal boon bo COD oUU boo dom eG bDOb GosagEOGoOUnS Modifiability of Behavior in its Relations to the Age and Sex of the Danc- ing Mouse. By Roperr M. YERKES. (From the Harvard Psychological Laboratory.) With Four Figures .......... fierie ws ashe. tanaiie, cue Mepersiane eke te Swerve The Reactions of the Dogfish to Chemical Stimuli. By RALteH E. SHELDON. (Contribution from the Woods Hole Laboratory of the United States Bureau of Fisheries.) With Three Figures Ce The Work of J. von Uexkuell on the Physiology of Movements and Be- lnvayatorey asi JEG [Sh GPowaAGNels) Gooccoconaccsmeoodcnd ideredocc hatoneiaie ert Number 4, July, 1909. Imitation in Monkeys. By M. E. Haccerry. (From the Harvard Psycho- logical Laboratory.) With Thirteen Figures ...................-.-- Number 5, November, 1909. The Morphology of the Forebrain Vesicle in Vertebrates. By J. B. JOHN- ston. (University of Minnesota.) With Forty-five Figures.......... Some Experiments upon the Behavior of Squirrels. By C. S. YOAKUM. (From the Psychological Laboratory of the University of Chicago.) , Visio Mba Mee 5 qo adoongocd cuconodoooUddcoddedeoonosenasccKoaC Tropic and Shock Reactions in Perichseta and Lumbricus. By His He Harper. (From the Zoological Laboratory of Northwestern Univer- sity.) With Two Figures .............2.05+--2 +02 cers eer eeeeenrns IGE, INVNGES gan goconcomooonodbonco Boo DDDDKO DOGO DOH dDUZUGGOOuCGONS Number 6, December, 1909. The Radix Mesencephalica Trigemini. By J. B. JOHNSTON. (University of Minnesota.) With Thirty-two Figures ...........- PEE Shh ome cic A New Association Fiber Tract in the Cerebrum with Remarks on the Fiber Tract Dissection Method of Studying the Brain. By E. J. CurRAN. (Harvard Medical School.) With Three Plates .........- Visual Discrimination in Raccoons. By L. W. Corr and F. M. Lone. With One: Migure’ |. . slevale cos eleye eyes cuees sicher Oks epee eRe) RR hme toon lean lennon A Statistical Study of the Medullated Nerve Fibers Innervating the Legs of the Leopard Frog (Rana Pipiens) after Unilateral section of the Ventral Roots. By Eizasrra H. Dunn. (From the Anatomical Laboratory of the University of Chicago.) With One JNAIRE Ata aa oo Factors Determining the Reactions of the Larva of Tenebrio Molitor. By Max Morse. With Two Figures .............2---+--e0ees--s seers 215 237 278 313 337 457 041 569 589 593 SUBJECT AND AUDHOR- INDEX VOLUME XIX. LTERATIONS in the spinal ganglion cells following neurotomy, 125. EHAVIOR, modifiability of, in its relations to the age and sex of the dancing mouse, 237. Of squirrels, some experiments upon, 541. Body length, relation of, to body weight and to the weight of the brain and of the spinal cord in the albino rat, 155. Body weight, relation of body weight to, and to the weight of the brain and of the spinal cord in the albino rat, 155. ARP, the nervus terminalis in, 191. Cerebrum, a new association fiber tract in the, 645. CoGHILL, G. E. The reaction to tactile stimuli and the development of the swimming movement in embryos of Diemyctylus torosus, Eschscholtz, 83. CoLp,L. W., and F. M. Lone. Visual dis- crimination in raccoons, 657. Color vision in monkeys, some experiments bearing upon, 1. CRAIG, WALLACE. tion in the pigeons. dove, 29. CURRAN, E. J. A new association fiber tract in the cerebrum, 645. The expressions of emo- The blond ring- OGFISH, the reactions of, to chemical D stimuli, 273. DONALDSON, Hpnry H. On the relation of the body length to the body weight and to the weight of the brain and ce spinal cord in the albino rat, 5. Dunn, HE. H. A statistical study of the medullated nerve fibers innervating the legs of the leopard frog after unilat- eral section of the ventral roots, 685. MOTION in the pigeons, the expressions E of. I. The blond ring-dove, 29. IBER tract in the cerebrum, a new asso- ciation, 645. Forebrain vesicle in vertebrates, the morph- ology of, 457. Formulas used for calculating weight of the brain in albino rats, 169. Sensations fol- I. The pres- II. The sen- FRANZ, SHEPHERD Ivory. lowing nerve division. sure-like sensations, 107. sibility of the hairs, 215. Frog, the nervus terminalis in the, 175. ANGLION cells, alterations in the spinal, S following neurotomy, 125 Hs M. E. Imitation in monkeys, 337. Harper, B. H. Tropic and shock reactions in Pericheta and Lumbricus, 569. Harar, SHINKISHI. Note on the formulas used for calculating the weight of the brain in albino rats, 169. Herrick, C. JuDSoN. The nervus terminalis in the frog, 175. The criteria of homology in the periph- eral nervous system, 203. Homology in the peripheral nervous sys- tem, the criteria of, 203. | fica in monkeys, 337. ENNINGS, H. S. The work of J. von Uexkuell on the physiology of move- ments and behavior, 313. JOHNSTON, J. B. The morphology of the forebrain vesicle in vertebrates, 457. The radix mesencephalica trigemini, 593. onG, F. M., and L. W. Coun. Visual discrimination in raccoons, 657. Lumbricus, Pericheta and, tropic and shock reaction in, 569. : (735) 736 legs of the leopard frog after uni- lateral section of the ventral roots, a statistical study of, 685. MU tee ot nerve fibers innervating the Modifiability of behavior in its relations to the age and sex of the dancing mouse, 92 mot. Monkeys, imitation in, 387. Some experiments bearing upon color vision in, 1. Morphology of the forebrain vesicle in ver- tebrates, 457. Morsk, Max. Factors determining the reac- tions of the larya of Tenebrio molitor, (eal Mouse, modifiability of behavior in its rela- tions to the age and sex of the danc- ing; 23%. PRVE division, sensations following. I. The pressure-like sensations, 107. Il. The sensibility of the hairs, 215. Nervous system, the criteria of homology in the peripheral, 203. Nervus terminalis in the carp, the, 191. In the frog, 175. Neurotomy, alterations in the spinal gan- glion cells following, 125. tropic and shock reactions in, 569. Physiology of movements and behavior, the work of J. yon Uexkuell on the, SiS: Pigeons, the expressions of emotion in the, I. Blond ringdove, 29. Pao and Lumbricus, Pinkus, nerve of, in the frog, 175. in, 657. the, ACCOONS, visual discrimination Radix mesencephalica trigemini, 593. Rana pipiens, a statistical study of the medulated nerve fibers innervating the legs of the leopard frog. after paheteeal section of the ventral roots, Ranson, S. Wawvtor. Alterations in the spinal ganglion cells following neurot- omy, 125. Rat, relation of the body length to the body weight and to the weight of the brain and of the spinal cord in the albino, 155. Weight of the brain in albino, note on the formulas used for calculating, 169. Reactions in pericheta and tropic and shock, 569. Of the larva of Tenebrio molitor, factors determing the, 721. lumbricus, Index. ENSATIONS following nerve divisions. I. The pressure-like sensations, 107. Il. The sensibility of the hairs, 215. SHELDON, R. E. The neryus terminalis in the carp, 191. The reactions of the dogfish to chemical stimuli, 273. Spinal cord in the albino rat, relation of the body length to the body weight and to the weight of the brain and of the, 155 Spinal ganglion cells, alterations following neurotomy, 125. in the, Squirrels, some experiments upon the _ be- havior of, 541. Stimuli, the reactions of the dogfish to, 273. The reaction to tactile, and the devel- opment of the swimming movement in embryo of Diemctylus_ torosus, Eschscholtz, 83. Swimming movement in embryos of Die- myctylus torosus, Hschscholtz, the re- action to tactile stimuli and the de velopment of, 83. HE stimuli, the reaction to, and the development of the swimming move- ment in embryos of Diemyctylus torosus, Eschscholtz, 83. Tenebrio molitor, factors determining the reactions of the larva of, 721. Trigemini, the radix mesencephalica, 593. PXKUELL, J. voN, the work of, on the physiology of movements and behavior, Biles Neeoe discrimination in raccoons, 657. sanece de_ l’intelligence, George Bohn, 589. Review of the animal mind. Second volume of the Animal Behavior series, edited by R. M. Yerkes, 211. WATSON, JOHN B. Some experiments bear- ing upon color vision in monkeys, 1. Weight of brain and of the spinal cord in the albino rat, relation of the body length to the body weight and to, 155. Weight of the brain in the albino rats, note on formulas used for calculating the, 169. W “sance M. F. Review of La nais- ERKES, RoBeERT M. Modifiability of be- havior in its relation to the age and sex of the dancing mouse, 237. Yoakum, C. S. Some experiments upon the behavior of squirrels, 541. The Journal of Comparative Neurology and Psychology VoLuME XIX APRIL, 1909 NuMBER I SOME EXPERIMENTS BEARING UPON COLOR VISION IN MONKEYS. BY JOHN B. WATSON. From the Psychological Laboratory of the University of Chicago. WITH FIVE FIGURES. For over a year, the writer has been experimenting with apparatus for obtaining large bands of spectral light suitable for use as stimuli in testing the color vision of animals by means of the discrimination method. In the process of constructing the apparatus finally adopted, the writer has availed himself liberally of the assistance of Pro- fessors Gale and Milliken of the physics department of the Uni- versity of Chicago, and of Professor R. W. Wood of the physics department of the Johns Hopkins University. He is also under deep obligation to Professors Angell and Carr for many valuable sug- gestions. The actual accumulation of the data bearing upon color vision in monkeys began March 12th and ended August 20, 1908. Two rhesus monkeys (J. and B.) and one cebus (H.), all gentle and accustomed to experimentation, were the subjects used in the investi- gation. The report is given in its present incomplete form, because of the fact that the writer's work could not be continued at the University of Chicago. The apparatus used there, however, has been duplicated in the Hopkins laboratory and the investigation will be continued in the latter place both upon two of the monkeys, THE JOURNAL OF COMPARATIVE NEUROLOGY AND PsyCHOLOGY.—VoL, XIX, No. 1. 2 Fournal of Comparative. Neurology and Psychology. J. and B., which served as subjects in the present work and upon other monkeys. In addition to continuing the tests upon color vision, steps are being taken thoroughly to test the delicacy of the white-light vision of the monkeys and their sensitivity to differences in the size and form of visual stimuli. In view of these further tests, which are concerned with the nature and delicacy of the visual reactions as a whole of this animal, it would seem to be premature in this preliminary report to enter into any general discussion of color vision in animals or to take up the related structural facts which bear upon color vision. The color, vision of several different species of animals is being tested in the various laboratories and for this additional reason the writer will confine his statements on the historical side strictly to the work of Kinnaman, which, so far as the writer is aware, is the only study of the color vision of this animal which can lay any claim to scientific accuracy. Kinnaman’s' work was certainly as careful and as exact as his method would permit. His stimuli were obtained by means of light reflected from pigmented papers. The method was roughly as fol- lows: A board, 1 inch by 7 inches, and 5 feet long contained ‘six holes pierced at regular distances. Each hole was large enough to admit the bottom of a cylindrical glass. The convex surface of each glass was covered with colored papers or with different grays. Food was kept with one of the colored glasses, the position of which could be varied at will after each test. The discriminations were made rapidly. ‘In order to determine whether brightness or color was the basis for discrimination four control tests were made. In three of these, I attempted to determine whether the monkeys could discriminate greys and colors varying by the same degree of bright- ness equally well. If blue and red, for example, with a difference in brightness of 15° (determined by the flicker method) were differ- entiated perfectly, and two grays differing by 15° very imperfectly, then color very probably was the basis of the discrimination in our first series of tests.” Both original tests and control tests were 1American Journal of Psychology, Vol. 13, p. 98. Watson, Color Vision in Monkeys. Z carefully made and so satisfactory were the results obtained from them that we find the author expressing himself somewhat extra- vagantly as follows: “1. There can be no doubt that monkeys perceive colors. “2. Two grays having a given degree of difference in bright- ness are not discriminated as well as two colors having an equal difference in brightness. “3. For accurate discrimination of difference in brightness a difference of about 35 degrees or 9 per cent of the white constituent of the gray is necessary. “4, The monkeys are able to distinguish colors from grays though the brightnesses are the same. “5. The male appears to have a preference for bright colors, but blue seems to be discriminated against.” “6. In two instances there were indications of at least a low form of general notion.” For detailed reasons, which can not be entered into here, the assertion is ventured that the use of colored papers can never give us a satisfactory test of color vision in animals.® Certainly, the writer has room for “doubt” both in Kinnaman’s work and in his own. The fact that there is a high percentage of correct choices of the positive color in both investigations cannot be doubted, but that the correct choices were made upon basis of ‘hue’ and not that of ‘brightness’ cannot be decided so easily as Kinnaman sup- poses. It would not be at all ridiculous for one, skeptical of the possibility of testing the color vision of animals, to assert that every discrimination which has been made between two colored papers by an animal could have been made if it had been totally color-blind. As a reason for such scepticism, a few of the defects of this method as a whole may be mentioned now: 1. The surfaces of the papers differ greatly, owing to accidents *Blue certainly was not discriminated against by the animals used by the writer. See Table VI. “The committee appointed by the American Psychological Associution at the 1907 meeting to report upon methods and apparatus for testing vision in animals will take up in its report the defects and advantages of the various ‘methods of testing color vision. 4 Fournal of Comparative Neurology and Psychology. in manufacture, dyeing, ironing, ete. In addition, it is extremely difficult to bend colored papers around glasses or to paste them upon doors so accurately that slight differences in form, size and depth do not appear. To sum up these defects under 1, we may say that colored papers afford numerous secondary criteria. 2. They do not reflect monochromatic bands, but overlapping bands. This is especially true of those reflecting the shorter wave lengths. | 3. Lhe range of intensity obtainable in them is so limited, that if any given region of the spectrum should offer to the animal a different order of intensity from that which the same region offers to our own eyes, the slight change which we could introduce in the brightness of a given stimulus, by substituting a paper of the same color only lighter or darken (to our own eyes) maght not at all reverse for the animal the intensity relation originally existing between the colors. Yerkes, in his work on the dancing mouse, mentions most of these objections and shows by experiment that for that animal the red end of the spectrum is probably extremely weak in intensity. Whether or, not the different parts of the spectrum possess the same intensity for the eye of the monkey, as they do for our own, is a question which we have at present no data for deciding, but cer- tainly in the present state of our knowledge we cannot assume such to be the case. Believing that many problems in the study of color vision in animals cannot be solved without the aid of a continuous spectrum, and being more or less disheartened by the failure of colored papers and filters (as they have heretofore been used) to furnish suitable stimuli for testing even the more elementary questions at issue, the writer began work upon a spectral light apparatus which it is hoped will make possible an experimental treatment of the following problems: 1. Has the animal the power to discriminate between any given color and any other selected color with equal ease, when the relative ‘Yerkes, R. M. The Dancing Mouse. The Macmillan Company, 1908. Watson, Color Vision in Monkeys. 5 intensity of either color, and the absolute intensity of both may he altered at will? In investigating this problem, we ought to be able to find totally color blind animals, red-green blind animals, animals with normal color vision, if such differences in sensitivity exist. After the problem has been solved, color theories based upon the phylogenetic development of a photo-chemical molecule will or will not have bases in fact. 2. How nearly identical in wave length may any two colors be and still afford a basis for the animal to discriminate between them (the qualitative ‘difference limen’, D.L.) ? 3. How nearly identical in intensity may two bands of the same wave length be and still afford a basis for discrimination (the D.L. for intensity) ? 4. Do the different parts of the spectrum possess different thresh- old values (stimulus limen, R.L.) ? 5. Is the spectrum of a given animal wider or narrower than the average width of the spectrum of man ? DESCRIPTION OF APPARATUS. As Figs. 1 and 2 show, the light apparatus is made upon the principle of a spectrometer. Fig. 1 shows, in order, the are, A, the condensing lens, L,, the slit, S,, the collimating lens, L,, the prism, P, the mirror, M,, and the lens, LZ, (all are enclosed in a system of dark boxes). The arc is an ordinary hand-feed are. The direct 220 v. current supplying the are is furnished by the university power-house. This current is very steady and uniform. The positive (cored) carbon is placed horizontally and in the axis of the optical system. The are is so arranged that it can be adjusted by the experimenter in the adjoining room at K (Fig. 2). Two long rods, RR, extending from the are into the experimental room permit this. AC in Fig. 1 shows the gearing system by means of which the long rods are brought into connection with the short feeding rods of the are. After practice the experimenter can control the are at K through ‘In the apparatus to be described, a double image prism would afford the conditions for this test, 6 Fournal of Comparative Neurology and Psychology. long periods of time without allowing sensible alteration in its intensity. Since the are is not in the dark-room where the animals react, the noise made by it is not a source of disturbance. When burning well, this are rarely makes a noise which is noticeable even near at hand, after the heavy wooden box (metal lined) has been closed. The lens used as a condenser is an ordinary 4” biconvex read- ing glass with 8” focus. This lens gives a clearly defined image of the crater of the positive carbon at its focus. The slit, S;, which is placed at the focus of [,, is a common op- tical slit with knife edge opening. Its width was adjusted once for all to give a spectrum of high intensity and was never thereafter changed. The collimating lens, Z,, is a heavy 314” achromatic lens with a focal distance of 18”. The slit, S,, is in the focus of this lens, consequently the face of the 65 mm. heavy flint glass prism, P, is filled with parallel light admitted by it. The now refracted beam falls upon the mirror, M, (the use of this mirror is necessitated by the narrowness of the room) which is silvered upon its anterior surface. This mirror reflects the beam through the achromatic lens, £3, (similar in all respects to LZ, except that D,’s focal distance "is 24”), The lens, L;, brings the refracted beam to a focus upon the double slit, S,, (not shown in Figs. 1 and 2, but shown separately in Fig. 3) in a series of colored images of S,. The solar lines are plainly visible and serve as a guide in the selection of particular wave lengths. The apparatus can be so arranged (by revolving M, upon its axis) that these lines shall coincide with definite mm. divisions of the slit.® In order that this double slit may be more easily understood, the following description is given: Fig. 3 shows the slit horizontally. The sharply outlined spectrum falls upon the polished surface of the slit between 0 on the SC scale, and 0 on the SC, scale. Two selected portions of the spectrum pass out between the knife edge *For example, the Na line can be made to fall upon 3 of the scale SC, ete. Watson, Color Vision in Monkeys. 7 openings, J-J,, and J,-J,. The width of these openings is controlled by means of the micrometer screw system Cal, HS, K, B, the mechan- ism of which is well known (i. e., turning Cal, e. g., forces the nut, K, backwards or forwards as the case may be and consequently the edge of the jaw at J advances or recedes from J/,). The two small jaws, J; and J2, must be moved by hand. They are held in place (i. e., in the grooved track of the slit) by means of a small bowed spring. Since they are cut accurately to fit the track in which J and /, slide, they are held firmly in vertical posi- tion by means of the spring. If it is desired to have the opening, J-J,, admit some other part of the spectrum, the apparatus easily permits it. Suppose we desire to have the opening at 6 instead of at 3.65, as it stands in the diagram (see relation of scale SC to index J). We turn the screw, Cal, until J falls at 6. The small jaw, J,, is then pushed up flush against J. J is then pushed forward, carrying J, with it for whatever, width of slit is desired ; J is then backed again to 6. This leaves the opening, J-J,, in its new position optically perfect. By means of the micrometer screw head this distance is made accurate to 1/1000 mm. This would leave a much wider space at 0 than before. This variable opening at 0 is closed with a strip of black cardboard. Four tiny points, D, Pi. Po» Pz, On the jaws, J, and Jo, facilitate this.” The position of this slit in the system can be inferred from an examination of Fig. 2: L, of Fig. 1 is shown on the right in Fig. 92; S, is at the focus of this lens, 24” distant. The two small vertical mirrors, M, and M,, placed at an angle of 45° so as to form a horizontal V (with apex directed towards S, and midway between the two openings) serve to catch the two selected beams (e. g., red and green) issuing from the openings, J-J, and J-,J, of S, and to reflect them to the mirrors, M, and M, respectively. These latter two mirrors in turn reflect the two beams in a parallel way down the room to the screEN. The width between these two mirrors can be adjusted to any desired distance. Since the rays as they issue from the openings of S. are diverging, we have a broad, diffuse, ™The slots, Sl and SJ, serve to admit bolts for attaching slit as a whole permanently in its vertical position. Fournal of Comparative Neurology and Psychology. N — | B J i | 4 | . | ! : | HS B ) S = \; 4 = S i} Nz " | 4 i aS. Iii ] ; ] ~~ ii a, \AZN % ] 5 \ 3 JG, Whe = —— : a - ‘| pu: Key ag i 1 7 oe & lohan a fe) . ‘] Z! - ty ro) COLOR VISION IN MONKEYS JOHN B. WATSON i | =| il i i ii aT TT en —? 7 i =| lilt Z WNMMLULA FLEES UIAAITES ED AKUDEDE BLD PALLET PIII AT UDRP EO BEEBE IT Via —e y. IN SSSR SESS SSS SS . | Irradiation sina CO SER TTT : A pn I | * ; Caines i a | @ i © Fic. 3. THY JOURNAL OF ComPaRATIVH NBUROLOGY AND PsycHOLOGY.—Vou, RIK, Nov 1: ,; Hi te haa Watson, Color Vision in Monkeys. Fie. 5. 10 Journal of Comparative Neurology and Psychology. faint band of red and of green light on the screen. In order to intensify and sharply define the bands on the screen, two small achromatic lenses, L, and L;, are interposed in the pathways of the two beams, #& and G respectively. These two lenses are of short focal length (6”). They project sharply defined and enlarged images of the two openings of S, upon the sereen (marked rED and GREEN in the diagram). These images are about 7” in height and 114” in width. . Four reversing mirrors, RM, are used in order to reverse the right and left positions of the two beams. These mirrors revolve in a vertical plane. They are mounted in bearings in such a way that the small weights, Wgt, pull them back to the 45° position whenever the cords, CRM, are slackened, as in the diagram. These cords are jointly fastened to a rod at X. A single forward pull upon this rod brings all four mirrors to the 180° position, in which position they no longer intercept the two beams. A glance at the apparatus will show that when, e. g., the red is on the left, the reversing mirrors have to intercept the beam; when it is on the right, they no longer intercept it and the beam is reflected directly from the mirror, M,, to the screen. It is clear from this that the absolute intensity of the two bands is slightly less in the “reversed position” than in the normal. But this redue- tion occurs in both bands equally.8 In order to compensate for this reduction, a vertical sliding bar, SB, is placed in the pathway of the beams. 1” x 2” windows are eut in this bar at the points where the beams impinge upon it. One half of each: window is “All the mirrors in the system are silvered on the anterior surface. They are kept highly polished by the use of “jewelers’ rouge.” When not in use, they are kept covered with silk handkerchiefs. The absorption of the light consequently is kept constant and at a minimum. [Since the mirrors used in the apparatus are a source of a great deal of trouble and care and since their use causes a certain variation in the absolute ‘intensity of the light, effort was made to find a substitute for them. After some experimentation it was found that total reflection prisms could be made to separate the beams, to space them properly, and finally to reverse them. In addition the dark room at Hopkins is large enough to accommodate the apparatus without the use of the mirror, M,, behind the large prism. The whole apparatus is now “self-maintaining” and completely constant so far as the absorption of the light is concerned. ] Watson, Color Vision in Monkeys. en left open, the other half is covered with a sufficient number of strips of plate glass to compensate for the absorption of the reversing mirrors. To one end of this bar a spring is attached. When this spring operates alone, the bar is held in such position that the beams of light have to pass through the open halves of the windows. A eord, CSB, runs from the opposite end of the bar around to the rod at XY, which controls the mirrors. Pulling upon this cord brings the sliding bar forward to such a position that the beams have to fall upon the halves of the openings which are covered by the plate glass. This bar is made to work synchronously with the mirrors in such a way that when the reversing mirrors are “out” the plate glass windows are “in” and vice versa. A simple forward pull or a release of the rod at VY adjusts both mirrors and bar. This compensatory device was used in certain of the control tests described below (all of the red-green), but since many trials showed that the reactions of the animals were not altered by its insertion or removal, its use was discontinued.? Behind the opaque screen, there is a vertically placed 12” x 24” pane of acid ground glass (acid ground is less granular than common ground glass; milk glass would have been used, could it have been obtained). The moment the screen is raised by pulling upon H, the two bright-colored bands (surfaces) appear. Immediately behind each of these bands (they are 8” apart)'® a hole is cut in the plat- form to admit the food-boxes shown in Fig. 4. A glass partition, GP, set in a low wooden base serves to keep the animal from opening both food-boxes at once. It also serves to force the animal to go clearly to the right or to the left and to keep a position habit from forming." *"However, if one were working with two colors approximately equal in intensity to the animal, it might very well happen that this change in absolute intensity would, owing to possible onset of the Purkinje phenomenon, alter the intensity relation for the animal. »This distance depends upon the distance of M, from M,. “Before this partition was at hand, one monkey, whose records are not given, went always to the right, then down the screen of ground glass to the left until he came to the box which contained the food. i) ‘fournal of Comparative Neurology and Psychology. Mernop or ConrroLiting THE INTENSITY OF THE Two Licuts. The diagrams (Figs. 1 and 2) show that the absolute intensity of the entire spectrum may be changed by opening or closing the slit S,. Since the present tests are not concerned with faint or weak spectra (as such), the width of this slit, as has been stated, was kept constant and as wide as was possible still to permit a sharply defined spectrum at S,. The intensity of the two selected bands can be altered separately in two ways: By attaching an iris diaphragm (or better, possibly, an Aubert) to each of the two projecting lenses, 1, and LZ, or by the use of an episcotister. In the experiments here reported, the intensity of the red and of the green was controlled by the use of the iris diaphragm, while that of the blue and of the yellow was controlled by the episcotister. The chief objection to the diaphragm lies in the fact that the width of the band changes slightly when the diaphragm is opened or closed. By an oversight, the spectroscopic reading was taken only when the diaphragm was completely open. It was my intention to return to the red-green discrimination during the summer months when it was possible to obtain sunlight, and to use both the diaphragm and the episcotister, but I found that the time at my disposal did not permit this. Accordingly, all the red-green tests reported below were made with the are-as a source; and all changes in intensity of the two bands were made by means of the diaphragm (in table of constants, e. g., “red maximum” was obtained by a wide open 5S diaphragm, “red minimum” by a fixed pinhole opening in the diaphragm). On the other hand, all changes in intensity in the blue-yellow tests were made by means of the episcotister. In the present state of color photometry, it is desirable to have some check upon photometric readings. The episcotister furnishes such a check by allowing us to increase or decrease in a constant way the angular opening through which the beam is allowed to pass. In the present work, the episcotister proved eminently satisfactory. In the table of constants given below, the maximum yellow, e. g., was the normal intensity of the beam as it came from S,; the minimum yellow was the intensity of the beam after it had been interrupted by the episcotister, set with a 30° opening (15° on each side). The angle Watson, Color Vision in Monkeys. 13 of 30° was chosen as the minimum after several preliminary trials. It was desirable to keep the minimum intensity of any beam always well over the human threshold. Any smaller angle did not permit this, with sunlight as the source. That this minimum was also well above the animal’s (reaction) threshold was tested in the following way: The episcotister, set at the minimum (30°), was allowed to interrupt the blue; the yellow was cut out at S, and the animal tested at XY in the ordinary way. When the screen was raised, only the blue band appeared. As a result, it was found in every case that the animal followed the hight regardless of its right or left posi- tion. The blue was then cut out at S, and the yellow interrupted, with similar results. There is the possibility, however, that the minimum intensity was over the ‘brightness limen’ but not over the ‘color limen.’ This objection cannot be met until extended threshold tests have been made. In conducting such experiments in the future, the following pro- cedure will be adopted: first, during the formation of the associa- tion, an episcotister opened to the maximum (320°) will interrupt each beam continuously during all tests;'* second, after the associa- tion has been established, the control tests will be made with the two episcotisters set at any desired angle; third, the iris diaphragm will be used as an additional control.® The episcotister was run at a very high rate of speed. As a test as to whether the beams were uniform for the animal, the screen 2In duplicating the apparatus at Hopkins, the motor and the two episcotisters were mounted upon a small revolving table. This table is so arranged that a pull on a cord (at X, Fig. 2, where the experimenter sits) will interchange the positions of the two episcotisters, thus making it possible to have the animal on the one trial react, e. g., to “minimum red,’ “maximum green,” and on the next to “maximum red,” “minimum green.” 8Tt will be remembered that there are three common ways in which intensity in the physiological sense can be altered: (1) by decreasing the amplitude of the ether waves of the beam which falls upon a given retinal area (increas- ing distance of source) ; (2) by lessening density of beam (use of diaphragms, ete.) ; (3) by interrupting beam (episcotister). In order to test whether the physiological effect, e. g., of distancing the source of the beam is the same as interrupting it, it is desirable to make changes in intensity by employing all three methods. The desirability of the use of the episcotister and dia- . phragm has been assumed in the present work. 14 ‘“fournal of Comparative Neurology and Psychology. was sometimes raised before the episcotister had gained full speed. The flickering ght never failed to frighten the animals. They would never leave my shoulder to make a choice until the lights appeared perfectly steady to my own eyes. Mertuop or Dererminine Invrensiry oF ILLUMINATION OF Mownocuromatic Banps. After vainly trying to obtain a photometric reading of the minimal intensity of the monochromatic bands,!4 with a photometer based upon the Joly principle, I finally, on the advice of Professor Milliken, abandoned the photometer and had resort to the simple apparatus, the ground plan of which is shown in Fig. 5. In the diagram, V is a band of monochromatic light visible upon the ground glass surface. C is a white surface (bristol-board; plaster- paris is preferable) equal in width to V, which reflects light to the prism, P, from a source the intensity, and distance from C' of which, is known. P is a 90° prism silvered on the two surfaces which reflect the images of V and C into the eye at H. The distances from V to P and from C to P are equal, 15.5 cm. The distance from P to EF is 20 em. The total distance from 2 to V approx- imately equals the distance of the color from the eye of the monkey when he reaches the partition GP in Fig. 2 and makes his choice as evidenced by his going to the right or to the left of the partition (the monkey B. often stopped at this point and turned his head first to the right, then to the left, ete., before finally making his choice). The photometric determinations were made in a dark-room under conditions as nearly as possible like those under which the animal reacts. The eye was dark-adapted (15 to 30 minutes). A com- fortable position was taken with the eye at H so that a clear reflected image of V appeared. An assistant then lighted a standard electric light which was screened from the observer’s eye by the opaque screen, OS, and mounted it upon the board, B, which was graduated in em. The distance of this light was varied until the observer at HL judged the two lights to be equal. The judgments made under “They are not really sources in the technical sense, but surfaces. Watson, Color Vision in Monkeys. 15 these conditions are introspectively similar to those made with an ordinary photometer. The images of V and C appear side by side, with no dividing surface between. Six judgments for each change in the intensity of the band were made, three ascending and three descending, and the results averaged. The units in which the results below are given are ‘“hefner- > In other words, the apparatus measures the “intensity meters.’ of illumination” of the variable surface, V, in terms of C, the ‘“‘in- tensity of illumination” of which can be ealeulated from the for- mula: ee r (@= 45° in all cases; r is read directly from the scale of B). The following table of constants gives the wave lengths of the four bands used in the present work and their “intensity of illumina- tion”’ under the various conditions. TABLE OF CONSTANTS. é | Standard White) Designation of Variable Light. Light. Average Distance from C. Hefner-Meters. 1. Max. Sun. Yellow LGSesip: 49.8 + 2.8 cm. 56.33 2. Min. ae os Hefner 294418 ‘ 8.18 ‘ Max. ss Blue GRE: 84.6 + 3.4 “ 19.51 4. Min. te ne Hefner 46.8 + 2.5 “ BES 5. Max. Arc Yellow 5) ©; 10: 810+ 4.0 “ 6.65 6. - Blue 5) © 19). 128.0 + 2.8 “ 2.665 Ue a ch Green ONCan: 86.0 + 5.2 “ 5.90 8 ue ad Red 5 ¢. p. 72.0 + 4.0 “ 8.42 9 Whi, Green | Hefner _— 130.0 + 3.6 “ .418 10. Min. Bt Red ot 69.0 + 2.8 ‘ 1.485 WiptH or MonocHromatic Banps.'® Red = A 6485 — 5790. Yellow = 2» 5750 — 5600. Green = A 5250 — 4825. Blue = A 4800 — 4650. *’The angle 6 refers to the angle at which the light from the source falls upon the cardboard C. 5 “These are broad spectral bands, mutually exclusive, but not ‘“monochro- matic’ in the sense in which the physicists use that term. \ 16 Fournal of Comparative Neurology and Psychology. Merruop or PRESENTING THE STIMULI. The animals were carried from their living room to the dark- room, and allowed to remain there for two or three minutes. I 3-27 6 0 100 ee | he os 3-28 6 | 0 100 i | ot Ss 3-29 6 0 100 A | ss ow 3-30 6 0 100 oe a | 3-31 5 1 83 e ee aes 4-1 5 0 100 ee Se emir 4-2 5 1 83 Scam ameacaT XN peact 4-5 | 5 1 Soe oie ae es hoe ts 4-6 6 0) 100 | an | a | ve Ts 5 1 | g3 | “ | “ | “ | i B—With Stimulus Variable. 4-8 0 100 Are | Min. Max. | 4-9 9 3 75 7 Misr a | 4-10 12 &} 80 | = Min. oe 4-11 8 1 89 | ae oe oo 4-12 14 2 Si 5) G ni 4-13 12 4 TED Wilt ete ss ‘« | Animal very hungry. 4-14 9 i 90 oe | Max. Min. 4-16 8 4 66.6 ee aN eabe ‘* | Fed too much. 4-17 9 1 90 Be ewe sone 4-19 5 1 83 gob ili ue en -~ Change in intensity made in middle 4-19 6 0 1000) | Min. | Max.) | of series. 4-20 14 0 100 | S K* % 4-22 9 1 90 ee | ~ se 4-23 7 2 TEL sane eV siexcs Min. 4-26 13 0 100 he E Max. 4-27 12 1 92 Max. | Min. | Only lower half of red exposed. *E designates subjective quality. The above tests were continued from April 27th to May 28th. An average of from 85-90 per cent of correct choices was maintained throughout the whole period. In these last tests, all variations in the presentation of the stimuli which could be thought of were introduced, such as presenting the red on the right and left alternately, red twice on right, once on left, then three times on right and three on left, ete. On account of the position error erftering into B’s reactions, he was fully one month behind J. Since it was desirable to keep J in practice, he was put through all the control tests with B. 24 Fournal of Comparative Neurology and Psychology. TABLE II. H’s Reactions TO ReD-GREEN A—With Stimulus Constant. Dat Red. | Green. |Per Cent-| source Intensit Remarks Elite : * | Correct. ye | aS | | Red. | Green. | 3-18 0 Ca as) Are | Max. | Max. | 3-19 1 3 25 Samet | eo 3 3-20 9: 4 33.3 oe ss tae 3-21 0 6 0) ie a os 3-22 4 2 66.6 4 oe 3-23 1 5 16.6 oe ts i 3-24 4 2 (Sd af ue 3-25 3 3 50 oe a3 Hy 3-26 4 2 66.6 4 a9 oa 3-27 4 2 (Aa) || = a 3-28 3 3 50 eee us Si 3-29 5 1 83 Uy of a 3-30 3 3 50 Ss oy Of 3-31 6 0 100 ey 7. sss Disturbance in general physical con- dition of animal for several days. | Position error became noticeable when tests were again started with animal, 5 days later, animal going | always to left. This error persisted until May 4. In the interim 6-10 trials per day were given. ‘The in- troduction of a partition broke up error. 5- 4 8 el 89 a i Hi 5- 5 8 1 89 i a a : B—With Stimulus Variable. 5- 6 6 0 100 Are Max. Min. 5- 8 8 2 80 ipa ia PES eal | Waa ie 5- 9 8 2 80 er He | i 5-10 8 2 80 23 ie ae 5-11 10 1 90 a a “y 5-12 8 2: 80 e e - 5-13 10 2, 83 ve eS st 5-14 9 1 90 ue es fie 5-15 9 1 90 a3 = oa 5-17 13 1 93 mie so me 5-18 6 1 86 oe ey | ‘* )| Change in intensity made in middle of 5-18 7 0 100 a EK. | Max.) series. 5-19 11 (0) 100 ag 2 S 5-20 u 5 58.3 we re os Animal too hungry; all 5 errors made in succession. 5-21 11 1 91 ie si os 5-22 14 2 87 6 Min e 5-23 9 1 90 ee es tS 5-24 9 1 90 “oe “ce oe 5-25 16 0 100 - Max Min. | 5-26 | 15 2 88 oa ue a Only lower half of red exposed. RO |) 1 4 75) a 13% 5 | Red 4 as wide as green. 5-28 | 12 0 100 a Min. Perl eT as ie | Watson, Color Vision in Monkeys. 25 TABLE III. B’s Reactions TO RED-GREEN. B was given test for test with J (see J’s record) with similar averages up to March 25th. Notice- able position error began to appear which grew steadily worse. This error, as in H’s case, was finally eliminated by the introduction of the glass partition. The error persisted for fifty-one days. Ten to twelve trials were given per day during this entire period. A—With Stimulus Constant. Per Cent. | Date. | Red. | Green. Gorrect! | Source Intensity. Remarks. | [fd SS ee ee ee ee ee ere SS) ee — — | | Red | Green. 5-15—5-20 Held 90 Are Max. Max. | This average will serve as_ basis for ¥ ay'age : | comparison with those in Ba B—With Stimulus Variable. — ~ ‘se = = . ——es — = Pore ies 93 Are | E. | Max. 5-22 14 2; 87 re ie | Ss 5-23 13 1 93 a Min oe 5-25 15 1 94 | mn Max Min. 5-26 15 2 88 SS ce of Only lower half of red exposed. 5-27 12 2 85 a er te Half vertical strip of red shown. Ani- | | mal frightened. 26 ‘fournal of Comparative Neurology and Psychology. TABLE IV. J’s REAcTIONS TO BLUE-YELLOW. A—With Stimulus Constant. Date. | Blue. | Yellow Fete eeut Source Intensity. Remarks. | | ] Blue. | Yellow 7-14 4 aa i) 90) | Are Max. | Max. 7-15 6 4 | 60 | My a i 7-16 8 3 Uae) = me Se eile) 9 4 Olea ees a 00 7-18 12 5 71 ys! os aS | 7-19 10 HW ioG ae iG sf 7-20 17 8 68 | o vy | “ 7-21 12 1 92 fs Me | i 7-22 12 2 86 Ko ee ff 7-23 | 13 2 8&6 a a } 7-24 11 0 100 a Me ss 7-26 16 6 as oes Te) Diam eae Disturbing noise. 7-27 18 5 78 Sun- os fae | light. 7-28 14 1 93 | Are 2 s 7-29 15 Ae | Oma \eSun= es me | light. 7-30 | 17 4) gi: | Are “s is 7-31 14 3 82 | Sun- a “ | | light. 8= 1 15 3 83 op re es 3- 4 13 1 | 93 | oe “ “oe 8- 5 13 3 81 | a se He eT Sig) 1 OS eal ot m2 % 8- 7 18 2 90 } a rs ne B—With Stimulus Variable. | 8-8 12 1 | 92 Sun- | Max. | Min. light 8-10 18 2 90 He HD oe 8-12 17 6 74 Are a Max. S213) || rots 1 | 94 | Sun- vs as light 8-15 | 8 4 GHG | = } ce ss Monkey growing very careless. I | | | | pulled him back vigorously so as to | | | punish when errors were made, hop- | | ing thereby to obtain a more careful choice. 8-16 Min. blue was thrown in for the first time. Monkey was entirely confused. After jerking him back vigorously for several trials, he began to go always over to left, out of range of beam; then, after remaining still for a moment, he would suddenly thrust out his paw to open left-hand box regardless of color exposed there. I tried in many ways to overcome this position error, even to the extent of allowing him to react as in the beginning, to max. yellow and blue with no changes, but the error was not overcome in the time at my disposal. Watson, Color Vision in Monkeys. 27 TABLE V. H’s Reacrions TO BLuE- YELLOW. A—With Stimulus Constant. Date. Blue. Yellow ieee: Source Intensity. Remarks. = 2 | acre ESS Se Th Pate | Blue. | Yellow a ; 2 60 Are Max. | Max. 7T- 5 44 oe ae oe 7-16 4 8 Bey eB) oe oe | +e 7-17 10 10 50 a = x 7-18 8 3 73 is . | > 7-19 4 4 | 50 | ae oe oe 720 1. 5 | 58 | oe oe oe (ot 8 6 57 oe oe ae as 2 | 11 | 5 69 “ec sé 6 T=93~ | 11 6 65 “e “ec “6 7-24 | 12 33 80 ag | wy ~ 7-26 | 12 3 80 a a | 4 7-27 | 18 5 | 78 | Sun |“ “s | | ight. | =2Ruie 29 6 | 60 Are us “s 7-29 | 15 4 | 79 | sue 2 = | light 7-30 | 14 2 Sigor ||) eAne ‘ a 7-31 | 14 4 78 Sur - “a First 3 choices wrong. | light 8- 1 | iP? if 92 1 ae ae 8- 2 | 17 2 90 s a Me 2 = sit ali 5 69 es ss * Made very angry by being jerked back aes | ae on errors. i 2 | 86 lgeres “ “ cE Gea 3 85 ‘: as “oe 8— Tf 10 | 0 100 oe ae ae | | qreelie tt ems | E B—With, Stimulus Variable. OS | oak 0 | 100 Sun- | Max. | Min. | light. |. | 8-10 | 14 OF} 100 hj See or locas 8-12 | 10 10) 100 }) Are: 7 Be Max. 8-13 15 0 100 Sun | - f light. | 8-14 | 10 1 | 90 | 1g “ on 8-16 | 21 5* si | ee PAN? alle 1 ee A 8-16 | 8 Oo | 100 ee Max. | Min. || Changes in intensity made in midst of | | | | eries | | series. S16) = 8) 4 89 | aes Mier: eye I Te | 2 86 ie Pe ae Changes in intensity made in midst of | 5 8-17 | 13 2 Son The Min. aaifleesericssmm ; 8-18 8 0 100 mene Vlas: Me Half vertical strip of blue shown. 8-18 8 0 100 ae come seers Lower half of blue shown. 8-19 10 0 100 ze ~ | o 8-20 15 0 100 | De Te ei kL ya Pa fete Surface value of colors altered by past- ing tissue paper over glass. *T wo of the five errors were made when the sun was overcast. 28 Fournal of Comparative Neurology and Psychology. TABLE VI. B’s REAcTIONS TO BLUE-YELLOW. A—With Stimulus Constant. : Per Cent Date. | Blue. ‘Yellow Garredt: Source _—_— Intensity. Remarks. | | | Blue. | Yellow 7-14 | 2 3 40 Are Max. | Max. 7-15 6 | 4 60 4 os * 7-16 | U | 3 70 se as He (17 8 4 66.6 “ so a TAS) || #IG | 8 2 89 se et 7-19 PAR} 0 100 ~ rie on 7-20 18 | 2 90 Se ~ io (EON UE Tl 93 fe z os (22 2 | 1 92 w Ke ae (POR) I) le Mec us ot a 7-24 ils} | 0 100 on oe Se 7-26 | 12) | 2 87 se we . 1-27 18 2 90 Sun- oy eee light. 7-28 17 5 Ue Are ae : Change to fainter spectrum seemed to | be noticed. 7-29 15) 2 88 Sun- ce we light. 7-30 14 Ons eLOO Are ne - 7-31 14 0 100 Sun- oS af light. Gai 9 0 | 100 zs of 26 8- 4 13 1 93 7 on Sa mole ih Wy 0 | 100 “ “ i B—With Stimulus Variable. = ts) 4) — ila! 0 100 Sun- | Max. | Min. light. | 8-10 16 1 94 sy os oe 8-12 | 13 0 100 | Are s | Max. 8-13 15 (0) 100 | Sun- ay o | 1 ght 8-15 10 OPP 00 | oS ae tg 8-15 13 Sy i, fehl es Min Oe By mistake on this day the episcotister, set in the tests on the other animals usually with a 30° opening, was closed to 10°. The animal dashed to the yellow on his first two trials, then | became steady and made only one | more error in the series. Filmy clouds were passing over the sun and at times the blue was barely over my own threshold. 8-16 17 1 94 os ae ea) 8-16 8 0 | 100 > Max ““ || Changes in intensity made in midst of series. 8-16 5 0 100 | Min ni f 8-16 | 8 0 100" |) | Max: 4) “Min:) 8-17 17 ) |) ilofo [tates Min. | Max. 8-17 | 17 OF 100 | - Max | Min. | | Changes in intensity made in midst of series. 8-17 14 (0) 100 re ne Max. 8-18 8 0 100 | es a : Half vertical strip of yellow shown. 8-18 8 0) 100 ey ee 5S el ie Half vertical strip of blue shown. S=(gn 5 10 0 100 a) ws a of = Oe OF too Ne ee ae st Surface value altered by pasting tissue paper over surface of ground glass. THE Introduction—Purposes of this study—Acknowledgements Description of sounds made by the blond ring-dove, and accompanying EXPRESSIONS OF EMOTION IN THE PIGEONS. I. THE BLOND RING-DOVE (Turtur risorius). BY WALLACE CRAIG. From the Department of Zoology of the University of Chicago. WitH ONE PLATE. CONTENTS. TINOV EMT SMES bese secre genic actopewars tel’ rotates ov cube caval RCE uae re ER eatin Prefatory remarks Ear RL OT CON opecten sino cane rite haste Parcrehe ays Chey ee cieVion. ala oC RAGE eT hie Ok RAST DMR Oa Taleestereweneerecrsiaw ens rove tetencuces vey seas Se toate aT bs Oca Pe eres Seo ee Spee AGL ATI TM aascersigstenel ogee Re) cua se iets ogee AR fe A) iiss ls Ni era RS oe Mere oar ad see 4. Cry intermediate between the alarm and the kah............ Sys) LEANN IREUI estsregeattrre OU s ORE R OUR ONE CU CRCRRE RRR ETCLEr Sr che EMEP Or (Ci eRe kordineamryelcale warrelscts omits eee sane eee Gill) ethesicalh=of-ExGibemienttins,. ar «suerte: tess cea coe) eee Gel) Pe hines copula OMmanO ler tee ani eicetemcwncee ein eee GSR CIT ATO CMe Stereos aes, orca tey elec orig teow ane Sn eisai cite aueven bom iatcew coo Sens an 7. Parent’s call when ready to feed the young................... 8. Cry intermediate between the kah and the coo................ GRATER COO Rr ererrteat ero he eee or meet Sl coay atin: TORTS Oi ae Bub ako a wo eee (Op) SeeRhesmerch=cooworstuherson gaara eee eee (GLeT) FN eR GO wall = COO. es toxcye iss ortole cre veveuciceaie enone, eee (GETS) Pe Rh eynes t= Calley eyes cs essen o chcla shotskerntercee ts Secor ie ikeshistorvaomwuheblondering-dovenarrs a: 2 asc cease aaeeeeee erica ASCSM TMIN OO une miewemGlew han octane ineinnrene DEevelopmentiOk WHERVOUNG Arce cas eon oer erten 1 02) a) esl OSA RES CRT Oe Chee MeN RCROISTa Cicero orc ice a oie THeEACHANe EVOL VOUCC aia ces sirens e naetevalene el meenioieie takers HitSteappearancemotuilencOOmaeracii erence IGauslevenves Or Olol lobrigoonodosodoodpoccoscocboaacece Hi Waley(el iW te uetnee Me cuhoricr eRe Co OO AatO.Oroiche cto Chae SWIMM Cre Ghenaoy oes Sooonogoscannceuavccugear . Beginning of 2 g AlvCy. Clee recy cre noe ie oe B.’ Beg g of the annual cycl L/ororb oven ao linden bol cence emIsnG casicia cora'e.o oto pio oe aero 6 Dance CYS BecinminerorwheWrooducy cles emcee eee Copulation, nest-building, and egg-laying............. IDS MUTE Rik xen ole een mre ier eomicectard pencick! do cla ab acy aoe Changinesplaceston the; mestc rs crsnsrteic) waiersreieeae ys Cz hem bRoodney.cle COnuUNIECOm demo eee en einer Incubation, brooding, and preparation for a new brood ON OO Cucina First appearance of the cries of the adult: the alarm-note and o 0 6 6 6 0 8:0 THE JOURNAL OF COMPARATIVE NEUROLOGY AND PSYCHOLOGY.—VoL, XIX, No. 1. 30 fournal of Comparative Neurology and Psychology. Ble’ The seasonalvey.cle; scombimilediierasrarieractereieis oiehss ysrelere eno eieieione ks) The lapse of brooding and loss of voice at the end of the breed- INS SEASON: 552. Eee ciene Oecactsnenso ce etelene ee lelenee hantial sae tayo tomea narelnlegeneas 7S A's Theclite: cycle; comune dee ereeereeicicee te oleeueruc ici cucicteisrsrereienel renin: 7S Thesprime of lite amdkoldlase ss ceqaryevccs euevsroaie roy yeas chiens iter 7S Summaryof the life-history cessor ae iac cero omnis tet) ters) suai Tis INTRODUCTION. This study of the behavior of pigeons was undertaken seven years ago with intention—first, to describe the various sounds produced by one type species of pigeon, and the bodily movements which in- variably accompany the utterance of the sounds; second, to compare these with the sounds and movements of all the species in Professor Whitman’s large collection of living pigeons; third, to throw light upon any problems which seemed naturally to connect with the study as it progressed. The present paper is drawn up to fulfill the first of these intentions. It is a descriptive account of the vocal and bodily expressions of emotion in one species chosen as a type. It may make this first paper more valuable if I indicate briefly the nature of the work by which it is to be followed. The following is a brief outline of the whole. 1. Description of the vocal and bodily expressions of emotion in the blond ring-dove. Life-history of this species, in so far as it concerns the use of voice and accompanying gesture (present paper). 2. Comparison of the sounds and gestures of different species; showing specific characteristics, homologies, and the possibility of voice and gesture throwing light on problems of phylogeny. 3. Inheritance. The forms of expression in pigeons are strictly he- reditary. They are not learned by imitation (copying). In hybrids the voice is intermediate, except when it is imperfectly developed. 4. Variation. Comparative study shows that the vocal utter- ances vary from group to group in a manner indicating determinate or orthogenetie variation. 5. Selection. Pigeons are subject to sexual selection of a kind more or less like that described by Hacker. But the theory of Hicker, Valentin. Der Gesang der V6gel. Jena, 1900. Hiicker’s state- ment is an improvement upon that of Groos, in his “Die Spiele der Thiere,” Jena, 1896. Craic, Expressions of Emotion in Pigeons. 2a sexual selection takes account of only a fragment of the great utility of the voice. Voice and gesture are of prime importance through all the cycles of the life-history. This is shown, but not at all fully explained, in the present paper. It is somewhat further shown in the following (No. 6). 6. Socrology. A preliminary account of the sociologie interpre- tation of pigeon behavior has already been published.? 7. Psychology. The psychologic conclusions are so numerous and so intimately connected with the details of description, that it is impracticable to summarize them in this place. My indebtedness to Professor Whitman is so evident from begin- ning to end of the paper that there is no need to speak of its details. I wish, however, to acknowledge in gratitude the two chief debts I owe to him. In the first place, Professor Whitman knows the emotions, the voices, and the gestures of the pigeons very much better than I do; he has told me a great many facts about the birds which my more limited experience has not afforded; and he has always given helpful answers to my questions as to what a bird is thinking about when it does a certain act. In the second place, more im- portant than the facts, I owe much to Professor Whitman for the influence of his spirit of research. Enthusiasm and steadiness of labor, sympathetic insight into the animal mind,. patience with details, yet a constant reference to general problems, I hope I have learned to some degree. I wish to express grateful obligations also to the University of Chicago and to the Marine Biological Laboratory at Woods Hole, especially for that freedom which allows a student to develop his own ideas. DESCRIPTION OF SOUNDS AND ACCOMPANYING MOVEMENTS. Preratory Remarks. There is among scientists a widespread impression that bird-songs are not susceptible of accurate description. But this impression is *The voices of pigeons regarded as a means of social control. The American Journal of Sociology, Vol. 14, 1908, pp. 86-100. \ 2 Fournal of Comparative Neurology and Psychology. certainly erroneous, at least so far as the utterances of pigeons are concerned. It may be granted that the qualities (timbres) and the intensities of sounds cannot be accurately determined outside of the physical laboratory; but in this respect the qualities and intensities of sounds are not very different from shades of color, feeling to the fingers, and many such vague impressions which are used in so-called accurate description. Those features of sounds which have to do with pitch and with time, on the other hand, are as susceptible of accurate description as are the forms and dimensions of visible organs. It must be remembered, too, that for the purpose of comparative study, a description need not go minutely into every detail. This study is to include a comparison of each utterance of the ring-dove with other utterances of the same bird, and with corresponding utter- ances of the opposite sex, of the young, and of different species. For all of these comparisons it suffices to have a general knowledge of each utterance as regards timbre and intensity and an accurate knowledge of each as regards pitch and time. For the study of expression, on the contrary, it is desirable that intensity and timbre, in addition to the other two sets of characteris- tics, be measured with extreme accuracy. Such a work of measure- ment, for one species of pigeon alone, would involve years of labor; indeed, such work is just beginning to be done, and its methods are just beginning to be developed, even for the human voice.* Hence we must be content for the present to describe the changes of ex- pression in the dove’s voice by means of non-quantitative musical signs and popular language; and though these means of description are broad and indefinite, they are full of meaning, and may convey a good idea of expression. There is only one point in which I have found it necessary to depart in any way from the regular musical notation. That point concerns the glide, or portamento. Pigeons’ notes very commonly glide with absolute continuity from one pitch to another. I have not been able to find any convenient musical sign which indicates such a glide, as distinct from a mere legato; hence, I have adopted the *. W. Scripture. The Elements of Experimental Phonetics. New York, 1902, pp. xvi + 627, Pl. xxvi. Craic, Expressions of Emotion in Pigeons. 33 double slur, thus, which must always be understood to mean a perfect glide, or porta- mento. 1. Srimence. Many birds, especially among the Oscines, are uttering some sound continually, being silent only when they are asleep. For example, the various species of American blackbird (as, Quiscalus, Agelaius) repeat their “chuck” so frequently, both while flying and while perch- ing, that the presence of a flock is always made known to the ear at a considerable distance. The Fringillide, similarly, are ever repeat- ing a short “chip” or “chirrup.” But the ring-dove has no such incessantly repeated note. ‘The dove’s notes are voiced only when prompted by some form of excitement. When engaged in any non- social occupation, such as eating, drinking, preening its feathers, or merely resting, the ring-dove is silent. And when on the wing, even in the midst of excitement, the blond ring-dove never utters any sound, except on rare occasions (only one occasion within my experience) an apparently involuntary grunt. Some other forms of birds even prostitute their most useful notes to purposes of play. The blue jay (Cyanocitta cristata), for example, often gives alarming cries when no danger is near, and seems to enjoy, so far as the limits of avian intelligence will allow, the consternation which it can thus produce among its feathered neighbors. But the pigeons, perhaps on account of their lower grade of intelligence, are incapable of carrying play to such a point; they never use the alarm-note except when really alarmed. Certain of the dove’s calls are given at times in a manner which might be styled half-serious, half-playful; but of the utter- ances of the adult ring-dove, there is only one (the song, p. 47) which ever appears to be given and enjoyed purely for, its own sake. 34. ‘fournal of Comparative Neurology and Psychology. De) aac: In the case of any object threatening or frightening a ring-doyve, the bird being, for one reason or another, disinclined to turn tail and flee, it exhibits attitudes and movements of terror and anger which we may call, for the sake of brevity, the expression of fear. The reasons why the bird may be disinclined to turn tail are numerous: it may be young and unable to fly; sick or wounded and hence unable to effect. a speedy retreat; it may be defending its nest or its mate; or it may be simply quarreling with a neighbor on equal terms. In all such cases the dove shows the expression of fear, which is now to be deseribed ; these cases are to be distinguished from those in which the dove uses its energies merely to escape and fly away, for then it shows a very different expression which we call alarm and which will be described later. The expression of fear in the ring-dove is not at all peculiar to the species; it is essentially the same as the expression of fear in all birds, and very similar to the expression of fear in reptiles and mam- mals. It consists chiefly in the erection of appendages—bristling the feathers, spreading the tail, lifting the wings—and in the emission of threatening sounds. It should be noticed that in the expression of this emotion all the feathers are raised to the utmost degree; in sudden fright the tail also is widely spread. The wing nearest to the feared object is raised and is used to strike with, dealing blows of great power and of such swiftness that, if a man allows his hand to be struck at, the hand feels the blow before the eye can see it. In some cases the near wing alone is raised, but in many cases the two wings are raised symmetrically. The head is drawn in close to the body, but is always turned toward the object of fear, ready to deal blows with the beak. The eye assumes a ferocious glare, utterly different from its ordinary mild look. This great change in the eye is caused largely by the following conditions. The eyelids are drawn back so as to open the eye to the widest limit; such wide-openness gives a staring appearance to the eye of any creature, and gives to the eye of the dove an especial glare, by exposing the maximum of the fiery red iris. The black pupil of the eye also remains large, not contracting to a pin-point as in the case of some of the other emo- Craic, Expressions of Emotion in Pigeons. 35 tions. Since the feathers of the head are slightly raised, they swell out around the eye and give somewhat the effect of a frown. Whether these changes are sufficient to account for the total change in ex- pression of the eye, I do not know. The important point, in any ease, is, that the change in expression of this one important feature is just as marked as the change in appearance of the bird as a whole. The sounds emitted under the influence of fear are various. Those ordinarily given in these circumstances are a hiss and a snapping of the bill. Both of these, however, are so feeble that they appear to be but impotent relics of a once powerful snap and hiss. They are so feeble, indeed, that an observer recognizes them by seeing a pufting movement and seeing the bill close, rather than by hearing either the hiss or the snap. Kah on alighting on perch. Pitch of the “s” not definitely determined. The ‘a’ is a hard chest-tone, impure. Each note begins with the sibilant sound but drops suddenly into the lower pitch. As the bird grows older the sibilant is reduced more and more, but many weeks elapse before it has entirely disappeared. I have observed a slight trace of it in the alarm-note, for example, at the age of seventeen weeks. The change of voice is due, no doubt, merely to the development of the vocal organs, just as it is in the adolescent man. This puts it on a different plane from the other developmental changes in ex- pression. The inception of the fear reaction, of the alarm-note, of the kah, or of the coo, and the changes in the form of the coo, must be due to the coming into play of fresh tracts and centers in the nervous system. But the deepening of the voice must be due to 62 Ffournal of Comparative Neurology and Psychology. ‘ changes in the syrinx. Observation of the birds leads me to believe that they have no control whatever over the breaking of the voice; it is purely mechanical. First appearance of the coo.—The first attempts at cooing usually appear much later than the alarm and the kah. Only in one ease, in a bird which showed other signs of precocity, did the song originate on the same day as the kah, the twenty-seventh day. In the nest- mate of this bird the coo was not heard until the fortieth day; in another bird not until the forty-seventh day, and not decisively until the fiftieth day. The first coo is thus very variable in the time of its appearance. And it is equally variable in its character. The variable character of the early cooing is shown in this quotation from my notes. The young bird “takes few hasty steps toward mother on perch, head directed toward her, giving kah in squeaky voice. He repeats this about three times, then stands up straight and stiff (in attitude of male in the up phase of the bowing-coo), then he bows, down and up several times, making not a sound. He goes through much the same performance two or three times. Little later he gives coo in a purely squeaky voice (pitch a”) without bowing.” In this ease, then, the perch-coo and the bowing-coo apparently ‘de- veloped at the same time; but, while the perch-coo was audible, the bowing-coo, if sounding at all, had not passed the threshold of audi- bility. In many eases, however, I think that the bowing-coo precedes the perch-coo by a day at least. This was true of the nest-mate of the bird just referred to: ‘After giving the kah in its squeaky voice, bird went through bowing motion, roughly, making just a single short note now and then, pitch 9’, sometimes two notes, with time- interval between, and second higher than first by one tone or less.” In another case the first coo was a “perfectly nondescript sound. Much resembled its attempts at kah, but notes more irregular, some of them more prolonged.” All these accounts go to show that the coo at its first appearance is not only variable but extremely imper- fect. While the alarm-note springs into being perfectly formed, as it were, and the kah, at its inception, is almost as perfect, the coo, at first, is an insignificant fragment which does not in the least suggest the sound it is ultimately to assume. It seems that in those Craic, Expressions of Emotion in Pigeons. 63 individuals in which the coo appears very late, it is correspondingly well developed when it does appear. Development of rhythm and melody.—The modulation of the alarm-note and of the kah being practically of the adult type from the very beginning, these utterances exhibit no development in rhythm or inflection. The coos are the only utterances which go through a long course of development in modulation. By the third month, the coo has been extended considerably. There is a greater number of notes, and the notes are, on an average, of longer duration. But even at this time the rhythm is so imperfectly developed, so irregular, that it often bears little resemblance to that in the adult song. Moreover, the rhythm varies so much, even in a single series of coos, that one must conclude it is largely accidental. The following are examples. NO.3e. 2) Bee eae Se SS I | | fin, @8 8 GS 6 6 TT (=i Bh Sree wl ua =o AN Psa eS aH ~~ ™ kukuku kuku uuu Q 56th day. Nest-call coo. Time: 5 crotchets per second. Attitude: nest-call. Tune very variable. In fact, no constant tune at all. NO.33 A ier ee ST _ = SEE nS ET ES GS Y OS, SS Bae ae | OS ie ee SS ee Re a Al Ve W—mikthweer WA ee eee ee) eee NU 4; (ej Ee Se AN A Ye Ee) eo ee a SS SS ka u kur ku ku ku ka iu NO. 331C: eS ka uU kur kau ku 36 82nd day. Though the coo is so formless at first, it very soon begins to show the general form of the adult coo. It comes to consist uniformly 64 ‘fournal of Comparative Neurology and Psychology. of three notes, and gradually this trisyllable comes to have exactly the accent, the tone-quality, and the melody of the adult coo. But even after attaining the trisyllabie form the early coo has three definite differences from the adult utterance, as follows: First. The different notes of the coo are separated more in the young than in the adult, often allowing a considerable rest between. This fact, together with the general character of the utterance, gives the impression that the young bird coos with difficulty and at the expense of considerable effort. Second. The rolling sound, represented by the letter r, is absent from the earlest coos, and develops rather slowly, for even when it does first appear it is a perfunctory performance. Third. The appendix to the coo, represented by the syllables ‘‘go 0,” is not given until the age of three or four months, and when it is first given it is only a monosyllable. In addition to the three features enumerated, the juvenile coo is characterized by poverty in the quality of sound and a hurriedness and lack of all beauty in the inflection. As the bird grows older, the coo becomes loud, voluminous, and mellow, and acquires a grace- ful, gliding inflection, which, without changing the general form of the melody, gives it an entirely new and improved character. The perch-coo and the bowing-coo develop at an equal rate and become practically of the adult form at the age of about seventeen weeks. But at this age the nest-call coo is still decidedly imperfect (at least in the male). All through the development of voice the nest-call lags behind the other coos. This is perhaps because the nest- eall is purely a sexual expression, whereas the other two forms of coo are used to express emotions which may be developed before sexual maturity, such as combativeness, or simple good-spirits. Influence of old birds.—Pigeons, young and old, are extremely sensitive to suggestion. The young ones often give a certain note when they hear the parents give it; this is noticed as soon as the first of the adult cries appears, i. e., the alarm-note. The more the young hear other birds, the more they eall. Thus the calling of other birds may lead the young to give a certain sound earlier than they would give it if left alone. But the young do not imitate the adults, in Craic, Expressions of Emotion in Pigeons. 65 the sense of copying them, or learning new sounds. The forms of utterance (herein the pigeons differ from many other birds) are strictly hereditary. The charge.—The charge is associated in its development with the kah, as has already been stated (page 40), and thus appears at an early age, even at the age of twenty-seven days. The charging activity in the young, as in the adult, includes chasing another bird, pecking her (or him), assuming the peculiar horizontal attitude, progressing by leaps as well as by steps, and uttering the kah (kah- of-excitement). But each of these acts is at first of a weak and gentle sort. The charging activity, like the vocal activities, passes through a prolonged and gradual development before it reaches the form seen in the adult. Development of certain other instincts—Since the first attempts at cooing appear at an age of from twenty-seven to forty-seven days, it might seem that they are too early to have any sexual significance whatever. Yet some activities connected with sex begin at an equally early age,—sitting on eggs, for example. In one instance, on the twenty-first day, a fresh egg having been laid by the mother, the young one entered the nest, observed the egg intently, and care- fully sat on it. At fifty-one days, a young one entered the nest, settled very carefully on the pair of eggs, sat for several minutes, and when the father tried to drive it off persisted for a considerable time in holding its position. ‘This sitting on the eggs is not an ac- cident, for the little fellow is very careful to have the eggs under him, and if there are two eggs he takes a great deal of pains in trying to get them both under, finally settling down upon them with that side- wise rocking movement always seen in the case of the adult. A young female showed the courting propensities of her sex, practising the art upon her father, at a very early age. It is difficult to say at just what age this began, because it is impossible to draw a sharp line between filial and amorous attentions. At fifty-six days this young female responded to the cooing of the father and some other pigeons by assuming the nest-call attitude, head down and wings shaking, and making an attempt at the nest-call coo. At sev- enty-four days she showed the typical courting behavior, for in the 66 Fournal of Comparative Neurology and Psychology. evening by lamp-light she huddled close to the father and preened his breast and neck, sometimes preening her own feathers in that spasmodic manner which is a sign of eros, and sometimes inter- rupting these proceedings to give the nest-call coo. From this day forth she did not cease to show her readiness and anxiety to mate. SUMMARY OF DEVELOPMENT. ‘Development of cries Development and decline of the voice and habits of the) The change of voice. and habits of the nestling. adult. Ist day. Voice just audible. Voice a stimulus to parent. Young and parent communicate also by touch, No fear. 3d day. Eyes begin to open. About 6th day. Slight expression of fear. 10th to 12th day. Young are left un- covered by parents all day. } Expression of fear reaches a maxi- mum. 9th to 14th day. Young first leave’ the|12th to 14th day. Ex- nest. pression of alarm. Begin to pick up food. < Alarm-note. | Begging begins to be accompanied by shaking of wings. Expression of fear begins to decline. 15th to 17th day. First night out of nest. 15th to 24th day and later. Weaning. /21ist day and later. Sit Maximum development of baby voice] on eggs. | and of begging behavior. , 27th day. The kah. 27th to 38th day. The charge. 27th to 47th day. The4 weeks. First im- coo. | purity in baby voice (?). 56th to 119th day. The Gth week. Distinet p nest-call coo. break in baby voice. i4th day. Female 3 months. Alarm and shows courting be) kah nearly as in havior. | adult. 4 to 6 months. Coos 17 i all differentiated and MR eee perfected. of the sibilant. 4 months and_ later. Begin to breed. B’. BreGinNING OF THE ANNUAL CYCLE. The age at which a ring-dove begins to breed depends upon the season, for the tendency is in all cases to begin breeding in the spring. Birds maturing in the autumn are delayed by the tendency Craic, Expressions of Emotion in Pigeons. 67 to sexual inactivity in winter; and birds maturing in the spring are accelerated, in comparison, by the tendency to begin breeding in spring. The autumn and the early winter are marked not only by inac- tivity in breeding but also by disuse of the voice; at least a disuse as compared with its copious use at other seasons. But as winter advances, long before warm weather has definitely set im, a change toward the musical life is noticeable. The voice is used more and more, and it gradually regains the volume of sound and perfection of form which characterize it in spring and summer. Whether the preliminary exercise of the voice aids at all in its devel- opment, it would be difficult to say. The fact is that the perfection of the voice and the tendency to use it arise gradually and coinci- dentally ; and it seems probable that each aids the other. Yet there are reasons for believing that practice has very little effect in de- veloping the voice of the dove. As the birds begin to coo, they naturally begin to coo to each other ; and while the whole pigtonry bombards the ear with an abundance of sounds, each pen presents to the eye an abundant spectacle of bowing and charging, wooing and fighting, love and jealousy. This may continue a long time before each bird secures a mate. But, to notice in detail the formation of a union between two birds, it is more convenient to study the case of two ring-doves isolated in cages. If a cage containing an unmated male ring-dove be suddenly brought alongside another cage containing another ring-dove, of un- known sex, the male becomes highly excited at once, and gives vent to his excitement in all possible ways. First he bows and coos with all his might, and he continues to do so for a long time. Then he charges about the cage, assuming the attitude peculiar to the charge, and frequently repeating the loud kah-of-excitement. At intervals he stops to glare at the strange bird and sometimes to peck at it through the bars, but soon he starts again to bow-and-coo and charge. After more or less of this display of aggressive impulse, he begins to show eros, by a certain spasmodic preening of the inside of the wing (a movement which invariably accompanies erotic activity), and by assuming the nest-calling attitude and sounding the nest-call. 68 Fournal of Comparative Neurology and Psychology. If left beside the stranger’s cage for some hours, the male must sometimes rest and be silent; but even the intervals of rest and silence are broken frequently by series of perch-coos. This behavior on the part of the male is useful in that it stimulates the strange bird to respond, and, in responding, to reveal its sex. If the strange bird be a male, it shows similar excitement and aggressiveness. And the two males are sure to fight if they can reach one another. But if the strange bird be a female, she acts far otherwise. She is at first very indifferent, unles she is particularly anxious to mate. And after some days, when she begins to show an interest in the male, she does not give the bowing-coo, nor charge up and down the cage, nor show other signs of pugnacity and aggressiveness. So far from tending to aggress upon the male, her conduct is rather an expression of submission to him. She shows a certain excitement; for instance when she utters the kah it is a kah expressive of gentle excitement. But she spends the greater part of her time in alluring the male by means of the nest-calling performance—the nest-calling attitude, seductive cooing, and gentle flip of the wings. She often tries to get through the bars of her cage to the male; and, failing to do so, she sometimes lies down with one side pressed against the bars. She shows eros by the usual method of preening inside the wing; she may even take the copulation position while the male is cooing and bowing to her. When the male sees the strange bird behaving in this submissive and seductive manner, he loses the intensity of his pugnacity ; though he always continues to be masterful. He spends less time now in the bowing-coo and more time in nest-calling and in trying to get to the female. If the doors are now opened and the birds allowed to come together, they become mated. The time it takes the doves to become mated varies greatly. In case of some old, experienced birds that are ready and anxious to mate, two or three days in contiguous cages may make them acquainted, and then as soon as the doors are opened and they come together, they are ready to copulate. In other cases, especially in cases of inexperienced birds, the male is so cruel to the female at first that it is not safe to leave her with him until after a CRATG, Ex pressions of Emotion in Pigeons. 69 long period—even weeks—of acquaintanceship. But once the birds have had their attention concentrated on each other and have become affectionate, the business of breeding proceeds smoothly and rapidly. C’. BrGginnInG or THE Broop CyctrE. In the preparation for a brood of young, whether it be the first brood of the season or a later brood, there is always first a period such as has been described, in which the male by means of the kah- of-excitement, the bowing-coo, charging upon the female and even pecking her severely, gains a mastery over the female that draws her attention to himself to the exclusion of all other males which may come in sight or which may be surviving in the female’s memory. The female on her part submits herself to the male and draws his attention to her. And both birds become worked up to a state of tense sexual excitement. This period is always followed by a second period in which the excitement, venting itself in copulation and in work upon the nest, becomes less violent, though perhaps not less powerful. The charge and the kah-of-excitement fall to a very low ebb, and even the bowing-coo is used much less than at first; but the perch-coo and the nest-call are in frequent requisition. Copulation is repeated a great number of times, there bemg many repetitions per day and continuance for a number of days. It is continued until near the time when the first egg is laid; and some- times even after the first egg is laid. The number of days of copula- tion seems to be ordinarily four or five; but there is at hand as I write, a pair of birds still continuing a series which they began ‘fifteen days ago. The number of copulations or attempts at copula- tion in one day, I have never determined under normal conditions. In certain abnormal experimental conditions, devised for another purpose, I counted on several different days from twelve to fifteen attempts per day. I should think that even in normal cage conditions the number of attempts might be equally great. The first day of copulation is a day of high excitement, and the divers expressions of this excitement may be divided into two classes ; namely, those that occur through a great part of the day in general, 70 = fournal of Comparative Neurology and Psychology. and those that occur immediately before each act of copulation. Throughout the greater part of the day, the male frequently gives the bowing-coo, the nest-call, or the perch-coo, the female gives the nest- eall, and both birds kah frequently and loudly. Preening of the feathers in a spasmodic manner, especially the preening of the wing on the inner side, and the preening of the head of the mate as the two birds sit side by side, are equally characteristic activities of the day. Immediately before copulation there is usually a special cooing and a special show of eros by preening inside the wing, and there is invariably the act of billing, the female putting her bill into the mouth of the male, and he disgorging a little of the contents of his crop for her to take. This is the signal, as it were, which is imme- diately followed by copulation. The search for a nesting-site and the building of a nest, which have been going on at the same time with the operations already described, are accompanied by a great deal of vocal performance, especially nest-calling. Both birds engage in the search for a nesting-site. When either bird has found a likely place, it sits there and nest-calls by the hour.. The mate, hearing this call, is drawn to the spot, and then both sit together and nest-eall, gently flip their wings, and preen each other’s heads for a long period. The construction of the nest is carried on by one bird, usually the female, sitting in the nest and building in the materials which are brought by its mate. Each time the male brings a straw the female receives it with the gentle flip of the wings and the nest-eall. When the eggs have been laid, the male and the female take turns most regularly in sitting on them. This fact gives to the daily program during incubation a complexity and definiteness which are not equalled at any other time in the dove’s life. The present point in the life-history is therefore a good place at whieh to insert a de- scription of the daily eycle. DY Dae Damay Cxcrn. The daily cycle of activities reaches its maximum complexity and its greatest definiteness at the time of incubation. At any other time it is indeed noticeable that the birds follow a daily program: Craic, Expressions of Emotion in Pigeons. aa they always wake with the sun; they begin to coo before leaving the roost; they are most active in the early morning, fighting and love- making, cooing and calling and running about; they rest, or even sleep, in the hottest part of the day; in the evening they are again active and musical; they go to roost as soon as the light has begun to fade; they frequently coo after going to roost; and, finally, while strictly diurnal in their habits, and helpless at night as are most diurnal birds, yet they may often be set cooing at night by the hghting of a lamp, by a bright moon, by the cooing of other birds, or perhaps by their own inward inclination. In order to obtain a more complete record of the daily cycle during incubation, I have more than once watched the birds continuously for half a day at a time, noting every movement and every utterance, now from before dawn until noon, and again from noon until night. Thus on July 2d I entered the room at 3.50 A. M. The birds were still in their nocturnal positions, female on the nest, and male on the perch. At 3.54, although there was not yet light enough to read by, the male cooed four times; the female in answer cooed four times ; he cooed once at the end of her song. Then he preened his feathers. Cooing and dressing the feathers always occupy the first part of the day. The coos given were mainly of the perch-eoo type, varied at intervals by series of bowing-coos; but at 5.09 A. M. the male turned on his perch so as to get his head in a corner, and gave the nest-call, continuing until he had repeated it twenty-four times. Not until 5.36, or one hour and forty minutes after I first heard him coo, did the male come down from his perch to feed. The female greeted him with gently fluttering wings, and as he flew up on the perch and down again, she again gave this sign of feminine affection. Thus matters continued, with but little interruption, until at 8.19 A. M. the male took his place on the nest. In these four and a half hours from the time of waking to that of taking his place on the nest, the male repeated his coo 487 times! He gave about 70 bowing-coos, 386 perch-coos, and 24 nest-calls. The bowing and perch-coos were given in 95 separate series, each series consisting of from one to ten coos. The female during all this time cooed only once; this once was when, in answer to the male’s song of four coos at dawn she 72. Ffournal of Comparative Neurology and Psychology. gave a similar series of four coos. After that she often fluttered her wings when the male happened near the nest, but she never cooed. The male’s 487 coos were pretty evenly distributed over the whole time. But they began at the hour of dawn with a somewhat slower rate than the average, rose to a maximum just after the bird had left his roost and breakfasted, and then declined somewhat until the time of taking the nest. The male takes the nest at 8.30 A. M. and keeps it until 4.45 P. M., when he yields it again to the female, who sits steadily until 8.30 the next morning. Of course the birds are not punctual to a single minute, but their regularity during early incubation, if nothing oceurs to disturb them, is remarkable. Towards the close of ineu- bation, and after hatching, they are much less regular. And at any time, the presence of other birds or of alarming objects is likely to throw them out of the regular order. Thus on the 2d of July which I have been describing, the female left the nest at 8.15 A. M., being alarmed by the barking of a dog, and the male entered the empty nest at 8.19, which was probably a few minutes earlier than he would otherwise have done. The most potent disturbing factor, however, is the presence of other birds, which arouses the jealousy of the male. Changing places on the nest.—When the male comes, at his due time, to relieve the sitting female, or when the female comes similarly to relieve the sitting male, there is always a little communication or ceremony. ‘There is little difference in behavior between the male and the female on this occasion. There is much variation in the ceremony, but the usual procedure is about as follows. The bird that is out, comes to the nest, giving the kah as it arrives; it Jumps on the edge of the nest-box, kahs again, flips its wings and tickles the head of its sitting mate. The sitting bird responds by fluttering its wings and showing evident satisfaction with its mate’s attention. This exchange of greeting is usually sufficient; after a few caresses, and sometimes cooings, on the part of each bird, the sitting bird gently rises and steps forward, and the other steps in behind and settles upon the eggs. It sometimes happens that the sitting bird leaves before the other comes, as in the case mentioned above when the sound of a dog’s bark caused the female to leave a little before Craic, Expressions of Emotion in Pigeons. mB her time. On the other hand, it is not uncommon for the sitting bird to be unwilling to leave, and for the bird that is due on the nest to paw the sitting bird’s back, probe with its bill all around the sitting bird, feeling for the eggs, and finally enter the nest and squeeze the former occupant until at last, slowly and deliberately, it leaves the nest. This happens especially when the eggs have just hatched, for the feeling of the young birds under the breast appar- ently is a greater attraction than the feeling of mere eggs; and so it often happens that both birds sit at once on the young, crowding each other, and each trying to cover as much as possible of the coveted nestlings. That the touch of the eggs or young and of the nest itself give pleasure and satisfaction to the sitting bird, is evident from many highly expressive acts; such as the manner in which the bird ar- ranges the eggs with its bill, touching them again and again, arrang- ing and re-arranging many times; from the complacency with which it finally settles down upon them; and from the absorbed interest it shows in arranging the straws and gently picking at anything about the nest (cf. p. 75.) When the male has taken the nest, all is quiet. The sitting bird always makes itself as inconspicuous as possible. Though this useful instinct has lapsed to some extent in the long-domesticated ring-dove, yet even the male of this house-bird rarely coos when on the nest. On this day on which the male cooed 487 times before taking the nest, he did not coo after that for three hours, and then he gave only one series of six coos. The female, after leaving the nest, goes first to breakfast at the seed-cup, after which she flies about the cage, preens her feathers, and busies herself with such small matters. Most of her activities have little interest for the present discussion, but it is worthy of note that she often alights on the edge of the nest- box, and on doing so she often sounds the kah. The female, it would seem, is always somewhat more attached to the nest than is the male. Although the female often uses the kah, she goes but little at any time; during the four hours I watched this female after her leaving the nest, she sang only once, a series of four coos. In the middle of the day, no matter at what stage of the brood 74. ‘fournal of Comparative Neurology and Psychology. cycle, the birds always sleep, or rest. The sitting male thus sleeps through the hottest hours. But the first and last hours of his brooding are spent in alert, though quiet, wakefulness. When the male is again free from the nest in the evening, he in- dulges in another period of cooing, though a much less noisy period than that of the morning. When first relieved by the female, having had a long fast, he goes at once to feed; then he usually performs an elaborate toilet; and only gradually does he rise to the evening musical performance. This performance, indeed, as has been men: tioned, is much less than that of the morning: for example, to compare with that morning’s performance of July 2d which has already been described, I made a similarly complete observation on an afternoon just four days previous, with the result that, during the whole time between leaving the nest and going to roost, 1. e. 344 hours (as against 414 hours for the morning), the male sang only 16 times (as against 95), making 65 coos (as against 487); moreover, while 70 of the morning’s repetitions were accompanied by bowing and 24 were of the suppliant type known as the nest-call, the evening per- formance was entirely of that less emotional and less elaborate type known as the perch-coo. The cooing generally reaches a maximum just before the bird goes to roost. After taking the roost, the male usually coos a number of times, but his songs become rapidly less frequent till all is silent. And this silence ensues while daylight is still much brighter than that by which the bird first begins to coo in the morning; which again makes the songster’s evening perform- ance inferior to that of the morning. C”. Tur Broop Cyciz, ContTINvED. After the laying of the eggs, pigeons in general spend their days in comparative quiet. This is not always very evident in the common ring-dove, as may be gathered from the foregoing pages; but in some of the wild species the change is sudden and almost complete. The birds go about with a haunted look, with a perpetual expression of alarm, as it were. The male sings only at sunrise and at sunset, and when singing he goes away from the nest as far as possible. Craic, Expressions of Emotion in Pigeons. 75 Even when it is necessary to sound the alarm-note while on the nest, the bird subdues its voice into something very like a whisper. The quietness of the brooding time is thus a forced quiet, an active silence, eaused by inhibition. In fact, in the tame ring-dove, which has so far lost its fear as to be at ease even during brooding, the inhibition is largely removed, and the birds are far more noisy during incuba- tion than are any of the wild species. Strong attachment of the mates to one another is shown throughout the brood cycle by tame and wild species alike. The notion that the comparative quietness of the birds during brooding is due to lack of conju mistake. gal feeling, is a Quietness and retirement form only one phase of a great alteration of disposition during brooding; another phase is a sudden defensive bravery and irascibility. The sitting bird, whether male or female, defends the nest as valiantly as a brooding hen. And even when off duty from incubation each bird is now more bold than ever in attacking and driving away enemies. The eggs hatch in 14 days; that is to say, on the 14th day after the laying of the second egg. The hatching of the eggs, the arrival] of the young, gives again a stimulus of the same sort with the first appearance of the eggs, and makes the parents again still more quiet, more jealous, and more devoted to their parental duties. It has already been shown (page 73) that the movement of the young under the breast of the parent is a stimulus to the latter. Professor Whit- man has found that when he needed a foster-parent for some valuable young pigeon, he could take a ring-dove whose eggs were not yet ready to hatch, and, by stroking her breast gently with his fingers in imitation of the movements of the young, he could induce her to commence feeding. Thus we see that the feeling of the movements of the young is a stimulus not only to the feeding impulse but at the same time to the secretion of “pigeon’s milk” in the crop. That the young are a greater attraction than are the eggs to the sitting bird, is evidenced by, the frequency with which the parents sit both together on the little birds, often crowding each other to get a larger share of the coveted nestlings. That the hatching-out of the young gives an additional stimulus to maternal jealousy, is shown by the 76 = “fFournal of Comparative Neurology and Psychology. fact that if, on the day of hatching, there are fledglings still in the cage from a former brood, the mother now ceases to tolerate the presence of those fledglings; her eye begins to glare and her feathers to bristle, and soon she attacks her fledged offspring with such fury that the owner is obliged, for humanity’s sake, to take them out of her cage. Within a few days after the hatching of the eggs, the birds begin to become irregular in their brooding hours. The young are still kept covered, to be sure, but the occasional desire of the parents to sit both at the same time, and the frequent necessity of their coming to feed their young, gradually breaks up the regularity of the sitting exchanges. Brooding ceases entirely, at least if the birds are kept indoors, in a period of 10 to 12 days. But while brooding has thus been gradually given over, the business of feeding has become rapidly more and more arduous, as a result of the rapid increase in size of the young and the enormous develop- ment of their begging powers. There may even be added a new note to the parents’ vocabulary at this time, a eall to the young to feed. But as soon as the young have reached their maximum importunity they, begin to pick some food from the ground, and the parents, tired and sore-mouthed from the feeding of youngsters almost as large as themselves, are ready to quit feeding; thus the young are gradually weaned, at an age ranging from about 15 to 25 days. The mother quits feeding before the father, for she is always more devoted to the next pair of eggs and young, while the father feeds the fledglings in the day-time and roosts beside them at night. A succeeding pair of eggs and young has already been mentioned. Preparation for such begins very early, in that the parents, while feeding young, commence another round of cooing and love-making and mating; the cycle of one brood is not finished before the cycle of the next is begun. Just how early the new eycle will be begun, depends upon the season and upon all the cireumstances. As to season: In the spring and early summer the succession of broods is more rapid than at any other time of year. As to other circum- stances: For example, the destruction of eggs or young at any stage sets the parents at once to cooing and love-making. Professor Craic, Expressions of Emotion in Pigeons. | C s Whitman, on removing the eggs from a nest, has observed the birds to begin fondling one another within half an hour afterwards. When only one egg hatches, so that the labor of feeding is only half what it usually is, the birds have more energy and come more quickly to the preparation for a new brood. The shortest interval I have observed between hatching and laying is, when only one bird is reared, 15 days; when two birds are reared, 14 days. In the normal course of events the inauguration of a new brood eyele is gradual, being a repetition, perhaps somewhat abbreviated, of the performance by which the birds first become mated. There is first a period of bowing and cooing by the male and a gradual rise of excitement in both birds; then a period of copulation, nest-calling, and nest-building, with a gradual decline in the excitement, followed by the laying of the eggs and the birds’ devotion again to incubation. Thus (even before the old brood cycle is finished) is a new brood cycle begun. It has been said that after the birds have begun a new round of mating they still foster the young of the last brood, but there is a limit to this fostering of the old fledglings; there comes a time when the parents not only refuse to feed them but cease to tolerate their presence. This desertion of the former brood happens much earlier with the female parent; so soon as the mother has taken to sitting again, she begins to acquire a hostile attitude towards her nearly grown-up-children ; so far from feeding them, she pecks at them when they try to share the seed-cup with her; and so far from brooding them, she keeps them always at a little distance from her body. Affairs generally continue in this smoldering condition for several days; but a day comes—and according to my observation it is almost invariably the day on which the new eggs hateh—when the fire of this maternal jealousy bursts forth and the mother perseeutes the fledglings with such fury that if they were not taken from the cage they would perhaps even be killed. The male, though not nearly so aggressive in this matter, has become more or less completely indiffer- cut to the old fledglings, and shows no regret at their departure. Thus ends the brood cycle. % 78 Fournal of Comparative Neurology and Psychology. B”. Tue Srasonat Crore, ‘Continue. Renewed efforts, as shown in cooing, kahing, nest-building, and a host of other activities, are necessary to initiate each brood cycle. If the birds be disinclined to effort, from any cause, such disin- clination will delay or prevent the commencement of another brood eycle. This is what happens in the molting season, beginning in the latter part of August. There may be a decrease of breeding power, especially in some pigeons, even before the molt; but most of the do- mestic ring-doves retain ample breeding powers up to the time when the molt begins to diminish their general vitality. The breeding powers lost at this time are not regained, by the wild species, until the following spring; and though the domestic ring-dove may be bred all through the autumn and winter, yet the frequency of repetitions of the brood cycle is lessened, the health of the birds may suffer, and it is evident that this extension of the breeding season is unna- tural. Coincident with the lapse of breeding propensity, in all spe- cies of pigeons, is a loss of voice, a loss especially of the more emo- tional, more musical notes. The loss of voice is not so conspicuous in the domesticated and unnatural ring-dove. The loss of song is not complete even in most of the wild species, for their coos may be heard at irregular intervals through the months of September and October at least; but the coo at this time is in some cases notably different from that of the breeding-season. And though the songs may be given thus sporadically, their sum total is exceedingly small. The comparative silence which reigns in the pigeonry is gloomy; the hushing of the birds in August is an annual surprise, a change so sudden and so great that one does not become accustomed to it. A”. Tur Lire Crcue, ContTINueED. Though ring-doves begin to breed at a very early age, even at four months, and thereafter continue to pass through the regular succes- sion of brood cycles and annual cycles, Professor Whitman has found that they do not reach their maximum breeding powers until the age of about three years. After that age, the breeding powers remain at the maximum for some years. Craic, Expressions of Emotion in Pigeons. 79 Professor Whitman has kept blond ring-doves till they were about ten years old. In one such case he knew pretty definitely that the bird’s death was due to causes other than old age. Yet he thinks that he has observed somewhat of a decline in the breeding powers in a bird about ten years old, SumMMary oF THE Lire-Hisrory. A’. Beginning of the life cycle. The voice of the young ring-dove is heard the first day, and is useful to induce the parents to commence feeding (page 55). The voice of the growing young is useful to cause the parents to give a sufficient amount of food, and to continue feeding until the young one is able to feed itself (page 57). Fear begins to be shown as soon as the young have the full use of their eyes (page 55). Alarm develops somewhat later, 12th to 14th day (page 59). The kah appears often on the 27th day (page 60). The charge appears on the 27th day or later (page 65). The coo appears from the 27th to the 47th day (page 62). It is at first very imperfect, and develops very slowly to the adult form. Development of the voice is of two sorts which may be referred to two causes; namely, development of the syrinx or vocal apparatus, and development of the nervous system (page 61). The young often give cries at the suggestion of the parents, but they do not imitate the cries of the parents (page 64). B’. Beginning of the annual cycle. The elaborate cooing and other performances of the spring season serve to proclaim the sex of each bird (there being no markings distinctive of sex), to bring the birds together in pairs, and to unite each pair by a firm bond (pages 66-69). CO’. Beginning of the brood cycle. The male and the female, by mutual stimulation and self-stimu- lation, work up a pitch of excitement sufficient to start them on the arduous, month-long labors of the brood cycle (page 69 ). 80 fournal of Comparative Neurology and Psychology. After days of copulation and nest-building, all of which are con- trolled by cooing and other ceremonies, two eggs are laid and the birds enter upon fourteen days of brooding, D. The daily cycle. The male and the female take turns very regularly in sitting on the eggs. Each time when one bird relieves the other, there is a ceremonial communication between them. C”. The brooding cycle, continued. After the eggs are laid, the birds are guardedly quiet when near the nest, but there is no diminution of conjugal affection (page 74). The hatching of the eggs, and the movements of the young under the breast, are strong stimuli to the parents (page 73). The parents may even add at this time a new call to their vocab- ularly, a call to the young to feed (pages 43, 44, 76). The parents, while still feeding the young, gradually work up, by cooing and other performances, to that pitch of excitement which is needed to start a new brood cycle (page 76). B”. The annual cycle, continued, At the end of the summer, especially when the molt begins, the birds have not sufficient energy to work up to the beginning of a new brood eyele. Thus brooding stops. A”. The life cycle, continued. Pigeons do not reach their maximum breeding powers until an age of about three years. The length of life is not definitely known. EXPLANATION OF PLATE. Fic. 1.—The alarm (page 35). Both are in the attitude of alarm; but the more extreme in attitude is the adult bird, distinguished by the black half-ring on the neck. Fic. -2.—The charge (page 42). Male. Fic. 3. The perch coo (page 47). The bird that is cooing is an adult male. The other bird is a young one. Fic. 4. —The nest-call (page 51). The bird with its head down in the nest is a male, nest-calling to the other bird, which is a female. Fics. 5 and 6.—The bowing-coo (page 48). Fig. 5 and Fig. 6 both show the same bird (an adult male) in a phase of the bowing-coo, Fig. 5 showing the up phase, and Fig. 6 the down phase. The bird is bowing-and-cooing to his own image in a large mirror which was placed close against the end of the cage for the purpose of getting these photographs. EXPRESSIONS OF EMOTION IN PIGEONS. WALLACE CRAIG, THE JOURNAL OF COMPARATIVE NEUROLOGY AND PSYCHOLOGY.—VOL. XIX, No. 1. TE REACTION TO TACTILE STIMUEEL AND. TEE DEVELOPMENT OF THE SWIMMING MOVE- MENT IN EMBRYOS OF DIEMYCTYLUS TOROSUS, ESCHSCHOLTZ. BY G. EH. COGHILL. Studies from the Neurological Laboratory of Denison University, No. XXII. WitH Six FIGURES. In 1906 I began a series of experiments upon embryos of Rana and Amblystoma with a view to determining whether there is any regularity in the earliest neuro-muscular responses to tactile stimuli in the amphibian embryo. During the season of 1907 these ex- periments were continued upon embryos of Diemyctylus torosus, Eschscholtz (Triton torosus). Although the work of the first year gave interesting results and convinced me that the field of investi- gation was a fruitful one, it was less exhaustive and critical in its methods than the later work has been, and there is no occasion to give an account of it in this connection. It will, therefore, receive no further treatment here and all the data and discussions of this paper will.relate exclusively to Diemyctylus torosus. These experiments were originally planned for correlated ana- tomical and physiological studies. As an introduction to such work upon Amphibia they form the basis for the anatomical part, since they reveal distinct phases in the development of neuro-muscu- lar response to the most primitive system of cutaneous receptors. But, apart from this significance to pure anatomy and physiology, “they are, of themselves, an interesting contribution to the science of animal behavior, for they deal with a most important phase of behavior, namely, its very beginning in the embryo. If, for in- stance, there is any such thing as a “simple reflex,” such as Sherring- THE JOURNAL OF COMPARATIVE NEUROLOGY AND PSYCHOLOGY.—VoL. XIX, No. 1. 84 Fournal of Comparative Neurology and Psychology. ton suggests,’ it must be found in the earliest reflexes of the embryo as observed in these experiments, and if it is possible to trace the development of a “simple reflex” into a form of acknowledged in- stinetive behavior, this link in the development of behavior would seem to appear in the development of the swimming movement as deseribed in the following pages. In view of this bearing of the experiments upon the subject of animal behavior certain results of the experimental part of my in- vestigations are here made known before the anatomical phase of the work has been completed. Merrnops. The embryos were removed from the egg membranes at various stages in development, ordinarily before they showed any sign of irritability to tactile stimuli. They were then placed in shallow Petri dishes, a single specimen in a dish, and tested from time to time for reactions. Usually an experiment continued until ‘the animal began to swim. The stimulus employed was a touch with the end of a rather fine human hair, mounted in such a way as to render the touch very gentle. The extreme sensitiveness of some very young embryos is remarkable. Even the touch of a fine piece of lint will at times evoke a vigorous response, as if it were a violent irritant. Without critical consideration the tactile nature of this mode of stimulation might be held in doubt. The touch of a hair such as was used in these investigations might easily cause a considerable pressure, so that there might be a question whether the responses were to a strictly tactile stimulus or to a mechanical stimulus upon the muscles or central nervous system. Indeed, in the very early phase of development, when the irritability was for some reason unusually low, some of the reactions, I believe, may have been to direct. pressure upon the muscles or central nervous system. But such instances, if they occurred at all, in these investigations, were, 1Sherrington, Charles S. “The Integrative Action of the Nervous System,” p. 8. CocuHILL, The Reaction to Tactile Stimult. 85 I believe, relatively rare. For instance, when the stimulus is applied to the under side of the head as the animal lies on its side, and the response is a movement of the head away from the side touched, it is inconceivable that this response is to a direct pressure upon the muscles effecting the movement, and it seems altogether iprobable that such a stimulus could be brought to bear upon the central nervous system directly in such a manner as to give rise to a constant form of response. Or, in case the stimulus is applied to the margin of the dorsal or ventral caudal fin and a movement ot the head only results, as regularly oceurs in certain phases of development, it is absolutely impossible for such a reaction to be given in response to pressure either upon the acting muscles or upon the central nervous system. As reactions of this sort occur here and there throughout nearly every one of my experiments, it seems to me certain that the stimulus employed was, with possibly rare exceptions, purely tactile, and that, so far as the mode of stimulation is concerned, my conclusions are valid. Ordinarily the stimulus was applied to the upper side of the specimen as it lay on its side on the bottom of the dish. Frequently, however, it was applied to the under side of the specimen from beneath, in order to determine whether contact with the dish had any influence on the mode of reaction, but it was impossible to detect any factor of this kind in the responses. Some embryos, also, were suspended in an upright position and tested for the same purpose, and with the same result. An individual record in detail was kept of each embryo from the time it was removed from the egg membranes till the end of the ex- periment. In the record of each trial, or application of the stim- ulus, the following factors were noted particularly: the region and side touched, the form of the response and the time of the trial. Tabulated schemes for rapid recording were tried in my first experi- ments of 1906, but it soon became apparent that such forms could not be adhered to, for they were necessarily based upon presump- tions of some sort and were, therefore, a hindrance rather than a help to alert observation. These methods were wholly abandoned and have no part in the records from which this paper is written, 86 Fournal of Comparative Neurology and Psychology. Reaction ro TactinEe STIMULI. A. Response to Stimulation on the Head. According to their reaction to a touch on the side of the head, - in the region innervated by n. trigeminus or n. vagus, embryos of Diemyctylus torosus may be grouped according to three types, as follows: Type I. Embryos which from the beginning and during a consid- erable period, respond regularly or almost regularly with a move- ment of the head directed away from the side touched. Type II. Embryos which for a relatively short period at first respond irregularly with movements of the head toward or away from the side touched, and then enter upon a relatively long period of response like that of Type I. Type Ill. Embryos which are at first asymmetrical in response, that is to say, they move their head in one direction only, regardless of the side touched, and then enter upon a short period of irregu- larity like the first period of Type II, and finally upon a relatively long period of response like that of Type I. Or individuals of this type may pass directly from the period of asymmetry to the regular form of Type I. The accompanying charts illustrate the behavior of typical specimens from each of these three types. The first column on the left in these charts records the serial number of the trials made, and the record of each trial is represented in the corresponding horizontal line to the right. The figures in the second column from the left record the time in hours and minutes that elapsed since the last preceding trial in each case. The diagrams in the third column from the left represent the form of reaction in the various trials. Where there is more than one diagram in a space these are to be read from left to right, and each represents a distinct phase in a series of movements. The arrow occasionally placed in these spaces indicates that a cephalo-caudal progression of the move- ment was distinctly observed. Where an “S” oceurs the specimen swam, and the following diagram in the same space indicates the composition of the swimming movement. It should be noted that these diagrams of the movements are simply free-hand representa- Cocuit1, The Reaction to Tactile Stimult. 87 tions of the reaction according to written descriptions made at the time of trial. They can not be considered as absolutely accurate in every detail, but they do represent truthfully the general order of the development of trunk movements in these animals. The curves of the charts represent the side touched and the direction of the initial movement in, the reaction relative to the side touched. The solid line records the direction of the movement of the head; divergence to the left from the vertical records a movement toward the side touched; divergence towards the right, away from the side touched ; conincidence with the vertical, undeter- mined. The broken line records the side touched; divergence to the left signifies a touch on the left side of the head; divergence to the right, a touch on the right side; a blank, no record. Obviously, where the two curves are parallel the movement recorded was to the left; where they diverge or converge the recorded movement was towards the right. The apparent incompleteness in the serial numbers of the trials in the first column of some charts is due to the fact that in these ex- periments alternate or occasional trials were being made with refer- ence to touch on the tail bud. The charts represent perfect series of trials with reference to touch on the side of the head. The charts presented here are selected from a series which, with descriptions, has been deposited with the Wistar Institute of Anat- omy and Biology, for the advantage of students who may be inter- ested in a more exhaustive report of my experiments than this paper affords. The accompanying table presents schematically some of the data upon which this classification into three types is based. It is the tabulation of the records of 36 specimens which have been selected solely upon the basis of completeness of the record and duration of the experiment. Owing to the difficulties in the manipulation of the work and unavoidable hindrances many experiments were not carried continuously through the entire period which is here under consideration, and, although contributing materially to the evidence on the problem as a whole, can not, on that account, be in- cluded in a comparative study of this kind. ournal of Comparative Neurology and Psychology. P &y y &y 88 woay ABAB osuMOdsaI UL Seties AUT JO Av SIQ}] Gorgas woay OAAQUO OTL ‘TT odd, SuteaAYsSnIIE 6 “potono} eprs ey [NSo1 JSOU OY} SBA OPRUT SBA PAODAT 9G, JUOWTIedxy—T]| “oI J ae! i SUIU/UAMMAUACCSAAASORRARSOOONK ERT : 5/585 50M Saas ola > 3 era rae " ‘[RAIOJUL SUOT BV TILA Jnq ‘potono} sprIs ay} wWory ABV osuOds -a1 ABpnse1 OUT AT}edIp AAO sossvd AaoUIUUASB oT} 91OT] A asvo B Ul ‘JIT sodA, Suynevaysnyyt ‘ogp Juewuldxy—’E ‘SIT ‘TT odd, SupVasnyt ‘Ep Juewp1edxg—s ‘91 eet SESE IEAL ler | PEEEEEeebeesteerre POUEEG EEECECEtt FEC H a ite jalia Z N > Seite erieae [enfin if fete a aT CASAS ie yy ~ tala here asymmetry case W ype III, in a fea illustratin abruptly into regularit 7, iment 3 y is preceded and the asymmetr y, ith it. iance w ar SSSeeeuee | Sata g(a atateseietal BOSE CECE Q as AIAN OAR del RAMU J RAUORA VIGUV UOC YN BGESHE GREER GREE REE GRBRGKERESERESEGRESGHRRARGRERRERGOKEGHUMLKENDREERDEMWBrWRREBGGEREECRNDREEREN cy q ay x ay is BY BH | 5] + iu OM] Ft] 0] A} 9] D/O] —/ MM] ST to] WIA] &| +] 5) fa) 9 x9] | 09) om} Co ke P 6 ~ N Ree ORR EREERREREERNRE RRO EER EECEEEREREE 'EEREE REECE RGEC ERE DEC REERBRRERROGEEEREECSEEEREOLGGEEE Fournal of Comparative Neurology and Psychology. go .—Exper vrements at v Fic. 4 in response passes by two moy I L tf mele Nea 5 NZ 4 + | |_|} st Z| Lf I { NZ HOS Viginie 9) | Slo RM ast . Uolsl9 5 = w a! HAAN SEL /0/0/40/9|0 Nao] 0/0) 8, 8/-) S185 ISPS P TIS SEIS TS) 8 PTS] as ysiay ys FMEA FEO! | SH ofa Ales n]ofolnfol af SS] SINT ala Mola] —1 SIs +: a 1d ds S a ~{ i] 2 Sls] fale SSERARREEAES Sal ala] al cs SEBERRE s awaae AUKS RIES [ t Pee T + cad asceen } a | T H 4 hott t { 1 ct 7] + ee ,, =e yet H ae AS roo => + is 7 X s 1 Into this center all impulses would seem to flow in erder to be directed in such a way upon the musculature of the trunk as to give rise to locomotion. Clearly the development of an eye or ear as such in its earliest functional condition has no part in deter- mining this region of centralization. The controlling factor in this centralization is the motor system: a cephalization in response to the prepotency of the requirements of effectors and not in response to the demands of the cephalic receptive fields. Phylogenetically, then, the most primitive cephalization of the nervous system may have occurred, also, in response to the demands for locomotion and have given rise to a center of control in the region corresponding to the lower portion of the myelencephalon or the upper portion of the medulla spinals. Quite in harmony with this suggestion is the convincing evidence that Johnston® presents for the migration caudad of the afferent roots of the cranial nerves. Such a change in their course would lead them more directly into this primitive locomotor center. Upon this hypothesis, also, the economy ““The Nervous System of Vertebrates,’ Chapter III. 102 ‘fournal of Comparative Neurology and Psychology. of the arrangement of the special cutaneous nerves of fishes and amphibians is obvious. It is not to be supposed that the cephali- zation of the locomotor effectors is, in any respect, a direct cause of the cephalo-caudal migration of the special cutaneous receptors and conductors, but such a cephalization would certainly favor the development of such systems, for, as already suggested, their peripheral conductors hold essentially the same relation to the cephalic part of the central system as do the most primitive central conductors from the trunk. It should be noted here that a certain amount of locomotion may be acquired by an amphibian embryo by other movements than the S reaction as described above. The body may be flexed, for in- stance, and straightened by a series of secondary, vibratory move- ments. Such a reaction propels the animal on its side in a circle or spiral path. Also, a rapid succession of reversed flexures, in which no S reaction can be detected, may give swimming in a zigzag, erratic course. But normal, upright swimming in a direct course is, according to my observations, attained only through the perfecting of the S reaction and its performance in series. As already suggested, this development of the swimming movement is of interest from the point of view of animal behavior. We now see that swimming, which may be regarded as instinctive in these forms, arises as the elaboration of the simplest known reflex in the embryo, the contraction of the most cephalic trunk muscles. Certain forms of the flexure, such as the U reaction and the coiled reaction, do not seem to be in the direct line of the development of the swim- ming movement, being simply intensive or tetanic forms of the or- dinary flexures. On the other hand, the other types of flexure develop in a regular order and in a remarkably constant manner into the movements of locomotion. Now none of these simple flex- ures can be regarded as having any value as trials, since the Diemyc- tylus swims perfectly upon leaving the egg membranes in the normal . course of development, and within them it can gain no practical ex- perience for swimming out of movements of any sort. Instinctive swimming, therefore, and the simplest reflex alike, are inherent in the neuro-muscular system of the embryo, and while the former de- CocuHit., The Reaction to Tactile Stimult. 103 velops in a regular order out of the latter, the movements themselves, which conform to this order, can have no selective value. The question naturally follows whether in forms which do not admit of such early experiments, such as birds, many quadrupeds and primates, the various forms of locomotion, as well as other forms of behavior, which, in a greater or less degree, appear to develop out of a series of trials, may not conform to the same law. It seems altogether pos- sible that im such cases, also, the so-called erratic movements may have only a trophic value. As such they would be essential to the perfecting of movements, but would have no directive value in the development of responses. If, moreover, this hypothesis is valid for the ontogenetic origin and development of instinctive behavior it would seem plausible, also, as a theory of phylogenetic development. Its application to phylogenesis, though, would clearly be in opposition to the idea, which is accepted by various psychologists, that instinctive behavior has somehow been reflected back into the race from the intelligent type,—or psychologically expressed, that instinet is a phylogenetic derivative of intelligence. "or the latter hypothesis, I am not aware that there is any direct, experimental proof, while we do see, in such vertebrates as Amphibia which admit of early experimentation, instinctive behavior (locomotion) developing directly out of the simplest known reflex. However, while we seem to have a definite conception of the psychic parallel of the former (instinct), the con- cept of the psychic parallel of the latter is much less definite, and largely disregarded by psychologists. Yet it would seem that in the ontogenetic developments of the psychie life of Diemyetylus there must be puite as definite a reflex-psychosis concomitant with the earli- est and simplest reflex as there is an Instinet-psychosis with the later instinctive behavior in the form, for example, of locomotion; for, although the neuroses of the simple reflex are evidently not as elaborate as are those of locomotion, they are quite as definite in form. But, however this hypothesis of the relation of the instinct to the reflex may appeal to the psychologist, an adequate knowledge of the behavior of Diemyct¥lus must take into account the origin and de- velopment of locomotion from the simple reflex; for this reflex 104 “fournal of Comparative Neurology and Psychology. represents the simplest known physiological unit of the somatic neuro-muscular system, or of the somatic “action system.” The relation of this unit to any of the more complex neuro-muscular processes is certainly an essential factor in the problem of behavior, or of physiology in the broadest sense. In presenting the mode of locomotion of the amphibian embryo it is not my intention to antagonize the current explanation of the propelling factors of the swimming movement of fishes, ordinarily described as being, in effect, the same as that of a sculling oar. The latter explanation, so far as I am aware, is offered with reference to the adult fish, and it might not apply to an embryonic or very young fish. Quite conceivably, the swimming movement might become modified during growth, in response to changes in body form, modes of feeding and other factors of behavior; and it is still quite possible that in the adult fish there is a cephalo-caudal progression of move- ment which is obscured by other factors of special adaptation. This contribution should not be submitted without reference to the splendid work of Paton’ on the reaction of vertebrate embryos. This is the only paper accessible to me that bears in any respect im- mediately upon the work in hand. Paton’s contribution, however, is chiefly upon the development of fishes, with merely a reference to Rana and Amblystoma, and is particularly devoted to the spontaneous movements. Such movements would seem to be much more common in embryos of fishes than in embryos of Diemyctylus. The latter, during the early phases of irritability to touch, may be under obser- vation for hours without making a perceptible spontaneous move- ment of the trunk, cardiac and branchial movements not being taken into account in my work. My approach to the problem of physiologico-anatomieal correla- tions in the development of the neuro-muscular system of vertebrates differs materially from that of Paton’s method. Paton undertakes “to determine in a general, but not in a specific way” how far the reactions are dependent upon “the functional activity of a nervous ™The Reaction of the Vertebrate Embryo and the Associated Changes in the Nervous System.” Mittheilungen a. d. zoologischen Station zu Neapel, Bd. 18, Heft 2 u. 3, 1907. CocuHitt, The Reaction to Tactile Stimult. 105 system” and dismisses the study of specific reactions as impracticable, on account of the “apparently conflicting” data; but my work clearly demonstrates that, in response to the stimulus employed in my experi- ments, embryos of Diemyctylus have a very definite and regular mode of response, during certain phases of development. In fact I have yet to find the first individual that, through any considerable period, reacts contrary to the mode described in this paper, that is to say, no embryo has yet come under my observation that regularly moves its head toward the side touched when the stimulation is on the head. Nor have I found a single embryo that, observed for a considerable period, has not fallen under one of the three types which I have here described. ‘ , 9 P| Wists , Dd dea. eA | G 3 Rly. 4 + ee : Nie | H =.) d aS is vt ‘ 4 : ; . a x if S 7 ' ‘ 4 f % > ; - ¥ - . SENSATIONS FOLLOWING NERVE DIVISION. BY SHEPHERD IVORY FRANZ. From the Laboratories of the Government Hospital for the Insane, Washington, D. C. With Five FIGURES. I.—Tur PRESSURE-LIKE SENSATIONS. Since the confirmation and elaboration by Goldscheider, by von Frey and by others, of the discovery of Blix that there are points or areas on the skin sensitive only to certain forms of stimulation, physiologists have assumed a form of punctate sensibility in the skin. ‘The work of these investigators has been taken to show that in the skin special nerves subserve the following sensations: heat, cold, pain and pressure (touch). On the other hand, the recent work of Head and his co-workers has not only cast considerable doubt on the validity of the broad generalization of punctate sen- sibility, but it is also plain that the earlier hypothesis is not in accord with the results obtained on man after injury or section of peripheral nerves. HTead, it will be remembered, investigated the sensibility to light touch, to different degrees of temperature, to pressure, to dual ‘stimuli, to pain, to size and to movement in patients following injury or section of peripheral nerves and he carried his inquiries up to the point of recovery for all forms of sensation. The criticism of the older hypothesis, which was made possible because of these recent pathological studies, may well be summed up in the words of Head: “When the median nerve is divided, sensation is entirely lost over a considerable part of both index and middle fingers. Over the palm, within the area said by anatomists to be supplied by this nerve, sensation is usually diminished and not completely abolished. THE JOURNAL OF COMPARATIVE NEUROLOGY AND PSYyCHOLOGY.—VOL. XIX, No. 1. 108 “‘fournal of Comparative Neurology and Psychology. In a similar manner, division of the ulnar nerve produces complete insensibility of the little finger, and of a variable part of the palm and the ulnar half of the ring finger. Such is the usual statement of surgeons and anatomists. . . . The most careful examination of the hand fails to show the shghtest diminution in sensation over the median half of the palm in consequence of division of the ulnar nerve. What has always been called the diminished sensibility produced by the division of a nerve is really a condition in which some kinds of sensibility are lost and others retained.” ? The results of careful examinations of about eighty patients, in whom the nerves of arm or leg had been divided or injured, led to the following conclusions: “The sensory mechanism in the peripheral nerves 1s found to consist of three systems: “(A) Deep sensibility, capable of answering to pressure and to the movement of parts, and even capable of producing pain under the influence of excessive pressure, or when the joint is injured. The fibers subserving this form of sensation run mainly with the motor nerves, and are not destroyed by division of all the sensory nerves to the skin. “(B) Protopathic sensibility, capable of responding to painful cutaneous stimuli, and to the extremes of heat and cold. This is the great reflex system, producing a rapid widely diffused response, unaccompanied by any definite appreciation of the locality of the spot stimulated. “(C) EHpieritic sensibility, by which we gain the power of cutane- ous localization, of the discrimination of two points, and of the finer grades of temperature, called cool and warm.” * The separate sensation elements in each of the three forms of sensibility may be tabulated as follows: “Loss of epicritic sensibility abolishes: recognition of light touch over hairless parts or parts that have been shaved ; cutaneous localiza- tion; discrimination of compass points; appreciation of difference in size, including the accurate discrimination of the head from the 1Head, Rivers and Sherren: The Afferent Nervous System from a New Aspect. Brain, 1905, Vol. 28, p. 100. With 104 dab Franz, Sensations following Nerve Division. 109 point of a pin apart from the pain of the prick (acuesthesia) ; dis- crimination of intermediate degrees of temperature, from about 25° C. to about 40° C. . “Loss of protopathic sensibility abolishes: cutaneous pain, espe- cially that produced by pricking, burning, or freezing, together with that of stimulation with the painful interrupted current; over hair- clad parts, plucking the hairs ceases to be painful; sensations of heat from temperature above 45° C.; sensations of cold from tem- peratures below 20° C. “After destruction of all cutaneous afferent fibers the part is still ‘endowed with deep sensibility, pressure can be recognized, and its gradual increases appreciated; pain is produced by excessive pres- sure (measured by the algometer); movements of muscles can be recognized ; the point of application of pressure can be localized; the patient can recognize the extent and direction of movement produced passively in all joints within the affected area.”* It will be seen, therefore, that the examination of patients in whom nerves have been injured or cut reveals many different degrees or qualities of sensation in addition to the four which, from the examination of normal individuals alone, have been supposed to be the only sensory elements. ‘The main points of difference between the old and the new view—so far as the enumeration of sensations is concerned—are as follows: There is apparently a difference between the sensations of hot and warm, and between those of cold and cool; touch is different from pressure; there are different kinds of pain. When, for example, the ulnar nerve has been cut, examination of the skin with various stimulating objects shows that over the hairless portions of the fourth and ring fingers light touches with a wisp of cotton wood or with a fine camel’s-hair brush are not felt ; the hair-clad parts may or may not react to such stimuli, depending upon the location of the lesion; parts of these fingers will not be sensitive to temperature stimuli, and perhaps not to pressure; there may, or may not, be sensations from pricks of a pin; and if the “Wead and Thompson: The Grouping of Afferent Impulses within the Spinal Cord. Brain, 1906, Vol. 29, p. 551. 110 ©“fournal of Comparative Neurology and Psychology. fingers can not be moved voluntarily there will be a loss of sensibility to movement passively produced. Some of these effects may be found over a variable extent of the palm and the back of the hand. In Fig. 1 is given the condition found in a man following an operation in which part of the ulnar nerve was excised. The area insensitive to light touch and to the intermediate degrees of temperatures in- cluded all the little finger, about three-quarters of the ring finger, and about two-fifths of the palm and back of the hand. Part of Fic. 1.—The extent of loss of sensation following the division of the ulnar nerve at the elbow. The part marked with horizontal lines was insensitive to light touches, and to intermediate degrees of temperatures. The vertical line area (cross-hatched on account of its being included within the area insensitive to touch), was, in addition, insensitive to pressures and to pain. Adapted from Head and Sherren. this area was insensitive to deep touch and no sensation was got from the vibrations of a tuning fork. This area was also analgesic. In this and other cases in which losses of sensibility were found, the ability to appreciate touch was tested with a wisp of cotton wool lightly brushed over the parts. When such a piece of cotton wool is carried over the skin of a normal individual, there is a distinct feeling of touch, which is magnified, perhaps altered, wherever the hairs are touched. The cotton wool should be very lightly grouped in a bundle, not tied, and I have found that on the hand a piece of cotton wool, weighing 55.5 mg., with a bending pressure of from 200 to 300 mg. will be accurately appreciated over the parts FRANZ, Sensations following Nerve Division. Tae which are not calloused.t On the lips and parts of the face, a wisp of cotton wool weighing 20 mg. and bending at 200 to 250 mg., was just sufficient to produce a sensation. At times it is more convenient to use a camel’s-hair brush, although Head has objected to the use of this instrument. I find it to be a more constant stimulus, in that it remains the same in bending strength for long periods. With cotton wool it is difficult, if not impossible, to select for each day’s series of experiments an amount equal to that used on previous days, and if the same piece be used on successive days, it soon loses its original strength. In the experimental results that follow, I have used both cotton wool and a camel’s-hair brush. I selected a long haired brush from which I cut off most of the outside hairs, leaving a brush 24 mm. long from the end of the hairs to the inser- tion, with about 125 hairs. As thus modified, the bending strength of the brush was 100 mg. for very slight bending, and about 200 mg. for more extensive bending. These figures are to be compared with the bending strength of the wisps of cotton wool mentioned above. I have found that the same results follow the use of the brush as the use of the cotton wool, and since, as mentioned above, it 1s more constant, it can be used for many patients as well as the same patient at different times. For further tests I have employed the touch instrument of Bloch, which is illustrated in the accompanying figure. To a piece of wood was attached a spring steel wire A which was bent at a right angle B; the long part of this wire A measured six inches. The area of cross section was about 0.1 square mm. no. 3; vol. 8, no. 11; Ref. Rev. of Newrol. and Psych., vol. PV Pok WARRINGTON, W. B., and GRIFFITH, IF. 704. On the cells of the spinal ganglia and on the relationship of their histological structure to the axonal distribution. Brain, vol. 2%, Dp. 297. Ranson, 8S. W. *°06. Retrograde degeneration in the spinal nerves. Jour. Comp. Neur. and Psych., vol. 16. ‘0S. The architectural relation of the afferent elements entering into the formation of the spinal nerves. Jowr. Comp Neuwr. and Psych., vol. 18. 150 ‘fournal of Comparative Neurology and Psychology. Fic. 1. Zeiss, Ocular 4, Objective 8—Drawing traced from photo-micro- graph of a transverse section through the control second cervical ganglion of a young white rat twenty days after the operation, showing the character- istics of the normal ganglion. In the center is a clear area representing the dorsal root fibers. Note the size of the large cells, also the large number of small cells. Fic. 2. Zeiss, Ocular 4, Objective 8.—Drawing traced from a photo-micro- graph of a transverse section through the “operated” second cervical ganglion of a young rat twenty days after the operation. ‘This section illustrates the alterations which have occurred in the ganglion as a result of division of the nerve. It can be readily seen by comparing Figs. 1 and 2 that there have occurred both an atrophy of the ganglion as a whole and a decrease in the number of the cells. It is apparent at a glance that the cells in the operated ganglion are predominately of the medium size. None are as large as the largest cells seen in the normal ganglion, due to the fact that the cells have already begun to show some atrophy. The most striking feature is the loss of the small cells. Ranson, A/terations in Spinal Ganglion Cells. (2 iva See bs % ay Mas 5, : Last Nest SN “8, SAS) a 227 FR CT weed - 6%, tA SY ate LO Pea ha 2 NSE Fy OER: Cy RE ore ee OR a PA aw LS) CPEB TQ QA OX ae ‘Spe MOND oi =Zy PES pn" S = io" * Sh oe aN = ie Ee o 23 (9 - a ‘ ‘ ote a . y . AS Sul Y o} ay/ f %, as . » s a ry 7 © ry SD i rs a > a N 19! 152 ‘fournal of Comparative Neurology and Psychology. Via. 3. Zeiss, Ocular 4, Objective 1/12.—Drawing of a small area from a transverse section through the “operated” second cervical ganglion of a young white rat five days after the operation. With the exception of one medium- sized cell, all of the cells show more or less extensive chromatolysis. The medium sized non-reacting cell is distinctively of the coarsely granular type. The usual features of chromatolysis can be seen in the reacting cells. Notice the extreme peripheral position of the nuclei of the small cells. Iic. 4. Zeiss, Ocular 4, Objective 1/12—Drawings of a small area of a section through the same ganglion as that represented in Fig. 2. It repre- sents the condition in the ganglion twenty days after the section of the nerve. The most striking feature is that the large cells have almost regained their normal appearance. The nuclei are centric and there is a large amount of chromatic granules distributed in the normal manner throughout the pro- toplasm. Mek tay the capsule, from the second cervical ganglion of a rat eight days after the operation. Fic. 6.—Showing a cell distended by a large vacuole, from the second cer- vical ganglion of a rat seven days after the operation. Showing a degenerated cell penetrated by proliferating cells from “Te Ranson, Alterations in Spinal Ganglion Cells. B53 dice, 5, Hie. 6: a 7 q PECG. ! The Journal of Comparative Neurology and Psychology Vo_umE XIX May, 1909 NuMBER 2 ON ite: REGATON OK SHE BODY ENG TE TO. EH BODY WEIGHT AND TO>-THE WEIGHT OF THE BRAIN AND OR VEE SPINAL -CORDSIN ELE ALBINO RAT (MUS NORVEGICUS Woalie wUbiclUhswe BY HENRY H. DONALDSON. Professor of Neurology at The Wistar Institute. WITH THREE FIGURES. In a recent paper (Donaldson, 08) the relations of the body weight to the weight of the brain and of the spinal cord in the albino rat have been described. In addition to the determination of the body weight it was stated in the paper just cited (pp. 346-7) that measurements had also been made on the body length (trunk and head) of some of the rats, but to avoid confusion the discussion of this character and_ its relations was reserved for the present paper. The reasons for making a series of linear measurements on the albino rat were briefly the following :— 1. To obtain a second general measure of the body growth of the albino rat in terms other than those of weight. 2. To gather data by which to determine the body weight and body length ratio for the variety measured. This ratio is valuable because it gives a notion of the general shape of the animal and also enables us to state whether there are THE JOURNAL OF COMPARATIVE NEUROLOGY AND PSYCHOLOGY.—VOL. XEN NOA 2) 156 “fournal of Comparative Neurology and Psychology. differences in this relation according to sex, as well as to make com- parisons with other forms. It also permits the determination of the influence of dwarfing and other modifying conditions on the weight-length relation. 3. Both the weight of the brain and of the spinal cord can be related to the body length, and the measurement on body length thus made to furnish an additional datum from which the weights of the brain and of the spinal cord can be inferred. As we shall see, this datum is a much better one than body weight, especially in those cases where, for one reason or another, the animal has become emaciated. 4. If we consider the body length of the rat to correspond in a general way with the sitting height in man, we have one more means of comparing the growth changes in the two forms. In the following pages we shall discuss these points, so far as they have been worked out. For the mathematical treatment of the results I am indebted to my colleague, Dr. Hatai, who is pub- lishing at this same time some notes on the formulas previously used by both of us (Hatai, *08; Donaldson, ’08), as well as giving a new and more general formula for determining the weight of the brain from the body weight (Hatai, ’09). The technique of weighing and measuring was that described in the earlier paper (Donaldson, ’08).. A number of complete rec- ords on the albino rat have been added to those on hand at that time. Moreover, for the relation of body weight to the body length alone, additional records have been obtained by weighing and meas- uring animals which had been anesthetized lightly. It was my first intention to print the full series of individual records (235 males, 173 females) in a general table at the end of this paper. I have, however, decided not to do so for the following * reasons :— First.—Printing such a general table would involve repeating a number of the records already published in a former paper (Don- aldson, ’08), and would in turn need to be again repeated in a forth- coming paper on the change in the percentage of water during the growth of the nervous system. BRAIN AND SPINAL CORD OF RAT, CHART 1. HENRY H. DONALDSON, 240 BODY LENGTH 180 160 140 120 100 BODY WEIGHT Gms. 20 40 60 80 100 120 140 - 160 180 200 220 240 260 280 300 320 To show in the abino rat, the body length in millimeters according to body weight in grams. Records for 170 males @, and 148 females X, The theoretic curve for the sexes combined is based on formula (4). Tue JournaL or Comparative NEUROLOGY AND PsycHOLOGy.—VOL. XIX, No. 2. ay 4 : . br, 7 i. ‘ yj ; \ 156 fournal of Comparative Neurology and Psychology. : } a Dona.pson, Brain and Spinal Cord of Rat. 157 Second.—The individual records have been tabulated and are on file at the Institute. They are therefore available for use by other investigators, and may be had by application to the Director of The Wistar Institute. Third.—It is hoped that this condition will be only temporary, and that when this group of investigations is completed, the entire series of individual records employed for them can be printed in the form of tables in a special brochure, thus making them generally available. At this time only the mean values of the observation are tabulated. We turn at once, therefore, to the consideration of the special questions :— 1. The body length of the albino rat according to body weight. On Chart I, so far as is possible without confusion, the individual records for body length (170 males and 148 females) are entered according to the body weight. The continuous line on the chart shows the theoretical curve. As can be seen, the distribution of the records is such as to fit a theoretical curve that rises with dimin- ishing rapidity, and so far as it can be plotted, is still bending towards the horizontal. A distinction between the sexes in the rela- tion of body length to body weight, though present, is hardly to be seen on Chart I. The mean values for the body lengths are given in Table 2. Making use of these data, the weight length ratios have been determined for the series in hand. Table 1 gives the numerical expression of the relations obtained by dividing the calculated body length (for both sexes combined, see Table 2) by the body weight. The ratios thus obtained are given in Table 1, and these show that the albino rat becomes relatively shorter as its weight in- creases. By means of a correlation table based on groups differing by 10 grams in body weight and 10 mm. in body length, the mean statures for given body weights have been calculated. This has been done for each sex separately, as well as for both sexes taken together, and the final values obtained are given in Table 2. When the means for the males are compared with those for the 158 fournal of Comparative Neurology and Psychology. females (see Chart II, based on 179 males and 160 females) it will be observed that the latter run slightly below the former. The difference, though small, has significance, as we shall show later. However, for the general discussion at this time the results are not separated according to sex, but are treated together. TABLE 1. THkr RATIOS OBTAINED BY DIVIDING THE Bopy LENGTH BY. THE WEIGHT I THE CASE or Mus Norvecicus VAR. ALBUS. Body length Body weight mm. Ratios gms. | Both sexes combined Ay (See Table 2.) 5 51.9 10.38 15 77.6 5.17 25 94.8 3.78 35 109.1 3.11 45 120.5 2.67 55 130.6 2.37 65 137.7 Py ailit 75 144.9 1.93 85 152.0 1.78 95 157.7 1.66 105 | 163.4 1.55 115 | 167 .7 1.45 125 173.5 1.38 135 UGE sv 1.31 145 180.6 1.24 155 } 184.9 TLS 165 | 189.2 1.24 175 | 192.0 1.09 185 | 194.9 1.05 195 197.8 1.01 205 200.6 seri 215 203.5 .94 225 206.3 -91 235 209.2 -89 245 210.6 85 255 | 213.5 -83 265 216.4 -81 275 217.8 -79 285 220.6 afte 295 222.1 75 305 224.9 -73 315 226.4 -71 325 227.8 -70 The theoretical curve which most closely represents the change in body length with increasing body weight, is given by the formula (4) y = 148 log (x 4- 15) — 134 where y represents the body length and x the body weight. This is a formula of the same type as those used for determining the weight of the brain and of the spinal cord in relation to the BRAIN AND SPINAL CORD OF RAT. CHART II HENRY H. DONALDSON. BODY LENGTH FEMALE BODY WEIGHT Gms. 20 40 60 80 100 120 140 160 180 200 220 240 260 280 300 320. To show in the albino rat, the mean values for the body length according to body weight, sexes separated; ————— males, ........ females. The theoretic curve is not drawn, as it would confuse the other lines. Tue JOURNAL or Comparative NeuROLOGY AND PsycHoLoGy.—VoL. XIX, No. 2. vy et i eee an Redd ance * gee jadert pee eee PO ™ The = ‘i. ye < : OT a earete pe Nhe? oes re povienestoen ee Donatpson, Brain and Spinal Cord of Rat. 159 weight of the body, and the type has been already discussed in a previous paper (Donaldson, 708, p. 350). In this connection, however, there are some points to be corrected and further discussed. The consideration of these points is taken up in a paper by Dr. Hatai which appears at this time. TABLE 2. Mean body length according to body weight in Mus norvegicus var. albus. The body length as given in the last column has been calculated by the formula (4), y=148 log (x + 15) — 134. Bopy LENGTH OBSERVED. | Body |—— = = - — = Saran Body length in Werent | Frequen- Mean Frequen- Mean Frequen- Mean by geet cies. mm. cies. mm. | cies. mm. M M F F M.+F. M.+F | 5 12 59.2 12 58.3 24 58.8 51.9 15 15 @6ie3 24 75.4 39 75.9 US 25 11 STL 8 96.3 19 97.0 94.8 35 8 108.8 8 105.0 16 106.9 109.1 45 5 121.0 ala 121.4 16 Zee, 120.5 55 Ul 130.7 13 125.0 20 127.9 130.6 65 9 134.0 5 131.0 | 14 132'-5 137 .7 75 9 141.6 5 139.0 14 140.3 144.9 85 8 152.5 4 147.5 12 150.0 152.0 95 6 156.6 10 154.0 16 535483 VST A7. 105 6 165.0 9 159.4 15 162.2 163.4 115 12 166.7 9 165.0 21 165.8 167.7 125 7 iA! 15 AsO 22 171.6 le @s3et9) 135 9 176.1 4 175.0 13 175.6 UOC Sf 145 5 181.0 4 180.0 9 180.5 180.6 155 9 184.0 6 185.0 15 184.5 184.9 165 Uf 188.0 6 183.3 13 185.7 189.2 175 4 192.5 P4 190.0 6 191.3 192.0 185 5 193.0 3 188.3 8 190.7 194.9 195 2 195.0 0 2 195.0 197.8 205 7 200.7 2 195.0 9 197.9 200.6 215 4 202.5 4 202.5 203.5 225 2 205.0 2 205.0 206.3 235 0 ) 209.2 245 7 210.0 2 210.0 210.6 255 3 215.0 3 215.0 213.5) 265 2 220.0 2 220.0 216.4 275 1 215.0 1 215.0 217.8 285 0 0 220.6 295 1 205.0 1 205.0 222.1 305 0 0) 224.9 315 0 0 226.4 325 1 225.0 i 225.0 227.8 The co-efficient of correlation between the body weight and body length, the records being grouped as stated above, is found to be .90. It is possible, therefore, to infer the body weight from the stature, and vice versa, provided the body weight is normal. At the same time it is evident that body weight is much more open to fluctuations than is the body length, and therefore the body length is the better standard. 160 ‘fournal of Comparative Neurology and Psychology. 2. The relation of the weight of the brain and of the spinal cord to the body length. We shall consider each division of the central nervous system sep- arately. (a) The relation of the weight of the brain to the body length. When the data on brain weight are plotted according to the body length, we obtain the distribution of individual entries (196 males, TABLE 3. CALCULATED BRAIN WEIGHTS AND SPINAL CorD WEIGHTS ACCORDING TO Bopy LENGTH IN Mus Norvecicus VAR. ALBUS, Data FOR BotH SEXES COMBINED. Body weight Brain weight Spinal cord weight Body length gms. | gms. gms. mm. Calculated by | Calculated by Calculated by Formula (4). Formula (8). Formula (3). 50 4.5* .204* .031* 55 6.6 | -409 047 60 Tots) | noe -059 65 9.7 . 660 077 70 Wa Se -827 088 75 13.9 .962 . 106 80 16.6 1.065 .129 90 21.8 | 1.191 .159 100 PAY 1.288 194 110 36.6 1.379 .235 120 44.7 1.442 .270 130 ais II | 1.504 305 140 67.4 | 1.561 346 150 81.6 | 1.612 381 160 102.4 | 1.675 .428 170 118.7 | 1.714 -463 180 142.0 1.760 -498 190 169.5 1.811 039 200 201.3 1.851 580 210 239.1 | 1.897 -621 220 283.6 | 1.942 656 225 324.0 | 1.977 -691 * Since the formulas do not allow of extrapolation toward the lower end of the curve, the averages of the observed values are here employed. 137 females as shown on Chart III. The difference between the two sexes is slight, and in this instance therefore the data for both sexes will be treated together. The theoretic curve which fits the means most closely has been obtained in the following manner :— For the body lengths given in Table 3, the body weights were calculated by formula (4) transposed as follows:— BRAIN AND SPINAL CORD OF RAT. Ai HENRY H. DONALDSON. Gms. 2.40 2.20 2.00 BRAIN WEIGHT 1.80 wlenee 1.60 1.40 1.20 1.00 -80 60 " ets ee : xy gt Roopa raat sre? : 40 Le . ax cea “: ox! ? me a. ee Ty 20s as arp x2 5 oe —exees—ne—2gs— 0 2o——-2- BODY LENGTH “mm. 50 60 70 80 90 100 no 120 130 140 150 160 To show in the albino rat, the brain weight and spinal cord weight according to body length. curve is based on formula (8). 170 180 190 200 210 220 (1) Upper entries, brain. Individual records, 196 males @, 137 females X, The theoretic (2) Lower entries, spinal cord, 189 males @, 187 females X. The theoretic curve is based on formula (3). THe JOURNAL OF COMPARATIVE NEUROLOGY AND PsycHoLogy.—Vou. XIX. No, 2: Dona.pson, Brain and Spinal Cord of Rat. 1601 where x represents the body weight and y the body length. On the basis of the body weights thus determined the weight of the brain can be caleulated by the revised formula (8) ey pene. =) Ie x 1.56 ) pyesd ae . x 158 : 2 pe Nleee anyone 1.424 J 1 [ age =F =] Beall wlelicluls) ah oie ldijerte! tefiet ia" serie ltaliaiancuiiae sella tel teiios (= ~iieiee.e (8) 1 + (log x) ™ 1 + (logx)” as given by Hatai, ’09, in this number of this journal, in which y represents the weight of the brain and x the body weight. The computation is simpler, however, if we use y = .554 + .569 log (x — 8.7)...(1) (Donaldson, ’08) when x > 10, and a special formula ve boo lor (s)) == .87 5 CG) lata 08) when x < 10. The results obtained from these two formulas are identical with those from formula (8), and are given in the third column of Table 3. The corresponding curve is shown by the continuous line on Chart ITI. When the means are determined by the aid of a correlation table, in which the records are arranged in groups differing by 10 mm, in body length and 0.1 gms. in brain weight, the co-efficient of correla- tion between body length and brain weight is found to be .86, which is high. (b) The relation of the weight of the spinal cord to the body length. When the individual records for the weight of the spinal cord are plotted in relation to the body length, we obtain results which are surprisingly regular. See Table 3 and Chart III (189 males, 137 females). As in the case of the determination of the brain weights, the 162 “fournal of Comparative Neurology and Psychology. body weights used were those caleulated by formula (4), and then the theoretical curve which fits these results most closely has been obtained by the use of the formula (3). (Donaldson, 708.) y = .585 log (x + 21) — 0.795 (3) where y represents the weight of the spinal cord and x the weight of the body. This curve apparently forms a straight line, though in reality it is a trifle convex towards the base line. From the correlation table based on groups differing by 10 mm. in body length and .04 gms. in spinal cord weight, we obtain a co- efficient of correlation which is .99, being almost perfect. It will be seen from the foregoing that the weight of the spinal cord can be inferred from the body length with a high degree of accuracy. In this connection an application of the foregoing data can be made at once. It was noted in a previous paper (Donaldson, 708, p. 360) that for rats of the same body weight, but of different sex, the central nervous system in the male was shghtly heavier than in the female. The question naturally arises, therefore, whether there is any somatic character with which this difference in the weight of the central nervous system according to sex can be connected. I shall endeavor to show that in the sex difference in body length we find such a character. It has been pointed out in the present paper (p. 158) that for the same body weight the males have a slightly greater body length than the females. It will be of interest, therefore, to determine whether this difference in body length is sufficient to account for the difference in the weight of the central nervous system. It is to be remembered in this connection that when males and females of like body weights are compared, the brain in the male is absolutely heavier, but the spinal cord is absolutely lighter. (Don- aldson, ’08. ) The relative difference is slightly greater in the case of the spinal cord, but the absolute mass of the brain is so much greater than that of the cord that as a final result the entire central nervous system is found to be heavier in the male. Dona.pson, Brain and Spinal Cord of Rat. 163 If we turn now to the preceding Table 2, we find the percentage difference between the body lengths for the two sexes (as deter- mined from the average of the percentage differences between the five pairs ranging from 155 to 205 gms. in body weight) to be 1.74 per cent. in favor of the male. That is, on the average, mature males of a given body weight exceed by 1.74 per cent. in body length females of a like body weight. If now we select the body length of 193 mm., which is that for the male having a body weight of 185 grams (see Table 2), and consider that this body length is 101.74 per cent. of the correspond- ing female body length, we find by calculation that the body length of the latter is 189.7 mm., thus giving an absolute difference of 3.3 mm. in favor of the male. In order to determine what difference in the weight of the central nervous system would correspond to this difference in body length, we may refer to the preceding Table 3, where the weight of the nervous system (both sexes combined) is given according to the body lengths. From this table it is possible to determine how much increase in the weight of the nervous system corresponds to an increase of 1 mm. in body length. Taking the en- tries from the body lengths of 180 to 210 mm., we obtain the fol- lowing :— Average increase in From Increase in body length the weight of the central nervous system 180-190 i Snomar 0092 191-200 1 mm. O81 201-210 1 mm. OO8T ali SINGS eRe Oreos ory See REO RR HA Sc Sn .OO87 gins. If the average difference in weight for 1 mm., as shown by the table, is .0O87 gms., 3.3 mm. would imply an absolute difference of .02871 grams. This amount is 1.20 per cent. of the weight of the nervous system for a rat 195 mm. in body length (this is the mid-value between 180 mm. and 210 mm., the limits taken in the foregoing table). In Table 6, of the previous paper, Donaldson, ’08, it appears as an average of all the groups taken in pairs, that for rats of like body weight, but different sex, the entire central 164 “fournal of Comparative Neurology and Psychology. nervous system in the male exceeds that in the female by 1.13 per cent.’ It will be seen from the foregoing that the increase in the weight of the nervous system in the female, when the body length is made equal to that of the male, is 1.20 per cent, and the anticipated difference is 1.15 per cent. It follows that the difference according to sex in specimeus of like body weight is accounted for by the differ- ence in stature, the female having the smaller central nervous system because the stature of the female is less than that of the male. When, therefore, the influence of body weight and of stature is taken into account, the weight of the entire central nervous system in the two sexes is similar. It still remains true, however, that there is a characteristic division of this total weight according to sex, whereby the male has a slightly heavier brain, but a lighter spinal cord. These results are in accord with the more recent observa- tions on the human nervous system. (Brain: Blakeman, ’05; La- picque, ’08. Spinal cord: Mies, 793; Pfister, ’03, and Donaldson, 08.) Comparison oF THE Bopy Lenetu or toe Asino Rar Wrru THE Sittinc Heicur or Man. The objection is often made that the length measurements on the lower mammals cannot be compared with the measurements of stat- ure in man because of the differences in the relation of the head to the trunk, and of the trunk to the legs. As a matter of fact, however, the body Tength (trunk + head) which we have taken in the rat involves measurements of the pelvis, vertebral column and the skull quite comparable with those made in determining the sitting height in man. The chief difference is in the case of the skull which is measured from base to vertex in man, while in the rat the measurement is along the fronto-occipital axis, and so includes the nasal bones. These latter grow a trifle more rapidly than the cranium, especially in the male (Hatai, ’07), but *The value of 8 per cent given in Donaldson, ’08, page 360, ninth line, is an error. The correct value is 1.18 per cent as given above. Donatpson, Brain and Spinal Cord of Rat. 165 the difference becomes insignificant in comparison with the other parts of the skeleton which contribute so much more to the total result. We, therefore, conclude that a comparison between the body length of the albino rat and the sitting height of man may be prop- erly made. The purpose of making such a comparison is to determine whether the rat is similar to man in the way in which this character changes with age. It is not a character which at the time needs to be studied in detail and so only very general statements are necessary. In his study on the growth of school children at Worcester, Mass., West (792) made records for the sitting height in both sexes be- tween the ages of 5 years and 21 years. The results are charted in his Fig. 1 (p. 32) and given in his Table 1 (p. 35). If we take the average values of the sitting height in man for the two sexes, first at 19 years of age and again at 5 years of age, we find the following :— Situeheieht ate19) years. vs. sas sama. oo 873 mm. Sumo mevediiate oD VEATS). 2 $s. <4 s.« sinle matey sess < 595 min. DimeneMCem lin os Si. die Fats cette bee ee 278 mm. Percentage gain, 47 per cent. For comparison it is necessary to determine the increase in body length in the albino rat during the corresponding interval. Computing from birth as the zero age, and taking the time unit for the rat on one-thirtieth of that for man (see Donaldson, ’06), we obtain the following :— _ Nineteen years of human age correspond with 220 days of rat age. Five years of human age correspond with 60 days of rat age. Table 9, in Donaldson, ’08, shows that 220 days correspond with an average body weight of 234 grams, and of 60 days, with 78 grams. The corresponding body lengths in the rat, as shown in Table 2, are for 166 “fournal of Comparative Neurology and Psychology. 2 OA» OTATIG! pone nce hcicN Noh sees aan ae Net area eee emt 209 mm. (8. SEAMS <\s 4.5 ee ee ere ee 147 mm. DitterenGes.sccc.s ee ee ee eee 62 mm. Percentage gain, 42 per cent. It appears, therefore, that while the sitting height in man increased 47 per cent during the greater portion of the active growing period, the body length in the rat increased 42 per cent during the cor- responding period. Though not exactly alike, these figures represent changes of the same order, and this is all that we desire to show at the present time. The value of this determination, so far as it can be foreseen, is to indicate that the spinal cord during growth is subject to approximately the same relative amount of passive lengthening in both man and the albino rat. ConcLUSIONS. 1. In the albino rat the ratio obtaimed by dividing the body weight by the body length diminishes as the body weight in- creases. 2. Among rats of the same body weight, the males have a slightly greater body length than the females. 3. The correlation between body weight and body length is high, being .90. 4. The correlation between body length and brain weight is high, being .86. 5. The correlation between body length and the weight of the ‘spinal cord is nearly perfect, being .99. 6. The greater weight of the central nervous system in male, as compared with female rats of like body weight, is completely explained by the greater body length of the males. This result agrees with the more recent observations on man. 7. The relative increase in the body length of the rat during active growth is similar to the increase in the sitting height of man during the corresponding period. Hence, in both forms, the Donatpson, Brain and Spinal Cord of Rat. 167 spinal cord is subject to a corresponding amount of passive length- ening. 8. The body length is a better datum than the body weight from which to infer the weight of the brain or of the spinal cord. This is especially true when there is any reason to suspect emaciation of the body. 9. A mean of the two determinations of the weight of the brain or of the spinal cord (1) from the body weight (when normal) and (2) the body length, will give better approximations than the determination based on either datum alone. BIBLIOGRAPHY. BLAKEMAN, J. 1905. A study of the biometric constants of English brain-weights, and their relationships to external physical measurements. Bio- metrika, vol. 4, pp. 124-160. Donawtpson, H. H. 1906. A comparison of the white rat with man in respect to the growth of the entire body. Boas Memorial Volume, pp. 5-6. 1908. A comparison of the albino rat with man in respect to the growth of the brain and of the spinal cord. Journ. Comp. Neurol., vol. 18, no. 4, pp. 3845-892. HatTat, S. 1907. Studies on the variation and correlation of skull measurements in both sexes of mature albino rats (Mus norvegicus var. albus). Amer. Journ. Anat., vol. 7, no. 4, pp. 428-441. 1908. Preliminary note on the size and condition of the central nervous system in albino rats experimentally stunted. Journ. Comp. Neurol., vol. 18, no. 2, pp. 151-155. 1909. Note on the formulas used for calculating the weight of the brain in the albino rats. Journ. Comp. Neurol., vol. 19, no. 2. Migs, J. 1893. Ueber das Gewicht des Ritickenmarkes. Centralbl. f. Nervenheil- kunde wu. Psychiatrie, S. 1-4. PFISTER, H., 1903. Zur Anthropologie des Riickenmarks. Neurol. Centralb., Bd. 22, S. 757 und 819. WEST, G. M. 1892. Anthropometrische Untersuchungen tiber die Schulkinder in Wor- cester, Mass., Amerika. Archiv f. Anthropologie, vol. 22, pp. 13-48. NOTE ON THE FORMULAS USED FOR CALCULATING TEE WEIGH OR TEE BRAN NS Ee ALBINO RATS. BY SHINKISHI HATAT. Associate in Neurology at The Wistar Institute. In previous papers (Hatai, ’08; Donaldson, ’08) the formulas for calculating the weight of the brain and of the spinal cord in rela- tion to the body weight were determined on the assumption that the amount of increment to the weight of these parts is proportional to the reciprocal of the body weight plus a constant or where y is the weight of the brain or spinal cord in grams and x the weight of the body in grams. Integration of (1) gives at once the value of y. Thus: ; 1 ea haliaelce yah |) can er eas (x +a) +e or in our previous notation (Donaldson, ’08) : y =A+C log (x + f) This type of logarithmic formula has been used by the present writer (708) and Donaldson (’08) and was found to be very sat- isfactory for representing the relation between the body weight and the weight of the brain or spinal cord. This type was further employed by Donaldson (’09) to repre- sent the relation between the body weight and body length and was proved by him to be satisfactory. THE JOURNAL OF COMPARATIVE NEUROLOGY AND PsycHoLoGy.—Voun. XIX, No. 2 170 © ‘fournal of Comparative Neurology and Psychology. The formula in each case was as follows: Brain weight Ory — -509Nog (a — 76.7) 120004 aes eae (3) Spinal cord weight or y= .585 log (k+21 )—0.795.......... (4) Body length or y= 143) logiGa4- 15 i oa ee (5) Although the formulas (4) and (5) are entirely free from theo- retical objections within the interval x = (5 grams, 325 grams), the formula (3), however, has two defects when we apply it to the case of x < 8.7. The first defect, which appears when x, the body weight, is less than 8.7 grams, is due to the fact that the resulting value of (x— 8.7) becomes a negative quantity and the logarithm of such a quantity is necessarily imaginary. The diffi- culty thus presented is, however, merely a theoretical one, since for the purpose of computation the following method may be em- ployed. Let us consider the two cases when x is greater than @ and when x is less than q then we have (A) dyn when x >@ dx (Gx) (B) dye eee when x (2) 9 Es 1 1 s= +3v@-¢@)|,5-pel Wetery ir (6) where W(z) and }(z) are (some) functions of z, The sum of the first n terms of S becomes obviously W (2) — 9 (4) S =¢ (2) + n f (2) sey 172 “fournal of Comparative Neurology and Psychology. When (z) < 1, the limit of 2" is zero for n = « and consequently s=w (z). On the other hand, where (z) > 1, 2° tends to « and therefore in this case S = d(z). (See Jordan “Course d’analyse,’ Tome I, p. 320.) We have shown already that the brain weights in rats in which the body weights are greater than 10 grams, can be calculated by the formula y=2569 log (x 8st Orab4 = Fo ln oho ee woe ee (3) Later we found that the brain weights in rats in which the body weight lay between 5-10 grams may be calculated by a special formula for this portion of the curve, namely: WO ORs Ga) = Sie ere a i eee ee (7) and therefore in the two formulas (3) and (7) y can be considered as the function of log x. The values calculated by the latter formula (7) agree perfectly with the ideal line which completes the brain-weight curve between 5 and 10 grams of body weight. As has been shown already, the formula (6) is perfectly general in its application when two conditions are satisfied ; namely, when z| < 1 in the other. We also found that not only are the two formulas (3) and (7) functions of log x, but that (1) is applicable to rats in which the body weights are more than 10 grams or | log x | > 1, while formula (7) is applicable to rats in which the body weight is less than 10 grams or | log x | < 1. This satisfies all the necessary conditions. Thus a combination of the two formulas (3) and (7) will enable us to calculate the brain weight for any given body weight from 5 grams to 320 grams. (Extrapolation may be used towards the upper end of the curve). The final formula is represented by the following :— > 1 in one ease and Z 1.56 -569 ~ logx (x — 8.7) — 0.316 ‘ x1.56 Vines ee IS (02 eae BACCO LE 1.424 . ° (x —8.7) 1 1 * [ ——— - = | PP (8) 1+ (logx)" 1+ (log x)" in which y represents the brain weight and x the body weight. Hata, Weight of the Brain. 173 As to the actual use of the above formula, I may add the follow- ing remarks. As was mentioned already, the series reduces to @ (log x) =.569 log (x —8.7) + 0.554 when (log x) is greater than 1; while, on the other hand, the series reduces to W (log x) =1.56 log x —.87 when (log x) is less than 1. Therefore it is only necessary to note whether we are treating rats in which the body weights are greater or less than 10 grams. If the body weight is greater than 10 grams, we can simply use ¢@ (log x) or y=.569 log (x —8.7) +0.554 and if it is less than 10 grams, the other formula w (log x) or y=1.56 log x —.87 Of course, one can determine the brain weight directly from the formula (8) after some laborious calculation; nevertheless such a procedure has no advantage over the simpler process described above. The present formula (8) is desirable simply, first, because it is free from the theoretical objections; and, second, because by it we can express the complicated relations existing between the body and brain under a single generalized form. BIBLIOGRAPHY. . DoNALDSON, H. H. 1908. A comparison of the albino rat with man in respect to the growth of the brain and of the spinal cord. J. of Comp. Neurol., vol. 18, no. 4. 1909. On the relation of the body length to the body weight and to the weight of the brain and of the spinal cord in the albino rat (Mus norvegicus var. albus). J. of Comp. Neurol., vol. 19, no. 2. Hatal, S. 1908. Preliminary note on the size and condition of the central nervous system in albino rats experimentally stunted. J. of Comp. Neurol., vol. 18, no. 2. THE NERVUS TERMINALIS (NERVE OF PINKUS) tN TEE: EROG: BY C. JUDSON HERRICK. From the Anatomical Laboratory of the University of Chicago. WitH TEN FIGURES. A ganglionated nerve connected with the forebrain and inti- mately associated with the nervus olfactorius has been described in nearly all groups of fishes. The first clear description of such a nerve is that of Pinkus (794) for Protopterus. It was termed the nervus terminalis by Locy, in 1905, and accurately described in twenty genera (27 species) of selachians, and it was mentioned by Allis (97) as occurring in Amia. Brookover (’08) has deseribed it more fully in Amia and Lepidosteus and at the meeting of the Association of American Anatomists in Baltimore, December, 1908, Brookover and Sheldon reported the presence of a similar nerve in the teleosts. Further literature on the subject is cited by the authors mentioned. Ernst de Vries (’05) described a transitory ganglion on the vomeronasal nerve of mammals and suggested that the nerve of the organon vomeronasale (Jacobson’s organ) of higher vertebrates is homologous with the nervus terminalis of fishes. Since, however, the organon vomeronasale of mammals is lined with sensory epi- thelium of the same type as the undoubted olfactory parts of the nose and gives rise to nerve fibers indistinguishable from other fila olfactoria (Read, ’08), it is probable that its innervation does not differ from that of the other parts of the olfactory organ. In this case it is difficult to see how the nerve of the organon vomero- nasale can be compared with the nervus terminalis of fishes, for the latter fibers are not known to connect with the specific cells of THE JOURNAL OF COMPARATIVE NEUROLOGY AND PSyCHOLOGY.—VobL. XIX, No. 2, 176 ‘fournal of Comparative Neurology and Psychology. the olfactory mucous membrane, they bear a ganglion on their course and centrally, in most if not in all cases, they do not connect with the olfactory bulbs but with the brain farther caudad in the vicinity of the recessus preopticus or lamina terminalis. It may, therefore, be concluded that, while the nervus terminalis occurs in fishes gen- erally, its presence has not hitherto been demonstrated in the adults of any forms above the fishes. In examining preparations of the brain of the frog prepared by the Golgi method I found an impregnation of nerve fibers which conform so closely to the central course of the nervus terminalis of selachians and dipnoans that I have no hesitation in consider- ing them homologous. In the first series of sections in which this nerve was seen its fibers were completely impregnated on both right and left sides from a position rostrad of the olfactory bulbs to their decussation in the lamina terminalis; and, since the olfactory nerves and tracts were for the most part unimpregnated, the course of the nervus terminalis could be followed with precision. These findings were subsequently verified in several series of adult and larval frogs, as follows: Transverse sections of adult Rana pipiens by the Golgi method (the series referred to above, Figs. 1 to 7). E Sagittal sections of adult Rana pipiens by the Golgi method. Transverse sections of the adult Rana pipiens by the Weigert method. In this series the process of decolorization of the sections was incompletely carried out, leaving considerable color in the back- ground, so that, though the intra-cerebral course of the nervus ter- minalis is unmedullated, the course of the tract could neverthe- less be followed with precision. Other series of Weigert sections permitted the nerve to be identified where it enters the brain, but not through the brain substance, on account of the complete decol- orization of its fibers. Transverse sections of a half-grown frog tadpole by the Golgi method, illustrating the. whole central course of the nerve and its free terminal arborizations in the lamina terminalis (Figs. 9 and 10): Horizontal sections of an old larva of Rana eatesbiana 30 mm. Herrick, Nervus Terminalis of Frog. 17 long, stained with Delafield’s hematoxylin and erythrosin (Fig. 8). In these five specimens all, or nearly all, of the intra-cerebral course of this nerve was followed on both the right and the left sides. In several other specimens, both larval and adult, smaller portions of the nerve were also seen. For all of the sections on which this work is based I am indebted to the skill of my assistant, Mr, P. S. McKibben. The findings are briefly these. In the series of transverse sections made by the Golgi method through the brain of the adult frog first referred to, at the level of the olfactory bulbs (Figs. 3 and 4), there is impregnated a com- pact fascicle of a few (probably less than 40) unmedullated nerve fibers on each side, lying between the meninges and the ventral surfaces of the olfactory bulbs. This is the nervus terminalis. When followed caudad the nerves of the right and left sides are found to be similar; but rostrad they exhibit slight differences. On the left side at the extreme rostral end of the olfactory bulb the nervus terminalis has separated from the meninges and joined one of the small fascicles of fila olfactoria on the ventral border of the nervus olfactorius (Fig. 2). None of the: fila olfactoria of this fascicle are impregnated, so that it is easy to follow the nervus terminalis separately on this side. Scattered fila olfactoria are impregnated in other parts of the nervus olfactorius and these are indistinguishable in appearance from the fibers of the nervus terminalis. Less than 1 mm. rostrad of the olfactory bulb (Fig. 1) the nervus terminalis passes from the ventral to the medial aspect of the nervus olfactorius, still embedded in its marginal layer, and here it disappears from view. At about this level the impregnated fila olfactoria also lose their stain, so that it is prob- able that the nervus terminalis continues fariiee rostrad in the nervus olfactorius, though its fibers are not farther impregnated in my preparations. On the right side the relations are essentially the same, but not so clearly demonstrable on account of the fact that impregnated fila olfactoria mingle in some places with the fibers of the nervus terminalis: and the latter is not so compact a fascicle. Its fibers 178 Fournal of Comparative Neurology and Psychology. lie somewhat more deeply embedded in the nervus olfactorius than those of the left nerve. They curve dorsad and mesad before dis- appearing, as on the other side. Passing caudad, the fila olfactoria enter the bulbus olfactorius, but the fibers of the nervus terminalis separate from them ven- trally and constitute a small compact bundle of fibers lying in the meninges ventrally of the olfactory bulbs (Figs. 3 and 4). This position is maintained until they have passed caudad of all of the formatio bulbaris (glomerular formation) of the olfactory bulbs, where they turn abruptly dorso-mesad and enter the substance of the cerebral hemisphere at its ventro-medial border. The point of entrance of these fibers varies somewhat in different specimens. It is in the adult aways farther caudad than any of the formatio bul- baris of the ventral and medial aspects of the olfactory bulbs, and in all but one of the observed cases farther caudad than the for- matio bulbaris of the bulbulus accessorius on the lateral aspect of the olfactory bulb. Having entered the brain, the nervus terminalis passes caudad (Fig. 5), turning slightly dorsad and laterad in its course, embedded in the ventral part of the hemisphere about midway between the ventral border of the lateral ventricle and the medial wall of the hemisphere. The fibers of the tractus olfactorius medialis le ven- trally and medially of it and those of the median forebrain bundle dorsally of it. Upon reaching the lamina terminalis (Fig. 6), it ascends more rapidly between the lateral forebrain bundle and the crossed portion of the medial forebrain bundle to enter the middle part of the anterior commissure complex dorsally of the pre-optic recess, where it decussates (Fig. 7). The fibers can be clearly traced across the meson in a compact bundle, but their exact place of termination has not been determined. These relations were con- firmed in every detail in the transverse Weigert series, and in every- thing except the decussation in the anterior commissure in the sagittal Golgi series. In one of my series of transverse sections through the brain of the adult R. pipiens prepared by the method of Cajal the nervus terminalis can be followed in its course through the cerebral hemi- Herrick, Nervus Terminalis of Frog. 179 sphere on one side. Neither the fila olfactoria nor the fibers of the nervus terminalis are stained, and therefore the peripheral rela- tions cannot be determined. The nervus terminalis (unstained) is seen to detach itself from the olfactory nerve under the olfactory bulb and to pass back to the lamina terminalis exactly as already described. Within the brain it is surrounded by a dense mass of deeply stained fibers belonging to the secondary olfactory and other systems, so that it can easily be followed back to its decussation as a clear yellow area surrounded by a dark field of impregnated fibers. Professor J. B. Johnston informs me that in 1905 he observed a similar nerve in Golgi sections of the adult frog brain; but since he had no control of this single observation, it was not published. In the Golgi sections of the young larva (Figs. 9 and 10) the nervus terminalis is seen to enter the lamina terminalis and there its fibers arborize, some of the free termini crossing the meson and some remaining uncrossed. Other histological preparations of the larva show that the cells in the region of these arborizations are much crowded, forming the nucleus medianus septi. It is probable that in the adult also the nerve ends in the nucleus medianus septi, either wholly crossed or partly crossed and partly direct, as in the tadpole. In the horizontal series of sections through the brain of an old larva of Rana catesbiana stained with hematoxylin and erythrosin very nearly the whole intra-cerebral course of the nervus terminalis is shown in four consecutive sections, as seen in Fig. 8. The nucleus medianus septi does not appear here. It lies immediately dorsally of the plane figured as a dense mass of cells which crosses the median plane in the lamina terminalis directly ventrally of the foramen of Monro. The relations of the nervus terminalis of the larva are essentially similar, so far as observed, to those of the adult frog, save that the nerve enters the brain relatively farther rostrad and farther laterad in the larva. Fig. 8 shows it penetrating the formatio bulbaris rostrad of the bulbulus aecessorius. It seems probable to me that the point of entrance of the nervus terminalis remains relatively 180 “fournal of Comparative Neurology and Psychology. fixed, the changed relations of the adult being due to a farther growth of the olfactory bulbs rostrad rather than to a recession of the nervus terminalis caudad. In the adult the olfactory capsules lie far rostrad of the olfactory bulbs, while in the larva they he in about the same transverse plane, the olfactory nerves passing out to them almost laterally. In none of my preparations have I been able to trace the fibers of the nervus terminalis distally more than about 1 mm. beyond the olfactory bulbs. I have not examined the peripheral relations of the olfactory nerve and nasal capsules in the adult frog. In several preparations of frog tadpoles (probably R. pipiens) I have found cells scattered along the peripheral course of the nervus olfactorius which differ from the sheath cells of the fila olfactoria. The clear- est case observed is a large tadpole taken just before the metamor- phosis, which was prepared by the method of Cajal and cut into horizontal sections. Scattered along the ventral surface of the olfac- tory nerve in the middle part of its course are more than 100 nuclei which -differ conspicuously from the sheath nuclei among which they lie, being round or broadly oval and twice as wide as the nar- rowly oblong sheath nuclei. They are seattered along the olfactory nerve from its foramen through the skull to the point where it breaks up to spread over the olfactory mucous membrane. From the similarity of these nuclei to those found by Brookover on the nervus terminalis of ganoids and teleosts I incline to regard them as belonging to ganglion cells of the nervus terminalis of the frog, though I have not been able to demonstrate their fibrous connee- tions. The exact central connection of the nervus terminalis also demands further investigation. The single impregnation of the terminal arborizations in the lamina terminalis of the young larva is not altogether conclusive and this observation must be verified and extended before much weight can be given to it. One is, however, struck by the similarity between this observation and the deserip- tions of Locy of the central relations of the nervus terminalis in the selachians. In conclusion, it seems clear that the nerve here described in Herrick, Nervus Terminalis of Frog. 181 the frog is morphologically similar to the nervus terminalis of fishes, so far as our information extends. LITERATURE CITED. ALLIS, E. P. 1897. The cranial muscles and cranial and _ first Spinal nerves in Amia calva. Journ. Morph., vol. Osmias LBRooKOVER, C. 1908. Pinkus’s nerve in Amia and Lepidosteus (Abstract.) Science, INS SS, vols 27%, no. 702 p. 913. Locy, W. A. 1905. On a newly-recognized nerve connected with the forebrain of selachians. Anat. Anz., vol. 26, nos. 2 and 3. Pinkus, F. 1894. Ueber einen noch nicht beschriebenen Hirnnerven des Protop- terus annectens. Anat. Anz., vol. 9, no. 18, pp. 562-566. READ, IFFIE, A. 1908. 60. Each olfactory nerve is broken up into numerous fasciculi, some of the larger of which are indicated by the dotted outlines. A small proportion of the fibers of the olfactory nerve (fila olfactoria) are impregnated, some of these fibers being scattered singly among the fasciculi of the nerve, others aggregated into more or less definite bundles. The fascicle marked n. terminalis on the left side is unmixed with fila olfactoria; on the right side the two fascicles so designated are probably of mixed character. vn. terminalcs Fic. 2. Through the olfactory nerves just at their entrance into the olfactory bulbs. >< 30. The most rostrally placed glomeruli occupy the dorsal part of the section. The fila olfactoria occupy the middle and ventral parts of the section and only a few of them are impregnated. The nervus terminalis is embedded in the most ventral part of each olfactory nerve and all of its fibers are impregnated. 4 Herrick, Nervus Terminalis of Frog. 183 as Ny hh. “gtinule- aN . \ Veelts al \ Salk Mi. terminalis’ Iria. 5. Through the olfactory bulbs at the level of the rostral ends of the rhinoceles. < 30. The stippled area surrounding each rhinoceele indicates the extent of the layer of granule cells, most of which are not impregnated in the preparation. A few impregnated granules are drawn on the right side. Several typical neurones of the mitral cell layer are drawn on the left side. Fibers of the tractus olfactorius pass dorsad from all parts of the mitral cell layer. The glomerular layer lies farther ventrally, while the layer of fila olfactoria occupies the extreme ventral part of each bulb. The nervus terminalis has separated from the fila olfactoria on each side and lies between the latter and the meninges. 184 “fournal of Comparative Neurology and Psychology. Smterminalis Fic. 4. Through the olfactory bulbs at the level of the rostral end of the bulbulus accessorius. > 30. The layer of granule cells is indicated by the stippled area, ventrally of which is the layer of mitral cells (unimpregnated). The dorsal half of the section is occupied by secondary olfactory cells, two of which are imperfectly impregnated. The fibers of the lateral secondary olfactory tracts are not im- pregnated. They lie chiefly along the dorso-lateral border of the section ex- ternal to the dotted line. The olfactory glomeruli are limited to the extreme ventral part of the section. The nervus terminalis lies still farther ventrally close to the meninges. Herrick, Nervus Terminalis of Frog. 185 tr olfactorius med. ied. forebrain bundle Wat. forebrain bundle -n.terminalis ventro-lateralts Fie. 5. Through the cerebral hemisphere between the olfactory bulb and the lamina terminalis. x 80. The lateral and medial basal forebrain bundles are impregnated and the _ heurones marked @ and b send their axones into these two bundles respectively. The tract marked tr. olfactorius medialis contains also other elements, par- ticularly olfactory fibers of the third order for the nucleus medianus septi. The ventral fibers of the lateral secondary olfactory tract are impregnated, but not the dorsal fibers of this tract. The nervus terminalis lies ventrally of the median forebrain bundle and laterally of the tr. olfactorius medialis. 186 ‘fournal of Comparative Neurology and Psychology. . _-tr olfactorius septi Se / vi ZF et Ge: forebrain bund le 30. The tract marked medial forebrain bundle on the right side is composed chiefly of the crossed portion of this tract (cf. fig. 7). Fibers of the uncrossed portion of this tract arise from the cells marked 0b on the left side. As in fig. 5, the tract marked tr. olfactorius medialis contains also tertiary olfactory fibers for the nucleus medianus septi. Herrick, Nervus Terminalis of Frog. 187 Vat. forebrain bundle Fic. 7. Section immediately rostral to the decussation of the nervus terminalis in the lamina teminalis. x 30. The decussation of the medial forebrain bundle (med. f. b. 6.) occupies the ventral part of the lamina terminalis. The commissura hippocampi (dorsal commissure) is approaching the lamina terminalis from the dorsal side. The other elements of the anterior commissure complex lie farther caudad. 188 fournal of Comparative Neurology and Psychology. Fic. 8. Composite drawing of horizontal sections through the brain of a tadpole of Rana catesbiana about 30 mm. long, to show the central course of the nervus terminalis. > 50. The specimen was beginning the metamorphosis when preserved, having the hind leg buds about 6 mm, long. Sections were cut in the horizontal plane 30 microns thick and stained with Delafield’s hrmatoxylin followed by erythrosin. In these sections the nuclei of the cells are clearly stained and some of the forebrain tracts. Among the latter is the nervus terminalis on both sides. Distally this nerve can be followed only a very short distance after leaving the brain, its fibers being mingled with fila olfactoria and indistin- guishable from them, both being unmedullated. Centrally the nerve can be followed back to the lamina terminalis, where it plainly decussates in the anterior Commissure. All of the details of this figure are taken from section 19 of the series, except parts of the nervus terminalis which are taken from the neighboring sections whose numbers they bear. Section 19 shows the nerve at its point of entrance into the rostral end of the hemisphere and also its decussation in the lamina terminalis. With the aid of the camera lucida I have projected upon the outline of this section the remainder of the intra-cerebral course of the nerve, which is all included within the three sections lying next ventrad (sec- tions 20, 21 and 22). Three elements of the anterior commissure complex are shown, the decussation of the nervus terminalis, the decussation of the lateral forebrain bundle and the commissure of the corpora striata. The decussation of the medial forebrain bundle lies in the plane of the section, but it is not stained in the preparation; cf. fig. 7. Herrick, Nervus Terminalis of Frog. 189 n.terminalis / if ide ral fore- if oi pl recessus preopticus optic chiasma 190 Fournal of Comparative Neurology and Psychology. trolfactoyxiu< _n.terminals \ ~\ N Fic. 9. A transverse section through the brain of a half-grown frog tadpole, taken just behind the olfactory bulbs. Golgi method. > 70. The nervus terminalis is shown immediately after its entrance into the cerebral hemisphere. A few fibers of the tractus olfactorius are impregnated ventrally of it. hucleus medianus sept ue p hervus Fic. 10. The section following immediately caudad of the one shown in fig. SE SS AOE : The section is very thick and shows almost the whole central course of the nervus terminalis and its ending by free arborizations in the nucleus medianus septi. The tractus olfactorius lateralis is also impregnated. THE NERVUS TERMINALIS IN THE CARP. BY R. H. SHELDON. From the Anatomical Laboratory of the University of Chicago. WITH SEVEN FIGURES. _ Fritsch figured in 1878 the stump of a nerve arising mesad of the olfactory from the rostral aspect of the brain of Galeus canis. This he called an “‘iiberzihliger Nerv.” Our present knowledge concern- ing it is, however, due almost entirely to the work of the last few years, during which its existence has been demonstrated in group after group among the lower vertebrates. Pinkus, ’94, 795, in Pro- topterus was the first to trace and describe the entire course of the nerve which he called simply, “ein neuer Nery.” Allis, ?97, found the nerve in Amia but added nothing concerning its structure and connections, naming it, however, the nerve of Pinkus. Locy, 799, in Acanthias deseribed a nerve, closely associated with the olfac- tory, as in the eases previously reported, but ganglionated. Pinkus -had founds cells in connection with the nerve in Protopterus, but hesitated to call them a ganglion. In 1902 Sewertzoff described in Ceratodus embryos a similar ganglionic nerve which he named the nervus preopticus owing to the fact that it appears to arise near the preoptic recess in Dipnoans. Later, Burckhardt, in 1905, found the same nerve in adult Ceratodus. Loecy in several papers, ’03, 05a, ’05b, takes up in detail its occurrence in different groups of sela- chians, its peripheral and central connections and its embryonic history. He pointed out its homology with the nerve of Pinkus in Protopterus and Amia and the nervus preopticus of Ceratodus, pro- posing for it the name of nervus terminalis. Brookover, ’08, demon- strated a ganglion for the nerve in Amia and Lepidosteus, adding also to our knowledge of its peripheral connections. THE JOURNAL OF COMPARATIVE NEUROLOGY AND PSYCHOLOGY.—VOL. XIX, No. 2. 192 “fournal of Comparative Neurology and Psychology. Until very recently, however, its presence in forms other than the selachians, ganoids and dipnoans has not been demonstrated. At the Baltimore meeting of the Association of American Anatomists, in a joint paper, Brookover and I reported its existence in teleosts, describing its ganglion, and peripheral and central connections. At the same time Herrick showed that it is also present in both the larval and adult frog. This paper takes up in detail the central course of the nerve in the carp, Cyprinus carpio. It consists entirely of a tract of un- medullated fibers which was traced by means of the following methods and material. 1. Weigert method. (a) Four transverse series of the entire olfactory crura, bulbs and nasal capsules of adult individuals about 50 em. in length. (b) Two longitudinal series through the olfactory bulbs and cap- sules of similar adults. (c) Two transverse series through the cerebral hemispheres of adults. (d) One transverse series through the entire head of a young carp about 2 em. in length. These were stained by a modification of the straight Weigert method which left the unmedullated fibers a reddish brown. This method was particularly valuable, as the tract is, for most of its extent, surrounded by medullated fibers from which it stands out quite distinctly. 2. Vom Rath method. One transverse and one longitudinal series through the olfactory erura, bulbs and capsules of an adult about 30 em. in length. This method also gave good results, as the unmedullated fibers appear hghter in color than the medullated. 3. Cajal method. Two transverse series through the cerebral hemispheres of an adult about 35 em. long. In my preparations the nervus terminalis is an orange yellow, while the medullated fibers surrounding it are nearly black. These series were of especial value in showing the decussation of the nerve in the anterior commissure. 4. Toluidin blue and thionin methods. SHELDON, Nervus Terminalis in the Carp. 193 (a) Two transverse series through the bulbs and capsules in an individual of 35 em. length. (b) One longitudinal series through the bulbs and capsules of a fish 30 em. long. (c) One transverse series through the hemispheres of an indi- vidual 40 em. in length. By these two methods one can easily demonstrate the peripheral ganglion and the nucleus in which the fibers end centrally. As noted in the paper reported before the Association of Amer- ican Anatomists, numerous scattered ganglionic cells are found on the ventro-median side of the olfactory nerve about half way be- tween the formatio bulbaris and the olfactory capsule. Cells of the same type are also found caudad to the glomerular region and rostrad to the nasal capsules, diminishing rapidly in numbers, how- ever, as one passes caudad or rostrad from the main group of cells. It was also noted that Cajal preparations show coarse fibers which ean be traced from these cells rostrad to the olfactory mucous mem- brane, where they are distributed to the epithelium with the olfac- tory nerve fibers. The tract which forms the central course of the nerve is easily distinguished about half way between the caudal and cephalic ends of the olfactory bulb on its ventro-median side. Here it is sur- rounded by the medullated fibers of the tractus olfacto-lobaris medialis as shown in Fig. 1. Rostrad of this point, however, it is soon lost, as the medullated fibers either end or seek a new posi- tion, leaving the nerve surrounded only by the unmedullated fibers of the olfactory nerve. Scattered among these fibers in this region are the cells of the peripheral ganglion. As the tract passes caudad from the bulb it continues to hold its position on the ventro-median aspect but migrates peripherally until it les next the meninges, surrounded on three sides, however, by the tractus olfacto-lobaris medialis (Fig. 2). This relation holds throughout the length of the olfactory crus (see Fig. 8). On reaching the cerebral hemi- spheres the tract turns dorso-laterad through the tractus olfacto- lobaris medialis to le, for a time, between the latter and the radix olfactoria medialis propria (Fig. 4). It holds this position for some 194 ‘fournal of Comparative Neurology and Psychology. distance lying partly enclosed by the tractus olfacto-lobaris medialis. When the anterior commissure is reached, however, it turns ab- ruptly mesad (Fig. 5) and largely decussates in the mid-line (Fig. 6) in the most rostral part of the commissure. Part of the fibers apparently do not cross but end on the same side. The exact termination of the fibers could not be demonstrated. It is certain, however, that they end close to the mid-line, without doubt in the dense nucleus of small cells shown in Fig. 7. As was stated in the earlier report, it should be noted that the connection between the central tract and the peripheral ganglion with its fibers has not been established and cannot be except by for- tunate Golgi or Cajal preparations. There can be little doubt, how- ever, that this is the nervus terminalis for the following reasons. The ganglion and peripheral distribution of the fibres are identical with the condition found by Brookover in Amia and Lepidosteus in connection with what is undoubtedly the nervus terminalis. It is also similar to that described for Protopterus by Pinkus, for Ceratodus by Sewertzoff and for selachians by Locy. The course of the central tract corresponds to that shown for the nervus ter- minalis of selachians by Loey, who worked out the central termina- tion in some detail. Still stronger support comes from the findings of Herrick in the frog. In this form the nerve takes the same course through the brain, including the decussation. Peripherally, however, the nerve leaves the brain to run in the meninges rostrad past the formatio bulbaris so there can be no question as to its char- acter. Summarizing: there is little doubt that there exists in the carp a nerve comparable morphologically to the nerve of Pinkus or the nervus terminalis of other fishes and the frog. Hutt LABoratory oF ANATOMY, The University of Chicago, January 18, 1909. SHELDON, Nervus Terminalis in the Carp. 195 LITERATURE CITED. ALLIS, EB. P. - 1897. The cranial muscles and cranial and first spinal nerves in Amia calva. Jour. Morph., vol. 12, no. 3, March, 1897, pp. 487-SOS, pls. XX-XXXVIII. BinG, ROBERT AND BURCKILARDT, RUDOLF. 1905. Das Centralnervensystem von Ceratodus forsteri. Semon, Zoolog. Forschungsreisen in Austral. wu. d. Malay. Arch., pp. 513-584, Taf. XLII, 35 Fig. in Text. BROOKOVER, CHAS. 1908. Pinkus’s nerve in Amia and Lepidosteus. Science, N. §., vol. 27, no. 702, June 12, 1908, pp. 913-914. IRITSCH, G. 1878. Untersuchungen iiber den feineren Bau des Fischgehirus. Berlin, 1878, pp. 1-94, i-xy, 13 pls. Herrick, C. JUDSON. 1909. The nervus terminalis (nerve of Pinkus) in the frog. Report at Baltimore meeting, Assoc. Amer. Anat., 1909; and Journ. Comp. Neurol. and Psych., vol. 19, no. 2. Locy, W. A. 1899. New facts regarding the development of the olfactory nerve. Anat. Anz., Bd. 16, no. 12, pp. 273-290, Fig. 1-14. 1903. A new cranial nerve in selachians. Mark Anniversary Vol, article IIT, pp. 89-55, pls. V-VI, 1908. 1905a. To assist in the definition of the area in which protopathie sensibility remained and from which the epieritic sensibility had departed, I carefully examined the hairy parts of 2H. Head, W. H. R. Rivers and J. Sherren. The Afferent Nervous System from a new Aspect. Brain, 1905, vol. 28, p. 108. 7H]. Head and J. Sherren. The Consequences of Injury to the Peripheral Nerves in Man. Brain, 1905, vol. 28, p. 241. “Head and Sherren. Op. cit., p. 242. ‘For an account of the lesions and general sensibility changes see my article in Jour. Comp. Neurol. and Psychol., 1909, vol. 19, pp. 107-124. Franz, Sensations Following Nerve Division. 21 , is ik the forearm and hand, plucking individual hairs, and also brushing the hairs with cotton wool or with the ight camel’s hair brush. The results obtained on this patient are sutticiently different from Head’s reports to warrant a rather full description of the sensibility of the hairs. Over the parts of the hand and forearm which are quite normal, immediately a hair is touched there is felt a sensation, apparently sunilar to that when the skin is hghtly touched with a blunt in- strument, such as a pencil. This sensation results, I find, from the movement of the hair, and it is emphasized when more than one hair is moved or when more than one hair is grasped with forceps. After a hair has been firmly grasped with forceps and slight traction is exerted upon it, the sensation becomes clearer, or more intensive, and when sufficient traction is exerted a distinct feeling or sensa- tion of pain supervenes. The pain appears to differ in character from that produced by extremes of pressure, as for example that produced by an algometer, for it is rather burning in character. Observations such as these were made by H. (the subject of the ex- periments to be reported here) and by me, and confirmed by repeated experiments at various sittings. On the other hand, as may be expected from the results of Head’s experiments, over parts which are not normal the hairs react in a totally different fashion. On the volar side of H’s forearm, near the bend of the elbow, I marked off an area which ineluded subareas in which the different forms of sensibility were altered. First I examined carefully the sensibility of the hairs to traction. I went -over the whole forearm and hand wherever hairs were found, and marked in red ink a line between the area which was sensitive and that which was insensitive to such stimulation. The hand and arm were then photographed and from the photograph was traced Fig. 1, which is here reproduced. The area in which traction on the hairs was not accompanied by a pressure-like or by a pain sensation is that marked with vertical lines. The upper part of the arm, beyond the elbow, I did not carefully investigate, for it seemed that some of the change to be found there might be, and probably was, due to the cutting of superficial skin nerves both at the time of the accident 218 “fournal of Comparative Neurology and Psychology. and at the subsequent operation. The extent of this area is much larger than the area of loss of protopathic sensibility as determined by the methods of Head, and the extensive character of the change led me to a more careful examination of parts of the area. Since time did not permit the careful mapping of the whole area, I selected for more careful work two areas, one on the volar side of the forearm near the bend of the elbow, the other on the dorsal-ulnar side of the arm near the wrist. Fia. 1.—Diagram of hand and forearm of subject in whom ulnar and median nerves were cut at elbow. Scars of original accident and of operation shown above elbow. Area marked with vertical lines represents area in which hairs are insensitive to plucking. The diagram was made from a tracing of a photograph, and the distortion of the hand is due to the peculiar position in which the patient holds the fingers. An illustration of the upper of these two areas as it was marked in red ink on the arm of the subject is given in Fig. 2. The line separating the area of pain-on-pulling-hairs from that of no-pain-on- pulling-hairs is approximately that between areas D and E. In general contour, however, it is more irregular and more nearly ap- proaches the shape of the lowest line on the figure. The extent of this area in which pulling the hairs did not result in a sensation of FRANZ, Sensations Following Nerve Division. 219 pain or in a pressure-like sensation was determined as accurately as possible. Once the general line was found, I went carefully over all the hairs within a centimeter of the line and mapped out the exact area in which plucking did not result in pressure-like or in pain sensations. Within the areas F and G, I found none of the hairs to be sensitive to plucking. Throughout this area and below G, the usual pressure-like and pain sensations were not produced by pulling the hairs, even when two or more hairs were pulled out with their roots. In area E the plucking of only a very few hairs on the upper Jee. Be Fic. 2.—Diagram of upper inner part of forearm. MHorizontal areas G, F, and E were insensitive to plucking of hairs. Below G pressures were not felt, likewise traction and brushing of hairs. Horizontal areas A, B, and C had all forms of sensibility intact. Areas E, F, and G were sensitive to brushing of hairs. * border of some of the squares resulted in sensations, while stimulation of some of the hairs at the lower border of area D was not accom- panied by the proper kind of sensation. The extent of the lack of proper sensations in the hairs may then be said to be in the horizontal areas K, F and G. Although the plucking of hairs in area E was not accompanied by a normal sensation of pain and the pressure-like sen- sation, such as are produced by plucking the hairs over normal parts, when several, say five or six, were pulled at one time a feeling of an indefinable (to the patient) character was obtained. The subject 220 “fournal of Comparative Neurology and Psychology. could not make any good comparison with any other known kinds of sensation, for it was different from pressure and was not like the sensations obtained by brushing or plucking the hairs over normal parts. The feeling (or sensation) could not be localized beyond the general (upper or lower) part of the arm. From numerous stimula- tions and repeated experiments the feeling or sensation was found to depend upon the movement of the skin in neighboring regions. When great care was taken not to move the skin in near-by regions, this unusual and ill-defined sensation did not result. Throughout the series of experiments the patient reported that, although he per- ceived the plucking of the hairs, it should be understood that at no time within the area of the diagram did the sensations have the same quality or character as that produced by plucking the hairs on the right arm. How much of this curious sensation difference was due to suggestion and how much to an actual change can not be determined, but it is fair to assume that the patient was ordinarily trustworthy. After locating as accurately as possible the area in which the hairs were insensitive to traction, I mapped out the area in which the hairs were sensitive to stimulation with cotton wool. When, on normal parts, the hairs were lightly brushed with cotton wool, the sensation was immediately perceived, and an accurate localization was made of the place where the stimulus was given. When the hairs are stimulated as they lie, we find that many hairs from widely separated regions may be stimulated. This is particularly so if the hairs happen to be long and overlap each other to any extent. To determine “with some precision the presence or absence of this form of sensibility, I carefully lifted the hairs overlapping any special part and stimu- lated only those which were immediately beneath those that had been lifted. In carrying on the experiment in this way it was possible to locate quite accurately the extent of this kind of hair sensibility and the error of determining the line of division between sensitive and non-sensitive parts was not more than a few millimeters. The results of this careful examination were quite unusual and wholly unexpected. T found the hairs to be sensitive to stimulation of cotton wool or of the light eamel’s hair brush in all the areas FRANZ, Sensations Following Nerve Division. 221 above the lowest line on the diagram, Fig. 2. It will be noticed, there- fore, that the fifteen square centimeters embraced in the horizontal areas E, I and G were sensitive to this form of stimulation but not sensitive to the stimulation of traction. When this large area was shaved, it was found that stimulation of the skin with cotton wool or with a camel’s hair brush was ac- companied by sensation only in the horizontal subareas A, B and C. In these same areas other forms of epicritic sensibility, e. g., the appreciation of two-ness, were also present. Ulna ee gs ee HiGae3st Fic. 3.—Back of forearm near wrist. Horizontal area A, loss of all forms of sensibility. Area B, hairs do not respond to cotton wool or to traction. Area C, hairs react to cotton wool, not to traction with pain sensation, but with only pressurelike feeling. Areas D and E, hairs react to cotton wool and to traction. Area I, brushing hairs and traction on them felt more plainly than in any other areas. Area D and radialwards, area in which epicritic sensibility retained. This unexpected result on the upper part of the arm led me to a further examination of another section of the arm ten days later. The second area was on the outer and ulnar side of the forearm beginning about 8 cm. from the fold of the wrist and extending up the arm a distance of 5 em. This part of the arm is shown in Fig. 3. Area A in the figure is the area insensitive to pressures. Area B is separated for convenience, but experimentally not sharply to be distinguished, from area A and is that part of the arm in which pressures usually, but not always, were felt. The wavy line separates 222 “‘fournal of Comparative Neurology and Psychology. the areas in which the hairs do (C) and do not (B) respond to stimulation with cotton wool. The dotted line separates the areas in which plucking the hairs produced pain and no pain. Areas D and E are areas in which the hairs reacted both to cotton wool and to traction. Area C is the area in which plucking the hairs produced a sensation similar to that of pressure or to that of brushing the hairs, but in this area no pain was felt even when the hairs were pulled out with their roots. In area B plucking of the hairs was ac- companied by no pain and the hairs did not appear to be sensitive to any form of stimulation. At times in this area, as happened with the area near the elbow, when more than three or four hairs were pulled simultaneously a sensation was obtained. This was poorly localized, but appeared to be in neighboring more normal areas. Here also it was found that the movement of the neighboring parts appeared to be the determining factor in the production of the sensation. The sensation from this form of stimulation was de- scribed as similar to that of moving the hairs with cotton wool or with the camel’s hair brush, and also partly like that of lightly touching the part with a blunt instrument. When this part of the arm was shaved it was found that cotton wool could be appreciated in areas D and E. In these areas different degrees of temperature could also be appreciated and the subject reported marked sensation differences between cool and cold, and between warm and hot stimuli.® The above facts may be summarized as follows: When the ulnar and median nerves are cut, over parts of the hand and arm it is found that the hairs are not sensitive to plucking at a time when lightly brushing the hairs on the same areas is appreciated as a stimulus. The parts in which there is insensibility to plucking the hairs are within the area that, in accordance with Head’s differential signs, may be described as possessing protopathie sensibility, but in which the epicritie sensibility has been abolished. It appears, therefore, that we have in the hairs two forms of sensibility, one for traction and the other for light pressures or °A full report of the temperature findings in this portion of the arm will be found in Section III of this article, beyond. FRANZ, Sensations Following Nerve Division. 2.23 movements. The results, I believe, cannot be explained, as von Frey has attempted to explain all of Head’s results,’ as a difference in threshold values; for in traction we deal with amounts of stimuli much greater (or at least on normal parts they appear much greater) than that of brushing the hairs with cotton wool. Such a two-fold function in the hairs is in accord with the findings of a two-fold nerve supply to the hairs,® although the results on animals have not been confirmed for the common (bodily) hairs of man. So far as I am aware, the only results to be compared with these on the hair sensibility are those on temperature sensations to be reported in the next section and the few results by Head and Sherren on temperature sensations. In one or two cases these authors, it will be remembered, found parts of the skin not sensitive to hot (above 45° C.) and to cold (below 10° C.) objects, but found the same parts when stimulated with temperatures of moderate degree to be sensitive and to give appropriate warm and cool sensations. III. Temprrature SENSATIONS. From examinations of normal individuals it appears that there are special points on the skin that react to stimuli by giving a sensa- tion of heat or cold, but that in the small areas between these temperature points no sensations of hotness or coldness can be evoked by stimuli. On the other hand, when we stimulate the skin with areas of heated or cooled objects rather than with points the sensa- tions of warmth or coolness result. The areal sensations appear to differ from those in which separate spots are stimulated in that only one sensation is obtained, and there is not an apparent mixture of heat and cold from the spots that may be stimulated in the area. For the understanding of these differences in sensation no explana- tion has been offered that meets with universal approval, but the analogy of the rods and cones in the retina has been made. It is ‘von Frey. The Distribution of Afferent Nerves in the Skin. Jour. Amer. Med. Assn., 1906, vol. 47, pp. 645-648. : °F. Tello. Terminaciones sensitivas en pelos y otros organos. T'rab. del. lab. de Invest. Biol. de la Univ. de Madrid (S. Ramon y Cajal), 1906, tomo 4, pp. 49-77. 224 ‘fournal of Comparative Neurology and Psychology. said that if the area of the skin which is stimulated has a few spots that ordinarily give an intense sensation of coolness or warmth, the areal stimulation takes this character and that the spots which normally give a weaker sensation help to fill up and to make the areal stimulation continuous. In other words, it is assumed that when more than one spot is stimulated, the general character of the resultant sensation depends upon the sensation that is most intense in the spots stimulated.° It will be remembered that in their examinations of sensations following nerve injury, Head and his collaborators found certain sensation losses that appear not to conform with the hypothesis of special nerve endings for warmth and coolness alone. Certain of their results appear unaccountable on the supposition of loss of certain numbers of the fibers that supply the end organs concerned with the sensations of warmth and cold,—supposedly Ruffini’s eylin- ders and Krause’s end bulbs, respectively. Their work has cast considerable doubt on the singularity of the sensations of warmth and coolness and it appears from their examination of cases in which nerves have been injured or cut that there are two sets of nerves, four different fibers, which convey temperature sensations. The two sets belong respectively to the epicritic and protopathic systems, the former being concerned with medium temperatures, which are appreciated as warmness and coolness, while the latter mediates the sensations which may be spoken of as hotness and coldness.”° The results of temperature experiments made by me on a patient (H.) in whose arm the median and ulnar nerves were cut confirm, in a general way, those reported by Head."' In some particulars, however, differences were found. Over parts which were insensitive to pressures no sensations of temperature were obtained, even from those temperatures which caused a burning of the skin. H. at one time placed his hand “On this see Titchener. Haxrperimental Psychology, vol. 1, part 2, pp. 87-91. “Head and Sherren. The Consequences of Injury to the Peripheral Nerves in Man. Brain, 1905, vol. 28, pp. 224-228. “For an account of the patient and for other sensation differences, see Franz: this Journal, vol. 19, pp. 107-124. FRANZ, Sensations Following Nerve Division. 225 against a steam radiator and produced a burn about 1 em. in diameter, without appreciating that his hand was in contact with a hot surface. This burn did not heal, as Head has noted in similar cases, as ‘rapidly as burns on normal parts, and it was over ten weeks before this area of the side of the hand took on a healthy appearance. During this time no pain or feeling of temperature could be ob- tained from this part of the hand. Temperatures as high as I dared use in the experiments, without causing a burn, were not felt. In the same way low temperatures were not appreciated on this part of the arm and hand. A test tube, the temperature of which had been lowered to —5° C. was not felt and during cold weather the patient had to depend upon the sensations from the thumb and the radial part of the arm to determine when the hand should be covered. In the areas of the hand and arm in which protopathic sensibility remained, the extremes of temperature were easily appreciated, al- though the medium temperatures did not call forth a sensation. In the area retaining also the epicritic sense, however, both extreme and medium temperatures were readily appreciated. In the experiments with my subject, I used heated or cooled test tubes, 12.5 mm. in diameter with hemispherical bottoms. These were filled with water and in each tube a thermometer was inserted so that the temperature could be read directly after or before stimulat- ing any part of the skin. The tubes were placed on the skin and pressed only with their own weight. It was found that at no place of stimulation did an area more than 8 mm. in diameter rest upon the skin. For cold sensations the test tubes were cooled by being placed in a mixture of ice and salt, and for the lowest temperature the tubes were cooled to —5° C. For lesser degrees of cold, for tem- peratures of about 20° C. the tubes were immersed in cooled water. For testing for sensations of warmth and hotness the tubes were immersed in a water bath that almost completely covered the test tubes. Irregular orders were followed in determining the temper- ature sensations, and no indication was given the subject what the next kind of stimulation was to be. Moreover, the same square cen- timeter was never tested twice by the same stimulus in succession. The tubes were allowed to rest on the skin for only one or two sec- 226 “fournal of Comparative Neurology and Psychology. onds. Immediately before the test tube was placed on the skin, the signal word “Now” was given and after the test tube had been lifted, the subject reported whether or not he had felt anything and also the quality of the stimulation. The usual procedure of request- ing judgments when no stimulus or when an indifferent stimulus was given was tried to see whether or not the subject guessed. After some preliminary trials, the subject used exclusively the terms “hot,” “warm,” “cool,” ‘cold,’ “pressure,” and ‘‘do not feel anything.” The subject also often voluntarily compared the temperature sensa- tions of two or more stimuli in order and these reports gave a clue io certain differences that will be reported later. Throughout the test it was impossible to keep from the subject the knowledge that he was being tested for temperature sensations, and whenever he reported that the stimulus was accompanied by a sensation of pressure only, he was asked to try to determine the character of the temperature. At these times he was able to make a judgment only when a stimulus was repeated, and after the test tube had been left on the skin from 10 to 60 seconds, until in all probability there had been time for radiation of the heat or transmission of the stimulus to more nearly normal parts. . A series of early experiments on the hand showed certain devia- tions from the results of Head and Sherren, and for this reason, two areas were selected for careful examination, those which had been previously used for the determination of the hair sensibility on the upper part of the forearm near the bend of the elbow, and on the lower part of the forearm near the wrist.” In the area near the bend of the elbow, each square centimeter was tested separately, but in irregular order that no suggestion might be given or obtained of the extent of the loss of temperature sensa- tion, and each area was carefully gone over three times with each form of stimulation. The resultant sensations that were reported were, with two exceptions to be mentioned, the same in all three tests, and the uniformity is a striking evidence of the accuracy of the observations. In experiments in this area, no attempt was made “Figures 2 and 38, pp. 219 and 221, illustrate the areas which were carefully tested. FRANZ, Sensations Following Nerve Division. 227 to keep the temperature of the test tubes constant beyond that of having them cold, cool, warm or hot to corresponding parts of a normal individual, but temperature degrees were always noted and records made at the time. In Fig. 4 is shown the area on the upper I 2 5B 4 -) NSSAS OF Tem Oo) Oh ol > Nera OE AG) ar Cis Cle a Hires As Wie. 4.—Ilustrating the sensations accompanying temperature stimuli on the upper volar part of the forearm. Each of the four parts of the figure represents the same area of skin. The lowest line is the extent of the loss of pressure sensations. The line above area D separates, but not so sharply on the skin, the area of epicritic sensibility (A, B and C), from that of pro- topathic sensibility (D, E, F and G). Vertical lines, sensations of warmth. Horizontal lines, sensations of hotness. Diagonal lines running from top to the right, sensations of coldness. Diagonal lines running from the top to the left, sensations of coolness. part of the forearm divided into its component square centimeters with the results from each kind of temperature stimulation. In only two cases, with very cold stimuli, 2°-10° ©., did the three answers differ and in both cases, once each in the two 228 “fournal of Comparative Neurology and Psychology. squares E4 and KE5, the stimulus was perceived as cold. The four parts of the figure represent the sensations from the stimuli which normally would be called cold (2°-10° C.), cool (15°-22° C.), warm (30°-40° C.) and hot (45°-60° C.). The lines running down- ward from the left indicate sensation ‘cold,’ the lines running down- ward from the right indicate the sensation ‘cool,’ the vertical lines indicate the sensation ‘warm,’ and the horizontal lines indicate the sensation ‘hot.’ With the lowest temperatures all the areas responded with some form of sensation. The areas A, B and C reacted uniformly with the sensation cold, and the square centimeter D5 also reacted in the same manner. Once each, as has previously been mentioned, K2 and K4 gave a feeling of cold, while in each of these squares two other similar stimuli were reported to feel cool. With the next grade of stimulus, cool, there was a wider distribution of cool feeling than there was of cold feeling with the lowest temperatures, the areas comprising A, B, C and D, and of E and F the square centimeters marked 4 and 5. The feeling of warmness was obtained over areas A, B and C uniformly with temperatures from 30°-40° C., and in D the subject reported a similar feeling but ‘only slightly warm.” With stimuli from 45°-60° C., areas A, B and C reacted always to the stimulus as “hot,” while areas D, EK, F and G gave the feeling of warmth. Areas A, B and © are areas in which hot is distinguished from warm and cold is distinguished from cool. Areas E, F and G are areas in which only hot or cold stimuli are invariably distinguished, although area D5 reacted to cold, areas K2 and K4 reacted once to cold, D1—5 and also E4, E5, F4 and F5 reacted to cool stimuli. Area ID reacted to warm stimuli with an appropriate reaction, but the intensity of the stimulus appeared to be less than in the neigh- boring area C. With the hot stimulus areas A, B and C reacted in a normal manner, while the other areas reacted to the same stimulus with the sensation of warmth. It will be noted that in the area in which the epicritic sensibility remained intact, coolness and cold- ness and warmness and hotness were always distinguished. This comprised the horizontal sections, A, B and C. FRANZ, Sensations Following Nerve Division. 229 The area near the wrist gave similar results, although in the ex- periments of this region I did not map out the area according to squares and have to offer only the general results on the horizontal areas. In this case, however, I was careful to keep the temperature of stimulation constant for each area, and each time the stimulus was applied it was of the intensity noted on the diagram. The results of these experiments are shown in Fig. 5. Five dif- 10°C. 20-G. ~—— One ee -——_— Se Ke we KK — res 15. Fie. 5.—Skin area on ulnar aspect of forearm hear wrist. Area A, no pressure sensations felt. Areas B and C pressures felt and protopathic sen- sibility retained. Areas D and BH, pressures and touch appreciated and epicritic sensibility intact. For explanations of sensations of temperature, see legend to Fig. 4. ferent degrees of the stimulus were chosen, —5°, 10°, 20°, 40° and 60° C. The coldest stimulus felt cold in areas D and E, cool in area C. A temperature of 10° C. felt cold in areas D and EK, but did not eall forth a sensation of temperature when placed on area C. 20° C. was reported cool in areas D and E, but was indifferent in the other areas. Stimuli of 40° C. were felt warm in areas C, D and E, but indifferent in areas A and B. In area C the subject 230 fournal of Comparative Neurology and Psychology. noted that this temperature was “just warm,” “only slightly warm,” ete., not so warm a sensation as that given by the stimulus in the areas D and EK. Likewise with a stimulus of 60° C. The area in which warm sensations were obtained were B and C, but the sensa- tions from area B were not so intensive as those in area C. This temperature was felt in areas D and E always as hot. The parts of the arm which did not retain their ability to appre- ciate pressures, area below G in Fig. 4, and area A in Fig. 5, did not respond at any time to any, temperature stimulus. Areas G, F, E and D in Fig. 4 and area B and C in Fig. 5 may be said to have retained the protopathic sensibility in addition to the deep_sensi- bility, and in these areas hot stimuli were felt to be only warm, while cold stimuli were felt to be only cool. In the upper forearm area D-G there was no response, as a rule, to the intermediate degrees of temperature. ‘The remainder of the area A-C responded accurately to all degrees of temperature stimuli. In the area near the wrist, only D and E reacted well to all degrees of temperature, and showed the presence of epicritic sensibility. Areas B and C failed to respond to medium degrees of temperature and their response to the extremes was not well marked. At different times during the examination of these two arm areas, the following experiment was tried: The test tube, instead of being placed with its end on the skin, was placed so that three to four centimeters of its length extended over the horizontal areas which gave such widely different results. In these experiments the subject described the sensations which were produced and the accounts are in accord with the observations made when the small horizontal areas were stimulated by the end of the test tube. Near the wrist when cold, —5° C., was used, the subject reported that toward the radius the sensation was of extreme cold, but near the axis of the arm the sensation was only cool. In a similar way the sensation from 60° C. was described as hot near the radius and warm near the axis of the wrist. The places from which respectively cool and warm sensations were obtained when cold and hot stimuli were given were pointed to by the subject and they correspond closely to the area OC. This area it will be remembered was found by the previous Franz, Sensations Following Nerve Division. 231 experiments to give these sorts of sensations with the extremes of temperature. On the upper part of the forearm similar results were obtained. In this case the subject was permitted to keep his eyes open and to mark with his other hand the places where the stimuli changed in quality. The marking of the division between the areas was not clear and distinct, but between two points, about two centi- meters apart, the difference in the sensations was reported to be marked. These results are in accord, therefore, with those of stimula- tion of the individual square centimeters and they strengthen the impression that the sensation difference that was reported in the first series was not due to radiation or conduction. These results are quite unlike the results in normal individuals, for in the latter Pressure Eres 6% Fic. 6.—Area insensitive to cold stimuli, compared with the loss of pressure sensations. Area insensitive to cold inclosed within heavy line. Area marked with vertical lines insensitive to pressures. we find the most intense sensations at the places where the warm or cool object leaves the skin, and the intermediate area appears to be stimulated uniformly with the same temperature. On the hand, experiments similar to those on the arm were made, but with not so many different temperatures. The results of these experiments are given in Figs. 6 and 7. In Fig. 6 is shown the area of the hand insensitive to cold stimulation. This included the ulnar quarters of the palm and the back of the hand, the whole of the fourth and ring fingers, the first two joints of the second and index fingers and about half of the thumb. For comparison, the area in- sensitive to pressure with a pencil is illustrated on the same diagram 232 “fournal of Comparative Neurology and Psychology. by the use of vertical lines. In only a small portion of the third finger does the area of insensitivity to pressure go beyond that of insensitivity to cold. Fig. 7 gives in a similar way the area insens- itive to hot stimuli. For comparison, with the areal loss of cold and hot sensations, the area insensitive to light touch is drawn. On the back of the hand and on the index finger the area of insensibility to light touch is greater than that to heat, while on the thumb the area for hot sensation loss is greater than for light touch. With the exception of a very slight portion of the thumb the area in- sensitive to cold is included within that insensitive to light touch. In a general way, therefore, these results, as has already been Wie, 7 Fic. 7.—Area insensitive to hot stimuli, compared with loss of sensations to light touch. Area insensitive to hot stimuli inclosed within heavy lines. Area marked with horizontal lines insensitive to light touches (cotton wool and camel hair brush). said, confirm the observations of Head. The facts that do not agree with those of Head are: There are different extents of areas respond- ing to different degrees of temperature; the area for the appreciation of hot stimuli, for example, not being the same as that for the ap- preciation of cold stimuli as such; temperatures of extreme degree are felt like medium temperatures or call forth sensations of warm- ness or coolness on areas which do not respond to the medium tem- peratures with a corresponding sensation of warmth or coolness. The results obtained from the arm may be taken to mean that whenever a cold or hot stimulus was given there was a radiation of effect from the stimulated area to the neighboring areas. This Franz, Sensations Following Nerve Division. 238 explanation would also account for the less widely felt sensations from the medium degrees of temperature when the individual square centimeters were separately stimulated. Such an explanation would not account, it seems to me, for the results which were obtained when an extended line was stimulated rather than a small area. It appears more probable that, as Head contends, for temperature sensations we have two sets of nerves, one of which responds to the extremes of temperatures and the other to the medium temperatures. These two sets of nerves correspond to the forms of sensibility that are called respectively protopathic and epicritic, but in a different way than that deseribed by Head and his co-workers. The results from my subject show that, although hot and cold stimuli produce sensa- tions in areas endowed with the protopathie form of sensibility, the sensations correspond to those produced in normal areas by stimulation with medium temperatures. This at first sight looks as if we were dealing solely with differences in threshold values, but this hypothesis does not account for another apparently anomalous condition which was found by Head and Sherren, viz., the loss of the ability to sense the extremes of temperatures with the reten- tion of the ability to sense medium temperatures. It seems to me that the temperature results, taken in connection with those which have been described in my articles on the ‘Pressure- hike’ and ‘Hair Sensibilities,’ show there is an overlapping of the nerves, or rather there is an overlapping of nerve supply, and that the sensation differences which were found in this case are to be explained as due to the presence or absence of certain nerve endings or of nerve fibers. It is possible that the sensations of coolness and warmth are protopathic, while those of coldness and hotness are epicritic. This, it will be observed, is contrary to Head’s belief. Received for publication December 7, 1908. Nore.—Since the articles on sensations following nerve division were written I received the number of Brain in which Rivers and Head give the results of careful examinations made on Head’s arm in which the radial nerve was divided near the elbow (W. H. R. Rivers and Henry Head. A Human Experiment in Nerve Division. Brain, 1908, vol. 31, Part 123, pp. 328-450). Some of the new observations reported by Rivers and Head indicate a con- dition on Head’s arm similar to that found by me on H., but some of these 234 “fournal of Comparative Neurology and Psychology. anomalies are not sufficiently discussed. A full account of this recent work, with criticisms of special parts of it, will be found in a forthcoming number of the Journal of Philosophy, Psychology and Scientific Methods. In this review I shall make a careful comparison of the present work with that of Rivers and Head, and with that of Trotter and Davies (Journal of Physiology, London, 1909, vol. 38, pp. 184-246), which has also appeared since my articles have been in type. At present I wish to call attention to only a few matters reported in these two recent articles. The recovery of sensibility to light touch is considered by Rivers and Head to be dependent upon the regeneration of the so-called touch spots. These touch spots are, according to these authors, grouped about the hairs and much, if not all, of the sensibility of the hairs depends upon the presence of the nerve endings which make up these touch spots. They say, for example, “Over hair-clad parts these touch spots are strictly associated with the roots of the hairs, they express the sensibility to mechanical stimuli to that part of the hair which lies beneath the surface of the skin. Almost every hair is a delicate tactile sense-organ; any movement of its tip is transmitted to its root with the increasing power of a lever, setting up tactile impulses.” We find, however, they report in one part of their paper that tactile sensations (sensations to light touch) could be evoked only after a period of 365 days following the operation, but in another place we are informed that “‘No sen- sations were obtained from the hairs until 86 days after the operation, when there were found four hairs that gave a sensation when they were plucked.” It is also noted that 161 days following the operation the hairs on the arm were sensitive to the stimulation of brushing with cotton wool, although the sensibility to light touch did not return for 365 days. Surface indications are that Rivers and Head were dealing with the same form of dissociation of the hair sensibility which has been described by me in the foregoing papers, but that they did not carefully investigate this matter. At any rate, from their account of the sensibility of the hairs we are justified in assuming that the hair sensibility was found by them to be independent of the presence of dis- tinct touch spots and that their observations upon the sensibility of the hairs, casually reported, support the view expressed in the foregoing paper. According to Rivers and Head all temperature sensations also depend upon sensation spots, and the differences in temperature sensations described by Head and Sherren in a former paper are now reported to be due to the presence of a greater or less number of cold or hot spots in the epicritic and protopathic areas respectively. Head still holds the view that in an area in which there is no epicritic sensibility only the extremes of temperature will be appreciated, while in the area endowed with epicritic sense there is the ability to appreciate the intermediate degrees of temperature. It is not clear that Head had sensations similar to those of my subject, but we read “It would seem that the number of spots stimulated is of greater importance than the intensity of cold by which the sensation is evoked.’ In experi- ments in which his arm was tested by cold areas of different size Head reported a cool large area to feel colder than an ice cold small area, in the former case there being a number of spots stimulated and in the latter only one. Franz, Sensations Following Nerve Divtiston. 2215 On account of the fewness of the spots in the protopathic area it is difficult to see how the protopathic skin would give the hot and cold sensations even from the stimulation of comparatively large areas and how the sensations could be of the same intensity as those from normal parts. Head’s observa- tions on the temperature sensations are at variance with those reported by me, but it should be remarked that Trotter and Davies were not only unable to confirm Head’s observations, but that, in fact, they obtained results similar to those reported here. The matter of the sensibility to light touch needs hold our attention for only a brief time. The gradual increase in sensitivity from the anesthetic area outwards has also been demonstrated by Trotter and Davies in even a more convincing manner than that given by me. They have shown that the sensi- bility of individual touch spots differs at different times during the period of recovery, and it is very plain that the sensory disturbances following the section of a nerve are more widespread than Head and his collaborators admit. It is of some interest to note that a criticism of the cotton wool method of testing light touch has also been made by Trotter and Davies and that for similar tests they used a brush. April 26, 1909. MODIFIABILITY OF BEHAVIOR IN ITS RELATIONS TO THE AGE AND SEX OF THE DANCING MOUSE. BY ROBERT M. YERKES. From the Harvard Psychological Laboratory. Witu Four FIGuREs. CONTENTS. PAGE I. Introductory statements: the dancer as material for the investiga- (GO Cie [ROSIN Ole JORIS Son condomoacescctgcocoooocuduoS 237 II. Relation of age and sex to rapidity of acquisition of a visual dis- CANE ORK es ety MMe anc ha hoceomuiod sicade ccs ose 238 III. Sensitiveness to electrie stimulus, in its relation to age and sex.... 249 IV. Strength of electric stimulus, in its relation to rapidity of habit- HOM GL OM = ctenctorses susie cise aia vile, &eolee oneas st oahetolion s telisKarsutel’e ue uek oheuel-e leltomete epeows 252 VY. Relation of difficultness of discrimination to rapidity of habit- LORIMAWON Ae GUHELEME BAeCSe acre seers sie ceehetelsl vie) cle ec-lekeneneuehenenerers 2559 1. Experiments with cardboards in discrimination box...... 255 2. Experiments with discrimination box in dark-room...... 259 3. Experiments with one side of discrimination box covered iN “VALYINS CESTEOCS 2 oes. 0c0 w) 2e Sic wie Fue xe) ses ae eae ones 263 VI. Relation of age to rapidity of acquisition of labyrinth habits...... 265 VES Conclusionssanas SUMMA, c/s elses oles + cleus e cisis oie) euch elel heel oie) eb etelienetehancite 267 I. INTRODUCTORY STATEMENTS: THE DANCER AS MATERIAL FOR THE INVESTIGATION OF PROBLEMS OF BEHAVIOR. The dancing mouse is well adapted, by its abundant and in certain respects peculiar activity, to experiments on behavior. Taking advantage of this fact, I have used it extensively as material for the development of methods and the revelation of problems, both physi- ological and psychological. That the results which have been obtained are typically mammalian I am not prepared to assert. This, however, is, for my immediate purposes, secondary in importance to the methodological values of the work. Animal psychology is urgently in need of exact methods of research. It is an appreciation THE JOURNAL OF COMPARATIVE NEUROLOGY AND PsycHoLoGy.—Vou. XIX, No. 38. 238 “fournal of Comparative Neurology and Psychology. of this fact that has shaped my experimental work during the past five years, and that now leads me to offer the following results of my study of the dancer primarily as a contribution to the evolution of method and as an aid to the profitable formulation of problems. This paper is a direct continuation of the studies in the behavior of the dancer which are described in my book, “The Dancing Mouse.’’* Although I have attempted so to write the paper that both methods and results shall be intelligible to those readers who are not familiar with the details of previous publications,’ it has been necessary—ain order to keep my account within reasonable space limits—for me to omit everything except the chief points, in connection with methods which I have previously described, and a concise statement of new results. In other words, I have been forced to assume much more knowledge on the part of the reader than I should if this were my first publication on the subject. Certain problems concerning the relation of age and sex to habit-formation which were proposed in my book, and either left un- solved or only partially solved, are brought nearer to satisfactory solution by the results herein reported, and a multitude of new prob- lems are revealed. To me, however, the investigation presents itself simply as another step toward a realization of the complexity of the phenomena of behavior and of the need for accurate analytic methods. II. RELATION OF AGE AND SEX TO RAPIDITY OF ACQUISITION OF A. VISUAL DISCRIMINATION HABIT. Can the dancer acquire a given habit with the same rapidity at different ages? ‘This question was the starting point of a study of plasticity which has already been reported in part.* Before present- ing the results of my experiments I shall very briefly, with the help of figures which are reproduced from an earlier paper, describe the method of work. Yerkes, Robert M. The Dancing Mouse: a study in animal behavior. New York, The Macmillan Company, 1907. xxi + 290. “Yerkes, Robert M. and Dodson, John D. The Relation of Strength of Stimulus to Rapidity of Habit-formation. Jour. of Comp. Neur. and Psy., vol. 18, p. 459-482, 1908. *The Dancing Mouse, pp. 270-275. YERKES, Modifiability of Behavior 2.39 The habit whose formation was studied quantitatively, in the case of groups of dancers consisting of five pairs each, for the ages of one month, four months, seven months, and ten months, may be called the white-black discrimination habit. It involved the discrimination resale liivel, D Fie. 1.—Discrimination box. W, electric box with white cardboards; B, electric box with black cardboards. Fic. 2.—Ground plan of discrimination box. A, nest-box; B, entrance chamber; W W, electric boxes; L, doorway of left electric box; R, doorway from right electric box to alley; O, swinging door between alley and A; JC, induction apparatus; C, electric battery; K, key in circuit. of the entrances to two boxes, one of which was white and the other black, and the entering of the white box. Any attempt to enter the black box was punished by an electric shock. Figures 1 and 2 show the experiment box in perspective and in 240 “fournal of Comparative Neurology and Psychology. ground plan, respectively. The subject, after being placed in the nest box, A, by the experimenter, was permitted to pass into the entrance chamber, B. Then a piece of cardboard, which was placed between the animal and the opening into A, was slowly moved toward L, R, of Figure 2. Thus the dancer was brought face to face with the two entrances, L and R, of this figure (B and W, 2. e., black and white of Figure 1). One of these it would soon attempt to enter in order to escape to the nest box, and thus find space for dancing. If it started to enter the black box (and this might be either the box on the left, L, or the box on the right, R, for the white and black card- boards, which were at the entrances and within the boxes, could be transferred readily by the experimenter) it was immediately given a weak electric shock by the closing of the key, K. This usually caused it to retreat from the box and to try the other entrance. In case it entered the black box in spite of the shock, it was not permitted to escape by way of E and O to the nest box, but instead was forced to return to B and again make choice of an entrance. This was continued until the white box was chosen, then the animal was allowed to return to A. After an interval of one or two minutes it was given another opportunity to select the right entrance. This was con- tinued until the white box had been chosen ten times. Such a group of ten trials constitutes what we shall refer to as a series. One series was given each individual daily from the beginning of experimenta- tion until the acquisition of a perfect habit of.discrimination and choice. The positions of the white and black cardboards were changed in precisely the same way for each individual according to an order which has already been described.* These shifts in the position of the white box were made in order to prevent the mouse from acquir- ing the habit of going regularly to the entrance at the left or at the right. An experiment (test or trial) was recorded as yielding an error of choice if the mouse entered the wrong box far enough to get a shock; as yielding a correct choice if, without first entering the black box, ‘Jour. of Comp. Neur. and Psy., vol. 18, p. 461, 1908. YERKES, Modifiability of Behavior. 241 it entered the white one and passed through to the nest box. In the tables appear the number of errors per series made day after day by the various individuals. At the outset of the experiments each mouse was given two series of what may be called “preference tests.” In connection with these tests no electric shock was given and the mouse was permitted to enter and pass through either the white or the black box, for it was the sole purpose of the experimenter to discover, by means of these series, any initial preference that the subject might have for either the white or the black box. A habit of discriminating between the boxes, and of uniformly choosing to enter the white one, was considered perfect when the mouse made no errors in three successive daily series. As a measure of the rapidity of habit-formation we may use the number of tests between the beginning of the first training series (following pref- erence series B) and the end of the series which preceded the three perfect series. This measure of rapidity of learning, which I have named the index of plasticity, proves to be extremely useful for pur- poses of comparison. To ascertain age differences in rapidity of white-black-habit forma- tion I used groups of individuals which, so far as I could tell, differed from one another constantly only in age. Five males and five females constituted each group, and four such groups were used. During their lives all of the animals were kept under the same conditions. They were paired at the age of twenty-five days, and thereafter a male and a female were kept in a separate cage and were placed in the ex- periment box for their daily training at the same time and given their tests alternately. We may now examine the results of the experiments. Table 1 con- tains records of the number of errors of choice made by each of the in- dividuals of the one-month-old group in each daily series. The num- bers at the top of the columns refer to the mice. Even numbers always designate males; odd numbers, females. The two preference series are indicated by the letters A and B. No. 210, it will be noted, made six erroneous choices in each of the preference series and also in the first training series; that is, he attempted to enter the black box instead of the white box six times in ten. In subsequent training se ~ fournal of Comparative Neurology and Psychology. 242 TABLE 1. RELATION OF AGE TO MODIFIABILITY OF BEHAVIOR HABIT WHITE-BLACK DISCRIMINATION Results for dancers one month old FEMALES, MALES. ONCONOMVOANAATHNOCOS ONMMMAMANTOTANOOO 254 | 410 252. m~o ann | 6 6 OANHHANRTANOCTROnOCCS OMdtAAMAAOnOoCSO od | | ie i) ) a} NO | pee (yo o wo - | < | wo ra mo bo or) | ne Hin | MAMMNOCHAANCOCOSO | = Yo) Volts) N for) } x INO | N in — fo alte ¢) RMIMARANROROCSO SONNOKDAWHNOSCOS IMANOAAAOOO ANMMAOnOTAOCS Ns MOAAANROCO MAHANAHIOAROSCS LAMM AHAAAAOOCO OAMMIMINAOSOO ‘ on Ld ANC HID Or OO Sm Sts NS TABLE 2. RELATION OF AGE TO MODIFIABILITY OF BEHAVIOR WHITE-BLACK DISCRIMINATION HABIT Results for dancers four months old FEMALES. MALES. At coin wag | oO =H os alive} | ADNONDNOCNOGCONAHOCCO WMAMAAANOROOOG QOAMANANANAOnROCOSO OAM AMOOAAAONOCnNCOO HHMOIMANAMOR AA HOOO MAMA ARONCOO NANTON OOoOnOOCSo ONOOHAHAHHOOCOOHHONOCCO MMHMMANNARHOSG © OwMwM MM AANNONOTOOCO Series. rrMOMMAAHANAAHOOS DIN MID AMO HANH MDOnOnOOCO YERKES, Modifiability of Behavior. 243 ries the number of errors made by this individual rapidly decreased until in the seventh series only one was made. Then followed three perfect series. For this individual, since he acquired a perfect habit as the result of seventy training tests, the index of plasticity is 70. The tables contain, in addition to the individual results, the average number of errors per series for the males and for the females. TABLE 3. RELATION OF AGE TO MODIFIABILITY OF BEHAVIOR WHITE-BLACK DISCRIMINATION HABit Results for dancers seven months old MALEs. FEMALES. ] | ] Series. | 92 96 98 | 116 | 120 | Average. 91 93 99 101 , 109 | Average. | } | | | A ) 6 5 7 | Boe 4 4 7 6 @ || ed! B 7 4 7 3 is 5.2 6 6 i 5 7 6.2 | | ie Wee hese eine, Ven yen S 4.8 Be een erat Sens 5.8 2 | 4 5 Gig led Si 2586 2 3 7 6 2 4.0 3 4 3 Zoi Bee liens A A 4 4 6 4.4 4 7 5 4 | 5 3 4.8 A 3 6 4 4.0 5 By oi) 5 | 4 BT eae 5 3 5 2 3 | 3.6 6 Bo 2 4 4 | 3.6 5 2 + 2 2 3.0 7 I eso, ral 1 4 4 | 226 1 4 4 3 2) oes 8 eG 2 3 2 4 3.4 2 4 2 pe? 3 2.6 9 2 1 30S 5 3.2 1 5 it i) oi 1 1.8 10 5 1 3 4 yy Men 8.03 0 2 2 0 3 1.4 11 fal 1 3 1 ill 0 1 a) ail 1 0.8 12 2 | & 1 4 1 2.0 1 1 Gi Bh 2 1.4 13 | 6) 1 3 Bi GG 2 Sy he O OQ i.e 14 eS 1 3 Pa aaa 1 1 | @ | € Oe | 37 0)4 15 hecvote te @ 4 2 OF) 2088 1 1 0 1 ON OkG 16 1 i o 0 0 2 056 0 i 1 | -@:2 17 10 0 0 2 0.6 1 0 0 0.2 18 lhe 0 1 0 0.4 0 0 0 Q 19 0 0 1 0.2 0 0 20 ha 1 1 0.4 0 0 21 Vest | 0 1 0.4 22 0 eco ee 0.2 | 23 1 i OF ih 0 0.2 | 24 0 | 1 0.2 | 25 0 | 0 0 | 26 0) | 0 0 | 27 0) 0 | | | | | Any one who compares this account of my investigation of the re- lation of age to rapidity of learning with my earlier account will discover that only two pairs of dancers of one month of age for which results were given previously? have place in the group under dis- cussion. This is due to the fact that I felt it highly desirable to re- peat the experiments with one-month individuals in order to make “Phe Dancing Mouse, pp. 243, 273. Len} b ’ 244 fournal of Comparative Neurology and Psychology. sure that in the interval which had elapsed between the beginning of this portion of my work and its completion no important changes in the plasticity of the race had oceurred.® As a matter of fact this precaution proved unnecessary, for no important differences appeared as the result of the interruption of the investigation. The condensed results for the four-month individuals appear simi- TABLE 4. RELATION OF AGE TO MODIFIABILITY OF BEHAVIOR WHITE-BLACK DISCRIMINATION HABIT Results jor dancers ten and twelve months old MALES. FEMALES. Series. 90 | 112*| 142 | 144 | 196 | Average. ] 97* | 113*| 119 | 123 | 141 | Average. \ | A | 6 6 5 5 6 5.6 5 8 i 5 4 5.8 B 5 5 6 6 5 5.4 5 7 6 5 4 5.6 ee ES fee || ee = & = he (2 eo eee us Ei | | | 1 ) 2 2 7 4 5 4.8 7 6 5 6 4 5.6 2 | 6 4 Ba, 24 3 4.4 4 3 7 4 5 4.6 3 ez 3 reel nad 4 5.2 4 6 8 7 3 5.6 4 3 4 Ey | 3 5 4.4 7 3 5 5 3 4.6 5 5 4 Ey |e ze 1 358 4 2 3 2 6 3.4 6 3 4 alt, EZ 3 3.6 ® 2 1 1 5 B® 7 3 5 3 3 4 3.6 2 3 1 0 7 2.6 8 3 2 2 5 4 She 2 1 2 1 5 2.2 9 4 3 2 4 1 2.8 1 1 1 1 4 1.6 10 3 4 1 1 0 1.8 1 1 0 1 1 0.8 11 1 1 1 1 0 0.8 1 0 0 1 2 0.8 12 1 2 1 1 0 1.0 2 0 Te eeO 2 1.0 13 2 1 0 1 0.8 0 0 0 1 0 0.2 14 0 3 1 1 1.0 0 1 0 1 0.4 15 1 1 1 0 0.6 0 1 0 0 0.2 16 0 i | 1 0.4 0 0 1 0.2 17 0 | il 1 0.4 0 1 0.2 18 0 0 On| mel 0.2 eo) | O 0 19 Pt || a 0.4 ant) 0 20 | (10. @ || a | + 0.2 | | 0 0 21 | | O | O | 0 | | 22 | o | o | 0 | | 23 | | © | 0 | | "Twelve months old. larly in Table 2, and those for the seven-month individuals in Table 3. In the ten-month group (Table 4) I have included the results for three mice which were twelve months old. Although it is not wholly satisfactory to do this, it seemed better than to deal with the three individuals separately. At any rate nothing is concealed by averag- ing the results for the ten mice, for the individual results are avail- able. To make comparisons easier, I have brought together in Table 5 °An epidemic which destroyed almost all of my mice caused a delay of over a year. YERKES, Modifiability of Behavior. 24.5 the averages for the males and females of each group. This table presents also the general averages for each sex. Inspection of these results reveals the following significant facts. (1) The females exhibit a stronger initial preference for the black box than do the males. Both, however, choose the black box more frequently than the white box, in the preference series. Since TABLE 5. GENERAL RESULTS OF THE STuDY OF THE RELATION OF AGE TO MODIFIABILITY OF BEHAVIOR. Each result in the table is either the average number of errors in ten tests for five individuals, or the general average for twenty individuals. MALES. " FEMALES. Series. 1 mo. | 4mo. 7 mo. |10 mo.) Gen. Av. | 1 mo. 4mo.| 7mo. 10mo. Gen. Av. AS oy See Sry), 5-89 | 5.6 5.60 G2 | Ose Nebel 58 5.90 Be |. 5.0) 6-2 | Ye.2)s 5 5.50 6-0) | 524° 1) 6-2) | 5.6 5.80 1 52 0F P5RG. | 428-1! 4.8 5.05 536) |, 5.2% | 35-8) a 526 5.50 2 SHO) jh pear | 25.6)-|), 454 4.55 3.2 | 2.8 | 4.0 | 4.6 3.65 3 Bide) ARG a| 8.8" 5.2 4.20 3.0 | 3.2 | 4.4 | 5.6 4.05 4 3.2) | 3.6 | 4.8 | 4.4 4.00 2.6) | 2.60) 4.0) | «4.6 3.45 5 2.6|) (3:4) |) 4.6) || 3.8 3.60 Ppa oa\| 2) NES le ea 2.85 6 M8) | 2h4y | 35651) 3-6 2.85 Tete WS Be) 52 2.15 7 Ne2> 2245 v2.6: || 3.6 2.45 14.) 2.25 | 2.8" | 256 2.25 8 0.4 | 42.8 | 3.4 | 3.2 2.45 108 |) 20.61.) 42:6) | 9252 1.60 9 Ore 220 5) Se aioe 8 2.10 1 2e pd soe | iS k Slo 1.45 10 Op2 a0) || 36.4} 1-8 1.65 Or2> | 0266 )| 14AS One 0.75 11 OP) £6. fe F058 0.95 0.4 | 0.6 | 0.8 | 0:8 0.65 12 OW | sOs6i24 9250) |= 00 0.90 Om |) 0.2") 0s io 0.75 13 Os i O%A- | 16> | 201.8 0.70 O22 -| 0:4. | 12) | “Ou 0.50 14 Os | 2:4) 1.0 0.85 0 0. | 0.4.) O:4 0.20 15 0.2 0.8 0.6 0.40 0 0 0.6 0.2 0.20 16 0.) 0.6 | 0.4 0.25 0 O | O82. On2 0.10 17 Gi 0.65} 7034 0.25 0.2 | 0.2 0.10 18 O | .0.4 |) 0.2 0.15 a0 0 0 19 0.2 | 0.4 0.15 30 0 0 20 0.4 | 0.2 0.15 oa 0 3 0 6 0 | aS STS LE IE 13 0 1 1* 72 1* 14 | 0 0 1 3 i) 15 0 0 1 0 16 2 1 1 0 17 0 0 2 18 0 0 0 19 1 0 1 20 1 0 21 i 0 22 2 0 23 0 24 0 25 0 “*At this point condition of discrimination was changed from “‘easy”’ to “difficult.” relation of age to rapidity of habit-formation is more complex than certain statements made by students of animal behavior would lead one to suppose. My experiments reveal the presence and importance of a number of variable factors in the white-black discrimination habit; and until we know accurately the values and relations of these several YERKES, Modi fiability of Behavior. 259 factors it would be rash indeed to make general statements concern- ing the relation of age to plasticity. We must limit ourselves care- fully to particular statements, for what holds of one condition of training may not hold at all of what appears to be a very similar condition. TABLE 11. INDICES OF PLASTICITY FOR DANCERS OF DIFFERENT AGES, TRAINED UNDER CONDITIONS OF DIFFICULT OR OF HASY VISUAL DISCRIMINATION. CONDITION OF DISCRIMINATION. No. of dancer. DIFFICULT. EASY. ae Ree ca Mn ee eee in a. aae | oT | Young dancers— Old dancers— Young dancers— | Old dancers— 1 month old. | 12 months old 1 month old. | 8 months old. 112 | | 160 113 | 100 204 | 40 121 | 40 292 60 | 291 80 430 50 432 50 136 | 40 166 20 416 60 105 90 Averages. 70 | 130 62.5 i 35.0 2. Experiments with discrimination box im dark-room. The results of the experiments which have just been described suggested to me the idea that ability to acquire the white-black visual dis- crimination habit depends largely upon two factors: capacity for visual discrimination and associative memory. The facts of plas- ticity thus far revealed might be accounted for, it would seem, by the assumption that in the young dancer capacity for visual dis- crimination was either greater at the outset or more readily devel- oped than in the case of old individuals, whereas associative memory is more highly developed in the old than in the young mice. This hypothesis I immediately attempted to test experimentally. If it be correct, young mice should develop the capacity to discriminate slight differences in luminosity more quickly than old mice. To test this matter I planned a series of training experiments with the apparatus which I have previously described'!! as the Weber’s “The Dancing Mouse, p. 118. 260 “fournal of Comparative Neurology and Psychology. law apparatus. It is a discrimination box in which the two boxes which have heretofore been referred to as white and black are illuminated by standardized incandescent lamps. There are no card- boards and difference in illumination, as desired, is obtained by shifting the position of the source of light for one of the boxes. This apparatus permits easy and fairly accurate measurements of the absolute and relative illumination of the two boxes, and in this respect it is more satisfactory than the cardboard method. Its chief disadvantage is that it compels experimentation in a dark-room or at least with artificial illumination of the boxes. In the Weber’s law apparatus two pairs of dancers were trained systematically until they had been given almost a thousand tests. The individuals represent the age limits of the plasticity experi- ments. The old ones, Nos. 170 and 95, were ten and twelve months, respectively; the young ones, Nos. 294 and 293, were one month old. Instead of a single series of ten tests per day, all these indi- viduals were given two such series each day. To start with, all the mice possessed perfectly formed habits of choosing the white box, in the old white-black discrimination appa- ratus. Experiments in the Weber’s law apparatus were begun with the two boxes illuminated the one by 80 hefners, the other by 20 hefners. The difference in luminosity in this case may be stated as three-fourths, since the latter value is only one-fourth the former. I have found it convenient to keep one of these values constant throughout a training experiment and to vary the latter as need dictated. The fixed value, which may then be known as the standard, is indicated in the table by the abbreviation S. The other value, which may be known as the variable, is indicated by the abbrevia- tion V. A habit was considered perfect in this experiment when a dancer succeeded in choosing without error in two successive series. As soon as ability to. discriminate a certain degree of difference in luminosity had been acquired, the amount of the difference was reduced and the training continued under the more difficult condition of discrim- ination. We may now examine the results of this experiment as they appear in Table 12. YERKES, Modifiability of Behavior. 261 At the outset the condition of discrimination was fairly easy and the old dancers learned to choose correctly with 110 tests, the young TABLE 12. RELATION OF DISCRIMINATING ABILITY TO AGE. EXPERIMENTS WITH WEBER’S LAW APPARATUS. Dancers 10-12 Mo. Dancers 1 Mo. OLp. Dancers 10-12 Mo. | Dancers 1 Mo. Op. Series No. 170 No. 95. Series No. 294 No. 293] Series No. 170 No.95 Series No. 294 No. 293 S.80h.V. 20 h. Difference three-fourths. S. 80 h.V.60h. Difference one-fourth. aha 8 1 9 6 AA Sia ges 41 AW lead 2 4 5 2 8 5 45 I. 7° 5 42 2 4 3 7 6 3 3 4 46. || A | 3 43 4 3 4 5 3 4 6 4 Al) Siem lhe et 44 3 7 5) a: 3 5 4 1 AS OLetlipes 45 8 6 6 1 3 6 5 0 AGE Ras 4 46 3 7 Z 1 3 7 iy) |e 50) ty) el 2 47 2 6 : 51 4 4 48 3 4 8 1 2 8 5 0 52 4 5 49 2 5 9 1 1 9 3 1 53 3 4 50 3 6 10 0 0 10 3 0 54 3 2 51 9 3 11 0 2 11 to} 6 55 Sh 3 52 2 3 | 56 oN 53 4 4 a 4 0 12 2 0 57 eas |e 54 3 4 13 (0) 1 13 1 | a 58 3 | 5 55 4 3 14 0 0 14 ie 59 2 | 5 Be tl Lae 5 15 On He 0) 15 Os! 60 | 3 i @ i |) ee 5 | Gly a eS 58 3 5 16 tee 16 One| 62 2 4 59 2 3 63 | 2 4 60 3 5 64 3 6 61 6 7 S. 80h. V.40h. Difference one-half. a Z 2 oz : y Lie eS ie Waly 0 2 67 3 4 64 3 va 1g || @ 0 18 1 2 68 2 3 65 Dateless 1G) i 28 3 19 5 4 69 2 4 66 fe k |o 20 2 1 20 1 1 70 2 4 67 QP 21 0 3 21 1 3 71 3 4 66 Ae ala BS 22 1 3 22 1 1 72 3 5 69 2 3 23 3 0 23 3 2 73 3 2 70 4 0 24 1 1 24 2 5 74 2 4 71 5 2 25 1 2 25 2 5 75 2 3 72 2 3 26 1 2 26 0 3 76 hale ee 73 3 4 27 1 1 Tee: 3 77 2 @ 74 2 5 28 0 0 Oe ieca. | 50 S. 80 h. V. 26.66 L. Difference one-third. 29 2 1 | 29 2 | 0 78 2 4 75 2 0 | 79 1 3 76 2 2 30 1 il 30) | 0 } 2 80 1 5 a7 9 3 31 i i) Sies| 1 | 0 81 9 6 73 1 9 32 0 1 32 | OP ee 82 1 1 79 0 1 33 PA 2 33 1 2 83 4 5 80 0 1 34 1 0 34 4 0 385 2 2 35 2 0 84 2 5 81 3 36 2 1 36 0 0 85 1 3 82 0 37 DB 2 ee] 1 Pi 86 1 2 geo 4 38 3 1 38 1 87 4 3 Sate 3 39 1 0 39 0 88 2 3 85 | 3 40 0 2 40 0 89 3 5 86 3 41 1 1 S. 80 h.V.32h. Difference two-fifths. 42 0 0 ay i ak fp il 87 ened By || 0) 0 | 91 10 alliances 88 3 ones, with 95. In view of the results of the previous section we might have expected the young individuals to learn more slowly than 262 “fournal of Comparative Neurology and Psychology. the old ones. But we must remember that the conditions of this experiment are markedly different from those in which cardboards were used to render the two boxes visually distinguishable. Next the amount of difference in luminosity was reduced to one- half, and the experiment continued. Again the young individuals ac- quired the habit more quickly than the old ones. The index of plasticity for the old is 250, for the young it is 165. With a difference in luminosity of only one-fourth, the training was now continued for several days, but as no one of the four mice succeeded in acquiring a perfect habit it was changed finally to one-third. It is noteworthy that in the thirty-four series (340 tests) that were given to the mice with the difference one-fourth, the old individuals did not succeed in making a correct series, whereas both of the young mice did. With the difference one- third, No. 294 quickly acquired a perfect habit, and No. 293 came very near to doing so, but failed in twelve series. At the con- clusion of the twelfth series, neither of the old individuals had learned to choose correctly, with the difference one-third. Although the results of this experiment are not as convincing as they might be, they do indicate that young dancers can ac- quire the ability to discriminate slight differences in luminosity more readily than can old individuals. It is conceivable, then, although by no means demonstrated as true, that the young indi- viduals in the plasticity experiments acquired the white-black habit more quickly than the old individuals did because they could dis- criminate better or acquired discriminating ability more rapidly and not because they acquired an association more readily. In this event, our experiment measures differences in visual discrim- ination, and in changes which it undergoes with training instead of associative plasticity. I have already shown'? that the dancer is capable, as the result of prolonged training, of developing the power to discriminate be- tween boxes which differ from one another in illumination by less than one-tenth. This fact becomes important at this point, for “The Dancing Mouse, pp. 127, 128. YERKES, Modifiability of Behavior. 263 we are forced to ask, Do the plasticity experiments reveal anything except age differences with respect to what might be termed the educability of light vision? With the hope of getting further hight on this problem, I carried out additional experiments, with the indi- viduals used in the Weber’s law apparatus, by a method whose form and results will now be described. 3. Haperiments with one side of discrimination box covered in varying degrees. For this work the cardboards were removed from the discrimination box which had served for the plasticity experi- ments, and difference in the illumination of the two boxes was ob- tained by covering, with a piece of black cardboard, the whole or a part of the top of one of the two small boxes. The total inside length of the boxes was 29 em. I have described the condition of the darker box by giving in terms of a fraction the amount of the top which was covered. Thus 18/29 means that the cardboard covered 18 of the 29 em., beginning at the entrance and extending toward the rear of the box. Shifting the lighter box (the one to be chosen) from side to side involved merely the moving of the black cardboard from the top of one box to the top of the other. After the experiments just reported had been completed, mice Nos. 170, 95, 294, and 293 were given training tests in the dis- crimination box under the above conditions. Table 13 presents the condition of discrimination as well as the results of the various series of tests. When, as at the outset, the whole of one box was covered, discrimination was extremely easy, because the boxes dif- fered greatly in illumination. From the first, as the data of Table 13 indicate, the young animals learned more rapidly than did the old ones. We have in these results, therefore, additional support for the belief that dis- criminating ability is more readily gained by the young dancer. It may not be out of place to remark here that the simple form of the lighter-darker discrimination apparatus which served for this series of experiments is precisely what should have been used throughout this investigation. It has taken me years to learn that it is not only possible, but also perfectly easy, to devise a con- dition of experimentation which should be readily and accurately 264 ‘fournal of Comparative Neurology and Psychology. describable as to the difference of brightness of the two boxes and satisfactory in its results. I cannot too strongly urge, from my present point of view, the avoidance of cardboards as means of testing visual discrimination. The conditions of many of my ex- periments are practically indescribable so far as absolute value of illumination is concerned, yet, as I now see it, they might perfectly well have been describable with a fair degree of accuracy. TABLE 13. RELATION OF AGE TO ABILITY TO DISCRIMINATE ON THE BASIS OF DIFFERENCE IN ILLUMINATION, AND TO THE CAPACITY FOR IMPROVEMENT OF VISUAL DISCRIMINATION. | | Dancers 10-12 Montus OLp. Dancers 1 MontH OLp. Series. | : 2 Portion of darker | Portion of darker box covered by No. 170. No. 95. box covered by | No. 294. No. 293. | card. card. 1 | Whole. 5 5 Whole. 5 3 2 2 6 4 4 3 2 1 0 1 4 1 0) 0 0) 5 0 0 0 0 es es 6 0 0 48 0 0 see | ee a (0) 0 32 1 1 8 2 | 1 0 0 9 3 | 0 | ) 0 10 3 | 0) 5 2 11 0 1 | 3 | 1 12 2 2 oy | 4 2 13 2 2 0 See 2 | 1 14 & By 0 5 4 2 15 2; 5 2 35 2 | 1 16 . ee 2 s | Ve al 3 17 2 4 1 25 0 1 18 fs 4 1 os 2 2 19 a5 4 2, 2; 1 0 20 $5 4 1 | 21 35 2 3 | | if | | | | | | Having now tested the first of the two important factors in the the acquisition of the white-black discrimination habit, namely, ability to gain visual discriminating power (the educability of white-light vision), we must turn to the second factor and in- quire whether associative memory changes with age. It occurred to me that since the labyrinth habit, as conclusively proved by Professor Watson’* for the white rat, depends more *Watson, J. B. Winezesthetic and Organic Sensations: Their Role in the Reactions of the White Rat to the Maze. Psychol. Rev. Mon. Supp., vol. 8, no. 2, 1907. vi + 100 pp. YERKES, Modifiability of Behavior. 265 largely upon kineesthetie sense data than upon vision or any other special sense, it might serve well to reveal age differences in asso- ciative ability. In this connection we may ask, therefore, Do old dancers learn labyrinth paths more readily than young ones 4 VI. RELATION OF AGE TO RAPIDITY OF ACQUISITION OF LABYRINTH HABITS. For the labyrinth experiments I selected two mazes which I had previously used for the study of educability in the dancer: they are designated as D and C in my book.'* D is what I have described as the regular type of maze, and C as the irregular. Training in labyrinth D was given first to each of ten dancers of from one to two months of age, and likewise to the same number of about ten months of age. About a month after the completion of the training in labyrinth D, the same individuals were trained in labyrinth C. In all cases the experiments were conducted as follows. Two mice, a male and a female from the same cage, were placed in the nest box of the labyrinth together. One at a time they were given first a preliminary test in which they were permitted to find their way from the entrance to the exit of the labyrinth without being disturbed, and then training tests in which they received a shght electric shock each time they made an error in the choice of a path. The tests were continued without interruption, first one individual then the other being tested, until each had perfectly learned the path. A habit was considered as perfect when an individual succeeded in traversing the maze twice in succession without a mistake. Records were kept of the number of errors in choice of a path and of the time consumed in finding the way from entrance to exit. As typical series of results I present in Table 14 the time and error records in labyrinth D of No. 416, a six-week dancer, and No. 166, a nine-month individual. The young mouse was slower than the old one in most of the tests, but he acquired a perfect habit “The Dancing Mouse, pp. 219, 222. 266 “fournal of Comparative Neurology and Psychology. no less quickly. I need scarcely state that the time records have little value for our present purposes, and are therefore omitted, except in the case of Table 14. TABLE 14. RELATION OF AGE TO RAPIDITY OF ACQUISITION OF LABYRINTH HABITS. TyYPI- CAL SERIES OF RESULTS GIVEN BY Two MALES IN LABYRINTH D. Dancer 6 WEEKS OLD. Dancer 9 Montus OLp. No. 416. No. 166 Number of trial. Time in seconds. | Number of errors. | Time in seconds. | Number of errors. Preliminary. 280 | 31 161 19 1 240 60 56 6 2 134 24 25 4 3 53 6 62 8 4 22 0 143 17 5 141 3) 62 8 6 15 0 57 6 Uf Ze 1 sil 3 8 14 0 15 0 9 63 6 46 3 10 58 2 29 2, Wi 15 0 20 0 ale 12 0 13 0 TABLE 15. RELATION OF AGE TO RAPIDITY OF ACQUISITION OF LABYRINTH HABITS. Results in the table indicate the number of trials up to the point at which no errors occurred for at least two consecutive trials. Dancers 1-2 Montus OLp. | | Dancers 10 Montus Op. ] | Number of Results for | Results for | Nuniber of Results for Results for animal. Labyrinth D. | Labyrinth C. animal. Labyrinth D. | Labyrinth C. 256 as | 27 : 92 15 22 258 8 22 | 96 6 6 396 15 26 | | 98 6 19 398 16 14 120 9 12 418 13 | ai 166 | 10 — Avy. for Males. | ile) | 19.2 Ay. for Males. 9.2 14.7+ 179 4 | 16 91 19 8 181 19 9 93 15 13 255 10 18 97 6 11 263 6 7 99 8 19 395 13 — 109 | 10 a Av. for Ay. for Females. 10.4 1255 | Females. LG 11.6 Gen. Ay. 10.9 16.2 Gen. Avy. 10.4 13.0 The general results of the labyrinth experiments appear in Table 15. Averages are given for the sexes separately, inasmuch as so YERKES, Modifiability of Behavtor. 267 often heretofore we have discovered sex differences to be of import- ance for the interpretation of our results. Comparing the young dancers with the old, we note that the males of one to two months of age acquire both the labyrinth D and the labyrinth C habits considerably less quickly, as measured by the number of tests, than the ten-month individuals. In the case of the two groups of females there is practically no difference in rapidity of learning. The general averages likewise show that the old dancers are some- what superior to the young in ability to learn these labyrinth paths. What evidence we have, favors the conclusion that the associative memory of the dancer improves somewhat during the first year of life. Possibly change in the ‘‘apperceptive mass” is responsible. It is only fair to admit, in concluding the presentation of experi- mental data, that I consider this problem unsolved, for I have presented insufficient evidence to convince the critical observer that associative memory improves with age. VII. CONCLUSIONS AND SUMMARY. This attempt to discover the relation of age and sex of the dane- ing mouse to plasticity, or rapidity of habit formation, makes possible interesting and important, albeit not altogether favorable, comments upon the methods of the investigation. As the. work progressed it became increasingly clear that the use of cardboards as means of producing different degrees of illumination of the two boxes between which the mouse was forced to discriminate was unsatisfactory. Chief among the objections to this method may be mentioned the practical impossibility of keeping the difference in illumination constant throughout even a single series; the im- possibility of determining accurately, except by very elaborate and time-consuming methods, either the relative or the absolute illu- mination of the white and black boxes; the impossibility of chang- *Photometric determinations of the amount of light reflected by the white and the black cardboards which were used throughout these experiments indicate that the white cardboard reflected about 10.5 times as much light as the black cardboard. For a careful measurement of these values of the card- boards I am indebted to Professor J. W. Baird. 268 “fournal of Comparative Neurology and Psychology. ing the amount of difference in the illumination of the boxes with ease and accuracy. These are only a few of the objections to white, grey, and black cardboards or papers that experience enables me to raise. Naturally I shall neither use them nor recommend their use hereafter in investigations of the visual powers of animals. Later, in connection with a report on “Methods of studying vision in animals” which is to be made by the committee on standardiza- tion of tests of the American Psychological Association, I shall pro- pose a substitute method. The investigation has shown, I believe, the great importance of choosing conditions of experimentation which may be readily and accurately measured and controlled, and of determining, as a preliminary to any experimental study of habit-formation, the value for the individual animal of the several important factors in the experimental situation. It has further shown that we should work with individuals, and with relatively simple and perfectly analyz- able situations; and that no treatment of our results is so likely to hide their real significance as the averaging of groups of ob- servations for different individuals. Evidently the best prepara- tion for an experiment is a thoroughgoing study of the character- istics of each animal to be used in the investigation; and the best result which an experiment can yield is evidence of the relation of experimentally controlled conditions to the particular traits of an individual animal. Averages are important, but we should not sacrifice individual facts for the purpose of presenting them. The primary aim of the investigation, it will be remembered, was the discovery of the relation of plasticity to age and sex. The data prove that dancers at the age of one month acquire the white- black habit more rapidly than do older individuals, and that the females, on the average, acquire the habit more rapidly than do the males. Of great importance is the fact (represented by the curves of Fig. 3) that whereas the females make more mistakes of choice at first than do the males they very soon begin to choose with a higher degree of accuracy, and ultimately acquire a perfect habit with considerably fewer training tests than the males. That these age and sex differences in the form of the habit- YeRKES, Modifiability of Behavior. 269 formation curves are not necessarily indicative of general differ- ences in plasticity is rendered evident by the results of the sections on sensitiveness, strength of stimulus, and difficultness of discrim- ination. For in the light of these results we are able to name as two important, and to a certain extent independently variable, con- ditions upon which the acquisition of the visual habit depends, (a) ability to sense the difference in illumination of the two boxes and (b) ability to associate the darker box with the electric shock. Evidently an individual which possesses, highly developed white light vision and is capable of distinguishing very slight differences in illumination may, at the same time, possess little ability to asso- ciate stimuli. The data of section V indicate that what we have called age and sex differences in plasticity are in all probability, to be referred to differences in visual discriminating ability and in “‘asso- ciative memory.” Admitting that the results of the experiments justify only tentative conclusions, we may say that the young dancer seems to be somewhat superior to the old individual in ability to dis- eriminate on the basis of difference in illumination, whereas the old individual seems to associate stimuli somewhat more rapidly, if anything, than the young dancer. This suggestion, for it is scarcely more than that, of the way in which a habit may break up into two relatively independent factors is one of the most interesting results of the investigation. It strongly emphasizes the importance of studying the various sense factors separately and of attempting to discover upon what exter- nal and internal conditions “associative memory” depends. To sum up the results of the investigation point by point, if ap- pears that— 1. The dancer at one month of age acquires a particular white- black visual discrimination habit more rapidly than do older indi- viduals. From the first until the seventh month there is a steady and marked decrease in rapidity of habit-formation; from the seventh to the tenth month the direction of change is reversed. These statements hold for both sexes. 2. Young males acquire the habit more quickly than young fe- males, but between the ages of four and ten months (at least) the females acquire the habit the more quickly. 270 ‘fournal of Comparative Neurology and Psychology. 3. Curves of learning for the sexes indicate that the female makes more mistakes early in the training tests than does the male, but that this condition soon gives place to greater accuracy of choice on the part of the female. 4. Initial preference for the white or the black box does not seem to be a very important determinant of the rate of habit-forma- tion. 5. Tests of sensitiveness indicate that the male dancer is some- what more sensitive to electric stimuli than the female. There are no evidences of changes in sensitiveness with change in age. 6. The strength of the electric stimulus which is used as an incentive for habit-formation is extremely important as a determi- nant of rate of habit-formation. For a given animal and condition of visual discrimination there is a certain strength of stimulus which is most favorable for the acquisition of the habit (the optimal stimulus). 7. It is extremely important that experimenters discover the optimal stimulus for habit-formation. 8. For the dancer the following law appears to hold in connec- tion with the particular habit under consideration and for the electric stimulus. As difficultness of visual discrimination in- creases that strength of electric stimulus which 1s most favorable (the optimal) to habit-formation approaches the threshold. The easier the habit the stronger that stimulus which most quickly forces its acquisition; the more difficult the habit the weaker the stimulus which most quickly forces its acquisition. . 9. A given difference in the illumination of the boxes which are to be discriminated cannot be detected with equal ease by old and young dancers. When the difference in illumination is slight the young individuals detect it more readily than the old mice; when the difference is great, the old individuals apparently detect it as readily as do the young mice. 10. The capacity for “associative memory” is greater, if any- thing, in the dancer of ten months of age than in the one-month individual. This is indicated by certain results of the visual dis- crimination experiments and by the results of labyrinth tests. YERKES, Modiftability of Behavior. 271 11. The results of this investigation indicate, then, that the ac- quisition of a visual discrimination habit depends upon two inde- pendently variable (within limits) capacities in the dancer: (1) the power to detect differences in illumination or to gain this power (educability of white-light vision), and (2) the power to associate the darker box with the electric shock (associative memory). The former of these capacities seems to be greater in the young than in the old dancer; the latter seems to be somewhat greater in the old than in the young individual. 12. Should the statements just made hold true for animals gen- erally, it is evidently important that the senses be trained early in life and that the development of associative memory be furthered later. Investigation of the problems suggested by our results should vield important practical data for the science of education. THE REACTIONS OF THE DOGFISH TO CHEMICAL STIMULI. BY RALPH BDWARD SHELDON. Contribution from the Woods Hole Laboratory of the United States Bureau of Fisheries.* WitH THREE FIGURES. TABLE OF CONTENTS. PAGE 1HayE Rove hb KCL Val namin taeate tots erie NO OaIe AO oOo too cio UD dob Dae taco ocd.c ooo as 273 CondiWonsPrOL.expPerimyESmlasLOy a atecs pees siete eieeeis eter econ Reenter Porat 276 IRGRKE NOMS COMUNNCO Go obcuen done oon og biah dooctb opp ceanoboasconos2o0€ 278 Sensitiveness to chemical stimuli. BESPELINTVEMGA] ESM i 5oe ooh. s. wl enssaces siete onset femasser oe) ey ous fe wake) wien sh ene en 281 ENTVAIIV SIS OL PES UES. «ors wo Sus coseign Sims. sion Sees his road urele totocer aoeus nen CRS Cue] Raa CRORE PEER 281 Operations. DESHRUCHOM Of theaspinalOcOrd, sss dacs ce ecu ere ick errno 288 ANNE HO OIE Tovey RjowTNMIGORO Geoe dn ubocodcoscnconoucosdaneusmacaa 289 IMMervanoneand Teachons of the) MOSES a-s soe ae eerie 291 Ghemicalssensationy as sa ssense=quality...2 ...)- ars ce cases eee citreus eirarenore none 294 (COTE TTESNOTORSE Wc. cintees Oc Se RICE Oe oae nce tear On sont. cle tiensscne 5 Gb d do dicid op aoe 294 SSUUMMMIT TENT fat syei role bese ye ke sect sieod atts ls apesh wig + guineas olane oe lone) opal Htetiaisnet ac tene feet 297 SOMOS ENT ys ge ecbeasets ve wseas ele tes es ceiee 4 cuss faye su toitaseieace sstauepeh oMeaehc homemade omete aural metas 299 JENSEN x Ae SPS Sh Ouch eRe CRG AEG Rh ER EMC ee AE REECE LOM ae Hoary chore er ot ence ity eens cites oo 308 In TRODUCTORY. The smooth dogfish, Mustelus canis (Mitchell), was the subject of experiment in an endeavor to find out the sensitiveness of the gen- eral body surface to chemical stimuli and the extent to which the nerves of general sensation share in the reactions called forth through stimulation of the mouth and nostrils. In the investigation of these problems certain accessory points, such as the spinal animal and the innervation of the olfactory capsules, are taken up. The substances *Published by permission of George M. Bowers, U. 8S. Commissioner of Fisheries. THE JOURNAL OF COMPARATIVE NEUROLOGY AND PSYCHOLOGY.—VOL. XIX, No. 3. 274 fournal of Comparative Neurology and Psychology. used as stimuli were those known to affect the gustatory, and to a less extent the olfactory sense, in higher forms. The work was-carried out at the laboratory of the U. S. Bureau of Fisheries at Woods Hole. I wish to express to the Director, Dr. F. B. Sumner, my appreciation of his assistance in furnishing me with every facility necessary for the successful prosecution of the work. The subject was originally taken up in 1907 under the direc- tion of Dr. G. H. Parker in the Zoological Laboratory at Harvard University. I desire to tender to him my thanks for many helpful suggestions made both at that time and at the Woods Hole laboratory. During the last few years much has been done, particularly among the invertebrates, on the reactions of animals to different kinds of stimuli. Part of this work has been concerned with chemical stimula- tion, the character of which is shown by the work of Pearl (’03), Bell (706) and Jennings (704 and’06). So far as the vertebrates are con- cerned, work on chemical stimulation has dealt almost exclusively with their two chief organs of chemical sense, smell and taste. A serious attempt has been made, however, to determine for the organs and their functions a physico-chemical basis. Haycraft (87) was one of the first to attempt seriously to deal with taste from a strictly chemical standpoint. He was followed shortly by Corin (88). The most im- portant work of this character appeared about a decade ago from sev- eral sources simultaneously. Overton (’97) considered osmosis, while Kahlenberg (’98, ’00), Richards (’98, ’00), Kastle (’98) and Hoéber and Kiesow (’98) have taken up critically the physical and chemical characters of substances which stimulate the gustatory ap- paratus in man, together with the chemistry of taste itself. Still more recently Herlitzka (07) and particularly Sternberg, in a series of papers published from 1898 to 1906, have made a detailed study of the chemical basis of sweet, sour, salty, bitter, metallic, electrical and alkaline tastes. Many other writers have made a physiological study of the action of this same series of chemical substances on the gustatory apparatus of man. This includes the work of Kiesow (94b), Haycroft (00a), Hinig (’01), Nagel (05) and Lemberger (’08), together with nu- merous others, practically all of whom, however, consider in this con- SHELDON, Reactions to Chemical Stimult. 27.5 mection only the taste buds and associated nerves. Other authors have argued that, in addition to these structures, the nerves of general sensation take part in the sense of taste in man. Such a view is sup- ported by the work of Camerer (’70), von Vintschgau (’79b), von Anrep (780), Adduceo and Mosso (’86), Hooper (’87), Berthold (’88), Oehrwall (91), Shore (’92), Kiesow (’94a), Vinci (’97, 799), Fontane (’02), Ferrari (04) and Herlitzka (’07). Extensive work has also been done on the sense of smell in man from the physiological, and to a less extent the chemical viewpoint, as may be seen by consulting bibliographies such as that given by Zwaardemaker (795) and Bawden (’01). There will be no general consideration of the subject here, as it does not bear directly on the problem at hand. It is to be noted, however, that in connection with the olfactory organ as well as the gustatory, the free nerve endings take part in the reactions secured. Physiological evidence is noted by Haycraft (00b), while the presence of such terminations has been demonstrated by a number of writers from Grassi and Castronovo in 1889 to Read (1908). In spite of the evidence presented by these authors, outside the single work of Griitzner (’94), little has been done on mammals to- ward a study of the reactions of the free nerve termini gener- ally to chemical stimuli. This has been due partly to precon- ceived ideas on chemical sense and partly to the feeling that nothing is to be gained by a general study of the chemical sense among the vertebrates,—even to the extent of including smell and taste under the same category. Such is the view of Zwaardemaker, who says (03), “Bei den Wirbelthieren jedoch sind Geruchs- und Ge- schmackssinn in vieler Hinsicht so grundverschieden, dass es meines Erachtens keine Empfehlung verdient, sie zusammen zu behandeln.” On the chemical senses of the lower vertebrates little has been done. Bateson (’90a, ’90b) discussed the senses of smell, taste, and touch in several fishes. The work is of little value in the present connection. In 1894 Nagel published his great monograph on smell and taste. Nagel repeatedly uses the term chemical sense, always meaning, however, the combined organs of taste and smell and not a general chemical sense. He stimulated selachians, teleosts, and am- 276 “fournal of Com parative Neurology and Psychology. phibia by means of solutions of different chemicals. The selachians were very sensitive to weak stimuli, reacting to dilute solutions of vanillin all over the body, but to quinine only about the head. Bar- bus failed to react to salty, sweet, bitter, or sour substances on the general body surface, while Gusterostend reacted to quinine only about the head. With Cobitus and Gobius he obtained reactions with meat juice and sugar solutions. Lophius was sensitive to chemical stimuli over the entire skin. Triton, the only amphibian tested, re- acted on stimulation of the head only. It is evident from later work, particularly that of Herrick (02, ’03c) and Parker (707, ’08a, ’O8b), that Nagel’s results did not permit the drawing of sound conclusions, partly because of the substances used as stimuli, and partly because he failed to differentiate between fishes with taste buds on the outer surface and those lacking such structures with their visceral sensory innervation. Herrick ((02) was concerned almost exclusively with the sense of taste, in a narrow sense, that is, the reactions to sapid solutions through stimulation of the taste buds. He performed a few experiments of a general chemical nature, insufficient, however, to per- mit any conclusions. Almost the only work on vertebrates which takes up the reactions of the nerves of general sensation to chemical stimuli is that of Parker (’07, 08a, 08) on Amphioxus and Ameiu- rus. His results will be ee later. Considering that a general chemical sense is probably more primi- tive in phylogeny than taste and smell and that a careful study of such a general sense may do much to make clearer the development of these two senses, as well as their physiology, it is strange that so little has been done on this topic. This is especially true in the lower vertebrates, where, in many eases, it is difficult to separate the re- actions due to stimulation of the organs of taste and smell from those due to the nerves of general sensation. Conpirions OF EXPERIMENTATION. The substances used in this work were hydrochloric, nitric, and sulphuric acids for acid stimuli; sodium, ammonium, and hthium chlorides for saline stimuli; sodium hydroxide for alkaline; cane sugar, dextrose, saccharine, and its carbonate for sweet; and quinine SHELDON, Reactions to Chemical Stimult. Qi hydrochloride, picric acid, ammonium and sodium picrates for bitter. All were made up in distilled water on the basis of the gram-molecu- lar solution. The inorganic acids were prepared as normal solutions, titrated against an alkali of known strength for accuracy. The other solutions were made up by weight, the concentration first used as a test depending partly on the solubility of the chemical used. The chlorides were prepared as 5n solutions, the sugars 3n, sodium hy- droxide as n, saccharine n/6, quinine hydrochloride n/10, picric acid and its salts n/15. In the experimental work all of these solutions were gradually diluted until the limit of reaction was reached. Sut- ficient time was given between tests at different degrees of concen- tration and with different substances to eliminate after-etfect. A large number of dogfishes were used in the experiments in order to rule out individual variation. Most of the normal fishes used were those caught in the fish traps and placed shortly in tanks about a meter and a half long, two-thirds of a meter wide and a third of a meter deep. A current of sea water was kept constantly running through the tanks. . After a few days in these tanks the fishes could be handled with little difficulty. For most of the work, individual adult dogfishes were removed and placed in a smaller trough about eighty cm. long, thirty em. wide and fifteen em. deep, through which a strong current of sea water was flowing. After a little handling the animals would lie quietly in this trough either on the dorsum or ven- ter, submitting to a certain amount of manipulation. In cases where it was necessary to have part of the animal out of water or where the fish was very unruly it was fastened to a frame. Application of the Stimulus.—The solutions were applied by means of a pipette and were, in most cases, ejected slowly with the tip ot the pipette about two millimeters from the skin of the fish. In such cases the fish was completely covered with water. Where it was es- sential that the region stimulated should be out of water absorbent cotton saturated with the solution was usually apphed. Occasionally, however, the solution was ejected directly against the skin and the time of the tactile reaction taken, after which the slower chemical re- action could usually be identified. In stimulation of the mouth or nasal capsules a metal guard, closely fitted to the snout, was placed 278 ‘Fournal of Comparative Neurology and Psychology. between the two sets of apertures preventing diffusion of the stimulus | from one set to the other. Regions Tested.—With each solution used, approximately fifty places on the body were tested at each concentration used, and the time of reaction always recorded with a stop watch. These regions included the mouth, nostrils, spiracles, anus, claspers, and selected places on the fins, dorsal, lateral, and ventral surfaces of the fish. For the location of these points see figures 1 and 2. REACTIONS OBTAINED. The reactions obtained varied according to the part of the body stimulated, as follows: Stimulation of the mouth or spiracles is fol- lowed by one or more violent gulps, accompanied, of course, by a quick ejection of water through the branchial openings. A more rapid respiration for a greater or less length of time, depending on the stimulus, follows this. This is the only reaction secured by chemi- cal stimulation of the mouth and spiracles, and it is secured by stimu- lus of no other region. When the nostrils are stimulated by any of the substances used, the reaction is likewise very characteristic. It consists essentially of a very quick jerk of the head. This reaction is likewise secured by stimulation of no other region. In the case of the paired fins the characteristic reaction is a quick movement of the fins, usually of a vibratory type. Often, particularly if the stimulus is weak, the first reaction is a turning or movement of the whole fin toward the stimulus, occasionally away from it, followed usually by vibration of the fin. With the median fins the reactions are very similar. ‘The more usual reaction, however, begins with the movement of the fin toward the stimulus. Often the small caudal finlets of the median dorsals and the anal fin will react by a rapid vibration, even though that part of the fin is not stimulated. If this reaction occurs, it usually begins by a quick movement of the finlet toward the stimu- lus, more rarely away from it. When the finlet of either the anal or second dorsal fin takes part in the reaction, the other does also, so that the action of the two is simultaneous and in the same direction. When the caudal fin is stimulated, the reaction consists in a rather slow sidewise movement of the tail either toward, or away from, the SHELDON, Reactions to Chemical Stimult. 279 stimulus. This is evidently the beginning of a swimming movement. Stimulation of the anus results usually in bending over ventrad of the pelvie fins. Occasionally the fins react alternately in an attempt to turn the body over. If the. stimulus is strong or long continued, these reactions are followed by a lateral squirming of this part of the fish enlminating in the swimming away of the animal. Stimulation of the claspers results in a quick lateral movement and vibration of the structures. The head responds by a rather slow movement away from the solution. In general, stimulation of the dorsal, lateral, and ventral surfaces, other than those already mentioned, results in a movement of the fish which is very evidently a part of the general swimming movement. In fact, stimulus of almost any region of the fins or body, if persisted in, will transfer the local reaction to one which forms part of the swimming movement of the animal. ‘This is shown especially in the case of stimulation of the fins or lateral surfaces. If the caudal, second dorsal, or anal fin is stimulated and the reaction is toward the stimulus, for instance, there will often be a movement of the first dorsal fin but in the opposite direction. The same relation holds true if the first dorsal fin is stimulated. Often a reaction of all the fins is secured. For example, there will be a movement toward the stimulus by the caudal, anal, and second dorsal fins, a movement away from the stimulated side by the first dorsal fin, an upward movement of the paired fins on the side stimulated and a downward movement on the opposite side. This reaction was first pointed out to me by Dr. Parker as a response secured by tactile stimulation of the same regions. Such reactions are unquestionably part of the general swimming movements of the fish, as may be seen by observing the animal in an aquarium. As caused by chemical stimuli, they are evidently of a kind to preserve the fish from injury, enabling it to remove itself from an injurious environment. Very similar reactions are secured by stimulation of the sides of the body and tail. In general it can be said that a slight stimulus ealls forth a local response, while a stronger or longer-continued stimulus almost invariably results either in a new reaction which is part of the swim- ming movement, or else in a gradual change of the local reaction into such a part of the swimming movement. The former occurs where 280 “fournal of Comparative Neurology and Psychology. the loeal reaction differs decidedly from the swimming movements, as in the case of stimulation of the mouth, while the latter holds in cases where the two are similar, as in the reactions due to stimula- tion of the fins. Certain interesting special cases are to be noted. If the dorsum or side of the fish near the small finlets of the dorsal or anal fins be stimulated, a quick movement of the finlets toward the side stimulated, usually followed by vibration of the finlet, occurs. This may be a reaction to remove an irritant, as is noted in the case of the frog when a drop of acetic acid is placed on the skin, or it may be part of the swimming movement, as seems more probable. Evi- dence against the former interpretation is offered by results which Parker obtained by tactile stimulation. He found that tactile stimu- lation of the dorsum near the finlet of the second dorsal fin caused this reaction; but he also found that if he now stimulated a point be- tween the finlet and the mid-dorsal line on the same side the finlet continued to wipe the skin, but ventrad of the point now stimulated. It would, therefore, appear that the reaction is called forth by stimu- lation of any part of the side in this region and is not a local response to remove an irritant. It might be argued, however, that the power of localization is not well developed in this form. The strongest evi- dence in favor of the second interpretation is that when the skin be- side both dorsal finlets is stimulated on the same side at the same time one turns to one side and one to another, as is the case when the animal is swimming. This reaction of the finlets was one of the most dcheate found. Reactions could be secured by stimulation of the skin beside the second dorsal finlet when all the remainder of the body was insensitive. If the claspers are turned to one side and the venter underneath stimulated, a quick vibration of the claspers over this point follows. This is probably part of the general swimming movement also, al- though its constancy and accuracy suggest the wiping reaction. In the case of the pectorals, however, when the ventral surface between them is stimulated there follows a quick seissors-like action of the two fins over the point stimulated. If one fin is held, the reaction takes place with the other alone. This reaction is not a part of the general swimming movement, is yery consistent and accurate, and apparently is of the same character as the wiping reaction of the frog. It is SHELDON, Reactions to Chemical Stomult. 28t probably purposeful only in the sense that such a reaction is of a gen- eral preservative character such as is discussed by Sherrington (706), This reaction often occurs, also, when the pectoral itself is stimulated, particularly on its median margin. SENSITIVENESs To Curmican SriMuLt. deperimental Results.—The least stimulus which will cause a re- action, the comparative sensitiveness of different parts of the body, and the time of reaction for the different substances used are shown in the tables. The data were obtained under the following conditions. Several animals were always used for the tests and the figures given are based on results obtained from two or more individuals. From three to five tests were made at each point stimulated, with each solu- tion used, and at each different degree of concentration of that solu- tion. When individuals were used as controls, however, fewer tests were made if these demonstrated that the reactions were in conformity with those first obtained. Before the solutions were applied, both distilled and sea water were used to make certain that no reaction would result from their use, exclusive of the test solution. So much variation in the reaction time between different individauls was oted that both upper and lower limit in seconds are stated for each point tested. These limits differ considerably in many eases, yet it is easy to see that there is a general difference in the reaction time for dif- ferent regions of the body and for different degrees of concentration of the solutions. J*or all tests on the dorsal or lateral surfaces, the fish lay on the venter; while for experiments on the ventral surface, it lay on the dorsal or dorso-lateral aspect. About ten of the points stimulated are omitted from the tables. Analysis of Results.—It will be noted that the same reactions are secured by the use of any of the inorganic acids used as stimuli, that is, the reactions are due to the hydrogen ions. The reactions to the acids in the more concentrated solutions are very strong and definite. In nearly every part of the body they take place as quickly as me- chanical conditions will permit, that is to say, almost instantaneously. With the decrease in concentration the reaction time becomes a trifle longer. Practically the entire body is sensitive to n/20 acid, the head 282 “fournal of Comparative Neurology and Psychology. PrcTrORALS. | PECTORALS. Basr. MARGIN. PrELvics Subs. Base. Mth. | Spir. | Nost. | Anus. PreLvics. MARGIN. and Cone. Dors. | Vent. | Dors. | ia Ve 1 Dors. Vent. Vent. 5 | Dors. 6 2 10 bo nt. 2 HCl, | | HNOs, | | | Ho2SO,, | n/1 n/10 n/20 n/40 n/50 n/75 n/100 Dore | = vo i anes rl ral eB a OO ) erieor) * *Nhe ! | Ou He O22 2 Wo GO HH bohobo bobo | * 1 AoOonnt co oS * HPI b LO So * #NNwry | o: — =) SCNNNNNe CNaAWhNh a NaOH, n/1 n/5 n/10 n/20 n/30 n/40 So *¥to bo on #dr bh ~~] n/120 15-37 Am. and Sod. | pier. n/15 0 0 n/30 Cane sug. and dext. 3n SHELDON, Reactions to Chemical Stimult. Dorsat. Sreconp Dorsat. ANAL CAUDAL. Heap. SNour. Dors. ere us a ‘Dors. or, oa / Vent Zi | Vent. wi | oa 7 tip. Finl. | Base. tip. Finl. | Base. | tip. Finl. | Base. | marg.| Tip. | Dors. | Vent. | Dors. Vent. 14 15 16 17 18 19 20 21 22) be Zo: 24 25 | 26 27 28 | lier | Tae | | | | | | | | 1-4 1-3) | 2-4 2-4 1-3 1-3 1-3 2-3 1-3 | 2-4 1-4 4-8 3-4 | 7-8 3-6 4-5 3-5 6-8 5-6 2-6 6-8 2-4 4-7 | 5-7 | o-t 0 5-7 5-0 7-0 3-8 4-8 3-8 6-0 5-9 3-8 10-12} 4-10; 4-11; 6-9 3-12; 0 | 5-8 0 |) 70) 0 4-0 | * 0 * * x ee 6-0 0 * 0 0 0 | O 0 0) 4) =0 ) * 0 0 * 0 0 ) 0) | O ) 1) 0 0 1 @ 0 0 0 0 0 0 0 0 10) 0 0 0) | | 5-0 | 0 5 | | 0 0 4-14 14-18! 7-14 4-12} 8-16} 5-7 7-9 6-14 9-16} 8-14! 6-0 5-0 0 0 8-13 10-0 | 8-13) 8-18] 12-13 6-10) 5-9 0 7-0 0 0 0 0 0 * * 0 * * 0) 0 0 0 0) 10) 0 0 0 | O 0 0 0 0 0 0 0 0 0 10) 0 | 0 | | | | | | 2 12-36 12-20} 12-19} 9-19] 8-16} 8-21} 6-24] 6-15] 11-23| 15-22} 8-25) 4-0 7-13| O 7-0 6-36 | 12-0 13-25, 19-0 6-32} 7-14) 18-20] 6-19 0 10-21) O 7-0 0) 0 0 LO) 0 |} O 0. 10) 0 0 0 0 0 0 0 0 0 0 | Oo 0 0 0 0 0 0 0 0 0 0 0 Gre 0 | (537 | TWP! O22e7|| alg. ale) See) |) Taka!) Gals 12-16] 8-22) 9-35/ 14 42 0 if ieee ag ft ft ft t it tt t 0 0 0 | 0 0 0 0 0 0 0) 0 0 10) 0 0 10) 0 0 | | se | | | i | | 0 0) | 0 0 0) 0 0 0 0 | O 0 i) | 0 0) 0 | | x | x ois x x x x 5s x | x x x x eg x = Es je | ike 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 * | | ieee | | 0 0 0 0 0 0 0 0 0 1 @) 0 | 0 0 0 0 284 ‘fournal of Comparative Neurology and Psychology. Dorsat Asprecr LatreraL ASPECT. VENTRAL ASPECT. Subs; |— SS —= = = = and Ceph. | 1D | Betw.| 2D | Vent.) Mid- | Anal.) Tail. | Betw.| Mid- | Clas- | Dors. | Caud Cone. | of 1D) Finl. | Dors. | Finl. | of 1D! dle. Finl. | pect. | dle. | pers. | of cl. | of el 29 30 | 31 32 33 34 35 36 37 38 | 39 40 41 - =| | HCl, | | HNO; H.SO, 2-3 1-4 1-3 1-2 1-3 2 n/1 1-4 1-4 3-6 1-2 1-4 2-3 1-2 2-7 3-6 a) 2-4 2-6 n/10 | 3-5 2-3 4-8 2-4 3-7 2-7 1-2 2-6 SI) 3-8 2-7 2-8 n/20 3-5 3-4 4-8 2--6 3-4 2-6 * 0 * * * * 0 n/40 | 4-0 | * 0 * 0 * * 17*0 | 0 0 0 | 9#78| 0 n/50 | O * 18*22 18*0 0 * 3 0 0 0 0 * 0 n/75 0 | 18*30 | 0 18*30) O 0 | n/100, 0 0 | NaOH | | n/1 | | | | n/5 2 | Baloo : n/10 6-9 4-10! 6-9 6-11| 8-17) 8-17) 5-8 5=8) 4-0) 0 5-15 ' 10-14 n/20 | 0 9-0 | 0 4-13 | 0 0 0 0 | n/30 0 * 0 * fe W) 0 aS 10) 0 0 | es * | = n/40 | 0 0 0 cr SPO 0 * 0 0 (ies ic 3 a ae n/50 | * | * 0 | 3% * | * n/60 | * 0 | eo alle a a) n/70 | | * | 0 0 lem | | NHC | | | | | 5n 0 | | 12-21| 5-24) 7-14) 15 7-26; 5-12| 7-27] 13-27) 5-14] 7-8 | 20-37 2n | 25-0 | 7-0 | 25-0 5-32 * 10-15] 11-0 6-16 | 10-22 8-18 n/1 | 0 OKO * 0 0 + 0 0 Oy Ome ee Quin. | h.chl. | | | n/10 | 6 0 0 0 0 0 0 0 0 0 0 0 | O n/20 | | p/30 | | | n/40 | = — i} Picr. | acid. | n/15 6-35.) (5-8 10-12} 5-8 0 2-10 | 6-10} 7-10) 8-11 3-10 12-15 n/30 | 20-40 | 20 | 20 20 it 23-29 | ft 9-27 if 12—42 | 22-45 n/60 | O * 0 * 0 0 * 0 0 0 = JP a3 * n/120 0 0 OF ail Ono 0 Am. and | | | Sod. lines | | } pier. | | n/15 0 10*20 | O- | 10*20| O 0 | 10*20! O 15*30! 0 0 | 15*20) O n/30 0 | 0 | O | } | 0 Sacch. n/6 | x: x x X. x x: x aX: x x x 3x x Carb. | of | Sacch. ; | | n/6 0 0 0 0 0) 0 0 0 po) 0 0 0 0 Cane | | | sug. and | | | dext. | | sn | O 0 10 0 0 0 1 0 0 0 0 0 0 | 0 SHELDON, Reactions to Chemical Sitmult. 285 SIGNS AND ABBREVIATIONS USED IN THE TABLES. Numbers at the heads of columns refer to the points stimulated as shown on Figs. 1 and 2. Other numbers indicate the reaction time in seconds. — is equivalent to the word to. * signifies that a reaction was secured with the region stimulated out of water, according to some of the methods already described. The reaction time in such cases was rarely taken. The exceptions are indicated by the replacement of — by *. ~ refers to cases in which the reaction was secured with the region tested under water, the reaction time not being taken. This sign usually signifies that the reaction was very weak and incomplete. x is used in one case where the reaction was very strong and definite, but where the reaction time was not taken. 0 indicates that no reaction could be secured with the fish, either under or out of water. When used after another number it signifies that no reactions were secured in two or three of the five tests made. Where blank spaces occur no tests were made. Am. and Sod. picr., ammonium and sodium picrates; Betw., between ; Cane sug. and dext., cane sugar and dextrose; Carb. of Sacch., the carbonate of benzylsulphonic amide formed by the neutralization of saccharine by sodium carbonate; Caud., caudal or caudad; Ceph., cephalad; cl., claspers; cone., concentration; Dors., dorsal or dorsals; /finl., finlet, caudal prolonga- tion of anal and dorsal fins; marg., margin; Mth., mouth; Nost., nostrils; pect., pectorals; Picr. acid, picrie acid, trinitrophenol; Quin. h.chl. quinine hydrochloride; Spir., spiracle; Subs., substances used as stimuli; Vent., ven- tral or ventrad; 1D, first dorsal fin; 2D, second dorsal fin. being the only part at all insensitive. At n/40 the body surface gen- erally reacts, although the reactions are less definite, particularly on stimulation of much of the dorsal surface. At 1/50 the mouth, spiracle, anus, nostrils, fins, claspers, and side are still sensitive, as is also the dorsal surface beside the finlets. Stimulation of the mouth, spiracle, claspers and skin beside the finlets by n/75 still calls forth a reaction. At n/100 no reaction could be obtained, although the fish in many cases seemed to perceive the stimulus. In all this work many observatious were made which indicate that the dogfish actually feels stimuli to which it does not react to an appreciable extent. One probably comes to a point in decreasing the concentration of the solu- tions where the stimulus is perceived yet not sufficiently strong to cause action of any kind on the part of the animal. All of the re- actions secured at n/40 or less were weak, although usually definite. The reactions to acids in general are characterized by their quickness 286 “fournal of Comparative Neurology and Psychology. and definiteness. There are rarely premonitory symptoms of any kind before the reaction takes place, even though the reaction time is long. In summary, it is evident that the dogfish is sensitive to acids of a solution of n/75, both in the mouth and spiracle in which are found taste buds, and also on the outer body surface when no such structures are found. NaOlIl is the only hydroxide given in the table, although experi- ments were made with KOH which indicate that essentially similar reactions would be secured by its use. To NaOH. the dogfish reacts quickly and definitely in the stronger solutions, although not quite so quickly as to acids. It will be noted likewise that the animal is not sensitive generally to so dilute solutions. Reactions are secured from the general body surface, however, to a solution of n/70. It is of special importance to note that the mouth and spiracle are almost in- sensitive to alkalis except im very strong concentration. Some shght chemical reactions take place between the hydroxide and the sea water, as was the case with the acids used also. This probably renders the solutions less powerful than they would otherwise be. | To salts the reactions are slower than to acids and alkalis. The responses to both lithium and ammonium chlorides are practically the same. No reactions to sodium chloride could, however, be obtained. This is due, doubtless, to its presence in such quantities in the sea water. The reactions to the chlorides are usually preceded by pre- mnonitory symptoms a few seconds before the definite reaction occurs. These consist of a local or general uneasiness. The reactions are also often prolonged, continuing for a few seconds after the stimulus is removed. It will be noted here, as in the case of the alkali, that the general body surface is more sensitive than the mouth. Definite re- actions are secured from the former to a 2n solution, while the fish shows a sensitiveness to a normal solution. The mouth and spiracle, however, react very weakly to solutions as strong as 5n, and almost never to a lesser degree of concentration. The general lack of effect of the solutions of a weaker grade than normal is probably due to the fact that the salts of the sea water make it about a 5/8 normal solution. This interpretation is supported by the results when NaCl was used and also by the tests made by Parker (O08b) on the fresh SHELDON, Reactions to Chemical Stimult. 287 water catfish (Aimeiurus), which is quite sensitive to salts. The re- actions secured by the use of salts are not due to osmosis, as sugar solutions of equal osmotic strength have no effect. Reactions to quinine hydrochloride take place only in the mouth and spiracles, that is, reactions to quinine take place only on stimu- lation of surfaces bearing taste buds, as Parker found for the catfish. The dogfish is extremely sensitive to pieric acid. In strong solution it is extremely distasteful and the animal responds vigorously. The response is slow, however, and there are usually premonitory symp- toms before the reaction, as noted for salts. Reactions are also often prolonged after the stimulus is removed. The mouth and body sur- face are sensitive to n/60 unquestionably, while an apparent uneasi- ness of the fish seems to indicate that it feels in the mouth, spiracles and nostrils a still greater degree of dilution. Picric acid was used by Parker (07, ’08b) as a bitter stimulus. It was with this idea in mind that the substance was used on the dogfish. It might be argued, nevertheless, that inasmuch as aquatic forms are very sensitive to acid stimuli, the reaction in this case is due to the acid radical in the trinitrophenol rather than to the base which gives to us the bitter taste of picric acid. To test this point neutral ammonium and sodium picrates were used. Both of these are as bitter to the human taste as is pieric acid. From the tables it will be seen that the fish is by no ineans as sensitive to these as it is to the picrie acid, although weak reactions can be secured by the use of strong solutions. The results show that it can not be assumed that the stimulus of picric acid rests entirely or even largely with the base, but that the acid radical is probably responsible almost entirely for its influence on fishes. It is evident, however, that the base does possess a stimulating power, as shght reactions were secured to the neutral picrates. It ean be stated, therefore, that the dogfish reacts to substances which give us a bitter taste. To stimuli of this character the mouth is more sensitive than is the remainder of the body. On the whole, the fish seems less sensitive to bitter substances than to the other kinds of stimuli used. Outside of picric acid the animal showed little distaste for the solu- tions used, even though reactions were secured. No reaction at all could be obtained to sugars. This holds true 288 “fournal of Comparative Neurology and Psychology. for all aquatic vertebrates, due probably to the fact that sweet sub- stances are substantially unknown to aquatic life. To saccharine or benzylsulphonie amide very quick and definite reactions were always secured. Thinking that this reaction was probably due to the acid radical, the saccharine was neutralized by sodium carbonate, the re- sulting product being as sweet to man as is the saccharine. No re- actions at all were then secured, proving that those first obtained were due to the acid radical. Comparing different regions of the body as to sensitiveness, it will be noted that the head is least sensitive, while the mouth, nostrils, paired fins, particularly the pelvic, the anal and dorsal, especially the second dorsal, are more sensitive. Areas of skin closely asso- elated with the dorsal and anal finlets are included with these fins. OPERATIONS. Some operations were next performed to find out what part of the nervous system takes part in these responses. After operations no fishes were subjected to experimentation until at least twenty-four hours had elapsed. Narecotization was accomplished by a mixture of ether and water from the effects of which the animals did not ap- pear to suffer in any way. Destruction of the Spinal Cord.—In this experiment the tail was cut off, the caudal artery and vein plugged with cotton and the cord entirely destroyed as far cephalad as the cephalic margin of the first dorsal fin by means of a small steel wire. This method was suggested to me by Dr. Parker, who had used it with much success. By this operation the peripheral innervation of all of the caudal part and middle of the body is destroyed, except that from the lateral Hne nerve. The individuals subjected to this operation lie in the water perfectly motionless so far as the caudal part of the body is concerned, occasionally trying feebly to swim by means of the pectoral fins. The skin caudally gradually turned white as is the case with dead dogfish. Such fishes lived, however, for some weeks. As was to be expected, no reactions could be obtained chemically either by stimulation of the general body surface or the lateral line. Parker (705) showed that the function of the lateral line is to respond to slow mass movements ’ SHELDON, Reactions to Chemical Stimutlt. 289 of water. The head is sensitive to chemical stimuli after the opera- tion as before. Section of the Spinal Cord—The cut was made about two centi- meters cephalad of the first dorsal fin. After the operation the fishes lived for weeks, differing from the normal individuals in their ordi- nary actions only in the fact that the caudal part of the body kept up a constant swimming. The fish would be propelled to the side of the tank, where it would remain for hours keeping up a vigorous swimn- ining movement of all the body caudad of the cut. This was never observed to cease until the death of the animal. Bethe (99) noted this same swimming after section of the cord in the dogfish, but found that there were intervals of cessation. When this fish was tested with the solutions used on the normal fish, it was found that the caudal part of the body was more sensitive than before. The reactions take place from a fraction of a second to a few seconds quicker than before, the reactions seemed more positive and definite and some of the fishes studied reacted to slightly weaker solution. The reactions of the head were practically normal. After most of the operations many of the fishes failed to react to quite so weak a solution as before, due, prob- ably, to the shock of the operation or to diminished vitality. The reactions of the spinal fish are almost never long continued after stimulation with salts or picric acid, contrary to the condition in the normal animal. Several points are involved in these experiments. These are, how does the spinal fish differ from the normal and why? Is the observed difference due to a lack of inhibition by the brain, the use of the new paths for the reflexes, or stimulation of the cut ends at the cephalic end of the cord? Taking up the first: Danilewski (792) on Amphi- oxus found that ‘voluntary’ movements ceased with removal of the rostral part, the only action recorded taking place in response to definite stimuli, He concluded that the centers for movement lie in the rostral part of the animal. Steiner (88) worked on the lamprey, selachians, ganoids, and teleosts. In the lamprey he found no volun- tary movement of the caudal part after section, while the other fishes acted about as usual after section of the cord except for the dead weight of the head. Steiner believed that the centers for equilibrium 290 “fournal of Comparative Neurology and Psychology. and voluntary motion lie in the cord. Loeb (91) found that a dog- fish with the cord cut was no longer able to orient itself, even though it could swim off readily. Bethe (99) observed no difficulty in the swimming movements of the spinal dogfish. Pfliiger (55) and Bickel (97) took up the study of the spinal eel. The former found that the eel had no difficulty in maintaining its equilibrium and could swim readily, while the latter says that the spinal eel is incapable of remaining in the normal position and can swim only forward. The normal individual can swim in both directions. The spinal, frog has been studied with great care. Pfliiger (53) found that the spinal frog still retained its musele tonus and could spring; Mendelssohn °83, 785) found the reflexes good after section of the cord; Steiner (785) observed no change in the reactions; Schrader (’87) obtained similar results. Moore and Oertel (°99): found, however, that the reflexes were increased, although fatigue occurred quickly. Babak (03,05, ’07) found that section of the cord in the larval frog resulted in no loss of power in the spinal part. In regard to mammals, particularly man, there is a mass of experimental and clinical evi- dence on the results of the separation of the spinal cord from the brain. This is presented mainly by Rosenthal (75, ’?84); Bastian (90); Burns (793, 796); Rosenthal and Mendelssohn (97) ; Senator (98); Brauer (99); Moore and Oertel (99); Walton (02); Wal- ton and Paul (’06), Sherrington (97, ’?00b, 705, ’?0Ga, ?06b); and Sherrington and Laslett (03). In mammals, as observed by most in- vestigators, the loss of reflex power is very great, often practically to- tal after a break in the functional continuity of the cord and brain. It is evident, therefore, as is emphasized by Sherrington, Walton, and Walton and Paul, that in the ascent of the animal scale the cord loses. its power as an automatic center and becomes less and less capable of responding to stimul. Tn the fishes under observation in these experiments no lack of the power of equilibrium was observed. This may easily be due to the fact that the innervation of the pectoral fins was left intact. This probably accounts for the different results obtained by Loeb and Bickel. The greater activity and sensitiveness of the caudal part of the SHELDON, Reactions to Chemical Stimult. 291 fish after section of the cord are due, in all probability, to a removal of inhibition from the higher centers, by the use of shorter paths, or to both. Pike (08) believes that the second is the true explanation. He considers that normally the long paths by way of the brain are made use of. After section, pathways through the cord must serve, giving a shorter are which requires less time for the transmission of the impulses. This is supported by the conclusions of Moore and Oertel who suggest, however, that the higher centers may have a regu- latory rather than an inhibitory influence over the cord. Walton and Paul think that the cord is particularly for active, instantaneous and violent reflexes. It is probable that use of shorter paths by the re- flexes has much to do with the result secured, but the evidence pre- sented im the literature is strong in support of the view that the brain possesses an influence over the cord of a regulatory type, at least. Wiping movements of the fins and finlets occur as in the normal animal, This is in line with the results in the spinal frog and ean not, in either case, be considered purposeful, as Howell (’05) and Sher- rington (06b) clearly show. Such movements in a spinal animal are simply such as are of a general protective nature for the normal animal or else form a part of the habitual action of the organism. No spinal shock was observed. This is as would be expected. Steiner, Loeb, Bickel, and Bethe do not mention it for fishes. Moore and Oertel show that it is shght in the frog and passes off quickly. In maminals, as is well known from the work of Bastion (99), Burns (93), Sherrington (797, ’00b), Senator (98), Babaék (07) and Pike (08) the influence of shock is greatest. These citations fall in line with the statement made above to the effect that the power of the cord as an automatic center decreases as one ascends the animal seale. Probably in the dogfish many of the reactions of the caudal part of the animal take place normally through the cord, so*that section of it brings about little change. The Innervation and Reactions of the Nostrils.—It will be noted from the chart that the nostrils are very sensitive to the solutions used as stimuli. The reactions obtained were also very decided. The question arises as to whether these are due to the olfactory nerves or the nerves of general sensation. Accordingly, different nerves were 292 ‘fournal of Comparative Neurology and Psychology. cut in order to clear up this pot. The only nerve hitherto known to supply the nasal mucosa in the dogfish is the olfactory. It has long been known that the trigeminus nerve supplies this region of the head for tactile sensation, but none of its fibers have been traced to the olfactory capsule. Aichel (95) found trigeminal fibers in the olfactory mucous mem- brane of embryonic teleosts, but was uncertain as to their derivation. Jagodowski (701) observed the same fibers. Later, Sheldon (’08) demonstrated the existence of fibers, derived from the trigeminal nerve, in the mucous membrane of the adult earp. Retzius (92) in the frog and Rubaschin (703) in the chick obtained similar results. Tn the lower mammals and man such an innervation has been demon- strated by Grassi and Castronovo (789), Ramon y Cajal (789, 793), van Gehuchten (790), von Brunn (792), von Lenhossék (792), Retzius (’92), Disse (94) and Read (’08). In the dogfish the trigeminal nerve is divided into four rami. These are the ophthalmicus superficialis to the dorsum of the snout, closely associated with the ophthalmicus superficialis facialis of the lateralis system; the ophthalmicus profundus to the side of the head and snout; the maxillaris to the upper jaw and the mandibularis to the lower. These latter two are bound together for some distance from the brain in close association also with the bucealis. All of these trigeminal rami are general sensory with the exception of the mandibularis which is combined general sensory and motor (Strong, 03). The relations of all these rami except the ophthalmicus super- facialis are shown in fig. 3. The olfactory crura were first cut in an endeavor to destroy the sense of smell. At first the section was made from the dorsum, but this operation was usually fatal. Such fishes would either die inside of twenty-four hours or else react very feebly. Finally the method of Lyon (00) was adopted. The fish was etherized, the mouth pried open as far as possible and the olfactory erura reached: by means of an incision in the roof of the mouth. The maxillaris and mandibu- laris were also reached in the same way (see fig. 3). Before the cut was made the mucosa was reflected as shown in the figure, then a piece of cartilage was removed, care being taken not. to cut any of the SHELDON, Reactions to Chemical Stimult. 293 blood vessels of the roof of the mouth, as this leads to serious bleeding. The location of the vessels to be avoided is shown in the dissection. The fishes seemed to suffer no ill effect from this operation and would live for some weeks thereafter. The profundus and superficialis nerves were cut by means of an incision into the orbit at the caudal marein of the eye ball. After section of the olfactory crura, stimulation of the nostrils causes the same reactions as were obtained for normal fishes. It is, therefore, evident that these reactions are not due to stimulation of the olfactory nerve. When the four rami of the trigeminal are cut and the olfactory left intact, no reactions are secured. This shows clearly that the reactions obtained are due to the nerves of general sensation. The associated nerves of the lateralis system have been proved to be insensitive to chemical stimulation. The following nerves were next cut without destroying the reactions; the ophthal- micus superficialis, the ophthalmicus profundus and the oph. sup. and oph. prot. together. When the maxillaris-mandibularis trunk is cut, however, all reactions cease. When normal fishes are taken, this trunk cut on one side and not on the other, stimulation of the nostril on the operated side calls forth no responses, while the nostril on the normal side is as sensitive as before. As the mandibularis nerve goes to the lower jaw exclusively, it is evident that the sensitiveness of the nostrils to the chemical stimuli used is due to the maxillaris nerve. These experiments also show that the nostrils in selachians are imner- vated by the trigeminal nerve, as is the case in most other vertebrates. Some odorous substances were also used in the nostrils although without results so far as the sense of smell is concerned. The sub- stances used were the oils of cloves, pennyroyal, thyme and aniline oil. a sOS, a San LalO=IEhOs BLiIx, MAGNUS. 1ss4. Experimentelle Beitriige zur Losung der Frage tiber die speci- fische Energie der Hautnerven. Zeitschr. f. Biol., Bd. 20, Sel4te156. Tat. 1. 1885. IExperimentelle Beitriige zur Losung der Frage tiber die specifische Energie der Hautnerven. Jbid., Bd. 21, S. 145-160, Taf. 2. BRAUER, L. 1899. Die Beziehungen der Riickenmarksquerlision zu den unterhalb derselben sich abwickelnden Sehnenreflexen. Jineh. Med? Wochenschr., Jahrg. 46, Hft. 1, Nr. 2, 10 Jan., S. 41-42. 3RUNN, A. VON. 1892. Beitriige zur mikroskopischen Anatomie der menschlichen Nasen- hohle. Schultze’s Arch. f. nvikr. Anat., Bd. 389,°8.°632-652, Tat. 29-30, Fig. 10-11. Bruns, 1893. Ueber einen Fall totaler traumatischer Zerst6rung des MRiicken- markes an der Grenze zwischen Hals- und Dorsalmark. Tin Beitrag zur Frage vom Verhalten der Lihmung und der Re- flexe speciell der Patellarreflexe bei hochsitzenden totalen Querschnittsliisionen des Ritickenmarkes. Arch. f. Psychiatrie, 3d. 25, S. 759-830, Daf. 15-16. 1896. Islinische und pathologischanatomische Beitriige zur Chirurgie der Ritickenmarkstumoren. /bid., Bd. 28, S. 96-168, Taf. 4-5. CAMERER, W. 1870. Ueber die Abhingigkeit des Geschimackssinnes von der gereizten Stelle der MundhGhle. Zcitschr. f. Biol., Bd. 6, S. 440-452. CoRIN, JOSEPH. 1888. Action des acides sur le gofit. Arch. de biol., T. 8, pp. 121- 138. DANILEWSKI, B. 1892. Zur Physiologie des Centralnervensystems von Amphioxus. Pfliiger’s Arch. f.-Physiol., Bd. 52, S. 3938-400. *DISSE, J. 1894. Ueber Epithelknospen in der Regio olfactoria der Siiuger. Nach- richt. der k. Gesellsch. d. Wiss. 2u Gottingen. SHELDON, Reactions to Chemical Stimutlt. 301 IraAno, G. 1903. Contribution a lVétude des réflexes spinaux. Arch. Ital. de Biol., Tom. 39, pp. 85-128, 1 pl. WERRARI, P. 1904. Comnient se modifie la sensibilité gustative pour les tres petite doses des anesthésiques locaux. Arch. Ital. de Biol., Tom. 42, pp. 411-420. KONTANE, ARTHUR. 1902. Ueber die Wirkung des Hucain B auf die Geschmacksorgane. Zeitschr. f. Psychol. u. Physiol. d. Sinn., Bd. 28, S. 253-260. I’rry, M. von. 1804. Beitriige zur Physiologie des Schmerzsinns. Ber. ber die Ver- handl. der kgl. Sachs. Gesellsch. der Wissensch. eu Leipzig. Math. physische Classe, Bd. 46, S. 185-196. 1S94. Geitriige zur Physiologie des Schmerzsinns. Zweite Mitth. /bid., S. 283-296. 1895. Beitriige zur Sinnesphysiologie der Haut. Dritte Mitth. Jbid., Bd. 47, S. 166-184. GOLDSCHEIDER, ALFRED. 1885. Ueber Wiirme-, Kiilte-, und Druckpunkte. Arch. f. Anat. w. Physiol, Physiol. Abth., S. 840-3845. GRASSI, BATTISTA AND CASTRONOVO, A. 1ss9. Beitriige zur Kenntnis des Geruchsorgans des Hundes. Schultze’s Arch. f. mikr: Anat., Bd. 34, S. 385-390. Taf. 21). GRUTZNER, LP. 1s94. Ueber die chemische Reizung sensibler Nerven. Pfltiger’s Arch, f. Physiol., Bd. 58, 8S. 69-104, Taf. 1. ANTE, 1D. P- 1901. Zur Psychophysik des Geschmacksinnes. Wiundts Philos, Studien, Baselines olG-02or un ice sim exc. HaAycrart, JOHN B. 1Ss7. The nature of the objective cause of sensation. Pt. 2, taste. Brain, vol 10, pp. 145-163. 1900a. The sense of taste. Schifers Text-book of Physiol., vol. 2, pp. 1237-1245. 1900b. The sense of smell. Jbid., pp. 1246-1258. HEAD, HENRY; Rivers, W. H. R., AND SHERREN, JAMES. 1905. The afferent nervous system from a new aspect. Brain, vol. 28, Dp 99S, ply de HEAD, HENRY, AND SHERREN. JAMES. 1905. The consequences of injury to the peripheral nerves in man. Jbid., pp. 116-338, Figs. 44. Heap, Henry, AND THOMPSON, ‘THEODORE. 1906. The grouping of afferent impulses within the spinal cord. Jbid., vol. 29, pp. 5387-742, figs. $8, pls. 1-3. 302 “fournal of Comparative Neurology and Psychology. HERLITZKA, A. 1907. Ueber den “metallischen Geschmack” und den Geschmack der metallischen Ionen. VII Internat. Kongress zu Heidelberg. Rey. Zentralbl. f. Phys., Bd. 21, S, 504-505. 1908. Sul ‘‘sapore metallico,” sulla sensazione astringente e sul sapore dei sali. Arch. di Fisiol., T. 4, Fasc. 2, pp. 217. HERRICK, C. JUDSON. ‘ 1899. The cranial and first spinal nerves of Menidia; a contribution upon the nerve components of the bony fishes. Jour. Comp. Neur., vol. 9, nos. 38 and 4, pp. 153-455, pls. 14-20. 1901. The cranial nerves and cutaneous sense organs of the North American siluroid fishes. Jbid., vol. 11, no. 3, pp. 177-249, pls. 14-17. 1902. The organ and sense of taste in fishes. Bull. of the U. S. Fish Commission for 1902; also Bull. of the Scientific Lab. of Denison Univ., vol. 12, 1903, article 6, pp. 39-96. 1808a. The doctrine of nerve components and some of its applica- tions. Jour. of Comp. Newr., vol. 13, no. 4, pp. 301-312. 1903b. On the morphological and physiological classification of the cutaneous sense organs of fishes. Am. Nat., vol. 37, no. 437, pp. 313-318. 1905e. On the phylogeny and morphological position of the terminal buds of fishes. Jour. Comp. Neur., vol. 18, no. 2, pp. 121-158. 1908S. On the phylogenetic differentiation of the organs of smell and taste. Jowr. Comp. Neurol. and Psychol., vol. 18, no. 2, pp. 157-166. HW6ser, Rupoir, AND IKiEsow, F. 1898. Ueber den Geschmack von Salzen und Laugen. Zeitschr, f. Phy- sik. Chemie, Bd. 27, S. 601-616. Tioorper, Davin. 1887. Examination of the leaves of the Gymnema sylvestre. Nature, vol. 35, pp. 565-567. HOowELL, W. H. 1905. A text-book of physiology. Philadelphia, W. B. Saunders and Co., 1905. JAGODOWSKI, K. P. 1901. Zur Frage nach der Endigung des Geruchsnerven bei den IKnoch- enfischen. Anat. Anz., Bd. 19, S. 257-267, 10 Ab. JENNINGS, H. S. 1904. Contributions to the study of the behavior of lower organisms. Pub. 16, Carnegie Inst. of Washington. 1906. Behavior of lower organisms. 14 + 2 + 366 pp., New York, 1906. SHELDON, Reactions to Chemical Stiniult. 303 JOHNSTON, J. B. 1902. An attempt to define the primitive functional divisions of the ; central nervous system. Jour. Comp. Neurol., vol. 2, joys 87-106, pls. 1-4. 1906. The nervous system of vertebrates. Philadelphia, 1906. IXAHLENBERG, L. 1898. The actions of solutions on the sense of taste. Bull. Univ. of Wise., Science series, vol. 2, no. 1, pp. 1-81. 1900a. The relation of the taste of acid salts to their degree of dis- sociation. Jour. Phys. Chem., vol. 4, pp. 33-87. 1900b. The relation of the taste of acid salts to their degree of dis- sociation II. Jbid., pp. 583-557. ICANGHibibsdls JB 1898. On the taste and affinity of acids. Am. Chem. Jour., vol. 20, pp. 466-471. IKtiEsow, FRIEDRICH. 1894a. Ueber die Wirkung des Cocain und der Gymnemasiiure auf die Schleimhaut der Zunge und des Mundraums. Wundt’s Philos. Stud., Bd. 9, S. 510-527. 1S894b. Beitriige zur physiologischen Psychologie des Geschmackssinnes. Tbid., Bd. 10, S. 329-368, 523-561. 1806. Beitriige zur physiologischen Psychologie des Geschinackssinnes. Iv. Jbid., Bd. 12, S. 464-473. 1898. Expériences gustatives sur diverses papilles .isolément ex- citées. Arch. Ital. de Biol., T. 30, pp. 399-425. 1903a. Ein Beitrag zur Frage nach den Reaktionszeiten der Geschmacks- empfindungen. Zeitschr. f. Psychol. w. Physiol. der, Sinn., Bd. 33, S. 453-461. 1908b. Zur Psychophysiologie der Mundhéhle nebst Beobachtungen tiber Funktionen des Tast- und Schmerzapparates und einigen 3emerkungen tiber die wahrscheinlichen Tastorgane der Zun- genspitze und des Lippenrots. Jbid., S. 424-443. 1. Fig. IKxirsow, F., anp Haun, R. 1901. Beobachtungen iiber die Empfindlichkeit der hinteren Teile des Mundraumes fiir Yast-, Schmerz-, Temp.-, und Geschmacks- reize. Jbid., Bd. 26, S. 383-417. LEMBERGER, I‘RIEDA. 1908. Psychophysische Untersuchungen iiber den Geschmack von Zucker und Saccharin (Saccharose und Kristallose). P/fliiger’s Arch. f. Physiol., Bd. 123, S. 293-311, 1 text fig. LENHOSSEK, M. VON. 1892. Die Nervenurspriinge und -Endigungen im Jacobson’schen Organ des Kaninehens.- Anat. Anz., Bd. 7, Nr. 19 u. 20; S. 628- 635. Mit 1 Ab. 304. fournal of Comparative Neurology and Psychology. LorB, JACQUES. 1891. Ueber Geotropismus bei Thieren. Pfliiger’s Arch. f. Physiol., Bd. 49, S. 175-189; 1 Holzsch: IGxONs -E)., (P- 1900. Compensatory motions in fishes. Am. Jour. of Physiol., vol. 4, no. 2, pp. 171-82. MENDELSSOHN, Morirz. 1882. Untersuchungen iiber Reflexe. Site. ber. d. kgl. preuss. Akad, d. Wissensch,. 2. Berlin. Ualtbd. 2, S. S9T-900. 1SS5. Untersuchungen tiber Reflexe. Zweite Mitth. J/bid., Halfbd. 1, S. 123-125. 1885. Untersuchungen iiber Reflexe. Dritte Mitth. Jbid., Halfbd. 1, Se lOratalle Moore, BENJAMIN, AND OERTEL, Horst. 1s99. A comparative study of reflex action after complete section of the spinal cord in the cervical or upper dorsal region. yvol. 20, pp 36-38: 1897. THE JOURNAL OF COMPARATIVE NEUROLOGY AND PSYCHOLOGY.—VoL. XIX, No. 3. 314. “fournal of Comparative Neurology and Psychology. 16. Die Physiologie der Pedicellarien. Zeitschr. f. Biol, vol. 37, pp. 324-405. 1899. 17. Der Neurokinet. Zcitschr. f. Biol., vol. 38, pp. 291-299. 1899. 18. (Breer, TH., BETHE, A., und v. UEXKULL, J.) Vorschliige zu eimer objek- tivierenden Nomenklatur in der Physiologie des Nervensystems. Biol. Centralot., vol. 19, pp. 517-521. 1899. Also in’ Centnalbl. fF. Phiysiot., vol. 13, pp. 1387-141. as99: 19. Die Physiologie des S@#relstachels. Zeitschr. f. Biol., vol. 39, pp. 73-112. 1900. 20. Die Wirkung von Licht und Schatten auf die Seeigel. Zeitschr. f. Biol, vol. 40, pp. 447-476. 1900. 21. Ueber die Errichtung eines zoologischen Arbeitsplatzes in Dar es Salaam. Zool. Anz., vol. 23, pp. 579-583. 1900. 22. Veber die Stellung der vergleichenden Physiologie zur Hypothese der Tierseele. Biol. Centralbl., vol. 20, pp. 497-502. 1900. 23.) Die Schwimmbewegungen von Rhizostoma pulmo. Aittheil. a. d. Zool. Stat. Neapel, vol. 14, pp. 620-626. 1901. 24. Psychologie und Biologie in ihrer Stellung zur Tierseele. Asher und Spiro’s Hrgebnisse der Physiologie, 1 Jaarg., 2 Abth., pp. 212-238. (Re- printed with the title “Im Kampf um die Tierseele.’’) 25. Studien iiber den Tonus. I. Die biologische Bauplan von Sipunculus nudus. Zeitschr. f. Biol., vol. 44, pp. 269-344. 1908. 26. Die ersten Ursachen des Rhythmus in der Tierreihe. Asher und Spiro’s Ergebnisse der Physiologic, 3 Jahrg., 2 Abth., 11 pages. 1904. 27. Studien itiber den Tonus. II. Die Bewegungen der Schlangensterne. Zeitschr. f. Biol., vol. 46, pp. 1-57. 1904. 28. Studien tiber den Tonus. III. Die Blutegel. Zeitschr. f. Biol., vol. 46, pp. 372-402. 1905. 29, Leitfaden in das Studium der experimentellen Biologie der Wassertiere. 130 pp. Wiesbaden, 1905. 30. Studien tiber den Tonus. IV. Die Herzigel. Zeitschr. f. Biol., vol. 49, pp. 307-832. 1907. 31. Studien tiber den Tonus. VY. Die Libellen. Zeitschr. f. Biol., vol. 50, pp. 168-202. 32. Die Verdichtung der Muskeln. Zentralbl. f. Physiol. Vol, 22, pps 34-3. 1908. . Die neuen Fragen in der experimentellen Biologie. Rivista di Scienza “Scientia,.. vol. 4, No.1, i pp: Some CnHaractrenristic Dicra. “We may quietly throw overboard the entire sum of our knowledge thus far acquired, so far as it comes from experiments on the frog’s leg. The nerve-muscle preparation has misled us in almost every point” (30> ip. 3a), Jenninos, Uexkill on Physiology of Behavior. 315 “Furthermore all electro-physiological experiments on muscles have shown themselves to be biologically worthless, so that we may here pass them over in silence” (29, p. 28). “The fate of biology in Europe, in spite of the efforts of excellent workers, seems to me to be sealed. One need not be a prophet to predict that in a few years biology will be an American science” (29, preface). “The question of the function of the nervous system in the animal body has aroused a strife between two sciences that must end with the annihilation of one of the two combatants,—and the champions of both sides are determined to carry the combat to the end. * gi While the comparative psychologists debated concerning the amount of sensation, memory, reflection, that one should attribute to these animals, there arose in the growing science of comparative physiology an enemy to the death of all comparative psychology” (24, pp. 212, 213). | “Before objective investigation the sensations, the memory, and thoughts of animals disappeared like fluttering forms of vapor. The iron chain of objective changes, which began with the stimula- tion of the sense organ and finished with the movement of the muscle, was welded together in the middle. Nowhere remained a smallest spot for the psyche of the animal. Basing itself on these incontesta- ble facts, comparative physiology pronounced the psychological con- clusions mere superstitious and denied comparative psychology the right to call itself a science’ (24, Dee 2aay)e “We stand on the eve of a seientifie bankruptey, whose conse- quenees are as yet incalculable. Darwinism is to be stricken from the list of scientific theories” (33, Paso “Concerning the origin of species we know, after fifty vears of unparalleled effort and investigation, only the one thing, that it does not take place as Darwin thought it did. A positive enrichment of our knowledge has not resulted. The whole enormous intellectual labor was in vain” (31, p. 168). “When in biology one has freed himself from the idea of develop- ment,—an idea which has at last been hunted to final death—vso that one is again in position to look upon each animal as a unity 316 “fournal of Com parative Neurology and Psychology. closed within itself, instead of as the last chance product of an an- cestral series that has been speculated together, then form and color gain a new interest and a heightened brilliance” (30, p. 318). “Driesch succeeded in proving that the germ cell does not possess a trace of machine-like structure, but consists of throughout equiva- lent parts. With that fell the dogma that the organism is only a machine. Even if life occurs in the fully organized creature in a machine-like way, the organization of a structureless germ into a complicated structure is a power sui generis, which is found only in living things and stands without analogy” (33, p. 9). “Tt is not to be denied that the vitalists are the victors all along the line. After having put an end to Darwinism, they have seized upon the entire field of the production of animal form, and now threaten the last positions of their opponents” (38, p. 14). “So there persists in the outer world of objects an unresolvable con- tradictoriness—” (29, p. 129). Inrropuctory CHARACTERIZATION. J. von Uexkill, of Heidelberg, has long been one of the most active and original workers in the physiology of behavior; his work will be found of the highest interest to all seriously concerned with these matters. It undertakes for the lower animals much the same sort of analysis that Sherrington gives us for higher ones, though with many features that are in the highest degree original. Further, an examination of his work has hardly less interest as a study in scientific ideals and method than for the concrete results attained. That he has been led to radical and iconoclastic views, and that he is not afraid to express these views with decision and picturesqueness will be evident from the quotations given above. The pessimistic impression given by the quotations taken together will not escape the reader; this pessimism appears rather an unintentional result than as characteristic of the animated and militant spirit of our author. The sweeping and perhaps ill-founded character of some of the propositions advanced should not prejudice the reader against the accuracy of the author’s work in his own field; this would be a serious mistake. Jennincs, Uexkiill on Physiology of Behavior. 317 INVESTIGATIONS. Von Uexkiill began as a student of the nerve-muscle preparation of the frog (1). He quickly determined to carry the study of the questions involved to the lower animals; this plan, carried out largely at the Naples Station, led to fundamental results. The work first undertaken was a study of the reflexes of the cuttle-fish (2, 3, 4, 7). This was followed by a study of certain sensory problems on the skate (9), by work on the muscle and nerve physiology of the worm Sipuneulus (11, 25), and by an extensive series of thorough and fundamental papers on the nerve-musele and sensory physiology of the sea urchin (10, 12, 18, 15, 16, 19, 20). Through these studies the author had developed the outlines of an original system of nerve- muscle physiology, having at its basis the concept of tonus. This system he developed farther in the series of Studies on Tonus (25, 27, 28, 30, 31), is still in progress, and which we hope may count many numbers a series dealing with various invertebrates, which besides the five that have appeared. Arising in connection with his systematic series of investigations there have come from his pen many incidental contributions, including notes on special points in nerve-muscle physiology (1, 5, 6, 8, 14, 17, 32), studies of rhythm (25, 26), and discussions of fundamental scientific questions (18, 22, 24, 33). In 1905 a brief textbook (29) was published, giving an outline of the views to which he had come, with directions for prac- tical work. Characteristic of v. Uexkiill is the intellectual working over of results at the time they are reached, so as to give a graphie and com- prehensible scheme of the way processes occur. This appears in the very first studies, on the cuttle-fish (2, 3, 4, 7); they show the author’s characteristic abhorrence of everything vague, and particu- larly his objection to psychic explanations in physiology. They contain much important detail on the physiology of muscle and nerve. The paper on the skate (9) is largely a polemie against the use of terms implying consciousness in the nerve physiology of lower animals, directed mainly against W. Nagel. The account of the skate is intended to illustrate the purely objective treatment of the facts, and is perhaps not in itself a strong example of the value of 318 Fournal of Comparative Neurology and Psychology. the method. In this polemic, as in later work, is seen the positive, definite character of the author’s thought, with no appreciation for shadings, transitions or compromises; sharpness of concept and of distinctions, black or white, is demanded everywhere. Tur Work on THE SEA URCHIN. It is in the series of papers on the sea urchin (10, 12, 18, 16, 19, 20) that the author “strikes his pace” and brings out clearly the important phenomena which form the basis for his peculiar system of concepts. The plan of investigation is to work out the structure of the muscular, nervous and skeletal systems and the way they act together to produce the characteristic behavior of the animal; to work out “the biological plan’ of the sea urchin. The paper on Reflexes in the Sea Urchin (12) gives a general sketch of this biological plan and the main reflexes which make up the behavior; the structure of the simple nervous system, the reflexes of the pedicellarix, the spines, the tube-feet, the teeth, are briefly described. Then followed in 1899 a thorough detailed paper on the physiology of the pedicellariz (16), a paper which must form the point of departure for all further work on these organs. The history of our knowledge of the pedi- cellarize, their structure, the different sorts, their uses and particu- larly the laws of their movements, are developed in great detail. In 1900 came a similarly thorough, and perhaps still more impor- tant, study of the movements of the spines (19). Special studies were made of the reactions of the spines to light and shadow (13, 20); for the purpose of obtaining species favorable for this work the author made a trip to the east coast of Africa; the outfit for this trip is deseribed in a separate paper (21). In general, v. Uexkiill found that the various organs of the sea ‘an perform a number ot urchin,—each spine, each pedicellaria, different acts or reflexes, and that these reflexes occur to a large degree independently of each other, so that they may occur when the organ in question is isolated on a small piece of the shell. Thus each organ is a “reflex person.” Yet all the differing reflexes of these various “persons” are of such a character that they work together in asystematie way to perform the necessary functions of the animal. Jennincs, Uexkiill on Physiology of Behavior. 319 They carry it about, toward food and away from danger, keep the sea urchin clean, capture prey, combat against enemies, and _ all together do the work of life in a competent way. This is brought about without any regulation by a “higher center,’ merely through the essentially independent activities of the various parts. Therefore the sea urchin is characterized as a “republic of reflexes.” The apparent unity of action is due merely to the way the different actions of the different parts fit into one another, by a sort of preéstab- lished harmony. “It is not that the action is unified, but that the movements are ordered; that is, the setting off of the different reflexes is not the result of a common central impulse, but the separate reflex ares are so constructed and so fitted together that the synchronous but independent setting off of the reflexes, as a result of an external stimulus, produces a definite general action of the animal, just as happens in animals in which a common center produces the actions” (16, p. 390). “When a dog runs, the animal moves its feet. When the sea urchin runs, the feet move the animal’? (19, p. 73). But the author finds that the separate reflex-persons are not abso- lutely independent; impulses may pass from one to the other, im such a way as to produce a unified action of all. But this is due merely to a set of nerve nets having a special arrangement; it is a matter of interconnections, not of regulation from a “higher center.” The question may be raised whether this distinction does not depend on an undefined and mystical use of the term higher center. If the higher center, in accordance with the illuminating ideas of Loeb, is after all essentially a place for complex inter- connections, then the difference between the sea urchin and higher animals is only one of degree. Von Uexkill finds that the separate reflexes of the various organs are not absolutely stereotyped, but the action of each part is changed at times in certain ways, often depending on the action of neighboring parts. We have then in the sea urchin an opportunity of studying codrdination in its most elementary condition, and thus perhaps of determining the fundamental nature of the phenomena involved. This animal is a sort of model, which can be taken to pieces without serious alteration of the action of the parts; yet in 320 “fournal of Comparative Neurology and Psychology. the animal as a whole the parts do influence one another. The author therefore took in hand a thorough study of the changes in the reflexes and the way they influence each other, undertaking to formulate and define precisely the underlying phenomena. It is here that we find the origin of the most important concepts in von Uexkiill’s highly original formulation of behavior. It will be worth while therefore to examine carefully a sample of the author’s method of analysis; for this purpose we select certain features in the physiology of the spines. TypricaL Exampie or Mrtruop or ANALYSIS. It will be recalled that the rounded shell of the sea urchin is covered with long spines. Each spine is a tapering calcareous rod, with a concavity at its base, by which it articulates with a hemi- spherical elevation of the shell. The spine is held in position by two circles of muscles radiating from the circumference of its base to the shell. These muscles, and particularly the inner circle, are steadily pulling upon the spine, thus holding it stiffly in position. This pulling takes place without external stimulus; it is due to a certain amount of tension which forms the normal condition of the muscles and continues without any such repeated contractions or trem- ors as are called tetanus. That is, the muscles have a certain normal fonus. This tonus becomes the central concept in v. Uexkill’s formulation of the physiology of movement. The imner layer of muscles is devoted chiefly to maintaining by its constant tension a certain position of the spine; it is an example of one of the two great types of muscular action——the “Sperrung” or tension, as dis- tinguished from actual contraction, involving shortening. The outer circle of muscles is more active in its changes; they shorten quickly and readily, thus moving the spine in various ways. They exemplify the other great type of muscular aetion,—* Verhkiirzung,’—shorten- ing or contraction. Tension and contraction v. Uexkiill shows occur quite independently of each other, and this independence, with all its theoretical and practical consequences, comes to play a very great part in the later development of v. Uexkill’s views; he finds it throughout animals (see 32). The neglect of the fact that we have Jennincs, Uexkiill on Physiology of Behavior. 321 here two entirely different functions has, the author believes, led all nerve-muscle physiology into false paths. The first reflex shown by the spines is as follows: When a cer- tain spot on the body is moderately stimulated, the surrounding spines bend toward it. The muscles of the side of the spine next the point stimulated contract; that is, their tonus is increased. Thus the points of the spines are directed, for example, toward an approaching enemy. But if the stimulus is very intense, the reaction just described is reversed; the spines bend away from the point stimulated. This result is produced by a decrease in the tonus of the muscles on the side of the spine facing the point stimulated. Thus from the same spot on the body opposite effects may be produced, depending on the strength of the stimulus. This phenomenon is called by v. Uexkill reversal of the reflex (‘‘Reflexumkehr”) ; it is observed in other or- gans of the sea urchin and other animals, under various conditions. The author holds it to be due to some sort of apparatus in the ganglion cells; an apparatus that he calls the “tonus switch” (“Tonusschal- ter”). This reversal is well seen in the spines when a strong chemical stimulus affects the body. Now, a further consequence of such a powerful stimulus is seen. After such a stimulus, even a weak stimulus, which formerly caused the spines to bend toward the spot stimulated, now causes them to bend away. So the same stimulus on the same spot may cause two different reactions, depending on what stimulus has preceded it. This phenomenon, very common in animals, v. Uexkiill calls the “switching” of the tonus (“Schaltung” ) ; he holds it due to the same apparatus as the reversal of the reflex. The two reactions of the spines serve, under natural conditions, certain functions. The bending toward a stimulated point serves for defense; the bending away under a strong stimulus, particularly a chemical one, preserves the spines from injury, while giving oppor- tunity for the action of certain large poisonous pedicellariz, which now bend their envenomed jaws toward the region attacked and seize whatever is there present. Certain other facts in the physiology of the spines are of extreme importance. A steady tension, not violent, exercised on the muscles 322 fournal of Comparative Neurology and Psychology. of the spines causes them to lose their tonus; they become limp. This effect of tension on tonus is common among animals. If then a spine is pressed steadily to one side by the fingers, or by the weight of the animal’s body, the muscles on the side pressed lose their tonus. The spine, therefore, becomes loosely movable in certain directions, but not in others. On the other hand, a sudden violent increase of tension, or a mechanical jar, increases the tonus, so that the spines stand out firmly. Now, the loss of tonus, caused in the way just described, is con- ducted, doubtless by the nervous network, to the neighboring spines. This conduction occurs in such a way that it is the muscles of corre- sponding sides of the neighboring spines that lose their tonus. (‘This involves complicated conditions in the nervous net; v. Uexkiill holds that it shows the existence of many independent nets.) Hence when a spine is pressed toward one side, the neighboring spines likewise bend in the same direction. This v. Uexkiill calls the chaining of the reflexes (‘‘Reflexverkettung”). It shows itself (an important fact ) most readily when the spine is bent toward the mouth; the other spines also bend toward the mouth. These facts have the following result. When a spot on the body is strongly stimulated, so that the spines bend away from it, the disturbance is not limited to those in the immediate neighborhood. The spines in bending away press upon the surrounding spines, tending to bend them down. They are more easily bent toward the mouth than elsewhere, so a new set of spines bend over in that direction. They again press on the next spines, bending them in turn toward the mouth. Thus a sort of wave passes toward the mouth from the point stimulated, the spines bending in turn far over toward the mouth, then back again. The entire phenomenon vy. Uexkiill calls the wandering of the center of excitation (“*Wanderung des Erregungsmittelpunkts” ). Another most important fact shows itself. Muscles that are not in tonus are much more easily stimulated to contraction than those which are in tonus. When the muscles have their usual strong fonus, it requires a powerful stimulus to cause them to contract further. But muscles which have lost their tonus as a result of Jennincs, Uexkiill on Physiology of Behavior. 323 steady tension (as deseribed above), contract readily in response to even a weak stimulus, tending thus to bend the spine toward that side on which there has been tension. From this a number of peculiar facts result. As we have seen, a moderate stimulus at a certain point tends to cause the spines to bend toward that point. If, as a result of pressure, the muscles that face the point stimulated have lost their tonus, they respond readily ; the spine at onee bends toward the side stimulated. But if the spines have been pressed over in the opposite direction, so that their muscles facing the point of stimulation are in strong tonus, no effect is produced; the spines retain their position. Hence, when a spot on the body is stimulated, certain spines will respond while others will not, depending on the previous tonus of their muscles. This phenomenon y. Uexkiill calls “Klinkung’’; those spines which are in such a condition or position that they can respond to the stimulus are said to be “eingeklinkt”; those which are not are ‘‘aus- geklinkt.”” These expressions may perhaps be translated by “in eireuit” and “out of circuit,” comparing the spines with instru: ments in an electric circuit. This condition of affairs has great importance for the functioning of the spines in locomotion and else- where, and parallel conditions are found in other organisms. A similar analysis is given by the author for the pedicellariz, tube-feet, teeth, ete. Thus by a close and thorough study vy. Uexkiill has been able to analyze and formulate a number of what have been called vaguely the varying “physiological states” of organs or organisms; such analysis is needed for all cases. By making use of the concepts of Reflexumkehr, Reflexverkettung, Wanderung des Erregungsmittel- punkts, Schaltung, Klinkung, and by observing the changes in tonus and the rules for its increase and decrease, one can explain some of the most important features in the behavior of the sea urchin under natural conditions ; locomotion, negative reactions to various stimuh, defenee from enemies, capture of food, ete. It is, of course, no disparagement of the value of this analysis that it does not exhaust the matter for the sea urchin. Thus, when the animal is turned on its back, its spines move in ways that would not be expected 324 “fournal of Com parative Neurology and Psychology. from the physiological analysis based on their other movements (19, p. 105); if they did the sea urchin would not regain its normal position. In the starfish the method of action may be changed by the formation of habits, and this is doubtless true also for the sea urchin. Thus any formulation that is complete must provide also for the laws of change of behavior; for its regulatory features. Possibly no complete formulation can ever be reached, but the most direct way to approach it is by such analysis as v. Uexkiull gives. Later INVESTIGATIONS. We have given this account of the spines as a type of v. Uexkill’s methods of analysis; by following carefully such a concrete case the reader will get a better idea of the nature and justification of his work than by any systematic survey of the concepts to which he finally comes. Let us now follow further the development of these concepts. As we have seen, the central concept is that of tonus, and the laws of the changes of tonus are the chief object of research. To research on this matter, to studying the properties of tonus in various organisms, and to devising schemata which shall help us to understand how it acts, and hence how behavior takes place, have been devoted the later researches of v. Uexkiill. He has thus far analyzed from this point of view, besides the sea urchin, the worm Sipunculus (25), the brittle-star (27), the leech (28), the heart- shaped sea urchin (30), and the dragon fly (31). In the latest contribution, on the dragon fly, v. Uexkiill attempts to make provi- sion for a modification of the machinery of behavior through the experiences of the organisms. It would manifestly be impossible to resume here these researches, filled as they are with minute and technical detail. V. Urxkiit1’s System or Concepts. A view of v. Uexkiill’s system of concepts can be gotten most directly from his “Guide to Experimental Biology” (29). But here one does not see the development of the ideas; the actual grounds that have given origin one after another to the peculiar concepts, so that they are likely to seem on first introduction bizarre and artificial, having little similarity to anything dealt with in orthodox physiology. Jennincs, Uexkiill on Physiology of Behavior. 325 The fundamental concept is fonus. Just what are we to under- stand by this? V. Uexkiill at first defines it merely as the sum of those manifestations of the life of the cell that produce effects on external things (as distinguished from the internal energy used in metabolism, ete.) (19, p. 78). As his work develops, he finds need for a more precise idea of tonus. It is defined as a “form of energy” which has the property of flowing in certain ways (20, p. 474). The concept of tonus gradually becomes more and more definite. For purposes of handling and imaging it with ease, it becomes convenient to think of tonus as a fluid, which flows through a set of tubes (the nerves). This fluid becomes at last identified with Bethe’s “Fibrillensiure,’—an actual chemical, visible under the microscope (27, p. 31). But this identification is not held to uni- formly. This fluid tonus is contained in a system of tubes, the nerves. “The structure of the nervous system may then be conceived as an ageregate of peculiar vessels united one with another, which inter- change and equalize each other’s contents with relation both to pressure and quantity’ (25, p. 305). From the nerves the tonus either passes into the muscles, or causes in them the production of a fluid with similar properties, giving rise to either tension (“Sperrung”’) or contraction (“Verkiirzung’). In dealing with tonus, either in the nerves or the muscles, we must distinguish its quantity from its pressure; these may vary independently, so that any given quantity may have high or low pressure. On the quan- ity of tonus depends the contraction of muscles; on the pressure, the tension of muscles. There are certain general laws 1. r the movements of tonus. In simple nerve nets it always flows into muscles that are extended (causing them to contract again). This is attributed to a change in the capacity of the muscles; extended muscles have greater capac- ity than contracted ones, so in extending they suck, as it were, the tonus out of the nerves. This property gives a remarkable degree of self-regulation to the action of the nerves and muscles. Tn most animals, further, the tonus shows a marked tendency to flow toward a-certain part of the body,—usually the anterior end,— 326 Ffournal of Comparative Neurology and Psychology. so that this part responds when any part of the body is stimulated. This part to which tonus flows as water flows into a valley is denominated, with poetic feeling, the vale of tonus (‘“Tonustal,” 25, p- 310; 29, p. 56). After the tonus has flowed into certain muscles, it is possible (in some cases at least) to capture and hold it there, by cutting the nerves leading to the muscles (‘‘Tonusfang,” 25, p. 302); the muscles then remain contracted. During rest the fluid tonus is gradually used up and disappears ; at stimulation it is newly manufactured. There exist, however, reservoirs of tonus in the nervous system, so that the lost tonus of the muscles ean be replaced without new manufacture. The nervous system then contains, besides a system of communi- eating tubes, reservoirs of tonus; at the same time it is an elaborate apparatus for controlling the distribution of tonus. Each muscle has somewhere in the nervous system an organ which is its “repre- sentative’ (25, p. 303; 29, p. 44). The office of this representative is to see that the tonus pressure in the muscle remains sufficient to ‘ause the tension of the musele to correspond to the weight which it has to bear. When the pressure in the muscle is insufficient, this acts on the representative (through the nerve) causing it to increase the pressure, until this raises the tension so as to support the weight. The increased pressure is produced by the fact that the representative uses up a certain quantity of tonus to increase the pressure of what is left. The author’s further development of this system consists in work- ing out in detail the structure and action of this system of tubes, reservoirs and other machinery, by which the distribution of tonus is controlled. Main tubes, feeders, reservoirs, valves, ete., are de- vised and represented by diagrams, till we finally get figures which resemble the plan for a dye-works or a flour mill (see for example the schema for Sipunculus, 25, Tafel 6). The method of presentation is in general the ideal construction of an apparatus which could produce the results shown by the organisms. In this construction no attempt is made to represent apparatus that actually exists in the organism; it is merely a figure or illustration; ‘fa mere schema in accordance with which one can group Jenninos, Uexkiill on Phystology of Behavior. 327 the experimentally found facts in a convenient way” (25, p. 287). “The schema of indirect investigation is not a theory at all, but merely a sign language by means of which it is possible to at once express new results in a graphie (‘anschaulich’) way” (27, p. 31). All emphasis is laid on making the illustration thus “anschaulich” ; that is, of such a character that one can “see through it”; see how it would work as a machine works. The author makes extensive use of this “fictitious schema” (25, p. 291), basing long discussions for the greater part of entire papers on its properties. Perhaps nowhere else in biology has a figure of speech, as it were, been worked out in such tremendous detail, through a long series of papers. . Regarded thus as a figure or illustration, the author appears very successful in constructing apparatus that would produce results simi- lar in their compheation and regulatory character to the processes observed in organisms. This has necessarily been done, of course, by attributing new characteristics to the various components when required. The tonus is sometimes given the characters of a definite material fluid, and much pains is taken to account for the entire quantity ; again, it may be produced or disappear as required ; some- times it is considered a form of energy; at times it shows the prop- erties of electricity in producing effects by induction (31, p. 195) ; at times we are informed that the figure of a fluid quite fails (25, p. 213). When the author attempts to show how his compleated machinery may become modified in a way corresponding to the pro- duction of what are called psychologically memory images (31), clearness has to be given up, and the entire figure becomes uncon- vincing. As to the value of this figurative and artificial method of pre- senting the results of work, opinions will, of course, differ. The point can be best discussed in connection with a review of the guiding principles and scientific ideals in the author’s work; to this we now turn. It is peculiarly true in the work of v. Uexkiill that the author’s concrete results cannot be understood without an apprecis- tion of the principles that have guided him. This will lead us to « consideration of his general and theoretical papers. 328 =fournal of Comparative Neurology and Psychology. TneroretTicaL Vrews Anp Guiprne PRINCIPLES. Perhaps the main characteristic shown throughout v. Uexkiill’s work is the abhorrence of anything vague, ill-defined or mystical. In his early papers he sets forth clearly the ideal of scientific work as the discovery and presentation of what is verifiable or demonstrable. “We have to do only with processes that can be objectively demon- strated, and to write the history of these processes in an animal from the moment of stimulation to the resulting reaction” (9, p. 559). This led him at once into a polemic against authors that used psychic explanations in work on animal behavior (see 7, p. 608; 9, ete.). The circle was soon widened, and in 1902 v. Uexkiill declares that a war of extermination has arisen between comparative physiol- ogy and comparative psychology, a war that spells annihilation for one of the combatants and “both are determined to carry the fight to the end” (24). His fundamental point is, of course, the fact that there is no way of observing or verifying the existence of psychic phenomena in animals, so that they cannot form a part of a strictly verifiable science; and a further postulate is that all ob- jective processes can and should be fully presented and accounted for without bringing in anything from outside. To substitute psy- chological interpretations for certain steps of objective experimental analysis is vicious and destructive of consistent science. V. Uexkiill’s polemic papers take extreme positions and are written with much picturesqueness of statement; they are of great value for rousing to a realization of the difficulties those who need such a spur. Apparently, however, all the valuable results that would be reached by utterly destroying the unhappy comparative psy- chologists would be equally well attained by keeping carefully sepa- rate the two fields of work. If the experimenter never substitutes a psychological explanation for a physiological one, he may also be interested, as a separate problem, in the development of mind, with- out injury to his objective scientific work. This same demand for objective verifiable results, without admix- ture of anything else, has led v. Uexkiill to take a part, with Beer, Bethe, and others, in trying to establish a purely objective nomen- clature for the processes occurring in the movements of animals (18; Jenninos, Uexkiill on Phystology of Behavior. 329 see also 29). This nomenclature has philosophical value and has “been used by a few authors, though its employment is by no means common. There is difficulty, as with all ideal system of nomen- clature, devised before investigation is complete, in the fact that its use often implies a precise knowledge of the nature of the phenomena, when such knowledge does not exist. To give precisely the correct name to a process implies that we know fully the nature of the process. The author’s abhorrence of the vague later becomes still more accentuated in the demand (which we have noticed above im our ac- count of his investigations) that work shall be presented always in a way that is “anschaulich”; that is, in such a way that one can see just how the processes would occur, as one sees how a machine works from knowing its structure. It is extraordinary to what an extent the author makes the attainment of this ‘Anschaulichkeit” the chief object of biological setence; he declares it plumply to be the “most essential character of all’? (“die allerwesentlichste”) for the science of biology (31, p. 184). ‘Biology is in its essence ‘Anschauung’ ” (33, p. 16)..) “Only the anschaulich structural diagram, not prov- ing, but showing the unified working together of different factors, is adequate to the requirement of bringing the life processes together into an intelligible unity without omitting life itself” (31, p. 185). It is only by grasping fully the fact that ‘‘Anschaulichkeit’? is the author’s ideal, that one can understand many of the peculiarities of his work. The first far-reaching consequence of this ideal arises from the fact that that which is “anschaulich” is not always that which is verifiable. The author is therefore sometimes compelled to a choice between the two, and in his later papers he at such times deliberately chooses the ‘Anschaulichkeit” in preference to verifiableness. He thus falls into contradiction with his own earlier requirement that we shall deal only with what is cbjectively demonstrable (“objectiv nachweisbar,” 9, p. 559), as well as with the procedure of other in- ‘The word intuition, by which “Anschawing” is commonly translated, cer- tainly fails to carry to most minds the same graphic idea as the German word, so that I do not employ it. 330 © fournal of Comparative Neurology and Psychology. vestigators to whom it is more important that scientific propositions shall be verifiable than that they shall be anschaulich. Let us here look in a general way at the contrast between the results reached by making ‘Anschaulichkeit” the ideal, and those which flow from making verifiableness the ideal. For many investigators the object of science is to prepare a system of verifiable propositions, in order that we may know what to depend on in our conduct; “to know what is true in order to do what is right,” as Huxley put it. Verifiable propositions are propositions that say “Under such and such conditions you will find such and such things to oceur or exist.”” Now, if one supples the conditions set forth, and does not find the predicted things to occur or exist, the proposition is not verifiable, and many would therefore hold that it should be stricken from science. A large proportion of the propo- sitions concerning machine-like structures in organisms, given by v. Uexkiill, do not even profess to be verifiable. One of the main objects of investigation is to find out what particular kinds of machines are present in animals, and how these actually present machines have arisen and how they are changing. This object is incompatible with the mere assumption of fictitious machines, for the first result of investigation with this object in view is to cancel these fictitious machines. This would indeed leave our science for a time less sharply formulated, but “at a certain stage in the develop- ment of a science a degree of vagueness is what best consists with fertility,’® for reference of the phenoniena to complete fictitious machines tends to cut off search for the real ones. When we have found out what really occurs in organisms and what machines actually exist there, then our knowledge will be as ‘‘anschaulich” as the facts warrant, no more, no less. To the present reviewer it seems that, even for practical purposes, the author has overestimated the value of a rather gross “Anschau- lichkeit.”. The bringing in of machine-like — structures,—tubes, valves, ete.,—that confessedly do not exist, seems rather to confuse “Or, put in a form which holds whatever one’s theories, “when you have such and such experiences, you will have such and such other experiences.” ‘Vim. James, Psychology, Preface. Jenninos, Uexkill on Physiology of Behavior. 331 than to aid the mind. It is not possible to conclude directly from the properties of the assumed machines as to what physiological properties one will find, for the parallelism is far from complete, so one must try to keep the system of machinery separate from the system of physiological facts; there are two systems to grasp in place of one. The reader finds it difficult if not impossible to disentangle statements which the author wishes to present as verifiable facts, from statements which are a mere necessity for carrying out the figurative schema. V. Uexkiill shows all through his work an astounding facility in coneluding as to the structures that must be present, from the functions which he sees performed. The reader wonders whether these structures are held to actually exist, or whether they are part of the fictitious schema. The reviewer finds that for his own use it becomes necessary in reading v. Uexkill’s work to ask “Now, what did the author here actually observe and demon- strate?’ It then becomes necessary to transform or almost. re-write a paper before the verifiable results can be disentangled from the figurative presentation. I believe that this condition of affairs has prevented the work of v. Uexkiill from exercising the great influence that it deserves from its importance. Nothing would be more help- ful to most readers than for the author, after putting his results together in the figurative language of his peculiar system as he has done in his Guide (29), to give us a new compendium of his experi- ments and results, making the test of admission that which is veri- fiable, or at least that which the author believes will be found veri- fiable. This would not involve, of course, the omission of his im- portant concepts of “Schaltung,” “Klinkung” and the lke, for these are hames for experimentally verifiable processes and conditions ; nor would it involve the omission of general laws, as verifiable statements that apply to whole classes of objects; nor would it exclude hypoth- eses, presented as such, for these are propositions which the author believes will be found verifiable on further investigation. It would involve simply the omission of what the author himself recognizes as fictitious. I believe it would be found that nothing of value had been lost; that the author’s important work would stand out with a clear- ness not before attained. Further, in the technical accounts of his 332 fournal of Comparative Neurology and Psychology. investigations, it would be extremely helpful if the author would at least segregate carefully his verifiable, experimental, results from his fictitious schema, if he finds that he cannot bring himself to totally abandon the latter. This demand for “Anschaulichkeit” rather than verifiableness in a scientific account is what has led to an apparent opposition between vy. Uexkiill’s work and that of some others. Such is the case, I judge, with the differences between his work and my own. He pre- sents his work in an ‘“anschaulich” form that is confessedly not verifiable, while I have tried to present strictly what is verifiable, whether immediately “anschaulich” or not. The results are bound to be different in the two cases. If my work should be presented by the aid of “anschaulich” fictions, or if v. Uexkiill should preseut his own results without these fictions, the two accounts would show a most gratifying agreement; this is especially true now that. v. Uexkiill has included, in his last Study (31) attempts to show how his machines could be modified by the influences which act on the organism. I have never argued against the existence of machine- like arrangements in organisms. My point was merely that these machines are not fixed and final, but that they are continually changed by the environment and by the action of the organism itself.* Personally I believe that even these changes occur in an essentially machine-hke way. The demand for “Anschaulichkeit” at all costs is apparently what has led the author to certain extreme views; to his separating and cou- trasting biology and physiology; and to his tendency to fall into vital- ‘In his recent paper on New Questions in Haperimental Biology (38) Vv. Uexkiill, in presenting a graphic picture of my exposition if carried to a logical extreme, has attributed to me extreme views which I have never held. He says that I “denied the existence of the reflex; denied the existence of any Structure in the central nervous system.” This statement I am sure is given as part or an “anschaulich”’’ fictitious schema, not as a state- ment of verifiable fact; I have made no such denial. Again he quotes me as saying that “the organism is only something happening,’ when what I said is that “The organism is something happening.” The difference is like the difference between black and white. I was trying to insist upon certain facts that had heen commonly left out of account,—not trying to substitute these facts for everything else known. Jenninos, Uexkill on Physiology of Behavior. 333 ism at certain junctures. Having abandoned (in favor of the con- struction of fictitious machines) the requirement of finding out what are the real forces at work in organisms, of finding out what machines actually do exist (as determined by the test of verification), the author finds himself in opposition to physiology, which searches pre- cisely for the real (verifiable) forces, materials and machines of organisms. To escape this opposition, v. Uexkiill renounces physiol- ogy and all its works; renounces finding out the causes of things, and ‘alls himself a biologist only; biology he maintains has an entirely different purpose from physiology. ‘We distinguish two sciences of animate nature; Physiology, which arranges her materials according to causality; Biology, which arranges her materials according to pur- posiveness (Zweckmiissigkeit )” (29, Vorwort). The purpose of biol- ogy is to work out the plan according to which the body is made up and acts (33, pp. 10, 11, ete.). The materials—the actual chemical and physical substances, properties and forees—used in realizing the plan, do not concern biology, but form the field of physiology (28, p- 376). Hence the biologist may content himself with schemata which reproduce what the organism does, even though the organism and the schema are operated by different forces acting on different materials in different arrangements. Thus “When I for example lay out the plan of structure of a worm, and in so doimg use any con- venient physical schema, it doesn’t occur to me at all to touch upon a physical problem. One may always think of any other force as at work in the same object. I am not concerned with that. I seek only for a fitting expression in order to make the plan of the animal anschaulich” (28, p. 377). The biologist need not concern himself with causal questions; with physiology. “It is therefore not to be complained of if we biologists, who are asking about the fune- tions of animals, look with much coolness at the end problems of physiology” (28, p. 377). ‘ Renouncing then a causal study for biology, and holding that “Anschaulichkeit” or the demonstration of the production of proc- esses In a machine-like way is the “‘most essential of all” things in biological explanations, v. Uexkiill naturally gets into serious diffi- culties when he confronts processes which he is unable to present as 334 ‘fournal of Comparative Neurology and Psychology. “anschaulich” by “searching about for a satisfactory mechanical scheme of structure” (31, p. 188). Such he feels that he finds in all developmental processes, both in development from the egg, and in the development of new features in movement and the organs of movement. “It is greatly to be regretted that we must give up the hope of finding an anschaulich structural schema for animal develop- ment. But there is no structure that could explain (veranschau- lichen) its own production” (31, p. 185). Since, then, it is impos- sible to bring development under the only poimt of view which seems to v. Uexkiill to give a satisfactory explanation, he finds it necessary to take refuge in vitalism. He is, however, under no illusions as to vitalism’s being an explanation; it is a mere renunciation; “when we therefore give over the production of form to vitalism, this giv- ing over involves a renunciation of all real understanding in this science” (31, p. 187). In his latest paper v. Uexkiill counts himself, if I understand him correctly, as a vitalist so far as developmental processes go, but as a “‘machinalist” so far as the functioning of developed organs is concerned (33, p. 14). If in place of making “Anschaulichkeit” the end to be reached, one takes verifiableness as his aim, a very different set of views will be reached in biology. There are many fields of exact science in which such ‘“Anschaulichkeit” as v. Uexkiill demands is not re- quired. To understand how water is produced from oxygen and hydrogen, most chemists do not construct a fictitious machine on the plan of a flour mill or a dynamo. They merely accept the fact as a datum, in connection with other similar facts. V. Uexkiill himself mentions a number of fields of science which are not ‘“anschaulich” in character (33, p. 16), so that it seems extraordinary to found vitalism on the basis that biology is similar to other sciences in this respect! The only condition that science requires in order that accepted principles of explanation shall apply is this; that differ- ences in resulting conditions shall always be found to be pre- ceded by differences in foregoing conditions, so that nothing shall happen undetermined. But why oxygen and hydrogen in the pro- portion of one to two should give the properties of water rather than those of aleohol we do not know any more than we know why Jenninos, Uexkiill on Physiology of Behavior. 335 in biology one combination produces a sea urchin, another a star- fish. Throughout both chemistry and biology we find unpredictable results produced by new combinations. The repeated changes shown by the development of an organism seem, as to intelligibility, quite on a par with a series of transformations due to recombinations of chemicals. If in either field the same combination under the same conditions should sometimes produce one result, sometimes another, then indeed science would be in distress, and it biology were the field in which this occurred, then the biologist might perhaps grasp at vitalism as a drowning man grasps at a straw. Our quotation from vy. Uexkill (given above, p. 316), in which he holds that Driesch has shown that the germ cell “does not possess a trace of machine-like structure, but consists of throughout equivalent parts” and that it is “structureless,” perhaps implies that he conceives this distressing condition to have been reached. But those who have spent years in working with the astoundingly complex machine- like structures and processes in the chromatin of the germ cell, and have considered the demonstrative evidence brought forward by Boveri, Wilson, Herbst and many others as to the distinctive func- tions of these various parts in development, will find the statement that the germ cell is structureless and composed of throughout equiva- lent parts so absolutely schematic and fictitious as to omit all the truth! Taking verifiableness as our aim will likewise leave biology and physiology resting peacefully in union. We shall be interested in the plan of the organism so far as it 1s verifiable; and to work out the verifiable plan we shall be forced to consider the actual forces, inaterials and arrangements, not fictitious ones. Doubtless physiol- ogy has in practice become narrowed; the remedy les in broadening it till it includes everything verifiable in the study of the processes of organisms. Criticism of theoretical points is not a proper close for a con- sideration of work of such solid value as that of v. Uexkiill. Though we may differ from him in theoretical ideal and in method of pres- entation, we must recognize the fundamental soundness of his methods of actual work. Never was a truer principle set forth for 330 © “fournal of Comparative Neurology and Psychology. successful biological investigation than that the first requirement is “The continued and accurate observation of the ving animal in its environment” (29, p. 75). And v. Uexkill has done more toward an analysis of the internal processes in the behavior of lower animals than perhaps anyone else. The Journal of Comparative Neurology and Psychology Votume XIX JuLy, 1909 NuMBER 4 IMITATION IN MONKEYS. BY M. EH. HAGGERTY. From the Harvard Psychological Laboratory. WitH THIRTEEN IIGURES. CONTENTS. PAGE Tee InGroductory, Stabeme>nntsy 2. s/n. saree serssene oncisi aes erenel clsteenehelopsl ook eteheeiene 337 hie Descriprion=ands Care of Animals Sitidledarsneyarmrrrereledenerste et aeieiee lel 340 PI Method Oe Invest atone cctesccs.o0e ain ie tcl edave, Suenerenel se acoeelleseusre aietere Rea 348 Vie HE XPELIMEeNntS andy RESULES! soars wisescareve eletey cuca ehatonen oreo clit Roteucisueber heute 355 V. General Summary of Results and Conclusions ................<.. 433 I. IntTRopDUcTORY STATEMENTS. 1. Statement of Problem. Popular opinion has generally attrib- uted to monkeys the ability to learn by imitation. As will appear later, experimental evidence on the matter has been of a conflicting nature, but in the main it has not supported the popular belief. The general problem of imitation presents itself in the form of two questions: Do monkeys imitate human beings? and Do they imitate one another? It is conceivable and, indeed, quite probable that an animal which fails to copy the acts of persons, may yet imitate indi- viduals of its own species. In the native state, monkeys must have innumerable opportunities to imitate one another, whereas they rarely, if ever, have opportunity to imitate human beings. Further- THE JOURNAL OF COMPARATIVE NEUROLOGY AND PSyCHOLOGY.—VouL. XIX, No. 4. 338 “fournal of Comparative Neurology and Psychology. more, a monkey lifting a latch is a very different stimulus for an observing monkey from a person lifting the same latch. In view of these considerations it is important In an experimental study of imitation in monkeys to deal separately with the two questions pro- posed above. The first question, Do monkeys imitate human beings ¢ is only indirectly related to the natural activities of the animals; the second, Do they imitate one another? is extremely important for an understanding of the behavior and mental life of monkeys. To discover in what ways certain species of monkeys are influenced hy one another’s acts has been the chief aim of the investigation which I have here to report. 2, History of Present Investigation. The investigation was be- eun in the Harvard Psychological Laboratory in October, 1907. From that time until June, 1908, the experimenter devoted himself (a) to studying the behavior of three Cebus monkeys; (b) to making experiments with these individuals for the purpose of developing methods of testing imitative ability, and (c) to devising and con- structing apparatus for experimental work. In June, 1908, the investigation was transferred to the New York Zodlogical Park in order to make use of the large collection of mon- keys available there. The apparatus which had been built in Cam- bridge, and two of the Cebus monkeys which had been used in the preliminary experiments were taken to the Park. Here, under peculiarly favorable conditions the investigation was continued until September. Well-prepared apparatus and methods of experimental procedure, the fine collection of animals and the excellent local conditions provided by those in charge of the Park, greatly facili- tated the work and within the short space of ten weeks much was accomplished in the way of results. 3. The Work of Other Investigators.—Noteworthy observations concerning the imitative ability of monkeys have been made under experimental conditions by Thorndike’, by Kinnaman’, by Hob- ‘THORNDIKE, Epwarp L. The Mental Life of the Monkeys. Psychological Review, Monograph Supplement, vol. 3, no. 5, 57 pp. 1901. "KKINNAMAN, A. J. Mental Life of Two Macacus Rhesus Monkeys in Cap- tivity. American Journal of Psychology, vol. 13, pp. 98-148 ; 173-218. 1902. Haccerty, Imitation in Monkeys. 339 house,* and by Watson.* In the main these observations are but indirectly related to the present investigation, for they are largely concerned with the animal’s ability to copy the acts of human beings. On this ground, the work of Hobhouse, which gave positive results, may be excluded from this discussion. The other three investigators, who studied the tendency of monkeys to imitate one another, used, in one form or another, the problem method. One monkey was taught to get food by manipulating a mechanical device; then another monkey was allowed to learn the act by watching the trained animal perform. None of the investigators has given the problem an extended study, since the observations in this particular were incidental to studies of wider scope. Thorndike reports a series of five experiments on a Cebus monkey. This animal, “No. 3,” was, at the time of the experiments, ‘‘on terms of war” with No. 1, the animal he was to imitate. In none of the imitation tests did ‘‘No. 3” learn to do the act. Thorndike concludes: “There is clearly no evidence here of any imitation of No. 1 by No. 3. There was also apparently nothing like purposive watching on the part of No. 3.” “This lack of any special curiosity about the doings of their own species characterized the general be- havior of all three of my monkeys and in itself lessens the proba- bility that they learn much from one another.’ Kinnaman observed two cases where the conduct of a male rhesus caused the female to learn an act. The problem was to get food by manipulating a mechanism—in one case, the pulling of a plug, in the other, the bearing down of a lever. In each case, the female was given opportunity to get food but failed. The male was then allowed to get food while she was present and watching. In each ease she went at once, after seeing the male get food, and operated the mechanism and repeated the performance numerous times later. Kinnaman says: “Here we have a copy in the form of an act. It was copied almost in detail, and that, too, so far as the place of *HosHouse, L. T. Mind in Evolution. Chap. X. London. 1901. ‘Watson, JOHN B. Imitation in Monkeys. Psychological Bulletin, vol. 5, pp. 169-178. 1908. Sar): sp. 42: 340 ‘fournal of Comparative Neurology and Psychology. laying hold of the plug and the direction of the pull were con- cerned, both requiring very radical changes from the monkey’s own previous efforts. He also says, “It seems to me that the two cases with the box are quite as good examples of imitation as could well 298 997 be gotten even with human beings. Watson’s contribution to this subject is the latest and agrees with Thorndike’s in giving negative results. He reports three imitation tests made upon two Macacus rhesus monkeys. In no one of these tests did the watching animal learn to get food by seeing another animal get it. He concludes, “I unhesitatingly affirm that there was never the slightest evidence of inferential imitation manifested in the actions of any of these animals.’”® If we group the work of the three investigators together, we have ten imitation tests in which four animals were used. One animal manifested imitative behavior in two different tests. None of the other three animals showed any tendency to imitate. From such fragmentary and conflicting evidence it is impossible to conclude what role imitation plays in the behavior of monkeys and the need for further investigation is apparent. 4. Acknowledgments.—In presenting this report of my investiga- tion, I gratefully acknowledge my indebtedness to the Harvard Psychological Laboratory and, in particular, to Professor Robert M. Yerkes, at whose suggestion I undertook the investigation. His sympathetic codperation at every stage of it has been invaluable. To Dr. William T. Hornaday, Director of the New York Zodlogical Park, I am deeply indebted for the opportunity to use the facilities of that great institution. His interest and generosity did much to further my work. The photographs which are here reproduced were made for me by Mr. E. R. Sanborn, the Staff Photographer of the Park. I am grateful for his services. II. DescriprioN AND CARE OF ANIMALS STUDIED. 1. Cebus Monkeys. (a) General Characteristics—In my experiments I have used ™. 144. Spe 22. es Us Haccerty, Imitation in Monkeys. 341 eleven animals from two genera and seven species. Eight of them represent five species of Cebus monkeys. This is the genus with which we are familiar as the consort of organ grinders. The home of these monkeys is South America, especially the head waters of the Amazon and northward into Central America, where they live a gregarious life in the tree tops, feeding on fruit, nuts and insects. TABLE 1. NuMBER, SPECIES, SEX AND PROBABLE AGE OF ANIMALS USED IN THE INVESTIGATION. | No. Species. | Sex. Age. Remarks. No. 1........| Cebus lunatus _ Female 3 years | Bought of dealer in New York. NOR oe | Male 3. Bought of dealer in New York. Nowe me re “ hypoleucus | “ | Dene Bought of dealer in | New York. INOF 4is.fo2s, 2". “ fatuellus | Female Gries Had been several | years in Park. INOR Osaaete: SS eecApucinus.: > | . Dee Had been several | years in Park. Nos 'Giaer, ace Pn lunatus Male aa Had been several years in Park. INOS Sen she “ hypoleucus oo dame In Park but eight | | weeks. NOZO.. es “flavus a 1 year In Park but eight weeks. No mlOtsy) fr: Macacus rhesus Female 4 years | In Park two years. Nome o gs | Male Sas In Park two years. INOS ore. 4 cynomolgus, | eee. In Park two years. They travel about by leaping from one tree to another; in this arboreal life their long grasping tails serve them better than a fifth hand would. The facial portion of the skull is small in com- parison with the cranial portion, and many specimens have quite prominent foreheads. Forbes notes that the cerebral cortex is almost as much convoluted as it is in the Old World Apes. The forehead, usually bare of hair, is often wrinkled, giving the mon- keys the appearance of being “burdened with sorrows, which,” as Dr. Hornaday remarks, “most captive monkeys certainly are!” The Cebus monkeys are cowards except toward those they can easily vanquish. One fight is usually enough to settle the supre- macy of a cage. The whipped animal seldom makes another effort 342 ‘fournal of Comparative Neurology and Psychology. to rule. The victor, however, often delights in continuing punish- ment which the vanquished receives with howls and shrieks of fear. The noise made by the victim is out of all proportion to the injury inflicted. A slap, a theft of banana, or even a threat often arouses piercing shrieks. No. 6 and No. 4 were together one day in a small cage. It was about feeding time and both wanted to be at the wire front. No. 6 was in the way of No. 4 and she slapped him with the palm of her hand. He retreated and doubled up in his characteristic fashion, moving his body up and down and yelling loudly. Any movement of No. 4, even so much as the turn of her head toward him, served to release another volume of shrieks. This continued for several minutes with no further demonstration on the part of No. 4. On another day, No. 4 was sitting on a brace in the experiment cage with her hands on the wire. Without allowing her to see me move, I touched my finger to the back of her index finger. As if struck by an electric current she leaped to the floor and began to yell vehemently and continued to do so for some time. IT am informed by Dr. Hornaday that the Cebus monkeys which are marketed in this country are obtained when quite young. The offspring rides about on the mother’s back and hunters shoot the mother, who falls to the ground with the young still clinging to her. The small animal is then caught and kept in captivity until the keeper desires to ship it to market. This makes it next to impossible for any one who buys these monkeys of dealers to know much about their previous experience. In a study such as this, however, it is desirable to know all that ean be known of each animal’s normal activities, so at the risk of multiplying words, I shall give a brief account cf each animal used. (b) Characteristics of Individual Animals.—No. 1, Cebus lunatus, female, and No. 2, Cebus lunatus, male, were obtained from an animal dealer in New York City. When they came to the Har- vard Psychological Laboratory in November, 1907, they were ap- parently about three years old and were in excellent physical condi- tion. No. 1 made herself at home from the start and on the third day Haccerty, Imitation in Monkeys. 343 would sit on my knee and eat her banana out of my hand. Within a short time she would ride on my shoulder as I walked about the laboratory, thus being sure to keep near whatever food I might have in my hand. No. 2, however, was more cautious, never coming near except when No. 1 preceded him, and retreating whenever he eot his food. His favorite position was sitting on the floor of the cage with No. 1 sitting in front, and his arms clasped tightly around her body. When No. 1 moved, No. 2 would start ner- vously and try to keep close to her, never once taking his sparkling brown eyes off the persons in the room. Gradually his fear wore off and with No. 1 he went curiously about the cage, biting at every projecting piece of wood, and poking his fingers into every crack and cranny. | _IwrratinG No. 26 | Number of times : j | No. 2 performed wamber of times | Number of times | ‘Time in Date—1908. ier No. 1 saw No. 2. |No. 1 saw in part.| Result. | Minutes. Jan. 30a 7 3 3 Failed. | 30 Jan. 30b 6 3 3 LL Be | 30 Jan oles 22. 11 8 3 Fos “30 Repeies. she 4 17 10 7 F 30 Heb. Gos... 10 10 | -F 30 Kebsl0. .. 12 10 2 F 30 Heb. U7 22s. 12 | 10 2 F 30 Feb. 18... . .| 13 | 10 1 aa 30 Hebetore .: 10 10 hee yee 30 Feb 20es< 12 10 2 F 30 Ben eots 10 10 F 30 Bebe 25.5... 12 10 2 Note 2 30 Rebsa26se0-4| 22 20 | F | 30 Heb 27. 20 20 1 F 30 Hebeose ee. . 23 20 3 he ar 30 Hebio9.... 33 | 20 3 F 30 March 2.... 23 | 20 | 3 oe 30 | } Motte... | 253 rae F204 35 ek 510 S means fine the imitator did ear fie i. ste ior of the imitatee. . F means that the imitator failed to repeat the behavior of the imitatee. The time is always given in minutes unless otherwise indicated. It was taken with an ordinary watch and where it is recorded in seconds the time was taken from the second hand. How No. 1 Learned.—March 8. No. 1 had been given more than two hundred opportunities to see No. 2 perform the operation and had profited frequently by getting food. It then seemed certain that she would not learn to work the device from seeing No. 2 do it. A stick, two inches wide, was placed from the wire front of the cage to the chute. Within eight minutes, No. 1 had climbed the side of the cage, had walked on the stick to the chute, had swung down and thrusting her hand up the chute had opened the door. The stick was then removed while No. 1 was on the floor. When her food was eaten she became very active, making long leaps all about 358 fournal of Comparative Neurology and Psychology. the cage, but never once to the chute. She then settled down to her usual behavior and after ten minutes the stick was replaced. She pulled the string once and the stick was again removed. She then resumed her usual conduct for almost ten minutes more. Then, when under the chute, she looked up and her eyes accidentally fell on the chute. She rushed up the front of the cage and leaped to the chute, swung herself down and worked the device. This she repeated several times, although in a much less skillful way than No. 2 was able to do it. 2. CHUTE EXPERIMENT B. A. Description of Device. This apparatus was a modification of the one used in Chute Experi- ment A. Frem the top of the experiment cage (fig. 3), 30 em. from the board side and 40 cm. from the board end, a hollow wooden chute, a, 5 cm. square, projected into the cage a distance of TO cm. Inside this chute, 40 cm. from the lower end was a trap door, 0, hinged to drop downward, but held up by a rubber elastic. To the bottom of this trap door was attached a string, ¢, which hung down to within 10 cm. of the lower edge of the chute. A coiled wire spring, d, was tied to the lower end of the string to serve for a hand-hold for the animals. In the top of the cage was a tube leading into the chute from a small feeder (fig. 19, f,) adjusted on top of the cage. By pulling a string, attached to the feeder, the experimenter could drop food (sunflower seed and chopped peanuts) upon the trap door.* Two horizontal rungs nailed to the sides of the chute helped the animal to support himself on the chute. The problem set for the animal was to leap from the wire front or end of the cage and, while holding to the chute, to swing his head and shoulders down, thrust one hand up the inside of the chute, grasp the coiled spring, and pull the trap door down. The food on the top of the door would then fall to the floor, unless checked by striking the body of the animal. In either case the animal could get it. B. Behavior of No. 2. No. 2 had learned to work the mechanism in Chute Experiment A. But when he was first put into the new cage, four months after his last experi- ence in the old one, he apparently had no memory of the chute. Only after several minutes did he go to it. He jumped to it from the front wire. He stopped to examine the opening in the end before he thrust his hand into it. In the old cage he usually thrust his hand into the opening without “The essential part of the feeder was a copper plate, 38 cm. wide and 6 mm. thick, arranged to slide back and forth in grooves beneath a food hopper. In the plate was a circular opening into which the food dropped from the hopper. When the string was pulled this opening, full of food, was drawn over a tube leading into the chute, into which the food dropped. Haccerty, Imitation in Monkeys. 359 looking at it. After examination, he pushed his hand up the inside and touched the string. Then he became very eager to work and would have worked continually (fig. 4). C. Behavior of No. 13. Preliminary trials.—VFirst trial, August 23. On entering the cage, No. 13 climbed to X where he sat for a short time. He then walked along the brace until opposite the chute, when he leaned out toward the chute and touched it with his hands, a little way above the lower edge. Drawing his body back from the chute, he walked along to X and went down to the floor. Several times he climbed up and down the wire parts of the cage and then tried to get out at the door. He went back and forth from the door to X. In his endeavor to get out of the cage, he pushed on the small doors repeatedly and even climbed the wire front to push on the upper part of the large door. Next he settled at XY and looked about. He moved down to the floor and up and down the wire. Moving along the brace to the door, he tried to push it open. He then descended to the floor. He tried to climb the corner post opposite X and then climbed the wire to the top of the front, going to the floor again and then up to XY. He was active during the entire time he was in the cage, but he took no notice of the chute after the first three minutes. Second trial, August 24. The movements of No. 18 on this day were as follows: Up the end of the cage; to the floor; up to the brace and along the brace to opposite the chute; leaned over to the chute and put hands on the lower rung; felt up and down the edges of the chute nearest him; tried to bite the edges of the chute; back to the wire and to the floor; again up to the brace and leaned over to the chute; again bit the edges of the chute; to the floor, about, and looked up at the chute from exactly under it; to the door; up to X and perched; to the floor and hunted about; bit the door; again to X and perched, looking out through the wire; down to the floor to get a roach and back to the brace; up the wire front, and back to X; to the door and back to X; along the brace; putting his hands over on the chute, he swung his body to the chute and climbed up on it to the top of the cage and looked at the screws, etc.; tried to bite rung; back to brace and perched at YX. Third trial, August 26. Behavior as follows: Climbed wire; perched at X; to the floor; to the door; up front and around to the top of the wire end; back to X where he sat for a little time; to the floor and hunted all about; to the place where L* had been and hunted for it; back to X and along the brace to the large door; turned around, faced the chute, but paid no attention to it; to the floor, all about and back up to XY; along the brace to the chute; leaned over to it; put hands on the near edges and looked up and down; grasped rung in one hand and pushed other up and down the edges of the chute; back to X; to the chute and clung to it while he examined the top of *An opening where he had gotten food in a previous experiment. 360 Fournal of Comparative Neurology and Psychology. the cage; on the chute two minutes and then slid down to the end and dropped off; again to the chute and examined the crack in the top of the cage; perched on the rung and looked out of the cage; bit at the edges and then jumped off. Fourth trial, August 27. Behavior as follows: To X and sat down; along the brace and back to X and to the floor; up the front and along the brace to X and up the end of the cage; to the floor and about all the edges looking for food; up to X, along the brace, and to the floor; about the floor and to the door, which he tried to open; to the end of the cage; up, around to the front and sat at XY; up the end of the cage and down to the floor; to the corner opposite X; to the front, up and sat at Y; up and down the wire; up the front, around to the end and back to the front; to the end and down to the floor; up to X and sat; up the front of the cage and shook it. FWitth trial, August 29. Behavior as follows: Up and down the end of the cage; to the door and up the front; to the chute and sat on the rungs; back to front and to X where he sat for some time; to the floor and to the door, which he tried persistently to open; climbed the front and looked about; to the floor and pushed on the door very hard; up the front and perched at X; up the end of the cage and back to X; along the brace and pushed at the chute; to X and down to the door; up the front and to the chute; back to the brace and along it to the end of the cage and back to X; to the floor and back to X; looked up at X from the floor; afterward climbed to X and sat there during the remaining part of the time. Imitation test—No. 13 imitating No. 4.* First test. No. 13 was put into the observation box and the box was set on the floor of the cage so that No. 15 could have a good view of the chute. No. 4 was put into the cage and, at once, began to get food from the chute. No. 13 was attentive to every movement. His record in seeing is as follows: 4 Performance 1. No. 15 saw perfectly and became very threatening and eager to get out of the cage. Pp. 2. Just as No. 4 thrust her hand up the chute, No. 13 looked down. As a result he did not see the pull. He saw her eating food and shook his box with such force that he moved it about over the floor. P. 3. No. 18 saw perfectly and sat on the floor of his box attentively watch- ing No. 4 eat her food. P. 4. No. 15 saw perfectly and was eager to get out. P. 5. No. 13 saw fairly well; he was eager to get out of the box; failing to get free he sat on the floor of his box and watched No. 4 eat the food. These performances did not occupy more than five minutes. No. 4 was now removed and No. 13 was released in the cage. At first he looked about over the floor for food and then climbed the front wire, stopping on the brace opposite the chute. He leaned over to the chute and while still standing on the brace with his feet, tried to thrust a hand into the bottom of the chute. Ivailing in this, he ran along the brace to X and back again to opposite the *The learning of No. 4 will be given later. Haccerty, Imitation in Monkeys. 301 chute; catching the rung of the chute in his hands he drew himself over to it; finding himself above the end of the chute he tried to let his body down, first on one side and then on the other, until in the most awkward manner he managed to get near enough to the end to thrust a hand up the inside far enough to reach the string. At once he pulled and the food came tum- bling down on his chest and to the floor. Dropping to the floor he picked up the food and ate it. The time from the removal of No. 4 was 40 seconds. Within one minute he climbed the front wire, reached the chute, and got food in the same manner. On reaching the chute the third time he did not pull himself above the end, but holding to the rung with his hands he dropped his body below the end and placing his feet against the back of the cage steadied himself while he thrust the free hand up inside and pulled the string. Time: 40 seconds. From this time on No. 13 repeated the per- formance as rapidly as his food was eaten. Within ten minutes he had gotten food eleven times and had eaten it all. From the moment he was released in the cage he seemed bent on getting the food. In his efforts, he made but one useless movement, namely, when he drew back from the chute after first putting his hands on it. This, however, did not indicate a wavering from the end in view. It was merely a drawing back for the renewed effort which he immediately made. Summary of Behavior of No. 13 in Chute Haperiment B. During the preliminary trials No. 18 was exceedingly active, but at the end of the time he had made no progress toward a solution of the problem. He had gone to the chute, but there was no evidence that this was more than a random act in his movements about the cage. He did not notice the end of the chute and in no way did he seem to connect the chute with getting food. During his last trial he was quiet much of the time. After his preliminary trials he saw No. 4 getting food at the end of the chute five times. He was confined in an observation box so that he could not follow No. 4 about. He did not get any food and he experienced the result in no way. However, when he was released in the cage his behavior was strikingly different from his behavior during any of his preliminary trials. He went almost directly to the place where he had seen No. 4 get food and within two-thirds of a minute he had gotten food for himself by doing essentially the same act No. 4 had done while he was watching her. TABLE 3. No. 13 Imiratine No. 4. Number of times rs pe EAS ; ¢ Date No. 4 performed Number of times | Number of times Result. Time in says. avert No. 13 saw. No. 13 saw in part. Seconds. 362 “fournal of Comparative Neurology and Psychology. D. Behavior of No. 4. First trial. No. 4 spent her time on the floor and the sides of the cage. She was fairly active. After four minutes of random movements about the cage, she hung by her tail and two feet to the front of the cage opposite the chute and swung her body around crane-like toward the chute, looking at it steadily. She then moved about the cage as if she had nothing to do; she either sat quietly or leisurely climbed the cage. Second trial. No. 4 walked about on the floor; then climbed the wire and looked about. Once or twice she examined the cracks in the floor and in the door. She looked at the chute twice and looked out through the wire toward the window. Third trial. During the third period No. 4 spent her time on the floor and in climbing the wire. Several times she pulled on the brace across the front of the cage and then remained quiet. She paid no attention to the chute during the entire time she was in the éage. Fourth trial. On the fourth day No. 4 spent most of her time perched on the brace. She varied this by climbing up and down, catching roaches, and looking out through the wire and the window. She displayed no interest in the chute during the entire time she was in the cage. Fifth trial. On the fifth day No. 4 spent her entire time on the floor, op the brace, and in climbing the wire. Most of the time she sat still, and when disturbed, simply changed her position and settled down again. No. 4 paid most attention to the chute on her first day’s trial. On the second day she gave it less attention, and on the third, fourth and fifth days none whatever. Imitation tests—No. 4 imitating No. 2—The animals were put into the cage together. At first No. 2 was afraid of No. 4, who walked about the floor and climbed the wire at her will. As No. 2 would not work at the chute because of his fear, No. 4 was put into the observation box and the box was placed on the floor of the cage. No. 2 was still afraid and refused to work for some time. After twenty minutes, he leaped to the chute and pulled the string. No. 4 did not see him, but some of the food fell into her box and she ate it. Fifteen minutes later No. 2 jumped to the chute, but he did not pull the string. No. 4 saw him on the chute. Later No. 2 jumped to the chute, pulled the string and caught a seed on his chest. No. 4 saw him on the chute, but did not see him pull the string. The next time No. 4 saw nothing, but got food. No. 2 then became more frightened at No. 4 and refused to jump to the chute during the rest of the morning. Since No. 4 had not seen the entire performance once, she was not given an opportunity to get the food. Second trial. This trial was made on the afternoon of the same day as the previous test. No. 2 was still much frightened and worked very slowly. The first time he pulled the string and got food, No. 4 was looking. He pulled the string again, but not hard enough to get food, and No. 4 saw him. No. 2 did the same thing again and No. 4 saw him. The fourth time No. 2 pulled the string he got food, but No. 4 did not see. In all, No. Haccerty, Imitation in Monkeys. 363 4 saw No. 2 at the chute and pulling the string twice; once she saw him with food at the end of the chute, and twice she got food which fell into her box. No. 2 was now taken out and No. 4 was released from the observation box. She at once climbed the wire front opposite the chute. Then she leaned toward the chute as far as she could while still holding to the wire with one hand. She drew herself back and descended to the floor, went to the door and then to the wire end, climbed the end opposite the chute, threw her head, shoulders and arms toward the chute, catching the lower part of it in her hands. Then she let go the wire with her feet and tail and drew her body over to the chute, catching it by her feet and wrapping her tail around it. She then swung her head down under the chute and looked up into it, at the same time thrusting her hand up inside. She rattled the metal hand- hold against the side of the chute and in a moment pulled it. The food fell on her chest and on the floor. The interval was less than one minute, from the time No. 2 was taken from the cage. She then dropped to the floor, ate the food, and climbing the front of the cage, leaped to the chute again and repeated the act in two minutes. She repeated it again in three minutes, and again in five minutes from the time No. 2 was removed, in the mean- time, eating all the food that fell to the floor. She repeated the act again in one minute and six times more within the next twelve minutes. In all she operated the mechanism eleven times in twenty minutes and ate all the food—about thirty sunflower seeds. She would now work the device as often as she got the food eaten. Her manner of solving the problem was direct from the first, and with one exception, without loss of time or motion. Summary of Behavior of No. 4 in Chute Experiment B. No. 4 was quite active during her first preliminary trials, but during the later ones she was more quiet and wholly indifferent to the presence of the chute. The conditions of her imitation test differed from the test of No. 13 in the fact that No. 4 herself ate some of the food that came from the chute when No. 2 pulled the string, whereas No. 183 had only seen without experiencing ae result of the act. The behavior of No. 4 after being released in the cage was like that of No. 18, in that there was a marked change from the behavior of the preliminary trials. She went directly to the chute and performed the act she had witnessed, securing the same result. TABLE 4. No. 4 Imiratine No. 2 |ieee Ton Dee Nae Number of times | Number of times Result Time in ; ” aye Giese. No. 4 saw. No. 4 saw in part ; Seconds. duly. 29 ...°.: 3 ) 1 No test. July 29.. 4 | 2 | 2 8 55 | Potals.2 25 7 2 | 3 S 55 364 “fournal of Comparative Neurology and Psychology. EH. Behavior of No. 11. Preliminary trials. First trial, August 24. No. 11 was very active and very hungry when put into the cage. He moved about as follows: Across the floor to the end; up the wire and down again; to the door and looked up at the chute; chewed and pushed at the door trying to get it open; to the front and to the end; up, and back to the floor; to the door; to the front and back to the door, where he was very vigorous in his efforts to get out; to the front; up to the brace and looked all about the chute; shook the cage; to the floor and to the end; looked out through the wire; up the cage and shook it vigorously; to the door and made frantic efforts to get it open; repeated this soon again; up to XY and perched; to the floor and about; tried the door again and walked about the floor; tried the door again and walked about the floor; up to XY and perched; to the floor and up to the wire front; shook the cage vigorously and returned to the floor; again made frantic efforts to open the door; up to X and sat on the brace; to the door and frantic to open it; to the end of the cage and up the wire; around to the front and reached one arm over to the chute and shook it; to the floor and about; again climbed the front of the cage and reached to the chute; to and perched. Second trial, August 25. Behavior as follows: Up to the brace and down; repeated; while on the wire looked at the chute; shook the cage; to the end of the cage and down to the floor; looked all about; up the front and shook the cage with great vigor; down to the floor and searched about for food; up the front and shook the cage again; perched on the brace and sur- veyed the chute carefully for some time; perched at XY; to the floor and sat near the end; to the front and sat looking out through the wire; to the door and tried to open it; up the front and perched at XY; carefully surveyed the inside of the cage and looked out through the wire; looked squarely at the chute; up the front and shook the cage vigorously; back to X and sat for some time; to the floor and about. Third trial, August 26. It was past feeding time and the animal was abnormally hungry. He was, therefore, fed on entering the cage, but not enough to satisfy him and he went about the cage as usual looking for food. His movements were as follows: Up the front of the cage and along the brace to the end and down to the floor; from underneath the chute he looked up at it steadily and then climbed the end to X where he perched and looked about the cage; shook the cage vigorously and perched again; to the floor and tried to open the door; looked toward the chute and then climbed the end of the cage to X; along the brace; back to XY and to the floor; up to X and down the end of the cage again to the floor; sat under the chute; jumped to wire front and ran along the brace to X; to the floor; picked up some hulls and smelled them; repeated this several times. Fourth trial, August 27. No. 11 behaved as follows: Up the end and down to the floor; about the floor and up the front to X, where he perched; to the floor and back; to the floor and sat, looking out through the wires; up the front and shook the cage; to the floor and sat at the end of the floor Haccerty, Imitation in Monkeys. 365 of the cage; walked about the floor and climbed the front of the cage; to the floor and sat near the end of the cage; crossed the floor to the door and climbed the front wire; along the brace to Y, where he perched; to the floor and looked about; to the door and up the front to VY; down to the floor, where he sat looking out at the end wire; very alert, but the chute apparently had no interest for him; to the door and pushed, in an effort to get out; he had been very eager to get into the cage, but was now just as eager to get out. Fifth trial, August 28. The behavior of No. 11 on this day was evidence that he did not expect to find food in the cage. Most of his efforts seemed to be directed toward getting out of the cage. There was no reason for his desiring to leave the cage other than the lack of interest on the inside and his desire to be back in the cage with his mate. He was not in the least frightened. His behavior was as follows: Up the end of the cage and down to the floor; up the front and down to the door; about the floor and to the top of the wire front; around to the top and down to X, where he sat for some time, uttering a cooing call; after a short time, to the floor and about; sat down near the wire end for a time; up the front and shook the cage; to X and perched; to the floor and grabbed the front wire, shaking the cage very vigorously; up to Y and perched for some time; down to the floor and back up to XY ; along the brace and back to XY, where he stayed during the remainder of the time. Imitation tests—No. 11 imitating No. 4. First test. No. 11 was placed in the observation-box, which was then placed on the floor of the cage; No. 4 was free in the cage. No. 11’s record in seeing was: Performance 1. No. 11 was distracted; saw No. 4 on the chute, but did not see the pulling of the string. He saw her eat the food on the floor. P. 2. No. 11 saw No. 4 on the chute; saw her swinging at the end; saw her pull the string and get the food. P. 3. No. 11 saw No. 4 leap to the chute, swing down, pull the string, and get the food. P. 4. No. 11 saw No. 4 on the chute; saw her swing down and get the food; he jumped at the side of his box in an effort to get out. No. 4 now spent some time on the floor getting food and No. 11 watched her attentively passing from one end to the other of his box as No. 4 walked about. Several times he jumped at her striking the side of his box; when she climbed upon his box he became very threatening. P. 5. With his eyes No. 11 followed No. 4 about the floor and up over his box to the chute. During the pulling of the string his eyes were riveted on her. She dropped one seed in his box and he ate it. P. 6. Same as P. 5, except that No. 11 did not get food. No. 11 was very threatening toward No. 4. No. 4 was now removed and No. 11 was released in the cage. His first movement was to work at the door in an effort to get out of the cage. He then went up to the brace, leaned over to the chute and placing one hand on the side of it, attempted to pull it toward him; he then grabbed the edge of the lower end of the chute in his hand and pulled. Letting go of the chute 366 © “fournal of Comparative Neurology and Psychology. he went to XY, where he perched for some time. Going to the floor he walked about and then looked up at the chute; he tried to jump to it from the floor, but, though he touched the rung with his hand, he was not able to hold. He then walked about the floor and climbed the front of the cage, walked along the brace and leaning over to the chute, pulled it as before. Going to the floor, he tried to climb the corner post near the chute. Failing in this, he jumped to the chute from the floor, holding with both hands. Pulling himself up to the chute, he bit the rungs and then worked his way around the chute biting at all the edges, but not turning his head down to the end of the chute. Leaping to the front of the cage, he descended to the floor and walked about. Once again he jumped for the chute, but failed to hold on. He then walked about the cage and climbed to Y, where he perched for the remainder of the time. Second test. Conditions were the same as in the preceding test. The record of No. 11’s seeing was as follows: Performance 1. No. 11 was looking at the experimenter and did not see. P. 2. No. 11 saw, though his attention was divided between the experimenter and No. 4. P. 3 to P. 5. No. 11 saw fairly well, but did not threaten as on the day before. P. 6. No. 11 saw perfectly. No. 4 was now removed and No. 11 was released in the cage. He found a seed on the floor and ate it. He jumped to the chute from the floor, but could not hold. Sitting down beneath the chute he looked up at it and then walked about the floor looking for food. He climbed to Y, but returned to the floor after a minute, going to the door, where he tried to get out. Failing to open it, he went to the wire end of the cage and sat on the floor. He tried the door again and then climbed to Y, but after one minute came to the floor and sat down. Turning toward the chute, he jumped for it, and catching hold, drew himself up to the chute. For some time he sat on the rung; then he bit his way around the chute. He then shook the chute so hard that the iron attached to the string on the inside rattled. He was then quiet, looking about the cage and at the sides and edges of the chute. Twice more he shook the chute with such vigor that he all but tore it from its fastening at the top of the cage. Becoming quiet, he sat for a moment and then leaped to the front of the cage. He went to Y and perched for a moment; he then went to the floor and sat near the wire end of the cage. Time: 25 minutes. Third test. No. 13 had by this time learned to get food and he was used as the imitatee since, in size and general behavior he was much more like No. 11 than was No. 4. No. 11 was put into the observation-box and the box was placed on the floor of the cage. No. 18 was not at first inclined to work, but moved all about the cage. He finally went to the chute and hunted along the top of the cage for roaches. Several times he jumped from the wire side of the cage to the chute and back to the wire. Performance 1. At last he went to the chute slowly and pulled the string, getting food. No. 11 saw every movement perfectly. No. 13 was suffering from a fall he had gotten a short time before when fighting with No. 10. Haccerty, Imitation in Monkeys. 367 He seemed afraid of No. 11. When he would work no more, he was taken out and No. 4 was substituted. No. 4 operated the chute eleven times. Nine of these performances No. 11 saw. He was alert and every muscle was tense. In all, No. 11 had seen ten times during this test and a total of twenty times in the three tests. No. 4 was now removed and No. 11 was released in the cage. He first looked over the floor for food and finding none, climbed the wire front and went over to the chute, shaking it with such vigor that he almost tore it loose from the top of the cage. Jumping back to the brace he went to X and to the floor. Passing to a position immediately under the chute he jumped up to it from the floor and climbed up on it. Without stopping to make examination he swung his body down, held to the rung with one hand, placed his feet against the back of the cage for support and, thrust- ing the other hand up inside the chute, pulled the string. The food fell onto his chest and on the floor. The time, from the removal of No. 4, was 60 seconds. Having eaten the food, he again jumped to the chute and in the same position tried to pull the string, but not being able to hold his weight with one hand he had to catch with both; he then pulled himself up on the chute, and having regained his equilibrium swung down and got food as before. Again he jumped to the chute from the floor, catching the rung in one hand and curling up so as to grasp the rung with his feet also. Then holding by his feet and one hand, he thrust the other hand into the chute as before and got food. He repeated this in exactly the same way, at once. Again he repeated this in the same way, except that he placed his feet against the back of the cage instead of on the rungs of the chute. From this time on he got food as rapidly as he could eat it, most of the time hanging below the chute with his feet braced against the back of the cage. Summary of Behavior of No. 11 in Chute Experiment B. No. 11’s preliminary trials were much like those of No. 13 and No. 4. ‘They ended with No. 11 not having got food and with his being indifferent to the means of getting it. The stimulus-complex was the same as in the case of No. 4, i. e., No. 11 saw No. 4 getting the food and experienced the result of her act himself. When he was released from the observation-box, his behavior was different from what it had been in the preliminary trials. However, it was not sufficiently like the behavior of No. 4 to bring the same result. His attention had been directed to the chute, but not to that part of it which would enable him to get food. - After his second observation his interest in the chute seemed increased, as evidenced by the great vigor with which he shook it. The third test seemed to direct his attention to the important part of the mechanism and he succeeded in getting food as No. 4 had done in his presence. The result, in the case of No. 11, differed from the result in each of the previous cases 368 + “fournal of Comparative Neurology and Psychology. in that No. 18 and No. 4 both repeated the act which they had seen imme- diately. No. 11, on the other hand, seemed to learn a part of the act at a time, and only after repeated opportunity to see it, did he learn fully to attend to the act as it was performed in his presence. TABLE 5. No. 11 Imrratine No. 4. | Number of times | | Ww Fc : Number of times _Number of times Time in DES [Ses pate Gores No. 11 saw. No. 11 saw in part. PSS: | minutes. | es — =e ee Se Z 3 Aug. 28..... | 6 5 1 F 10 AUIS S28 eee | 6 5 F 10 Aug. 29. 12 10 S 1 Rotel 24 20 1 S) 21 IF’. Behavior of No. 6 Preliminary trials.—First trial. The first few minutes were spent on the floor. After four minutes No. 6 climbed the cage front and reached to the chute with his hands. He repeated this a minute later. A minute later he looked at the chute from the floor, climbed the front of the cage and grabbed the lower edge of the chute in his hands. This he repeated once, and then spent the rest of the time on the floor of the cage. Second trial. On the second day No. 6 climbed about the cage, then reached to the chute and put one hand slightly into the end of it. He gave no further attention to it and went to the floor. Later, he climbed the front and while holding with tail and feet to the wire reached to the chute, clasping a hand on each side of it about four inches from bottom. This he repeated after eight minutes, and once more before the close of the time. Third trial. On the third day, No. 6 reached to the chute as on the previous day, after five minutes in the cage. Later, he reached to the chute and tried to get his hands and feet on it while holding to the wire with his tail. Fourth trial. On the fourth day his only attention to the chute was to look at it once and to attempt to get to it as on the previous day while holding to the wire with his tail. Failing, he spent the rest of his time on the floor. Fifth trial. On the fifth day No. 6 once climbed the wire and looked at the chute. Later, after running about the floor, he climbed the front of the cage and jumped to the chute to get a cockroach on the back of the cage. While there he explored the top of the cage and jumped back to the side. Once more he leaped to the chute, but he leaped back immediately. During the latter part of the time he remained quietly on the floor of the cage. Imitation tests.—No. 6 imitating No. 2.—First test. No. 6 was put into the observation-box, which was set on the bottom of the experiment cage. No. 2 was free in the cage. No. 2 was interested in No. 6 and pretended fight. Once he ran up the wire, jumped to the chute and leaped to the wire again Haccerty, Imitation in Monkeys. 369 at once. Then both animals pretended fight toward another animal which was making a noise behind a curtain. Performance 1. No. 2 jumped to the chute and jumped back to the wire without pulling the string. No. 6 saw. P. 2. No. 2 jumped to the chute and pulled the string. No. 6 saw No. 2 on the chute and saw food fall. P. 3. No. 6 saw as before. An empty shell bounced into the box and No. 6 got it. P. 4. No. 6 saw No. 2 on the chute, looked away, heard the sound of the trap door, looked back and saw No. 2 at the end of the chute and the food falling to the floor. No. 2 now jumped to the chute twice, but he did not pull the string. No. 6 saw him jump. P. 5. No. 2 jumped to the chute, pulled the string and the food fell to the floor. No. 2 now jumped to the chute and jumped back to the wire. No. 6 saw nothing No. 2 did. P. 6. No. 6 saw the entire performance. No. 2 was now taken out. No. 6 was released from the observation-box. He climbed the cage at the front and reaching over to the chute pushed a hand up inside. He could not reach the string. This occurred only 30 seconds from time of release. He then went down to the floor. Immediately, he climbed the wire opposite the chute, jumped to it, threw his head and shoulders down, reached up inside and pulled at the string, but, though he gave what seemed a strong pull, it was not sufficient to open the trap door. He raised his body up, but at once bent down again and looked up the chute. He then leaped to the floor. All fhis happened within two minutes from the time of his release from the observation-box. Four minutes later he repeated the entire performance, and then dropped to the floor. Four minutes later he leaped to the chute, but did not go to the end of it. He explored the top of the cage instead, leaped back to the wire, and went down to the floor. He did not seem as vigorous as usual. Three minutes later, he jumped to the chute and in attempting to get in position at the bottom of the chute, lost his hold and dropped to the floor. This he repeated five minutes later. He held with one hand to the rungs on the chute and allowed his feet and body to hang below. Holding thus -with one hand, he tried to put the other up the inside of the chute and being unable to hold himself longer dropped to the floor. Five minutes later he jumped to the chute and pulled the string, but not hard enough to get the food. Second test. Conditions were the same as in the previous test. No. 2 was now more active and worked rapidly. Performance 1 to P. 2. No. 6 saw No. 2 jump to the chute, then looked away, heard the rattle at the chute and looked back to see No. 2 at the end of the chute and food falling to the floor. Six times more No. 2 operated the chute. No. 6 saw the entire perform- ance each time but one; this one he saw in part. When No. 2 was out No. 6 found a grain of food on the floor of the cage and ate it. He then climbed the wire, jumped to the chute, and swing- 370 “fournal of Comparative Neurology and Psychology. ing down to the end of the chute pulled the string, but failed to get food. Then he swung down to the floor. Time: one minute. He tried again imme- diately, but failed to hold and dropped to the floor. After six minutes he jumped to the chute, touched the string with his hand, but did not pull it. Third test. Conditions were the same as in the previous test. Performance 1. No. 6 was playing and saw only in part. P. 2. No. 6 saw the entire performance, though not steadily. P. 3. No. 6 saw the entire performance. No. 6 now became angry at No. 2 and tried to get out of his box. No. 2 became frightened and ceased to work for some time. He lay stretched out on the floor and after repeated efforts to get him to work he was taken out, and No. 6 was released in the cage. No. 6 immediately climbed the front of the cage, leaped across to the chute, swung with one hand to the rung, looked up the chute, pushed his other hand up, lost his grip and fell to the floor. He repeated this within two minutes. Twice again within two minutes he jumped to the chute. Then he jumped to the chute, hung by one hand and looked up inside. He looked at the chute often. He tried again to hang by one hand and look up the chute, but dropped to the floor. He later jumped to the chute twice and looked at the top of the cage. Fourth test, No. 6 imitating No. 4. Same conditions as before, except that No. 4 was substituted for No. 2. No. 4 got food fourteen times. No. 6 saw the entire performance five times; seven times he saw the performance in part. After No. 4 was removed and No. 6 was released, the latter went at once to the front, climbed the wire, jumped to the chute, held by his right hand and touched the string. Then he changed to hold by his left hand and thrust his right hand up to touch the string. After this he dropped to the floor. He repeated this in less than two minutes, not changing hands while on the chute, however. Five minutes later he leaped to the chute, but did not swing down. He did not seem to “get the hang” of holding to the chute with his feet as some of the other animals did. He gave no further attention to the chute. Fifth test. The conditions were the same as in the preceding test. No. 6’s record in seeing No. 4 was: Performance 1. No. 6 only saw food strike floor. P. 2 to P. 10. No. 6 saw the entire performance. P. 11. No. 6 saw in part. P. 12. No. 6 saw the entire performance. No. 4 was taken out and No. 6 was released. No. 6 found a seed on the floor and ate it. After two minutes, No. 6 jumped to the chute, but only examined a crack in the cage door. At the end of five minutes, No. 6 jumped to the chute and searched the inside of the chute with his hand, but he did not pull the string. He then took his leisure about the cage till the end of 10 minutes. Haccerty, Imitation in Monkeys. By Summary of Behavior of No. 6 in Chute Experiment B. No. 6 differed from each of the previously mentioned animals in his pre- liminary trials. He gave some attention to the end of the chute, on the second day, putting one hand into the end of it a short distance. On the later days, however, he ignored the end of the chute entirely. The stimulus- complex in the first test was the same as in the case of No. 13, namely, the sight of another animal performing an act and getting food thereby. The effect on No. 6 was evident, for within thirty seconds after being released in the cage he had repeated a part of the act he had seen; within a minute he had tried again and repeated the act in every particular, except in the amount of force with which he pulled. This difference, however, kept him from getting the food. Although he failed, he repeated the act entire or in part several times during the next few minutes. After his second series of observations, he repeated the entire act again, but failed to exert sufficient strength to accomplish the result. During the succeeding tests he persisted in going to the chute, although he ceased to pull the string. He did not cease to investigate the inside of the chute with eyes and hands, although his only means of connecting the chute with food had been his observation of another animal getting food at the chute. TABLE 6. No. 6 Imrratine No. 2. | | | Date: Ne Number of times | Number of times Resuli | Time in | fhe act. No. 6 saw. No. 6 saw in part. | | minutes. | | JNO, 1) ction 6 5 2 3 S 2 Magee se | 8 5 | 3 Ss 10 / SUR ee | 3 2 1 S 18 No. 6 Imrratine No. 4. ap eOh cua: 14 | 5 7 Sea 10 NS ey de ea 12 | 10 1 | S | 10 Total.....| 41 24 15 Papas | 50 G. Behavior of No. 5. Preliminary trials.—First trial, July 30.—Within one minute after entering the cage, No. 5 had climbed to the chute and had found the string with her hand. She was able to reach the chute from the side of the cage by help of her long legs and tail, which supported her while she grasped the chute in her hands. Later she reached the chute from the end of the cage. Then she swung to it from the side and looked up at the end. She then tried to use her foot to pull the string, and failing, climbed up the chute and examined the top of the cage. Then she braced her feet against the 372 fournal of Comparative Neurology and Psychology. corner post and pushed the chute with her hands. She was taken with sneezing and, descending, rubbed her nose on the floor. Her time was up soon after this. Second trial, July 31. On the second day No. 5 reached to the chute with her hands, put one hand up the inside and pulled at the string, but not hard enough to cause the food to drop. Five minutes later she did the same except that she did not pull the string. After the next five minutes she climbed up on the chute and examined the top of the cage. Then she swung her head and shoulders down, touched the string with her hand and dropped to the floor. Third trial, July 31. On the third day No. 5 was quiet about the cage as if nothing interesting were present. She spent most of her time on the floor. Fourth trial, August 1. No. 5 ran up and down the wire several times. Then she surveyed the chute and the whole top of the cage from below. She climbed to the chute and examined the top and back of the cage. Then she remained quietly on the floor for ten minutes, after which she looked up at the chute, climbed the cage, reached the chute and struck at the string several times with her hand. Fifth trial, August 1. At the fifth trial No. 5 was indifferent in the cage for five minutes. Then she climbed to the chute, examined the top of the cage, threw her head down, reached to the string and played with it but did not pull it. Later she jumped to the chute again and examined the back of the cage. The remainder of the time she spent on the floor. ~ Imitation tests.—No. 5 imitating No. 2.—First test. No. 2 and No. 5 were put into the cage together. After a little time No. 2 jumped to the chute. No. 5 climbed the wire opposite the chute, leaned over, put her hand up the inside and touched the string, but did not pull it. No. 5 then went to the floor and No. 2 pulled the string. Two seeds fell to the floor and No. 5 got them. No. 5 did not see the string pulled. The second time No. 2 pulled the string, No. 5 did not see. She heard the seeds drop to the floor and got them, jumping down from the wire front ahead of No. 2. The third time, No. 5 heard No. 2 at the chute and looked up just in time to see him pull the string and to see the seeds fall. No. 5 then went up the wire, jumped to the chute, and tried to pull the string, but did not pull hard enough to get food. No. 2 then became excited and refused to work. Second test. No. 5 was put into the observation-box and the box was set on the floor of the cage. No. 2 was put into the cage and at once went to work. Performance 1 to P. 5. No. 5 saw No. 2 on the chute and saw the food drop. She did not see him pull the string. P. 6. She saw the entire performance. P. 7. She saw it in part. When No. 2 was out No. 5 was released. She climbed the front of the cage and leaning over to the chute tried to put one hand up the inside, but could not do it. Time: 50 seconds. She went down to the floor and at once climbed the wire again. She jumped to the chute, wrapped her tail around it, put her feet on the rung, threw head and shoulders down and Haccerty, Imitation in Monkeys. Big looked up inside the chute. She put her hand up and was in the midst of an interested examination when a sneeze from the side cage startled her and she dropped to the floor. Her examination was more direct and longer than at any previous time. She climbed the cage, jumped to the chute and repeated the examination a minute later. Then she dropped to the floor and wandered about. Six minutes later she jumped to the chute, but only examined the top of the cage. Third test. No. 5 was not attentive to No. 2. The latter got food seven- teen times. Of these performances, No. 5 saw but five; six times she saw in part. No. 2 was then removed and No. 5 was released from the observation-box. For a time she searched the floor and edges for food. After three minutes she climbed the wire, reached to the chute with her hands and tried to put one hand up the inside, but failed. Then she climbed down to the floor and sat in the corner. Later she vomited grass which she had eaten out of her bedding and then went about the cage quietly. Her lack of activity was probably due to sickness of stomach. Fourth test. The box containing No. 5 was fastened to the side of the cage on a level with the lower part of the chute. No. 4 was used instead of No. 2. Performance 1 and P. 2. No. 5 saw all except No. 4’s putting his hand up. P. 3 and P. 5. No. 5 saw entire performance. P. 4. No. 5 saw nothing. P. 6 and P. 7. No. 5 was not interested in No. 4 on the chute; she bowed her head and slept while No. 4 got food. P. 8 to P. 10. No. 5 saw the entire performance. No. 4 was now taken out and No. 5 was released from the observation-box. She went at once to the front, climbed the wire, reached to the chute, put one hand up and pulled, but not hard enough to get food. No. 5 was some- what frightened by the demonstrations of anger which No. 4 made when she was taken out of the cage. No. 5 gave no further attention to the chute during the ten minutes. Fifth test. No. 5 was put into a box on the floor where she could see No. 2 at the chute. Her record in seeing No. 2 pull the string was as follows: Performance 1. No. 5 saw nothing. P. 2. No. 5 saw entire performance. P. 3 and P. 4. No. 5 saw in part. Here the apparatus gave some trouble and the test was delayed. P. 5. No. 5 did not see. P. 6 and P. 7. No. 5 saw the entire performance. P. 8 and P. 9. No. 5 did not see the pull; saw food strike floor. P. 10 and P. 11. No. 5 saw perfectly. No. 2 was taken out and No. 5 was released. Immediately she climbed to the chute and pulled the string, but not hard enough to get food. Time: 40 seconds. A minute later she threw her head down and looked up the inside 374. Journal of Comparative Neurology and Psychology. of chute. Later she climbed to the chute, but paid no attention to it. Nor did she pay any more attention to it during the entire time in the cage. Sixth test. The conditions were the same as in the previous test, except that the string in the chute was lengthened four cm. Of seventeen per- formances No. 5 saw five completely, six in part, and six not at all. Toward the end of the time she seemed sleepy and paid but little attention. When No. 2 was out and No. 5 was released she at once climbed to the chute and took a long, steady look up the inside of it, but did not put her hand up. Time: 40 seconds. She then took her leisure about the cage, caught a roach and perched on the brace. Summary of Behavior of No. 5 in Chute Experiment B. No. 5 did not present the same problem in the imitation tests as the animals previously discussed. She had already performed every part of the act necessary to get food. She had evidently failed because of not exerting sufficient strength. Her interest in the chute seemed to wane in the fourth and fifth preliminary trials and to be accentuated after observing No. 2 in the first and second tests. During the later tests she repeated the act of the animal seen, but she never got the food, and in the fifth test she merely looked up the inside of the chute without putting her hand in. It seems fair to infer that the increase of interest manifested in the first and second tests and the continuation of interest in the chute through the successive tests was due to No. 5 seeing the other animals getting food at the end of the chute. TABLE 7. No. 5 Imrratine No. 2. No. 5 Iwrratine No. 4. | | Dates | None aes | Number of times | Number of times | Regult. Time in | the act. | No. 5 saw. | No. 5 saw in part. minutes. AT eA | 3 | 1 0 | F Aug. 5...... | 7 | 1 6 lees 10 AUER cea ily 5 6 | Ss 10 HN Obooecs 10 5 2 S 10 AUTO Wis oars 11 5 2 | S 10 PND iF 5 Bt | 17 | 5 6 | S 10 Totals... .| 65 | 22 22 | S 50 H. Behavior of No. 8. Preliminary trials.—First trial. No. 3 moved about slowly in the cage during the entire fifteen minutes, but gave no attention to the device for getting food. He spent his time on the floor and the wire parts of the cage. Second trial. No. 3 spent the first few minutes on the floor and then climbed the wire and came back to the floor several times. Once when on Haccerty, Imitation in Monkeys. 375 the wire front he looked steadily at the chute. Then he climbed about the cage and played on the floor. Once again he took a direct look at the chute from the front of the cage and then played about on the floor. Third trial. On the third day No. 3 climbed the wire and played on the floor, but paid no attention to the chute. Fourth trial. On the fourth day he climbed the wire and then spent his time on the floor, going from one corner to another and crouching with his face toward the center of the cage. Occasionally he surveyed the top of the cage. Then he climbed to the brace (across front of cage) and perched. Later he went to sleep on the floor in the corner. Then he climbed the cage and looked about, but took no notice of the chute during the fifteen minutes. Fifth trial. On the fifth day he remained on the floor for a minute and then climbed the wire. He then sat in the corner of the cage for five minutes before he climbed the wire again. Then he went to rest in another corner. No attention to the chute. Imitation tests —No. 3 imitating No. 2.—¥irst test. No. 2 and No. 3 were put into the cage together. No. 3 was attentive to No. 2 from the first, partly in order to escape punishment. Each time No. 2 pulled the string, No. 3 got food, and when he got a grain of sunflower seed the second time, No. 2 punished him. No. 3 cried and saw only in part the next time. Three times he saw the whole performance from the floor at different angles and twice from the front of the cage on a level with the chute. After No. 2 had been taken out No. 3 busied himself on the floor for a few minutes picking over the hulls No. 2 had left. Then he surveyed the chute from the four corners (on the floor) of the cage. Once he climbed the front wire and looked at the chute from its own level. Then he went to the floor and rested in the corner of the cage. Second test. Conditions were the same as in the previous test. No. 3 was again afraid of No. 2 after the second drop of food. He saw the first two times perfectly from the floor, but missed the third because of his fear of No. 2. The next three times he saw the entire performance from the floor. With No. 2 out No. 3 went hunting among the empty hulls as before. Then he looked upward toward the chute several times from different posi- tions on the floor. Later he climbed the front of the cage opposite the chute and looked back over his shoulders at it. Then he went down to the floor and remained there. Third test. Conditions were the same as in the preceding tests. No. 3 was attentive to every move of No. 2 and saw him jump to the chute and pull the string each time but one. He did not get food, however, because of his fear of punishment. At the seventh time, No. 5 got food. Although No. 3 looked steadily at No. 2 when he pulled, it was difficult for him to see No. 2’s hand go up the chute because No. 2’s body often got in the way. The first five minutes after No. 2 was out No. 3 was on the floor hunting over hulls dropped by No. 2 and fingering the cracks in the floor. Several 376 Fournal of Comparative Neurology and Psychology. times he looked up at the chute. Then he climbed the cage wire, but did not look at the chute. Later, when under the chute, he looked at it steadily and then started for the front as if to climb, but was turned away by seeing a hull on the floor. Fourth test. The conditions were the same as in the preceding tests. No. 2 got food fourteen times. Ten of these performances No. 3 saw com- pletely; the other four he saw in part. He kept away from No. 2 because No. 2 slapped him. When No. 2 was out No. 3 spent his time on the floor hunting over empty hulls and paid absolutely no attention to the chute during the entire time. No. 3 was so little influenced by seeing No. 2 obtain food that it seemed useless to continue the tests longer. They were, therefore, discontinued. Summary of Behavior of No. 3 in Chute Experiment B, No. 38 was not nearly so active in the preliminary trials as the animals pre- viously discussed. In the imitation tests he seemed to see what was done. What he saw, however, did not seem to influence his behavior in any way unless it was to increase his looking at the chute. He failed to make any effort to get the food for himself. TABLE 8. No. 3 Inrratine No. 2. : Number of times Date | No. 2 performed Number of times | Number of times | Resuii, 9 |) uimenm : “Fis, eves. No. 3 saw. No. 3 saw in part. | yp a | minutes, PNW Bo 8 5 4 F | 10 AUIS Siete es 6 5 F == 10 Aig Sita. 11 1 | once pest) Aug. 5.. 14 10 4 Tema ine 2) Rotalesser | 39 30 8 F | 40 General Summary of Results of Chute Hauperument A and Chute Huperiment B. Taking Chute Experiment B as a whole, we have to consider six animals, no two of which exhibited exactly the same behavior. In the cases of No. 13, No. 4, No. 11 and No. 6, there is a similarity in that each animal showed a decided change of behavior after wit- nessing another animal get food from the chute. Each of these animals repeated with more or less exactness of detail the act which it had seen the other animal perform. Without meaning to imply Haccerrty, Imitation in Monkeys. 277. anything as to the mental processes accompanying it, I shall call such behavior imitation. As I shall use it in this paper, the word imitation is a conceptual short cut to describe a complex form of behavior. It always implies these things: (a) The animal which imitates observes an act of another animal; (b) More or less directly thereafter its behavior is modified in the direction of the act ob- served; (c) This modification is usually sudden; (d) The behavior is changed to a considerable degree and, when wholly successful, to an exact copy of the act observed. In every case of behavior which T shall call imitative, the animal had abundant opportunity to learn the act by himself so that his repeating the act of the imitatee was apparently due to his observation of that animal performing. In the case of No. 3 and in the case of No. 1 in Chute Experi- ment A, there was almost no evidence that the act of the performing animal influenced the animal which saw. The case of No.5 is unique. Before seeing another animal per- form the act, she had herself done every part of the act necessary to get food. The only way in which she could have been influenced was by being stimulated to exert more force on the pull or by being stimulated to a repetition of the act. She was not influenced in the first way, but the regularity with which she went to the chute after seeing the other animal get food, suggests that she was influ- enced in the second way. In her habits, she was much like No. 4 and No. 6, and the clear evidence for imitation in the conduct of each of these animals furnishes some ground for a similar interpretation of the behavior of No. 5. However, the evidence on the point is not conclusive and remains rather a conviction in the mind of the experimenter than an established fact. TABLE 9. RESULTS OF CHUTE EXPERIMENT A AND CHUTE EXPERIMENT B. ie INUMbeH! OF ANIMAS TISEGS act ot 5S ao 0's sp cvecols. ova vn ee eve miel ovace, ola) cuatecere eave teusia ole oretoders 7 Casest Of SUCCESS HUST AGIONS ao. 5a 2.cldicvs a: Srepeveronetatecs, setae! ccctesateeietre tener arelavele pateve S CASsesTOn partiallvasiccesstiuleimibatiOni .vrsaeer eee ciel eee eres 2 Cases of failure to imitate............ EEE DO CREO Ie 2 378 ‘fournal of Comparative Neurology and Psychology. TABLE 9. (Continued.) Resuuts or CHuTE EXPERIMENT A AND CHUTE EXPERIMENT B. II. Cases of imitation when the imitator was confined during the activity OF the simMitALES Sa psve. cst leroterste ee doe ooo ehetetone he teiel elovalls tetetepel enon tereRere MeierovoLere cherie +) Cases of imitation when both animals were together in the cage............ 0 ee Cases Of Immediate amita lio meter tercnererscter tele tel stele er onciotelsctemenene: ole eter onokeiauehetsn els + Gases ofseradallcimil tail Oneal eryeteyorietereteretnereretletlel foteeNeteh consist ts oheiclorne Rca ot IV. Cases of imitation in which the imitating animal did not himself experi- ence the result of the act before performing..................)..++.+.- 3 Cases of imitation in which the imitating animal did experience the result oLtheyact before pertorming sterner elemental teroete eerste aeieiete 2 3. RopE EXPERIMENT. A. Description of Device. For this experiment a hole 5 cm. square was cut in board D, 26 cm. from the top of the cage, fig. 5. A door was hinged to one side and opened outward. It was cut so as to fit snugly and when closed was flush with the inside of the board. The only evidence of an opening was the sharp line around the square where the door fitted the board. Before this door, and 27 cm. from it, an inch rope, b, hung from a screw eye in the top of the cage to the floor. In order to get food the animal must climb the rope and, while supporting himself on the rope, push the door open, reach through it and get the food on the outside of the cage. The food was supplied by means of a two- inch leatherette belt connected with the experimenter’s table, which stood four feet from the cage. This convenience, together with a string by which the door could be closed after the animal had opened it, made it possible to manage the entire apparatus from the experimenter’s table. There was nothing on the inside of the cage to denote the food on the outside. B. Behavior of No. 2. Preliminary trials—The following preliminary observations were made in the old cage at the Harvard Laboratory. Each trial lasted 30 minutes. First trial. At first No. 2 walked about the floor and climbed the front of the cage. He then went about the cage, and once, in passing, touched the rope with his hand. Again he touched it as he went about the cage. He was very active and ran about the cage very rapidly, but made no effort to climb the rope. Haccerty, Imitation in Monkeys. 379 Second trial. On the second day the experiment was disturbed by No. 2 getting out of the cage. During the time he was in the cage he made no effort to climb the rope. Third trial. On the third day he went rapidly about the cage as on the first day. Once, in passing, he picked up the end of the rope. During the thirty minutes, however, he made no effort to climb it. Fourth trial. On the fourth day No. 2 was quite active and kept moving during the entire thirty minutes, passing over every bit of floor space many times and being repeatedly on every part of the front and end of the cage. He noticed the rope only to touch it momentarily in passing. Fifth trial. On the fifth day No. 2 was active and eager to get out of the cage. Once he grabbed the rope in his tail and ran up the front of the cage. Later he bit the end of the rope once. Sixth trial. On the sixth day he was very wild, possibly due to the death of his mate the day before. He made no effort to climb the rope. Seventh trial. On the seventh day No. 2 behaved as usual. His only notice of the rope was to push it aside in passing. Highth trial. No. 2 behaved as usual. Once he stood on his feet and grasped the rope with his hands one above the other as if to climb. Ninth trial. On this day No. 2 grasped the rope once in the same manner as on the previous day, but let go of it at once. He repeated this several times, but showed no other intention of climbing. He did not look up when holding the rope. Tenth trial. On the tenth day he went about the cage in his usual way. Once or twice he hooked his tail around the rope and ran up the front of the cage, dropping the rope when half way up. Hleventh trial. On the eleventh and twelfth days he went about the cage as usual and displayed no interest in the rope. Imitation tests.—No. 2 imitating No. 3.—These tests were made in New York. Both animals were in the cage together. First test. No. 3 got food twelve times. Ten times No. 2 saw him eating the food and once saw the entire performance. Twice when watching No. 3 eat while on the rope, No. 2 climbed the front wire and leaned toward the rope as if trying to get to the door. During the rest of the time No. 2 was distracted by the other monkeys in the living cages nearby. Second test. No. 8 was very active and climbed the rope often and rapidly. No. 2 was not accustomed to watch No. 3 and did not look at him, but tried to see out the window and into the other cages. Four times when No. 2 saw No. 3 on the rope he leaned out from the side of the cage toward the rope. Once when No. 3 pushed the rope toward the front of the cage No. 2 caught it in his hands and swung his weight on his hands, but held on to the wire with his feet and tail. Several times when No. 3 was up the rope No. 2 caught the end of it from the floor. No. 2 saw five times in fourteen. When No. 3 was taken out after performing the trick fourteen times, No. 2 tried to climb the post in the corner next to the rope and got two feet from the floor by the help of small sticks nailed to the post. Then he stood on the floor and grasped the rope in 380 “fournal of Comparative Neurology and Psychology. both hands as if to climb, but although he looked up, he did not lift him- self from the floor. He tried to climb the post again as before, but when he failed he did not turn to the rope at once; he did, however, a moment later. During the last five minutes of the fifteen he remained in the cage after No. 3 was out, he sat in the corner near the rope, part of the time holding it in his hands. Third test. No. 2 was more or less distracted by other monkeys in the room and not being very hungry did not observe No. 3 closely. He saw the whole act five times in twelve. He did not watch from first to last. It was counted if he saw No. 8 climb the rope and also get the food, even though his attention was not continuous. In no case, however, did No. 2 watch No. 3 continuously from the time he left the floor until he got the food. No. 2 saw by glances only. Three times he swung out from the wire front and twice he tried to climb the post as in the previous test. When No. 3 was taken out No. 2 ran about the cage. He grasped the end of the rope when on the floor. He looked up at the door and tried to climb the post. Then he grasped the rope with one hand above the other as if to climb. Dropping the rope he turned to the post, then back to the rope, grasping it in his hands and bearing part of his weight on it. It swung and he took a few steps. Again he grasped the rope and bit the end of it. Then he grasped it with two hands and one foot. Then he turned to the post and put his hands and one foot on it. Then on the other foot he turned as on a pivot and grasped the rope with the three members he had placed on the post. Then he ran to the front of the cage and back to the rope, grasping it again with two hands and one foot and bearing some weight on it, but not enough to lift the other foot from the ground. Then he grasped the rope in his hands and rushed to the wire to climb. This he repeated, wholly or in part, several times more in the next two minutes. He was then taken out twenty minutes after No. 3 had been removed. Fourth test. When No. 2 and No. 3 came into the cage together No. 2 was very attentive to everything No. 3 did and looked often toward the food door. When No. 8 climbed the rope, No. 2 climbed the wire front of the cage on the first and second trials. When No. 3 pushed against the door with his hands his feet pushed the rope over toward No. 2, who was on the wire front of the cage. No. 2 was eager to grasp the rope and once did grasp it in his hands, but would not let go of the wire with his feet. When No. 3 got the food the rope swung back to a perpendicular position, and No. 2, holding with feet and tail to the wire, threw his body out toward the rope. He repeated this motion several times, and when No. 3 got food the third time No. 2 was able to grab it out of his hands. This he did on the fourth and fifth trials also. No. 38 was then removed and No. 2 became very active on the wire, throwing his body vigorously toward the rope, and failing to grasp it, he repeated the act at once. His motions increased and he seemed frantic to catch the rope. Finally he jumped, catching the rope and holding. At once he jumped back to the wire, and ran down to the floor; he quickly climbed the rope Haccerty, [mutation in Monkeys. 381 twice and then jumped from the rope to the wire. Then he climbed the rope, opened the door by pushing on it with his teeth as No. 3 had done, and got food. Next he tried to climb the post in the corner, and then ran up the rope again and got the food. During the next ten minutes he climbed the rope twelve times, getting food most of the times. Summary of Behavior of No. 2 in the Rope Haperiment. No. 2 during a long series of preliminary trials did not climb the rope nor pay any attention to it or to the food door. During the first imitation tests he was not attentive to No. 3, but gradually became so as he saw him getting food, and in the last test his attention was riveted on No. 3 during the whole of the time No. 3 was in the cage. His learning to do what No. 3 did was also a gradual process. When his attention was directed to the food door his first effort to get to it was by climbing the wire front of the cage. His next step was to learn to bear a portion of his weight on the rope. When he got food from No. 3 there was a notice- able increase in his apparent desire to get to the door. The stimulation seemed to increase steadily until finally it forced him to leap to the rope and back to the wire and then to climb the rope from the floor. Once on the rope, he repeated exactly the act of No. 3 and got food in the same way. TABLE 10. No. 2 Imrratine No. 3. | eee | | Date ee ea eae Number of times | Number of times Result Time in : | "Raa Deh. | No. 2 saw. | No. 2 saw in part. . minutes. July 12.05 | 1 | 1 | 10 F 10 alivy 2s | 14 | 5 8 F 10 July 3...... 12 5 | 7 EF 10 uiliye Aen = oy: 5 5 S | | Total... 49 | 16 | 25 Wass 30 CO. Behavior of No. 4. Preliminary trials.—First trial. No. 4 spent the first few minutes on the floor picking over nut hulls. Then she became very active about the cage. She climbed the rope and investigated the top of the cage and the cracks between the boards. She spent the most of the remaining time on the floor. Second trial. On the second day No. 4 spent most of the time on the floor swaying back and forth before the door. Nothing inside seemed to interest her and she wanted to get out. Twice she climbed the wire slowly, but paid no attention to the rope. Third trial. On the third day No. 4 was not so active as usual and perched on the brace most of the time. She gave no attention to the rope. 382 ‘Fournal of Comparative Neurology and Psychology. Fourth trial. On the fourth day No. 4 behaved as usual, spending most of her time on the wire and brace. When on the floor she swayed back and forth before the wire and gave no attention to the rope. Fifth trial. On the fifth day No. 4 spent most of her time in the corner of the cage farthest from the rope and gave no attention whatever to it. Imitation tests.—No. 4 imitating No. 2.—¥irst test. No. 4 was placed in the observation-box which was fastened to the front of the cage on a level with the food door. No. 4 was on the floor of the observation-box. She was swaying as usual and this somewhat frightened No. 2, so that he climbed the rope only after three minutes and then jumped to the wire without getting food. This he repeated three times. Then he climbed the rope, tried the door, but failed to push it open. He jumped to the wire at once, after pushing. Then he climbed the rope and opened the door. No. 4 saw in part. No. 2 then tried the door four times unsuccessfully. Climbing the rope brought him close to No. 4 and his fear did not allow him to make a good effort. Then he climbed the rope and opened the door, getting food. No. 4 saw the entire performance. No. 2 was then removed and No. 4 was released. She climbed the wire on the front of the cage and then on the end. Then she climbed the rope and reached to the hole in the top of the cage. She looked at the door, put her ‘nose to it, and jumped to the front wire, and went to the floor. She then climbed up and down the front and climbed the rope looking at the door and jumping to the front of the cage. Again, she climbed the rope and looked all about the door more intently than before. She re- turned to the floor, climbed the end of the cage and perched on the brace at X. Again she climbed the rope, examined the top of it, and looked all about the door. Then she became interested in out of doors and soon her time was up. Second test. No. 4 was in the box as before and No. 2 was somewhat slow and fearful. No. 4 saw five of No. 2’s twelve performances com- pletely ; four other times she saw a part of the performance. No. 2 was removed and No. 4 was released. She climbed the rope at once (5 seconds) and smelled and licked the door. ‘Then she returned to the floor. She again climbed the rope and examined the top of it. She looked at the food door carefully, but after coming to the floor she gave no further attention to the rope or the door. No. 4 imitating No. 6.*—Third test. No. 4 was in the box on the floor. Performance 1. No. 6 climbed at once to the door, pushed it open with his hand and got food. No. 4 saw him smacking his lips when the food was gone, but saw nothing more. P. 2. No. 4 saw No. 6 reach through the open door and get food, but saw nothing more. In the sixteen following performances No. 4 saw the entire performance *The behavior of No, 6 will be given later, page 385. Haccerty, Imitation in Monkeys. 383 twice; she saw No. 6 climb the rope three times, and eleven times she saw nothing. No. 6 was removed and No. 4 was released. She at once climbed the rope and looked about the food door, but made no effort to open it. She examined the hole in the top and returned to the floor. She climbed the front of the cage and leaped to the rope. She looked at the door, put her palm against it and rubbed her hand over the door. She then fingered the crack around the door, but did not push the door open. She returned to the floor and worked about the edges of the floor for some time. She then climbed the rope, but gave no attention to the door. After a little more wandering she perched on the brace at X and remained quiet. Fourth test. No. 4 was put in the observation-box on a level with the food door. Performance 1. No. 4 saw the entire performance. P. 2. No. 4 saw the entire performance. No. 6 hesitated to climb for fear of No. 4. P. 3. No. 6 was slow at the door because of watching No. 4 and No. 4 saw perfectly every move. P. 4 and P. 5. No. 4 saw all except the push on the door. She looked down just as No. 6 pushed. ; P. 6 and P. 7. No. 4 saw perfectly. No. 6 was removed and No. 4 was released; she ran up the rope and put her nose to the food door, but gave no push. She returned to the floor and climbed the wire end and front and examined the edge of the cage door. She spent the remainder of her time without attention to the door or the rope. Later she perched on the brace at X and “hunted fleas.”’ Fifth test. No. 4 was in the box on a level with the door. No. 6 climbed the rope and opened the door nine times. No. 4 saw every part of the performance five times; three times she saw in part. When No. 4 was released she at once climbed the rope and looked at the food door. Then she jumped to the wire front and returned to the floor. She examined about the floor edges and climbed up and down the wire. Then she climbed the rope and passed her hand over the food door, but did not push with any force. She then looked at the top of the cage and leaped to the wire front. She perched on the brace at X for the last five minutes. Sixth test. No. 4 was in the box on a level with the food door. No. 6 got food seventeen times. No. 4 saw the entire performance eight times; once she saw in part. No. 6 was removed and No. 4 was released. She walked across the floor to the wire end and climbed half way up. ‘Then she leisurely climbed down and walked over to the corner opposite the rope, turned to the rope and climbed it. She stopped exactly at the door, put her right hand against the upper edge of the door, her fingers striking the board above, and pushed. Failing to open the door, she put her left hand lower down on the door, her palm this time striking the board below the door. She pushed again, but failed to open the door, probably because more of the 384 Fournal of Comparative Neurology and Psychology. force of her effort affected the board than the door. Then she changed back to the right hand and planted it squarely in the center of the door, neither her fingers nor palm touching the board. She gave a hard push and the door opened. ‘Time: two minutes. She got the piece of food near the door and thrust her arm farther out and got the piece of banana on the place next below. Then she pushed the door open as fast as I closed it and got food three times. This she repeated ten times as rapidly as the device was reset. Her hand when placed flat against the door reached from the top to the bottom and was almost as broad as the door. ‘To open the door she must place her hand in the center of it, in order not to strike the edges of the board in some place. Summary of Behavior of No. 4 in the Rope Experiment. The problem as it came to No. 4 was different from the problem of No. 2. She did not need to learn to climb the rope. She did this as if it were a familiar act during her first few minutes in the cage. What she had to learn was to open the food door. Her own unaided efforts helped her not at all, and during the last four preliminary trials she kept entirely away from the rope. Her first observation of the imitatee attracted her again to the rope and to the food door, but she did nothing except to nose about the door. Her second test again directed her attention to the door and possibly increased that attention. The third test augmented her attention to the door and she rubbed her palm over it and fingered the edges. After the fourth test her interest seemed to lag, but after the fifth her attention was as great as after the third. In both the third and fifth tests she used her hand at the proper place and in much the same manner as had the performing animal. ‘The sixth test served to make TABLE 11. No. 4 Imiratine No. 2. | ee | Number of times | | | Date | No. 2 performed | Number of times | Number of times | Recut. | : : d | N | : Time in RD Ze. o. 4 saw. | No. 4 saw in part. minutes. Aug. 10..... 2 1 | 1 Pe | 10 ANISe LOR eee | 12 | 5 | 4 | FE | 10 No. 4 Imrratine No. 6. Aug. 11..... 18 2 A al ask a at PN Wl 5.5 ae 7 5 2 | F 10 AIS aD Se 9 5 3 F 10 ACI ELD eas 17 8 1 Ss 2 Totals....| 65 26 15 | 'S) 52 Haccerty, Imitation in Monkeys. 385 definite the imitative behavior. She repeatedly tried to do what she had seen done and finally succeeded. At no time, until she performed the act herself, did she experience the result of the act, the stimulus-complex being the other animal performing an act and getting food. D. Behavior of No. 6. Preliminary trials. First trial. No. 6 was very active climbing all about the wire and running about on the floor. He caught the rope in his hands and later in his tail. He climbed the rope and in attempting to jump from the rope to the front of the cage he put his foot against the food door and the door opened. He did not notice it, however, and it was closed before he saw what he had done. He climbed the rope several times after and took no notice of the door. Second trial. On the second day No. 6 was very active on the wire and the rope, but took no notice of the door. Third trial. On the third day No. 6 behaved as usual, climbing about the wire. He took no notice of the rope. Fourth trial. On the fourth day No. 6 climbed the wire several times, each time carrying the rope up in his tail. Later he climbed the rope, swung back and forth on it, and after two or three oscillations he leaped to the wire front. He took no notice of the food door. Fifth trial. On the fifth day he grasped the rope and ran in a circle on the floor. Then he swung on the end of the rope twice. Then he grasped the rope in his tail three feet from the floor and allowed his body to swing. He dropped to the floor and caught flies. He climbed the rope and examined the cracks in the top of the cage. During his entire time in the cage he took no notice of the food door. Imitation tests—No. 6 imitating No. 2.—No. 6 was put into the observa- tion-box and the box was put on the floor of the cage. Performance 1. No. 2 spent the first few minutes on the floor of the cage, on the box, and in climbing the wire. He climbed the rope, but came down without any attention to the food door. No. 6 saw No. 2 on the rope. No. 2 again climbed the rope and worked at the door slightly, but did not open it: No. 6 saw all the movements of No. 2. No. 2. then became frightened at No. 6 and did not work. P. 2. No. 6 saw No. 2 with the food, but nothing more. P. 3. No. 2 now became angry and pretended to fight, hanging over the box by his tail and shrieking loudly. No. 6 on the inside of the box jumped and threatened. No. 2 retreated to the corner by the rope and shrieked. Suddenly No. 6 stopped jumping, put his head on one side and purred. No. 2 had done this just before and now repeated it. His fear was gone; he shot up the rope, opened the door and got food. As No. 2 climbed the rope No. 6 looked out through the wire, and when he turned again toward No. 2 the latter was eating his banana. At once No. 6 began to jump up and down in his box to show his anger. No. 2 was again frightened and for several minutes the shrieking was renewed. 386 “fournal of Comparative Neurology and Psychology. P. 4. At once No. 2 jumped to the rope from the front of the cage, but came to the floor without opening the door. He then walked about on the floor for several minutes. Then he climbed the rope, gave one push on the door, but failed to open it. He soon climbed again, opened the door and got food. No. 6 saw him with the food and threatened him; No. 2 shrieked; No. 6 folded his arms; No. 2 lay down on the floor. No. 6 jumped up and down; No. 2 came near the box, and seemed to have no fear. P. 5. No. 2 climbed the rope, opened the door and got food. No. 6 saw all. P. 6 to P. 8. No. 2 did as in P. 4. No. 6 saw No. 2 on the rope and at the door, but did not see him open the door. No. 2 then sat on the floor quietly for several minutes. P. 9. No. 2 climbed the rope and opened ‘the door, but did not get the food which had dropped off the belt. The food was replaced and No. 2 got it. No. 6 saw all but the opening of the door. No. 2 was now removed and No. 6 was released in the cage. At once he climbed the rope, put his hand against the door, but failed to open it. He then swung down, hanging by his tail to the rope, and dropped to the floor. He then climbed the rope and examined a hole in the top of the cage. He came to the floor again. Again he climbed the rope and examined all about the door; pushed on the door, but did not open it; he bit at the edge of the door and again pushed on it, opening it. He got the food and descended the rope, immediately afterward climbing the wire. When the device was reset No. 6 climbed the rope and examined the door with his teeth and fingers; he worked at the edge with his fingers. He then jumped to the wire and in so doing put his foot against the door pushing it slightly; he leaped back at once and pushed the door open with his hand, get- ting the food. When the device was reset No. 6 tried to open the door with his fingers and after one effort leaped to the wire. Leaping back he tried to bite the edge of the door and then by a vigorous push with his hand forced it open and got food. The device was reset and No. 6 climbed the rope at once. Placing his palm flat against the door he opened it with the first effort. He repeated the act as soon as the device was ready to operate, and four times more within a few minutes. In all No. 6 opened the door and got food nine times within sixteen minutes. Summary of Behavior of No. 6 in the Rope Experiment. No. 6, like No. 4, was free on the rope from the first. He became indif- ferent to it during the later trials and made no progress toward getting food. When he was placed in the observation-box to watch No. 2 he was very attentive to what No. 2 did and seemed quite excited by the conduct of the latter. He saw the entire performance once and in part three times. When he was released his behavior was markedly different from what it had been in the preliminary trials. His attention was directed to the proper place to get the food, and after a few random movements he succeeded in getting food for himself in a manner similar to that which he had seen. Haccerty, Imitation in Monkeys. 387 TABLE 12. No. 6 Imrratine No. 2. Number of times Date. No. 2 performed Number of times | Number of times Result. Time in the act. No. 6 saw. No. 6 saw in part. | minutes. Pulyeoo 9 1 3 S 5 EH. Behavior of No. 5. Preliminary trials.—First trial. No. 5 as usual was very active. She climbed the wire side and end of the cage. She climbed the rope and thrust her arm through a hole in the top of the cage. She then hung to the rope at the top of the cage and tried to see through all the cracks in the top. She did not notice the door. Second trial. On the second day No. 5 behaved as before, going all about the cage, on the wire, up the rope, etc., but did not observe the door. Third trial. On the third day No. 5 showed no interest in the rope, but spent her time going about the cage. Fourth trial. No. 5 climbed the wire; returning to the floor, she pushed on the door where she had entered. Once she grasped the rope in passing. Later she climbed the rope, but displayed no notice of the food door. Fifth trial. On the fifth day No. 5 was active on the floor, on the wire and at the entering door. She grasped the rope in her tail several times ; three times she climbed it, but took no notice of the food door. Imitation tests —No. 5 imitating No. 2.—First test. Both animals were free in the cage. Performance 1. No. 5 did not see. P. 2. No. 5 saw No. 2 get food and climbed the rope after him. P.3. No. 5 saw the getting of food, but did not see the push on the door. No. 5 saw the entire performance from the brace at X. No. 5 saw in part. No. 5 saw from the front and opposite the rope. No. 5 saw the entire performance and climbed the rope before the door was closed. P. 8. No. 5 saw the entire performance. When No. 2 was taken out No. 5 climbed the rope at once and put her hand on the door. She then became interested in the top of the cage. Three minutes Jater she climbed the rope and put her nose to the food door, but she did not push on it. Two minutes later she repeated this performance. Between the two performances and afterwards she went about the cage, climbed the wire and roamed about the floor for food. Apparently she was very hungry. Second test. No. 5 was put into the observation-box, which was set on the floor of the cage. eh Sae I oR Yb 388 “fournal of Comparative Neurology and Psychology. Performance 1 to P. 4. No. 5 saw No. 2 push the door open with his hand in his usual way. P. 5 to P. 7. No. 5 did not see. P. 8. No. 5 saw No. 2 at the door with the food and tried to get out of her box. This frightened No. 2 and he became quiet. P. 9. No. 5 did not see No. 2 until he jumped back to the wire with food. P. 10. No. 5 saw perfectly. No. 2 was taken out and No. 5 was released. She walked about the floor and made some efforts to push her hand through the wire to get a paper and to reach the belt. Then she climbed the rope, stopped when half way up to spat a fly, climbed to a small hole above the food door and tried to see out. Then she went down to the floor, walked about and climbed the wire. She jumped to the rope, but took no notice of the door. This she repeated later. Then she went to the floor, became quiet and curled up to sleep. Third test. No. 5 was put into the observation-box on a level with the upper part of the rope and the food door. Performance 1 to P. 8. She saw perfectly. P. 4. No. 5 saw nothing. P.5. No. 5 saw perfectly and jumped at the side of the box as if to get food. P.6. No. 5 saw in part. P.7. No. 5 saw perfectly. No. 2 was removed and No. 5 was released. She ran up the rope, but took no notice of the food door. She examined the top of the cage with the eye and hand and returned to the floor. She then climbed the rope and examined the food door with her eyes and hand, but did not push on it. She then climbed the wire from the floor and returned to the floor and looked about. She crouched near the door and slept, curled up in a characteristic fashion of her own. Fourth test. No. 5 was again put into a box on a level with the food door. No. 2 was free in the cage. He got food sixteen times. Five of these performances No. 5 saw completely; five other times she saw in part. No. 2 was removed and No. 5 was released. No. 5 at once climbed the rope, but a noise frightened her and she jumped to the wire front at once. Then she climbed about the wire and walked about the floor. Next she climbed the rope and looked and smelled about the food door. She returned to the floor and crouched in one corner. Then she lay down and slept. Fifth test. No. 5 was put into a box on a level with the food door. No. 6 was used instead of No. 2. At no time did she seem interested in No. 6. Her seeing was accidental and passive. At times when she saw No. 6 going up the rope or at the door she would turn away to look at the floor of her box. No. 5 saw five of fourteen performances completely ; five other times she saw in part. No. 6 was then taken out and No. 5 was released. She spent the entire ten minutes on the floor and in climbing the wire parts of the cage without once going to the rope. At the end of ten minutes she climbed the rope and looked at a crack in the cage door and at a hole in the top of the eage. She took no notice of the food door while on the rope. Haccerrty, Imitation in Monkeys. 389 She climbed up and down the wire and perched at the brace until she was removed. Sixth test. Both animals were free in the cage. Performance 1. No. 6 pushed the door open and got food. No. 5 saw and jumped to the rope from the front of the cage. P.2. No. 5 again saw from the front of the cage and jumped to the rope. She put her hand against the door, but did not push. P. 3 to P. 6. No. 5 saw only in part, the eating of food. P.7. No. 5 saw perfectly and started up the rope, but when the food door was closed she came down. P.S. No. 5 saw and climbed to the food door, but did not push. P. 9. No. 5 saw perfectly. P.10. No. 5 saw, climbed to the food door and looked through before it was closed. P.11. No. 5 saw and jumped to the rope while No. 6 was still getting food. Before the door could be closed she had her nose at the opening and was looking out. When it was closed she immediately pushed it open. In pushing she put her palm squarely against the door. There was a marked difference between the behavior of No. 5 at this time and her previous conduct. Before, as noted, she had not been inter- ested. Now she became interested and No. 6's movements directed her attention at once to the food door and kept it there almost all the time until she had learned. After her first effort she could do the trick perfectly, and she repeated it six times within a few minutes. TABLE 13. No. 5 Imrratine No. 2. No. 5 Imiratrine No. 6. Date eee ee Number of times | Number of times Result Time in : Pinciaem o. 5 saw. | No. 5 saw in part. ; | minutes. = . Ls Ss aa _— —_— | = AL OO era: 8 5 2 F 10 Pe MDs cae 8 5 1 iy 10 Aug. 10..... 7 | 5 1 Pa tO ANUS NPB SSes| 16 | 5 | 5 F 10 No. 5 Imrtatinea No. 6. Aiea Vor eae: | 14 5 5 F | 10 ‘While No. dita eee 11 7 4 S | 6 was present. otal. vc; 64 32 18 8 55 390 “fournal of Comparative Neurology and Psychology. Summary of Behavior of No. 5 in the Rope Hxaperiment. Like No. 4 and No. 6, No. 5 had to learn to open the door only. In the first test her attention was directed to the door; she went to it, nosed about it and put her hand on the door. The second test did not add any- thing to her learning, but in the third test she repeated her behavior of the first test. The fifth test added nothing to her ability, but in the sixth, when she was free in the cage with No. 6, his conduct directed her attention to the food door and kept it there until she had learned to get food. At no time did she get food for herself in connection with the performance of No. 6, and the stimulus was never more than No. 6 performing an act and getting the result. General Summary of the Results of the Rope Experiment. Considering together the four animals used in the Rope Experi- ment we may note a similarity in the general behavior. First, no animal failed to learn. Second, in the preliminary trials there was a total indifference to the food door and either total or increasing inattention to the rope. Third, without exception, the first imita- tion test served to direct attention to the door and to the rope. In the case of No. 6, imitation was complete in the first test. Fourth, in the cases in which imitation was not complete in the first test, the successive tests augmented the imitator’s attention and in no case were more than six tests needed to perfect the learning process. Here, as in the chute experiments, we have attentive watching on the part of the imitating animals, followed by an abrupt and radical modification of behavior in the direction of the act observed. This is imitation as we have defined it. TABLE 14. RESULTS OF THE ROPE HW}XPERIMENT. ie Number: Of vamimlall SeSCCctencnsiee cncisiaercecesietsteilsuonstteietclotel sieuckolenetent tenet t-eencRel ll oionet=ia 4 CASES OF (SUCEESS Hulse HaELOMY weyers etereesesiete sieve eee suave) norekcuc tou stete Netto tet oi-d bale ialet ot olsic 4 Cases of partially successful imitation.........5........+.2-0s+ +22 esos eeeans 0 GasesiOf Sailers cies cise ote crete eae tine elcier ole ncliel sie Aeeonoatonel ohele its Morano coo 0) 10K Cases of imitation when the imitator was confined during the activity of thie SMITA TCO Leg ccie orev s etna, oleteie poise Beever eneust noua honeneMoitoustsvoneuhoeacienstel aneteNe reel -Rekenets 2 Cases of imitation when the two animals were in the cage together........ 2 Haccerty, Imitation in Monkeys. 391 IMO Cases. of Immedinte- imitations ssj.ccrsvacieiecos ele ete evere ue aes orenaliore too vets sels eleuene 1 Cases’ of crad wads imitatl Omeaerciecvemlers sien wreverous atern aejes, eeeealctom hone Sere eemabre ce 3 Vi; Cases of imitation in which the imitating animal did not himself experience che result-of thesact=before) performing ibs .nrem | oeente se cele aeleielcre reise) crore 3 Cases of imitation in which the imitating animal did experience the result of the) act-before pertor mim or tte ity ac. oltarcisiecttrens 6 cnet © cua cis clever eieweinioney eters 1 4, PAPER PXPERIMENT. A. Description of Device. For this experiment board # was used. An opening 17 cm. square was cut, the lower edge 30 cm. from the floor of the cage. The opening was covered on the outside by a hinge door. In the center of this door a hole 5 cm. in diameter was cut and on the outside of the door, just at the lower edge of the circular opening was fastened a food box. With the door open, a sheet of ordinary writing paper was laid over the opening and the door was then closed upon it. The hole in the door and the food in the box were thus hidden by the paper (fig. 6). The animal could get food by breaking the paper and reaching through the circular hole. On the inside of board H was a wooden screen which, when dropped down, covered the whole device. When the paper and food were in place, and the animal or animals in the cage, this screen could be lifted by the experimenter by means of a string. When an animal had broken the paper, the screen was lowered by the experimenter and a new piece of paper was inserted. Then the screen was lifted and all was ready for a second test. B. Behavior of No. 2. On five different days, from April 6 to April 10, No. 2 was in the cage alone for thirty minutes each day. He did not get food from the box and made but little investigation of the paper. The most he did toward getting food was on April 7th, when he went to the paper, put his hands on the lower edge of the opening and bit at the paper, but did not tear it through. On April 21, in order to help No. 2 to learn, the experimenter punched a hole in the paper with the point of a lead pencil and the monkey thrust one finger through and tore a larger hole. This was repeated a number of times and No. 2 learned to tear the paper by biting when no opening was made. On April 22 he got food in this way ten times in seven or eight minutes. C. Behavior of No. 3. Preliminary trials—No. 3 was given four trials of thirty minutes each in the old cage in the Harvard Laboratory. In each of the first two trials 392 Fournal of Comparative Neurology and Psychology. he went to the opening and put his hands on the lower edge of the frame. In the third and fourth trials No. 3 was wholly indifferent to the device, not going to it once. He was given a fifth trial of fifteen minutes in the new cage at the New York Zodlogical Park. During this trial he went about the cage leisurely, but gave the paper no attention. Imitation tests—No. 3 imitating No. 2.—The two animals were in the cage together in each of the following tests. First test. No. 2 tore the paper and got food, and No. 8 got some of the seeds which No. 2 dropped. No. 3 did not go to the paper. This was repeated twice; the third time No. 3 went to the paper and looked. The fourth and fifth times No. 3 did not see, but the sixth time he went to the paper before the screen was lifted and turned away as No. 2 tore the paper. The seventh time No. 3 got food through the hole after the paper had been torn by No. 2. No. 2 was now taken out. No. 3 looked at the paper, but became inter- ested in the other monkeys, who were chattering in the nearby cages. He paid no further attention to the paper during the fifteen minutes. Second test. No. 2 tore a hole in the paper and stepped back. No. 3 went up, thrust his hand through and got food. No. 2 was now taken out for five minutes and No. 3 went to the paper and examined it. He did not bite or push. This he repeated four times. Third test. No. 2 was now put back and immediately got food. No. 3 searched the box after him, but got nothing. When No. 2 opened it again No. 3 got food, but he failed the next time. No. 2 was now removed. No. 3 went to the place and bit at the paper, but not hard enough to break it through. This he repeated three times. Fourth test. No. 2 opened the paper and No. 38 grabbed the torn paper and pulled it away. This was repeated twice and then, while No. 2 was eating, No. 3 went to the paper, put his nose against it and pushed. He did not, however, use his teeth. After No. 2 bit through the paper the next two times No. 3 used his hands to tear a larger opening. When No. 2 was removed from the cage No. 38 went at once to the paper and bit through and got food. This he repeated four times, getting food the last time in ten seconds. TABLE 15. No. 3 Imrratine No. 2. vi 8) i ES | ase is . : . Date eG aee | Number of times | Number of times | Rect, | Time in pots hrs aie, No. 3 saw. No. 3 saw in part. é minutes. | | | J ee See Bs SE | ee Jail yale eee 7 | 5 F 10 July alee 1 | 1 Fr 10 Sly Onan 2 2 F 10 wih Siocon os 3 3 S) t Motalieacer 13 11 S 305 Haccerty, Imitation in“Monkeys. 393 Summary of Behavior of No. 3 in the Paper Experiment. The case of No. 3 is a process of gradual imitation similar to that of No. 11 in Chute Experiment B and of No. 2, No. 4 and No. 5 in the Rope Experiment. The first test directed his attention to the paper and each test thereafter increased that attention and its attendant activity. During the tests he got food a number of times; finally, he repeated the act of No. 2 in the fourth test, after haying seen No. 2 get food eleven times. D. Behavior of No. 10. Preliminary trials—First trial, August 13. No. 10 at frst was frightened, due to some disturbance in getting her into the cage. She went about the floor rapidly and up and down the wire as if looking for some way of escape. Once she went to the paper, examined the lower edge of the frame and climbed up on it. Going to the side of the cage, she reached through the wire and tried to pick up straws on the floor outside. She climbed the wire and returned to the floor at once. She now became very persistent in trying to get the straws on the outside, stopping in her efforts only to walk about the cage. She found a hole in the floor which had been used in a former test; she worked at this for a moment; then grasping the frame at the paper in both hands, she shook it vigorously. Then she returned to the straws again. Climbing to X, she perched for a moment and then went to the floor and examined the cracks in the floor and in the door. Then she climbed the wire and remained quiet during the remainder of the time. Second trial, August 17. No. 10 was on the upper part of the wire end during the first eight minutes. Then she was driven to the floor, where she sat in the corner near the paper. Several times she climbed up on the frame about the paper. Then she sat with folded hands near it. Shortly she climbed the cage front. She went to the floor again and sat near the paper. She climbed the wire front and returned to the paper, surveying it with her eyes. She climbed upon the frame and then climbed the front of the cage. She returned to the floor and walked about. Third trial, August 18. No. 10 went at once to the wire in her usual excited manner and remained near the top for two minutes. Then she came to the floor; she walked to the door and back to the end of the cage, climbing the wire end. This she repeated several times immediately and continued to repeat it during the next five minutes. From the upper part of the end she surveyed the floor and sides of the cage. She went to the floor and for a little time sat in the corner near the paper. Then she moved over and sat near the wire end. Then she mounted the wire end. Fourth trial, August 19. Behavior as follows: Up wire end and looked about; around to the front of the cage; back to end and surveyed floor from upper part of it; around to the front and back to end; to floor and walked over to the door; about, looking through the wire and up the end; again to the floor and to the door, back to end and up; around to front and down to the door; glanced at the paper in passing; up the end and back and forth about the wire; to the floor and the door, about the floor; quite free 394 fournal of Comparative Neurology and Psychology. to go about the floor; put hand on frame near paper in passing; up end of cage. Fifth trial, August 20. Behavior as follows: On floor and then turned to door which was still open; looked out intently; climbed the cage end; to the front and down to the floor; across to the end and down; around to front; down to the floor; to door and up the end of cage again; to floor and up end of the cage; to the floor; to the door; sat near wire end and climbed cage again. ‘To the floor and to the door; sat near the paper; to door and back to wire end, sat on the floor and then climbed the cage. Sixth trial, August 21. Behavior as follows: Up the end of the cage and remained for some time; to the front to catch a cockroach and back to the end; surveyed whole cage from end; around to front and down to the floor to look at the door; back up the end of cage; down again; about the floor and up the end of the cage; about the wire. No. 10 was given a sixth test because this was her first experience in the cage and in the earlier tests she had seem disturbed. Imitation tests—wNo. 10 imitating No. 11.—¥First test. No. 10 was in the observation-box on the floor of the cage. No. 11 was free in the cage. No. 10 was attentive to what No. 11 did. P. 1 to P. 5. No. 10 saw perfectly. No. 11 was then removed and No. 10 was released in the cage. Imme- diately she climbed the end to the top and looked back to the floor, to the door, and to the screen. She went to the front, still looking downward. She went to the floor, to the door, and looked at the paper. She turned and climbed the end of the cage again. She went to the front of the cage and looked down at the door and the paper. She went down as far as the brace and looked at the screen. Then she went to the floor, put her hands on the lower edge of the frame and looked at the paper carefully. She turned back to the end of the cage and climbed the wire. Again she went down to the brace and surveyed the floor. She went down to the door and then climbed the wire end again. She repeated this within one minute. She seemed more interested in escaping than in getting food. Again she went to the floor, looked about beneath the frame, and again climbed the end of the cage. Second test. Conditions same as in previous test. No. 10 was very atten- tive to the movements of No. 11 and saw as follows: P. 1 to P. 3. No. 10 saw perfectly. She tried to get out of the observation-box and shook it vigorously. P. 4. No. 10 saw perfectly and shook the box. P. 5. No. 10 saw fairly well. When No. 11 was released, she ran up the end of the cage. She went to the front and looked at the door and floor. She went to the floor and about to the door. She climbed the end of the cage and returned again to the floor; she went to the paper and put her left hand on upper part and pushed; then she put her right hand on the lower part and pushed. Then she climbed the end of the cage and held fast to wire with head turned back toward paper; she went slowly down to the floor and walked across to the paper; she put two hands up on lower edge of frame and bit a hole at Haccerty, [mutation in Monkeys. 395 exactly the right place. She then put her hand in and got food. Time: two minutes. She tore away all the paper and tried for some time to find more food. Failing in this she climbed the end of the cage again. She remained there until the device was reset. Then she looked around at it for some time. Finally, she went slowly around to the front, climbed down to the floor, tried to look through wire at animals in the living cages, stopped an instant at the door, went on to the paper and with feet on the lower edge of the frame bit a hole in the paper. She thrust her fingers into the hole and tore the paper all off, getting the food. She then climbed the cage and waited until the device was reset. At once she went to the floor and across to the paper. She bit at it, but the paper did not break. Again she tried it with the same result. Then she tried to break it with her hand. She climbed the end of the cage and remained there a minute. Again she went to the floor and tried to bite through the paper, but failed as before. She walked about the floor and again returned to the paper. This time she bit at the edge of the hole and literally wore a hole in the paper by rubbing her teeth over the wood. When she had made a small hole, she poked one finger through the opening and by a very hard pull tore the heavy bond paper. Summary of Behavior of No. 10 in the Paper Experiment. In the preliminary trials No. 10 gave almost no attention to the paper, merely looking at it once and passing over it in climbing upon the frame which surrounded it. In the first test she watched No. 11 intently and when he was out of the cage, she manifested an increased interest in the paper. The second test increased this interest and she repeated exactly the behavior of No. 11 within two minutes after his removal. TABLE 16. No. 10 Imrratine No. 11. = Number of times e one eae hn 2 ; : Date: Nan all merisemed | Number of times | Number of times Result Time in Reet | No. 10 saw. No. 10 saw in part. ‘| minutes. | Aug. 23..... 5 | 5 2 12 Aug. 24..... 5 5 S) 2 Rotalesen. 10 10 N) 14 E. Behavior of No. 9. Preliminary trials.—¥irst trial, July 2. No. 9 was active about the cage. He went to the paper and put his hands on the lower part of the frame. He repeated this soon again. Then he climbed the wire in front; then he 396 fournal of Com parative Neurology and Psychology. climbed upon the frame at the paper. He whined and called most of the time he was in the cage. Second trial, July 3. No. 9 took no notice of the paper during the entire fifteen minutes he was in the cage. He climbed about the cage and tried to push the door open. Third trial, July 4. No. 9 was very active about the cage, but paid no attention to the paper during the first minutes in the cage. Later he went to the paper, bit at the frame and climbed upon it. Fourth trial, July 5. In the fifth trial No. 9 climbed about the cage and upon the screen frame about the paper. He made no effort to tear the paper. Fifth trial, July 5. The behavior of No. 9 in the fifth trial was similar to what it was on the previous days. He gave no attention to the paper. Imitation tests—No. 9 imitating No. 2.—The two animals were in the cage together in each of the following tests. First test. No. 9 was not at first inclined to be attentive to No. 2. It was not until No. 2 got food the fifth time that he apparently saw the act. Then he put his hands on the bottom of the screen frame and reached one hand through the hole, but he got no food. Several times before the device could be reset No. 9 went to the screen and bit it and climbed upon it. He had not been near the screen that day. During the sixth, seventh and eighth. manipulations by No. 2, No. 9 was beside him and saw what was done. Each time he put his hand into the opening, but got no food; each time he climbed upon the lifted screen. After No. 2 had been taken out No. 9 was quite active, running all about the cage. He went to the screen several times and bit at the edge of the frame. Once he pushed his hand up over the paper and at another time he bit at the inner edge of the frame next the paper. Second test. No. 9 saw each time and was near No. 2 in the corner of the cage. During the third, fourth, and fifth performances No. 9’s hands were on the lower edge of the frame and after the paper had been torn No. 9 got food along with No. 2. When No. 2 had been taken out No. 9 went to the paper, climbed upon the frame and jumped to the wire. He returned to the paper and bit at the edge of the frame, but not at the paper. Several minutes later he went to the paper and put his nose to it. This he repeated three times. At the last time of the three, he sat on the bottom of the frame and tried the. paper with his fingers. He finally tore it and got food at the end of fourteen minutes. When the device was reset No. 9 went to it and sat on the lower edge of the frame. He tried to tear the paper with his fingers, but failed to make a hole. He later went to it and bit at the edge of the frame, but not at the paper. A little later he examined the paper with his nose, but did not bite it. Third test. No. 9 saw perfectly five times in six and got food twice. When alone he went to the paper, examined it with his nose, and went away. Later he went to the paper and fingered the edges. He then went away, returning once more during the fifteen minutes, but doing nothing. Haccerty, Imitation in Monkeys. 207) Fourth test. This test was made forty-seven days after the preceding one. The conditions were the same as in the preceding test except that No. 6 was used instead of No. 2. Performance 1. No. 9 saw and was just back of No. 6 when he tore the paper. P. 2. No. 9 was on the back of No. 6, but his head was turned away. Because No. 9 insisted on riding on the cab back of No. 6, the latter was removed and No, 2 was substituted. P. 8 to P. 7. No. 9 saw No. 2 at the paper from front wire and came down to it. He reached his hand in to get food, but No. 2 had taken it all. When No. 2 was out No. 9 came down from the wire, climbed the screen frame, and sat on the edge. He jumped to the wire front, but at once returned to the corner by the paper and sat on the floor for some time looking at the paper and at that part of the cage. He then climbed the front of the cage. Twice he came to the floor, climbed the frame at the paper and jumped back to the wire front. After spending some time about the cage and on the floor, he climbed the screen frame and tried to bite the paper. He was too small to reach the hole from the floor and when he got upon the lower edge of the frame his body covered the place where he should bite the paper. Fifth test. No. 9 and No. 2 were in the cage together. Since No. 9 was so small, a box was placed on the floor below the paper so that he could climb upon it and thus have a more nearly equal chance with the larger animals in exerting his force against the paper. P.1. No. 9 saw from the middle of floor. Pp. 2. No. 9 saw in part. P.3. No. 9 saw from the wire front above the brace. P.4. No. 9 did not see the paper torn and did not come down for some time. He saw No. 2 eat the food, sitting on the box. During each of the previous times he had searched the hole for food and got none. He now paid no attention to the place. No. 2 was allowed to continue eating food at the opening. No. 9 ran all about the cage, but paid no attention to No. 2 and the paper. Finally No. 9 went to the box and got sunflower seed and a piece of banana. P.5. No. 9 saw from the top of the wire front. He came down for food and No. 2 punished him; he ran up the cage crying. P. 6 to P. 10. Did not see. Watching the experimenter. P. 11 to P. 12. No. 9 saw from above X. No. 2 was now removed. No. 9 ran up and down the wire and about the floor. Then he went to the box. He looked about for a moment and then pushed his hand over the upper part of the paper above the hole and around the upper edge of the paper. He looked at it and then climbed the wire and went about the cage. After climbing about the cage, he came back to the paper, and put his hands against it. He did not get his hand over the opening, although he rubbed them about the paper considerably. He then played about the cage. 398 “fournal of Comparative Neurology and Psychology. Sixth test. No. 9 was put into the observation-box and No. 6 was free in the cage. RP. 1 and P22. No. 9) saw in’ part P. 3 to P. 7. No. 9 saw the entire performance. No. 6 was taken out and No. 9 was released. No. 9 ran up the wire, and came back to the floor and to the paper; he looked at it and climbed up on it. He then ran up the wire. Again he went to the paper and bit at the edge of it. Then he climbed up on it and jumped to the wire front. He repeated this performance twice. Then he ran all about the cage and came back to the screen. This he did repeatedly. He seemed more bent on getting out of the cage than on getting food. Several times he put his nose to the paper, but was not persistent about it, looking away at once. Later he bit the lower edge of the frame. Still later he bit at the frame next the edge of the paper. Seventh test. The box was below the screen. No. 9 and No. 2 were free in the cage. P.1. No. 9 saw and got a sunflower seed through the opening. P.2. No. 9 did not see, although he sat near and must have heard the paper tear. He seemed indifferent to No. 2’s getting food. P.3. No. 9 saw from post above Y, but did not seem interested. P.4. No. 9 saw from upper part of wire end. P. 5-P. 6. No. 9 saw from the wire above X. P.7. No. 9 saw from the wire above X, and going to floor got a grape skin No. 2 had dropped. P.S. No. 9 saw from wire above X. He went to the paper and put his hand in after No. 2 had left. P.9. No. 9 saw from above X. He went to the floor, put grape skin in his mouth, and went to the paper, where he put his hand in the hole. No. 2 jumped at him and struck him. P.10. No. 9 did not see. P.11. No. 9 saw from upper part of wire end; he got grape skin and went to the hole as before. P.12. No. 9 saw from wire above X. No. 2 was taken out and No. 9 was left alone. He climbed the cage front at first. He came down to the box beneath paper, and looked all about the paper; climbed frame, jumped to front and ran to end of the cage. Again he went to box and looked all about the paper. He fingered the lower edge of the frame and then put his left hand flat against paper above the hole and pushed; he shoved palm over upper left-hand part of paper. Then he pushed his right hand over lower right-hand corner in the same way. Then he sat on the box and looked. After some time at box he climbed the wire to the upper part of the end. He soon went back to the box, where he sat before the paper and looked all about it. He climbed the frame, jumped to the front and ran up to the top of the wire end. He repeated this entire performance. He went to the box again, looked at the paper and the frame, and returned to the front and end of the cage. Haccerty, Imitation in Monkeys. 399 Summary of Behavior of No. 9 in the Paper Hxaperiment. During the five preliminary trials No. 9 gave the paper no attention. The first test brought forth imitative behavior in that No. 9 put his hand through the hole to get food after seeing No. 2 get food. After the removal of No. 2, No. 9 pushed his hand up over the paper as if to tear it, a thing he had not done in the preliminary trials. The second test increased this attention and after repeated fingering at the paper he tore it off and got food. In the later tests he did not succeed in breaking through the paper, but he repeated the movements of No. 2 and gave persistent attention to the paper. His failure was possibly due not to the absence of the tendency to imitate, but to the lack of muscular power to exert sufficient strength to break the paper. TABLE 17. No. 9 ImitaTtina No. 2. I ber of times : Ag ; : Date | NE eee a Number of times | Number of times | Result Time in ; Keaec tieaet | No. 9 saw. No. 9 saw in part. "| minutes. | : | | | July eG: 8 1 2 Fr 10 ule. fe 1 5: 5 5 S 15 ithye"8:.-5 2: 6 5 Fr 10 No. 9 Imrratine No. 6. Aug. 24.....| 7 | 5 2 F 12 PAN B24 re | 4 | 2 il F 10 No. 9 Imrratine No. 2. Aug. 24..... 8 3 He al) 10 Aug. 24..... 7 5 2 F 10 NUR PAD oo oc 12 10 F 15 Total 57 | 39 i S | 82 General Summary of the Results of the Paper Experiment. That the problem set in the Paper Experiment was one easy of solution is evidenced by the fact that of eight animals all but three learned it alone, most of them in the first trial. Of the three animals which did not learn it alone two learned it by a process of gradual imitation. The other one was never more than partially successful, but his failure seemed due to a lack of physical strength rather than to a failure to repeat the act which he saw performed. 400 ‘fournal of Comparative Neurology and Psychology. Here again, we note attention on the part of the observing animal and a subsequent marked change of behavior (somewhat sudden) in the direction of the behavior observed in the performing animal. TABLE 18. RESULTS OF THE PAPER EXPERIMENT. I NHN Ae Loe Chemin WSCOl whey avdMMlHOKON IWESHES) Koadcacdoadcooconoocdecosddode 3 CASES Of SUCCESS ly IMITATION epee eee crs oesey eliotn euoueeale shel crehcnepel tclek stat elceteye ersten 2 Cases of pantiallliys SUCCESSiml MiMitationmesew eel tieriier rer etor titel acict metre al Cases Of Sfailure iO: MTtAlews mele hate tae eters ceoretemener orci eaten iis Rosters sicusceroma) ie Cases of imitation when the imitator was confined during the activity of ENE TMNICATEE «6 iss sc5 Sis So wet ena rede are cli ndere eee ere ne at ouch ome ee rare ee thoteae to ane: ol eioreienes aft Cases of imitation when the two animals were in the cage together ........ 2 IHU CaASes- OL MMEeda te MM DAGON News reas eetete sects hele eel aren a ener terete tetereteiens 0 Cases ofseraduall amibacionee cna siete ereraicke is erorctets ote rekon relist teotteneler keine 3 EVe Cases of imitation in which the imitating animal did not himself experi- ence the result of the act before performing it ...................-.-+-- 1 Cases in which the imitating animal did experience the result of the act before. performine, iG sige sd sect onc el ccem ere eee hee Racer rem eaeiaat Melero 2 5. SCREEN EXPERIMENT. A. Description of Device. The device in this experiment was a modification of the one used in the Paper Experiment. The paper was not used. The string which lifted the screen (fig. 7, @)was removed. The act which the animal had to perform was to push the screen up with one hand and with the other reach through the hole and get food. No animal was tried in this experiment which had not previously gotten food in the Paper Experiment. B. Behavior of No. 4. No. 4 first pushed the screen up when the paper was being adjusted in the Paper HExperiment. She did not, however, tear the paper. The screen dropped back in place and she lifted it again. The fourth time she pushed the sereen up, it stuck and did not drop back. She then tore the paper. Haccerty, Imitation in Monkeys. 401 When the device was reset, No. 4 pushed the screen up and tore the paper. Thereafter, she lifted the screen and got food when she wanted to. C. Behavior of No. 6. Preliminary trials —First trial. No. 6 was active, climbing up and down the wire, and upon the screen. He fingered about the edges of the screen, but made no effort to raise it. Second trial. The second day No. 6 ran all about the cage, climbing the wire and upon the screen. He examined the screw eye where the string had been attached, but made no effort to lift the screen. Third trial. On the third day he seemed interested in all parts of the cage, examining every crack and hole in it. He fingered the top of the screen as if to move it. Six times he climbed upon the screen. The remainder of the time he busied himself catching flies. Fourth trial. On the fourth day No. 6 paid no attention to the screen during the fifteen minutes. Vifth trial. No. 6 paid but little more attention to the screen on the fifth day. Three times he climbed upon it to jump to the wire front and three times he examined the hole above the screen. Imitation tests—No. 6 imitating No. 4.—The two animals were put into the cage together in each of the following tests. First test. No. 6 was at first indifferent to the movements of No. 4. He usually saw No. 4 get the food, but failed to see him lift the screen. In the six times No. 4 lifted it in the test, No. 6 appeared to see twice. After No. 4 had been taken out, No. 6 paid no attention to the screen for ten minutes. Second test. The first four times No. 4 lifted the screen No. 6 did not see. He was picking over the hulls left on the floor. The fifth time he saw from the opposite corner of the cage, and while No. 4 was up on the wire front eating, No. 6 went to the screen and looked. The sixth time the screen stuck when lifted, and No. 6 put his hand in and got food. After the seventh time No. 6 went to the screen and pushed on the lower edge of the frame. Then he pulled at the top and went away. He went back immediately and putting both hands on the screen pushed. He then went away, but when No. 4 lifted the screen he saw and went at once to it. Putting his hands on it he pushed it up one-third of the way. Then he pushed it up so as to reveal the hole, and got food. No. 4 pushed it up again and No. 6 saw. Imme- diately No. 6 lifted the screen and got food. After No. 4 had been taken out No. 6 lifted the screen eight times in ten minutes. He could do it perfectly. Summary of Behavior of No. 6 in the Screen Haperiment. No. 6 had seen the screen go up in the Paper Experiment and he had experienced getting food when the screen was lifted. However, his five pre- liminary trials in the Screen Experiment did not lead him to get food. When first in the cage with No. 4 he was not inclined to be attentive. When he saw No. 4 getting food in the second test he at once became inter- 402 “fournal of Comparative Neurology and Psychology. ested in the screen. When once his attention was centered on the screen he very soon repeated the behavior of No. 4, not at first in a perfect way, but in his fifth effort he did it in exactly the way No. 4 had done the act in his presence. TABLE 19. No. 6 Imiratine No. 4. Dates None oes | Number of times | Number of times | Result. Time in ” F55 Biers, | No. 6 saw. No. 6 saw in part. | minutes. ze | | July Gt - 6 | 2 4 | F 10 Wulys Tae 11 | 6 S 12 | | Totals <0. 17 | 8 4 | aS 22 D. Behavior of No. 56. Preliminary trials.—First trial. No. 5 examined the cage all over, but she manifested no particular interest in the screen. After ten mintues she looked it over slightly. Second trial. On the second day she looked out through the wire, poked her fingers through the hole in the door, and then went to the screen and pulled at the screw eye. She was quite active, climbing about the cage rapidly. Once more she went to the screen, and then spent the remainder of the time catching flies. Third trial. On the third day she bit at the screen frame and pulled at the screen during the first few minutes. She then spent the rest of her time as on the previous day. Fourth trial. On the fourth day No. 5 showed more interest in the screen at first. She tried to shake the screen frame. Later she fingered the screen and bit at the frame. Fifth trial. On the fifth day No. 5 bit at the screen frame several times and climbed upon it twice. Most of the time, however, she spent in other parts of the cage. Imitation tests.—No. 5 initating No. 4 and No. 6.—In all of the following tests the two animals were in the cage together. First test. No. 5 was somewhat wary of No. 4 and did not come near. She saw No. 4 open the screen once in three times. When No. 4 was taken out No. 5 went to the screen and examined it, but she gave it no persistent attention. Later she fingered the lower edge of the screen. Second test. No. 5 was attentive and saw the lifting of the screen five limes. When No. 4 was out No. 5 went at once to the screen and pulled at the top of it. She then ceased to be interested in it and examined other parts of the cage. Haccerrty, Imitation in Monkeys. 403 Third test. No. 5 was very hungry. She saw No. 4 open the screen once in the first three times. After the third trial, while No. 4 was up the wire front, eating, No. 5 went to the screen and examined it. She put her fingers into the cracks and climbed upon it to examine the top. The fifth and sixth times she saw plainly, and after each went to the screen and examined it. The seventh time No. 5 saw and hurried to the screen, but No. 4 let it drop and No. 5 turned away without trying to manipulate it. 10 aes: 10 July 14.. 20 | 15 F 10 July 15.. 30 20 he Pad 10 July 31.. 26 | 15 3 sank 10 aes ot 10 5 F 10 Total..... | 147 106 5 (ee 134 H. Behavior of No. 2. Preliminary trials.—First trial. No. 2 was not active. He examined the sereen with his nose and hands and bit at the screw eye in the top of it. Second trial. On the second day No. 2 pushed the screen, but did not lift it; later he climbed upon it and examined the top of it. This he repeated twice. Third trial. The third day’s behavior was similar to that of the previous day. No. 2 pulled and gnawed at the screen and the screen frame. Part of the time he worked vigorously. Most of the time, however, he was in other parts of the cage. Fourth trial. On the fourth day No. 2 was more vigorous than ever. He spent five minutes without intermission chewing at the bottom of the screen frame. He then quit and looked at a hole in the door. He made another brief examination of the top of the screen and went away. Several times he returned and examined the screen, once lifting one corner of it by pulling on the screw eye at the top. Fifth trial. No. 2 was active at the screen, pulling at the top and biting the lower part of the frame. He made no progress, however. Imitation tests—No. 2 imitating No. 4.—During all of these tests No. 2 and No. 4 were together in the experiment cage. ; 406 fournal of Comparative Neurology and Psychology. First test. It was the first time No. 2 and No. 4 were together. They caressed at once and then No. 4 went to the screen. She lifted it five times and got food each time. No. 2 sat by her and seemed to see every movement, although his excitement may have kept his attention from centering on what she was doing. When No. 4 was removed No. 2 displayed no more than usual interest in the screen. He climbed to it, picked at it with his fingers and used it as a stand to climb up the post. Second test. No. 2 was excited and was attentive to No. 4. He ate the apple crumbs which she dropped after the first opening. At the second and fifth opening he was “picking fleas,’ and did not see the screen go up. While she was eating her fifth feed he went to the screen and examined it. The sixth and seventh times she lifted it, he saw perfectly. After the seventh he went to the screen and pushed against the lower edge of the frame and then bit at it where it joined the cage post. After No. 4 had been taken out No. 2 began to work at the screen. He picked at the lower edge of the frame where it joined the post and then climbed upon the screen. He was too excited to work persistently, running to the side of the cage and starting at every noise. He worked intermittently for six minutes. Then when at the opposite side of the cage he started, ran to the screen and gave a single push upward on the lower edge of the frame. He then put both hands on the frame, but did not push. Because of his activity the period was prolonged to fifteen minutes. Third test. No. 2 saw No. 4 the first time she got food and climbed the cage to get food from her. He saw the second time and got food from the hole while the screen stuck. When his food was gone No. 2 went to the screen and pulled at the top and pushed at the lower edge. He saw the third time, but became interested in picking fleas from No. 4 instead of getting food. The fourth and fifth times he saw perfectly. After No. 4 had been taken out No. 2 ran about the cage, but paid no attention to the screen for the first four minutes. Then he worked at the lower edge of the screen a little. Fourth test. After the third performance No. 2 went to the screen and touched it. After the fourth he went to it, put his hands against it and pushed, but he failed to lift it. Each of the five times he saw very well. After No. 4 had been removed No. 2 went to the screen a number of times and put his hands on it. He also bit the lower part of the frame and climbed upon the screen. Fifth test. No. 2 saw No. 4 well each time. He went to the screen after No. 4 had opened it and remained there while No. 4 ate her food. His efforts to get food were feeble. After No. 4 had been taken out No. 2 climbed up on the screen a number of times, but in no case did he seem bent on getting the food. Sixth test. No. 2 was interested in No. 4 and saw her get food each time in twenty. July 1 (2: Cee 20 10 F 10 Tialiys Wisse 2 | 25 10 | S 20 Total..... 68 30 kegees 50 Summary of Behavior of No. 2 in the Plug Experiment. The Plug Experiment set a different problem from any of the experiments already described. The food was obtained at one place, but the door could be opened only by working at a place removed from the door. No. 2 made no progress toward a solution of the problem during his pre- liminary trials. The first three tests did not aid him. In the fourth test No. 5 made more rapid trips between the door and the plug. She seemed 414 ‘fournal of Comparative Neurology and Psychology. quite excited. Probably her increased activity served as an increased stimu- lation to No. 2, for after her removal he gave more continuous attention to the door, and then went from the door to the plug and pulled. He repeatedly tried the plug and finally succeeded in pulling it out. After his first success, however, he did not go to the door, although he did after the next. The one experience, however, did not establish a perfect act, for when the device was reset he did not go at once to the plug, but worked at the door instead. He gave up trying to get the food and went about the cage. He went to the plug again only when his eyes accidentally (so it seemed) fell upon it. In the third experience there was apparent an element of accident, but after he got food the third time, he seemed to know the trick perfectly. D. Behavior of No. 6. Preliminary trials.—First trial. No. 6 climbed the cage and then went to the door and pushed at it. He examined all about it and then climbed the wire. He grasped the plug three times. Then he went back to the door six times and pushed it. Later he bit the plug. Second trial. No. 6 was very playful. He leaped about the floor and up the wire. Once he went to the food door and later he went to the plug and bit it. He then went back to the floor and to the food door. He pushed at it and then played about the cage. Third trial. No. 6 showed no interest in either the door or the plug. Fourth trial. On the fourth day No. 6 was very active. He looked at the door and later perched at the plug, but he made no effort to pull it out. Fifth trial. On the fifth day No. 6 went to the door and looked at the food. Then he ran about the cage. He was totally indifferent to the plug, and although he had been eager to get into the cage he was eager to leave it at the end of the period. Imitation tests —No. 6 imitating No. 5.—The two animals were in the cage together in each of the following tests. First test. No. 6 soon discovered the food outside of the glass door and when No. 5 opened it No. 6 got the food. No. 5 punished him several times and No. 6 cried so much that his howling compelled his removal. Second test. No. 6 was in the cage with No. 5 while she opened the door twenty times. He rarely saw—not more than five times in the twenty. No. 6 learned that the door opened and was inclined to sit in front of it. This turned his back to the plug and he did not see No. 5 pull it. Finally No. 5 drove him away from the door and he saw her pull the plug a few times. When No. 5 was removed No. 6 went to the door and examined it. Then he ran up to the plug, bit at the end of it, and tried to pull it out. He ran down to the door at once. He climbed to the plug and worked at it with his hands. Several times he repeated this trip from the plug to the door and back to the plug. Third test. No. 5 was very eager to get the food. No. 6 saw only occa- sionally—five times in fifteen. When No. 5 was removed No. 6 gave his first attention to the door. Then he climbed to the plug, but did not work at it. He worked persistently at Haccerty, Imitation in Monkeys, 415 the door. Finally he opened the door by working at it directly and got the food. Fourth test. No. 5 was exceedingly active and after getting food would pull the string a number of times before the door was reset or the food replaced. No. 6 saw her ten times in fourteen. When No. 5 was out No. 6 became very active about the door, working continuously to get the food. Once he ran up the wire and bit the plug. Fifth test. At first No. 6 was quite indifferent to No. 5, but he became attentive as he saw No. 5 getting food. When No. 5 pulled the plug the eighth time No. 6 climbed the post and pulled it after her. As he pulled it he looked at the door. This he repeated after the ninth performance also, and again after the tenth. When No. 5 was removed No. 6 went at once to the door; then he played up the wire and came down again to the door. Then he ran up to the plug and pulled it until he got the door open, two minutes after the removal of No. 5. When the apparatus was reset No. 6 began to work at the door. After one minute he climbed the cage, took one look through the wire, and went back to the door. Then in the midst of vigorous pushing at the door he suddenly stopped, fairly flew up the wire to the plug, and pulled it vigorously until it came out and the door opened. He came down quickly and got the food. He repeated this twice in two minutes. The next time the plug stuck, and although he worked at it vigorously, it was ten minutes before he succeeded in pulling it out. He had run about the cage somewhat during that time. He opened the door once afterward. TABLE 26. No. 6 Imrratine No. 5. | Number of times : ae . : : F Date: | No. 5 performed Number of times Number of times Result. Time in ihe eye No. 6 saw. No. 6 saw in part. minutes. ‘ill Sear 8 | 1 F 10 Palais ee 20 5 F 10 Dulyatse ge 15 5 F 10 Dialy. Tass. 14 10 F 10 Awl WSyee 5 oe 16 10 S 10 Totals. 73 | 30 | 1 S 50 Summary of Behavior of No. 6 in the Plug Experiment. In his first preliminary trial No. 6 gave some attention to the food door and to the plug. His interest in the plug disappeared in the later trials. The second imitation test, which may be reckoned as the first, served to direct his interest to the plug. The third and fourth tests did not seem to increase this interest nor to make it productive of profitable results. The fifth test did show a decided increase of attention to the movements of No. 5, and, 416 “fournal of Comparative Neurology and Psychology. finally, a repetition of those movements. The association between the door and the plug, however, did not seem perfect until after No. 6 had succeeded several times in pulling the plug and in getting food. The tendency of No. 6 was to center his attention on the door and, after failure there, to resort to the plug. This may have been due to the fact that he once got food by working directly at the door. General Summary of the Results of the Plug Experiment. In nicety of imitative behavior, the Plug Experiment furnishes less satisfactory results than do some of the other experiments. This is no doubt due in part to the fact that the food door and the means of its opening were in different parts of the cage. The two things could not well be within the range of vision at the same time. In transferring attention from the door to the plug the animal usually lost sight of the door. He did not see the imitatee pull the plug and at the same time see the result of the pull. In case he saw the plug pulled, his eyes must follow the imitatee back to the door in order to see the result. Despite this difficulty, the experi- ment yielded two cases of behavior in which the influence of the imitatee was sufficient to guide the behavior of the observing animals to a successful issue. In the successful behavior there seemed to be an element of accident. It is impossible, however, to explain the conduct of either No. 6 or No. 2 as a case of random movement and accidental success, for prolonged opportunity to solve the problem in this way resulted in failure for each of them. Nor does it seem possible to think that No. 2 or No. 6 repeated the movements of No. 5 merely from seeing her perform the act and without connecting with her act the result which followed it. Each of the animals failed to pull the plug after seeing it pulled, until there had been abundant opportunity to see the performing animal get food. TABLE 27. RESULTS OF THE PLUG EXPERIMENT. 1k Number of animals ised sintimitationetestsee cece eee ioiieitinene ee Spelt Cases: of Successful -imilitationiccs aceon cove aohoie cow iene ek IOP IGiSIcreear Shoe Cases) of “partially successtull imitations eee oer reece 0 Casesvofetailures to imitates eee sane is lataaaeneshoke etnias Greene ln ckave een ne Haccerty, [mutation in Monkeys. 417 1a Cases of imitation when the imitator was confined during the activity of thre AMIR CCS Se srate. crs secre eeetee tater seatals orc re or olelw eine! oh eee ren seek tay Soe ern tere 0 Cases of imitation when the two animals were together in the cage........ 2 Tee Cases sof “immediate pamita toms cate oa, oo shcudaves.cretersbotere a Setenere aoe eee aie 0 @ases) ofseraduall imitated Ome crac cre sare sce oe og frais sic seeds ee a ce aioe eee Zs IW Cases of imitation in which the imitating animal did not himself experience the result of the act before performing it.................. pete ee eees al Cases of imitation in which the imitating animal did experience the result OLothevactybeLone speLrORmMimn flit ccs ects cis cero custard orlerreneeia htehel cee Retort: al 7. BUTTON EXPERIMENT. A. Description of Device. In this test the slide door (fig. 8, @) used in the Plug Experiment was the place where the animal could get food. It could be opened by a button (fig. 10, b) in board D, which must be pushed to the right. This button was S cm. broad at the largest breadth of its pear shape and 14 ecm. long. Its lower edge was 22 cm. from the floor. A string, ¢, fastened to the back part of the button passed through a hole, 5 cm. in diameter, in board D, and along the outside of the cage to the slide door. The button was fastened to Board D at the top by a small bolt. A small knob fastened to the middle of the button enabled the animal to grasp it easily. A screw eye in the board prevented the button from being pushed to the left. The animal could get food by pushing the button to the right and then passing to the slide door in board A which had been opened by the movement of the button. B. Behavior of No. 3. Preliminary trials.—First trial. No. 3 worked at the door, biting the edges for ten minutes. He then walked to the button, gave one bite at it, and came back to the door. Later he repeated this, biting the knob on the button. He climbed the cage a number of times and then sat in the corner of the cage near the food door. Second trial. On the second day No. 3 went to the door and bit at the edge, but not so vigorously on account of nails that had been driven into the edges of the opening to protect it. He went twice to the button and bit the edges. Then he ran about the cage, and finally rested in the corner near the food door. Third trial. The third day No. 3 again worked at the food door, biting the edges. Then he went to the button, bit at it and came back to the door. He repeated this behavior four times in three minutes. 418 “fournal of Comparative Neurology and Psychology. Fourth trial. On the fourth day No. 3 paid no attention to the door or to the button. Fifth trial. On the fifth day he worked at the door for a short time. Then he climbed about the cage and ended the period by sitting in the corner near the button. Once he bit at the button. Imitation tests.—No. 3 imitating No. 2.—Both animals were put into the cage together in each of the following tests. First test. No. 3 was not attentive to No. 2 at first and was somewhat afraid. He saw four times fairly well. Several times the experimenter pre- vented No. 2 from opening the door because No. 3 was not watching. When No. 2 was out No. 3 went to the door and worked vigorously for three minutes. He then went to the button, bit it and pulled as No. 2 had done. He came back to the door at once. Then he returned to the button, bit it, and came back to the door. Later he went to the button a number of times. Second test. No. 8 was afraid and avoided the door and button while No. 2 was present. No. 2 was very active and opened the door much oftener than No. 3 saw. He saw five times in nineteen. When No. 2 was out No. 3 worked at the door intermittently for several minutes, going once to the button and biting it. Third test. No. 2 and No. 38 were on good terms and No. 3 kept near No. 2 and-watched him most of the time. No. 2 worked very rapidly, but No. 3 saw him ivell, five times in ten. When No. 2 was out No. 3 worked at the door for a little time, and then went to the button and pulled it back with his teeth. This movement was different from his previous acts at the button, which were mere bites with no effort to pull. He looked out at the opening behind the button and then went to the door and got food. Time: two minutes after the removal of No. 2. He repeated the entire performance within one minute and six times more within ten minutes. TABLE 28. No. 3 Imrratine No. 2. Number of times Date No. 2 performed Number of times | Number of times Result Time in : RS aie No. 4 saw. | No. 4 saw in part. ; minutes. ¢ | duly 2enee 9 4 2 Bs eal July, 276s 19 5 4 ‘ee 00 uly 28ers 10 5 3 Sa WW Motalaeee | 38 14 9 S 30 Summary of Behavior of No. 3 in the Button Experiment. At first No. 38 manifested an interest in the door and in the button, but this interest waned as the preliminary trials were continued, and seemed Haccerty, Imitation in Monkeys. 419 entirely gone in the fourth and fifth. It received a decided accentuation in the first test, after No. 38 had seen No. 2 get food four times. In the second test it seemed about the same, but in the third test it led No. 3 to repeat in detail the movements of No. 2 and to secure the same result. C. Behavior of No. 4. Preliminary trials.—Virst test. After four minutes in the cage No. 4 went to the button, put both hands on it, bit at the knob and bottom of the button and turned away. She spent the remainder of the time on the floor of the cage and on the wire. She returned to the door a number of times, but made but little effort to get food. Second trial. On the second day No. 4 went to the door frequently and occasionally to the button, but she made no effort to manipulate either. She was anxious to get out of the cage. Third trial. On the third day No. 4 went to the door, but made no effort to get food. Later she smelled at the button, but made no effort to move it. Fourth trial. On the fourth day No. 4 paid no attention to either the door or the button. TVifth trial. On the fifth day the behavior of No. 4 was as usual. She bit at the button once or twice in passing and went to the door twice. Imitation tests.—No. 4 imitating No. 2.—In each of the following tests both animals were in the cage together. Tirst test. No. 4 at first was not inclined to notice No. 2. .She saw five times in twenty-three. No. 2 was frequently prevented from opening the door until No. 4 was looking. She often saw the door open, but paid no attention to the button or to No. 2. When finally she saw No. 2 push the button, she went immediately and did the same thing. She did it three times more while No. 2 was present. When No. 2 was out No. 4 worked two minutes at the door and then walked over to the button and pushed it back. This disclosed the opening behind the button and she thrust her hand out. She withdrew it immediately and came back to the door and got food. When the apparatus was reset she went to the button immediately; pushed it back, thrust her hand out and came at once to the door and got food. She repeated this four times. Then she ceased to thrust her hand out, and came immediately to the opened door. Within five minutes she had gotten food ten times. Summary of Behavior of No. 4 in the Button Experiment. The behavior of No. 4 was decidedly changed by seeing No. 2 push back the button. For five days, fifteen minutes per day, she had had the oppor- tunity to get the food by pushing the button, but had not done so. Yet she pushed the button within five seconds after seeing No. 2 do it. There is no evidence as to whether she connected the button with the food at the time. The directness with which she later went from the door to the bution, pushed it back and came back to the door to get food would indicate that she had made the connection, That the association was complete after the 420 ‘fournal of Comparative Neurology and Psychology. second experience is evidenced by the directness and rapidity with which she continued to perform the act. TABLE 29. No. 4 Imiratine No. 2. | | | Number of times | : | | F Number of times i Hera 2 ; Date. 2 4 Number of times | Result. Time in Ne Ne eee No. 4 saw. | No. 4 saw in part. | ee | minutes. . | | Jalye dice see 23 5 | 3 8 10 D. Behavior of No. 5. Preliminary trials.—First trial. No. 5 worked at the door; then she climbed the cage and came back to the door. She went to the button and spatted it with both hands. Later she bit at the screw eye, held the button, and bit it. She then turned and pushed it with her feet. Later she grabbed the screw eye in her hand and bit it. After twelve minutes she placed herself opposite the door and plunged against it twice with great force. She went to the button, and, placing herself opposite it, plunged against it twice in the same manner. Then she went to the door and looked. She then plunged against the button and spatted it several times; she went once to the door and looked. Later she bit again at the screw eye; she then went from the door to the button and back to the door. Second trial. On the second day No. 5 worked at the door for the first four minutes. She then went to the button and spatted it. Later she hocked her tail in the wire about three feet above the door, placed her feet on the board A, and, with head down, she lifted her body out from the board and threw her weight on her hands against the door. Third trial. On the third day No. 5 went to the door, climbed the cage, and after several minutes went to the button and spatted it. During the remainder of the time she went about the cage in the usual way, paying no attention to the door or the button. Fourth trial. On the fourth day No. 5 went to the door once. Once she went to the button and taking the screw eye in the left hand spatted the button with the right. She gave no further attention to door or button. Fifth trial. On the fifth day No. 5 went about the cage mostly indifferent to the door and button. Once she spatted the button and bit it. She spent a little time at the door when she was first placed in the cage. Imitation tests —No. 5 imitating No. 2 and No. 4.—The animals were in the cage together in each of the following tests. Wirst test. No. 5 watched No. 2 closely and saw the entire performance five times in ten. The first time she saw No. 2 push the button she followed and pushed it back herself. She did not follow to the food. She did this three times on seeing No. 2 do it and usually missed seeing him get food. Haccerty, Imitation in Monkeys. 421 When No. 2 was out No. 5 worked vigorously at the door and then went to the button and pulled at the screw eye and the knob, but not in such a way as to open it. Three times she went to the button and back to the door. Second test. No. 5 saw No. 2 push the button five times in thirteen, but did not follow him to the food door. After No. 2 had been taken out No. 5 became quite eager about the door for a minute. She went to the button, but made no effort to move it. Third test. No. 4 was used instead of No. 2. No. 5 saw No. 4 five times in fifteen. She did not follow her about, but kept out of her way. When No. 4 was removed No. 5 worked a little while at the door and then played about the cage for ten minutes. Fourth test. No. 5 was afraid of No. 4 and kept away from her, owing to the punishment No. 4+ had given her. She kept close watch on No. 4, however, and saw her move the button back and get food. This she saw ten times in nineteen. With No. 4 out No. 5 looked through the door at the food, but did not work vigorously. She took her leisure about the cage for ten minutes. Vifth test. No. 2 was again used. No. 2 pushed the button and No. 5 followed and pushed it, and coming to the door got food. No. 5 retreated to the front of the cage and kept her eyes on No. 2. Each time when No. 2 pushed the button No. 5 came to the door for food. She thus prevented him from getting any, for he was afraid of her. Once (fifth trial) she got food, and immediately went to the button and pushed it back. No. 5 became much more attentive to No. 2 than at any time previously and her eyes flashed froin the door to No. 2 and from No. 2 to the door, and always when he pushed the button she came to the door. After seeing him three times more she went to the button again, pushed it back and went directly to the door for food. This she repeated three times, the first two times coming directly to the door, and the third stopping for a moment to examine the opening behind the button. This she repeated once and then performed the entire act six times, not stopping to make any examination of the button. Summary of Behavior of No. 5 in the Button Hxperiment. Despite her unusual and persistent activity No. 5 did not once push the button during her preliminary trials. Yet she did push it at once after seeing No. 2 do it. Three times she repeated this, but not once did she go to the food door after doing so. What she had learned seemed to avail her nothing until the fifth test. She then followed No. 2 through his entire act of pushing the button and coming to the door to get food. She maintained a heightened interest in No. 2 during the whole of his ten performances, and by the end of the time was able to get food for herself as she had seen him do it. It should be noted that after pushing the button back and securing no result she ceased to push it during the second, third and fourth tests. Where this conduct reappeared it was connected with the getting of the food. 422 “fournal of Comparative Neurology and Psychology. TABLE 30. No. 5 Imrratine No. 2 anp No. 4. | | co . | Number of times } Date No. 2 and No. 4 | Number of times | Number of times Result Time in . performed the act. No. 5 saw. No. 5 saw in part. ‘| minutes. July 2 7eoee 10 5 | 1 ae a (0) F 10 Julys28e-eee| 13 5 | 4 No. 5 Imtratine No. 4. July 28.....| 15 | 5 | 3 eons 10 July 29... 19 | 10 | paw 10 K as 3 ate oe oe. oe 8) | ¢) Ee Wests No. 5 Imrratine No. 2. | | | While No. July. 30's .5 10 | 10 | | 5 | 2was be a | present. 35 | a Tolls 6 on - 67 E. Behavior of No. 6. Preliminary trials.—First trial. No. 6 was very frantic about the food door. He rushed to where the plug had been in the Plug Experiment and worked at the hole in the post. He looked at the button and put his hands on it, but made no effort to move it. Later he bit at it. He then gave up his efforts. Second trial. On the second day No. 6 tried the door as before, but on account of nails which had been driven in the edge he could not bite it. Once he went to the button and put his hands on it, and ten minutes later bit at it in passing. Third trial. The third day there was the usual behavior about the cage. No. 6 worked at the door for several minutes and once pulled at the screw eye. Then he took his leisure about the cage. Fourth trial. On the fourth day No. 6 was very active at the door, biting and pushing it. Once he grabbed the screw eye in passing. He then played about the cage, going to the door frequently, but not working at it. Fifth trial. On the fifth day No. 6 tried the door a few times, but not vigorously as on previous days. He bit at the button in passing. Imitation tests.—No. 6 imitating No. 4.—Both animals were in the cage together in each of the following tests. First test. After the first few minutes No. 6 became afraid of No. 4 and kept away from her. In fifty-five perfomances No. 6 saw only three times. The first time he saw No. 4 move the button, he followed and did the same thing himself, but did not follow No. 4 to the door. When No. 4 was removed No. 6 went at once to the food door and worked incessantly and with great vigor for five minutes. Once during the time he went to the Haccerty, Imitation in Monkeys. 423 button and bit it. He did not come back to the door and he went to many other places in the cage as well as to the button. At other times, when near it he paid no attention to the button. Second test. No. 6 was not so frightened as in the previous test and remained on the floor near No. 4 five times in nineteen. He saw her push the button and get food. The second and fourth times he saw No. 4 push the button he went to it and looked out at the opening. When No. 4 was out No. 6 became very active with his teeth and hands at the door. After nine minutes he went to the button and pushed it with his hands, but as the push was directly toward the beard, and not to one side, as was necessary to open the door, he did not succeed. Later he put his hand on the screw eye. Third test. No. 6 watched more attentively than in the first test and saw five times in twenty-seven. When No. 4 was removed No. 6 worked incessantly at the food door with his hands, feet, and teeth. He used his tail to thrust through the wire when he could not reach around with his hand. Once he thrust his tail around the corner of the cage and caught the string to which the banana was attached. He was not allowed to get food in this way. He paid no attention to the button during the time. Fourth test. No. 6 was more attentive than on any previous day and saw five times in eleven. After seeing the fourth time he went to the button and pushed with his hand, but not in such a way as to open the door. After the fifth time he bit at the lower edge of the button. After No. 4 was out No. 6 did not go to the button. He worked more or less intermittently at the door for ten minutes. Fifth test. No. 4 worked slowly and gave No. 6 a good opportunity to see, but he was not attentive and saw only seven times in forty-one. Then his look was not direct. The first time he saw, he went to the button and looked through the hole behind it. When No. 5 was out No. 6 got food by fingering at the door. Then he worked at the door and once bit the button, but he was not at all active. Sixth test. No. 6 kept his attention on the door during thirty-two per- formances, but rarely turned his attention to the button even though No. 4 went from the door to the button and back to the door for food repeatedly. His attention was almost wholly on the door. Ten times in the thirty-two he saw No. 4 push the button. When No. 4 was out No. 6 became very eager at the door and continued so for ten minutes, but did not once go to the button. Seventh test. No. 6 was quite indifferent to all the movements of No. 4. He often looked at the door as No. 4 left it to go to the button. It was not clear whether No. 4’s leaving the door suggested that the door was about to open, or whether No. 6 was all the time interested in the door and showed his interest only when No. 4 left the way clear. He saw five times In thirty- five and then only by glances. When No. 4 was out No. 6 became at once interested in the door and worked at it most of the time for five minutes. Then he ran about the cage, but paid no attention to the button. 424 ‘fournal of Comparative Neurology and Psychology. Highth test. No. 5 was used instead of No. 4. No. 6 was not at all attentive to No. 5. At times he watched the door when No. 5 went to the button. He was not afraid of No. 5, so did not run away. No. 5 worked rapidly and moved the button with wide movements of her arms. No. 6, however, showed no interest in the button. When No. 5 was out No. 6 paid no attention to the button and but little to the door. Ninth test. No. 6 was very indifferent to the movements of No. 5 except at the door. He often saw No. 5 get the food at the door and twice got food there himself. The movement of No. 5 from the door to the button and back to the door apparently meant nothing to him. When alone No. 6 worked intermittently at the door for ten minutes, but did not notice the button. Tenth test. No. 6 saw five times in the twelve that No. 5 opened the door. When No. 5 was removed No. 6 became busy at the door, but paid no attention to the button during ten minutes. Summary of Behavior of No. 6 in the Button Experiment. What No. 6 saw in the Button Experiment seemed to profit him nothing. Once he repeated the movement of No. 4 in pushing back the button, but he did not at that time nor later connect the button with the food door. At no time did he give good attention to what was done in his presence, in the ten tests seeing only fifty-one out of three hundred and three performances. TABLE 31. No. 6 Imiratine No. 4 anv No. 5. Date Number of cimes | Number of times | Number of times Result. Time jn performed the act.| No. 6 saw. No. 6 saw in part. minutes. — = Si — = =| | July 2725. 55 | 3 | 2 F 10 Julie 2 Seer 19 5 F 10 July 28 Dil | 5 F 10 July 29..... 11 | 5 3 F 10 Julive29 eee 41 | 7 F 10 Jalkhy S0)5c00 < oo 10 Je 10 July 30eeee 35 | 5 F 10 No. 6 Imrratine No. 5. | July 30.2. 50 3 | F | 10 INN OD) ee cae 21 3 | 10 F 10 Aue 18), 12 5 | F 10 Total 303 51 Ee F 100 Haccerty, Imitation in Monkeys. 425 reneral Summary of the Results of the Button Kxperiment. Taken as a whole the Button Experiment gives three cases of unitation, no one of them immediately successful in detail. In the eases of No. 4, No. 5 and No. 3, there was an immediate modification of behavior, but in no ease was there an exact repetition of the behavior of the performing animal. It did not require many repeti- tions of the act, however, for each animal to learn to perform the act perfectly. F TABLE 382 RESULTS OF THE BUTTON EXPERIMENT. if NAUK e Cus EhoudiaUks; Weel Tol UV MOMN Wess Pooboouossooenb Oden oocod doco 4 CASCSHOLESUCEESS HUG AIM TCA TOME sop. ve hel aches cleuel en cher ceonel anche iol, soe elie/sfave ssi cis tmenetene ehene 3 Caseshormpartiallliyay SUCCESS Ue Trini tag Olle ecrciena cic cieiie es lel) ete tevelcietelsletenctelercccl O CasesrOherallure nto l Cater accesses ose cris sesiete sieve cle elcl. ei sie stelersi nn Bosca Sconce mel: Tele Cases of imitation when the imitator was confined during the activity of NCSI ACO persteiea sue cir take ooh asa Siete aie avai lsjaio elle a setts wieu wel enene. tote heaelts, #) Mpanete: « keLereiere 0) Cases of imitation when the two animals were together in the cage..... a3 Terele WaASESHO CaM INEGI ACS SUNIL OMe y2 cee o, oueleral ane: sss cereic « ;euews #1 akey2) a evel eeitehe ro ahiers1e cusltererocemels 0) CASCSHO fae Vadiail mI CAGLO Myers, aie «ickoiesuallets sie cqaisjersien si tiesepe ers, eearereheversuohoroneberere ts tere 3 TVG Cases of imitation in which the imitating animal did not himself experi- lence: the resultcor therack, DeLoresperLOrmillNes lice ccelsuctes seereeele euene eieiee 3 3 Cases of imitation in which the imitating animal did experience the resuli Of UhEex act? DELONes PELLOTMIM GUE. J aisvers scckeye wie econ ee) se cue vepaeehe ve ieus. oarsaeels eeecne 0 8. STRING EXPERIMENT. A. Description of Device. From the top of the experiment cage (fig. 11) strings 1, 5, 6, and 7 were dropped downward along each of the corner posts to within 18 cm. of the floor of the cage. Along the back of the cage and 15 cm. apart were sus- pended three other strings, 2, 3, and 4, in like manner. To the lower end of each string was fastened a small knob, k. In the following observations on the behavior of the animals 2¢ indicates the second string at the place where it enters the cage, and 2k indicates the knob attached to the end of the second string. 426 fournal of Comparative Neurology and Psychology. In the lower part of board B, 6 cm. from the floor, was a circular opening, L, 5 cm. in diameter. On the outside of the board was a square chute (fig. 12, a), the bottom of which, 0, was level with the bottom of the circular opening, LZ. In the chute, a little way above the opening, was a trap door, ¢, which could be opened by a lever, d, to which could be fastened any one of the seven strings above described. In this experiment, string 2 was so attached. By pulling this string the animal on the inside of the cage could open the trap door in the chute and thus cause the food on the door to fall to the bottom of the chute or roll out into the cage through the opening in board B. In either case the monkey could get it. B. Behavior of No. 13. Preliminary trials.—First trial, August 14. When No. 13 enetred the cage he went at once to L and looked at the opening. He then went to 6% and, taking it in his hand, bit it, dropping it after the first bite. Recrossing to 1k he did the same thing, immediately afterward climbing the wire end of the cage. Returning to the floor he went to 5 and to L, thence crossing the floor to 6, biting Gk and recrossing to 2k and biting it. He then moved about as follows: up the wire; around to 6; back to 1; on the wire; to the floor; to 2k and bit; thrust his hand into L; to 5k and bit; to 4% and bit; to the small door and pushed; to 7k and bit; to 2 and started to climb, but turned his attention to 1 and 8; thrust his hand into LZ; to the door and to 5; across to 7 and up the wire to X. The time for the above described behavior was four minutes. He then continued as follows: perched at X; to the floor; to the door and chewed the edges; up to X and perched; to the floor; thrust hand into L; up to X, perched and played with 7; to 1; back to 7 and to the floor; to 6 and looked at it; around to 1 and looked; on around to 7 and up to X; to the floor; to 5; up to the top of the cage and examined the hole in the top; to the floor; to 4 and to 7; up to X; to the floor, the door, and back to X; pulled 7 up to him; about the wire front; to the floor and about the floor to 35k, which he took in his hand; up to XY; to the floor; to the door and car- ried 6k up to X; to the floor and to 2k; carried 2k in his hand up to X and looked it over; then to the floor and around the cage to 5, 4, 3, and 2 in succession. He gave a jerk at 2 and dropped it. Second trial, August 15. No. 13 was active as on the previous day, but spent more time in looking and less in running about. He first ran up and down the wire several times; then he went to L and looked in. His later movements were as follows: up the wire; to the floor; to the door and bit at the edges; to LZ and looked; up the wire to the brace and back to the floor; to the corner post at 5k and bit the post; up the wire and bit at it; down to the floor and again up the wire on the front of the cage; perched at X for some time and then went to the floor and bit 7k; carried 1k up to X and worked with it; bit the knob and the end of the string; pulled 7% up to X and chewed it; then went to the floor and walked about; carried 1k up the side of the cage; after some time returned to the floor, bit at 5k; carried 2k up the wire to X and farther up the wire. Haccerty, Imitation in Monkeys. 427 In none of his movements did he display the same eagerness and expectation as in the previous trial. Third trial, August 17. On this day No. 18 was even less active than on the second day. He went up the wire to the brace and returned to the floor, going to the door, pushing on it and passing around to L. He then carried 2k up to X and bit at the string, dropping it almost immediately. Then he climbed the wire, but returned to XY, where he perched and remained for some time looking about the cage. Later he went to the floor and examined all around the edge of the floor, but soon returned to Y, where he remained for some time again. Twice later he climbed the wire to the top, but spent all the rest of his time at Y, looking about the inside of the cage and out through the wire. Fourth trial, August 18. The behavior of No. 18 was about the same as in the preceding trial; he climbed to X and returned to the floor; he touched dk and 4k; then he carried 2k up to Y and bit at the string, dropping it after a moment. For several minutes he sat at Y. Then he drew 7k up to him and worked at it for some time. At first he worked directly at the knot and made some progress toward untying it. Then, as if discouraged, he began biting and pulling the protruding end of the string. At the end of several minutes’ continuous work he dropped 7k and went to the floor, tried to climb the post at 5 and passed on to 2k, which he carried up to X, where he chewed at the knot in the string. After a moment he let it drop and swing back to its place. Two minutes later he went to the floor and to L, at which he looked intently. Climbing the wire to XY, he perched for the remaining few minutes he was in the cage. Fifth trial, August 19. No. 13’s behavior was about the same as on the previous day. He climbed up and down the wire several times and examined around the edge of the floor. Carrying 2k up to X, he bit at the knot once and dropped it. Then he played up and down the wire, and going to the door tried to open it, afterwards carrying 6k up to X, biting it and dropping it when he climbed higher up the wire. He climbed to the upper part of the wire and chewed at the edges of the cage frame, but quit when spoken to. Again he carried 6k up the side and end of the cage, dropped it, and settled at X for several minutes. Later he went to the floor, to L, to the door and carried 6h up to X. He dropped it at once and remained quiet. Once again he went to the floor and carried 2k up to X, where he chewed the string and licked the knob. Imitation tests.—No. 13 imitating No. 5.—F¥irst test. No. 18 was put into the observation-box and the box was placed on the floor of the experiment cage exactly in front of L. This position enabled No. 13 to see all the move- ments of No. 5 in getting food. No. 5 was free in the cage. The two animals had never been together before and No. 5 was much frightened. Instead of working at getting food she crouched in the corners of the cage and occa- sionally dashed at No. 13 as if to frighten him. Once No. 5 ran up the wire end and leaning over to 2¢ pulled it with her teeth. No. 13 did not see the pull, but he saw No. 5 leaning over to the place. He became demonstrative, and No. 5 did nothing but crouch for several minutes. The observation-box 428 “fournal of Comparative Neurology and Psychology. was then moved farther away from L, and No. 5 went to Z and got food. No. 13 saw her get it. He was very impatient and tried repeatedly to get out of the box, working at the door and shaking the box vigorously. No. 5 again waited and after some time went cautiously to 2k; she took it in her hand, but did not pull with sufficient strength to drop the food. No. 18 saw her do this. For some time No. 5 refused to work on the floor, but she attempted to get to 2f several times. This she was prevented from doing. Finally she became accustomed to the presence of No. 18 and moved about the floor freely. She then became very eager to get food. Within a few minutes she had operated the mechanism seven times. The record for No. 18 was as follows: Performance 1. No. 13 saw well. Pp. 2, P. 5, and P. 6. No. 13 did not see. P. 3, P. 4, and P. 7. No. 13 saw and was eager to get out of his box. If we count the times No. 5 pulled the string, but did not get food, No. 15 saw the performance five times, four of which he saw entire and one in part. No. 5 was now removed from the cage and No. 13 was released. At once he climbed the end of the cage to 2¢ and taking it in his teeth, he pulled it several times. Then he went to the floor and walked about; after some time he went to L and got the food that had dropped when he pulled 2¢. When the food had been eaten he climbed to 2¢ and pulled the string with his teeth. The dropping of the food made a noise, but No. 18 did not notice it. After several more pulls he went down to XY and sat there for a short time. Then he went to the floor and walked about, later going to L and discovering the food. Stowing it in his cheeks he went up to X and ate it. When it was gone he wanted to go up to 2¢ again, but was not allowed to do so. He then perched at Y, and looked about for some time. Going to the floor, he stopped at LZ, looked in, took hold of 2k, dropped it, looked at LZ again and walked away. He climbed to XY, and returned to the floor after a little while. He took 2k in his hands, dropped it and looked into L. Then he carried 2k up to X and played with it. Second test. The conditions were the same as before. No. 5 was not so frightened and worked at once. The record of No. 13 was: Performance 1. No. 13 saw very well. PAP ee eG NO» osGldanounsee: P. 3 and P. 4. No. 13 saw well. P. 7 and P. 8. No. 13 saw No. 5 get food, but did not see the pull. P. 9. No. 13 saw, but did not seem attentive. Of the nine times that No. 5 got food No. 13 saw the whole performance four times and in part twice. No. 5 was then removed and No. 15 was released. At once he went to L and got a grain of sunflower seed that No. 5 had left, and carrying it up to Y, ate it. He then wanted to go to 2t, but was not allowed to do so. He went down to the floor and to L, merely looked at it and passed on to the door. He returned to Z at once. Searching about he found another seed, which he carried up to X and ate. When the food was gone he attempted to go up to 2¢, but was prevented. He went to the floor and walked about, going to LZ twice. Once he looked at 2k, took Haccerty, Imitation in Monkeys. 429 it in his hand, but did not pull. After looking at it a moment he carried it up to XY, where he bit at it and dropped it. Then he went to the floor, walked about, and climbed back up to Y, where he remained during the remainder of the time. Third test. This test was made immediately after the previous one and the conditions were the same. The record of No. 13 was: Performance 1 to P. 5. No. 18 saw the whole performance and was very say- age in his demonstrations toward No. 5, jumping at the side of the cage with’ wide-open mouth. Only twice did No. 18 turn his head away from No. 5 and then only for a moment each time. No. 5 was now taken out and No. 13 was released. He was very slow of movement walking about the floor. Twice he went to L. and then climbed the wire to X, where he perched for a little while. Going to the floor, he passed the door and thence to L, looking into the hole several times. In one hand he took 3k and in the other he took 2k, but did not pull at either. Dropping both he climbed the end of the cage, but returned at once to the floor. Stopping at L, he clawed the opening with his hand and then climbed the wire to the brace. Going to the floor he tried the door where he had entered the cage and went to L, returning again to the door. For several minutes he worked at the door trying to open it. Once he stopped to turn about and look in at ZL, but renewed his efforts at the door immediately. Giving up opening the door, he went to L, took 3k in his hands, dropped it and started to carry 2h up the wire, but dropped it; clinbed to X and perched, playing with 7 and 7k. Fourth test. Conditions were the same as in the previous test. The record of No. 18 was as follows: Performance 1. No. 18 saw plainly. P.2. No. 13 did not see and did not seem to be interested as on the day before. P. 3. and P. 4. No. 13 saw. P.5. No. 13 saw very well. P.6 and P.7. No. 18 did not see. When No. 18 was released he at once climbed to XY, returning to the floor immediately. Going to LZ he looked in and put his hand into the opening. Passing up by YX, he tried to get up to 2k, but was not allowed to do so. He returned to the floor and walked about; glimbed to Y again; and, returning to the floor, he went to 1 and looked in. He then climbed to XY and up and down the wire end of the cage. Once again he went to Z and put his hand into the opening. Fifth test. No. 5 and No. 13 were put into the cage together this time. ~ Performance 1. No. 13 was at X when No. 5 pulled the string the first time; he saw her pull and saw her get food. Climbing down he got the food which No. 5 had not yet eaten. He then became very threatening and No. 5 was frightened. P.2. No. 18 saw again from X and got the food as before. No. 5 was still afraid of No. 13, who was threatening. When No. 13 had eaten the 430 ‘fournal of Comparative Neurology and Psychology. food he did not climb the cage as before, but kept near Z. Only once did he go up, and then to chase No. 5. ». 3. While No. 18 was at the brace after chasing No. 5 down No. 5 pulled the string. No. 13 saw this and after a moment he went to Z and got the food. From this on he came to the floor whenever he saw No. 5 near L. In her turn No. 5 assumed a threatening attitude toward No. 153. P.4. No. 18 saw perfectly from the brace and came slowly down and got the food. No. 5 was not inclined to eat the seeds, having an appetite only for grapes. P.5. No. 18 saw while on the floor. No. 5 got the grapes and No. 138 got the seeds. P.6. No. 5 pulled the string while No. 13 was climbing the wire. He jumped to the floor and rushed to L; No. 5 fled up the wire. P.7. No. 5 pulled the string when No. 18 was eighteen inches away. He rushed to Z and got the grape. A moment later when No. 5 went near the opening, No. 18 rushed to the place and kept such a close watch that for some time No. 5 could not get near 2h. P.Sto P.10. No. 18 saw from XY and drove No. 5 away before she could get the food. P.11. No. 5 got the grape and No. 13 got the seeds and went up to X to eat them. No. 5 was now removed. No. 13 finished eating the seeds he had gotten and then went to the floor, to L, and back up to XY. He spent almost the entire ten minutes at Y. Near the end of the time he went to 1 and examined it carefully. Then he looked up at knobs 2k and 3k, put his hand on 2k, took it off, and looked back at ZL. He then climbed to XY, returned to L, and went back to YX. Sixth test. No. 13 and No. 5 were put into the cage together again. No. 5 was afraid of No. 18 and kept away from him. Performance 1. No. 5 pulled the string and got the food. No. 18 saw from the wire near the top of the end of the cage. Coming quickly to the floor, he searched L vigorously. P.2. No. 5 pulled the string. No. 18 saw plainly and went to Ll. He tried to get food, but No. 5 had taken it. .5. No. 18 saw while on the floor near L, and going to the place searched a long time for food. Twice he put his hand on 2k. P.4. No. 18 kept near No. 5 at LZ, and when she pulled the string she had to reach her arm over the head of No. 13. His whole attention was on her movements and he saw perfectly. P.5. No. 13 was beside No. 5 at LZ and saw perfectly. He got the grape and frightened No. 5 away. P.6. No. 18 saw perfectly, got the food and sat by ZL, so No. 5 did not return. Once he put his left hand on 2k and straightened out his arm as if to pull, but he did not exert much force on the string. Immediately he thrust his other hand into LZ. Again he took 2k in his left hand, straight- ened his arm as before, immediately afterward thrusting his right hand Haccerty, Imitation in Monkeys. 431 into Z. A third time he put his left hand on 2k, straightened his arm and followed this action by thrusting his right hand into LZ as before. He then went away from L. P.7 to P.10. No. 18 watched No. 5 carefully and drove her away from the food, which he ate. He then went up to X and watched No. 5. When she went near 2k he dashed for L. No. 5 was now removed and No. 13 was left alone in the cage. Not all the food that No. 5 had brought down had been eaten, and No. 13 con- tinued eating, going to L to get the seeds and climbing to X to eat them. When he could find no more food he sat at L and scratched the edge of the opening with his hand. Then he grabbed 2k and pounded it against the board; taking 3k in his right hand and 2% in his left, he pounded them together; afterward he did the same with 2k and ijk. He then went up to X and perched for a moment, but almost immediately went to the floor and to L. Thrusting his hand in he searched for food and then looked into the opening intently. Looking up, he took hold of 2k with his left hand and pounded the board with it vigorously, then bit it and dropped its Taking 2k in his hands he went up to Y, dropping the string as soon as he was settled on the brace. His eyes turned at once to ZL and he went down to it and searched for food; he picked up 2h in both hands and looked at it carefully; then he pounded the board with it. Dropping it he went up to XY, returning at once to L; he grabbed 2k in his hand, put it gently against 7k and dropped both of them; he returned to Y, and, coming down to L, he did the same thing over with 2k and ik. He went up to X and tried to go up to 2t, but was pre- vented. He then perched at Y, looking at L for one minute. He was intent on the getting of food at L, but he seemed puzzled. After looking intently at L and the strings he went to the floor and tol, stopping to sit down and look the string and opening all over. Then he again went up to X. Again No. 15 left his place at Y and went to the door, pushing on it in an effort to get out. Being unable to get out, he turned away from the door to L and sat down in front of it. Quite slowly he looked it all over and, in the same deliberate manner he looked up to 2k, took hold of it with his left hand and gave a steady and vigorous pull. The food dropped to the bottom of the chute and his right hand shot into the opening and pulled it out. The ‘food was soon eaten and No. 13 immediately pulled the string again with his left hand, getting the food in the same way as before. Without once leay- ing his place, he pulled the string six times, eating the food between the pulls. While eating the food the third time, he put his hand up to 2k several times, but he did not pull hard enough to get the food. When his food was gone, however, he pulled the string with a jerk and the food came. Repeatedly he dallied with the string in this manner while eating the food, but he never failed to give a vigorous pull when the food was gone. For fifteen minutes he sat before L, getting food repeatedly. He pulled the string fourteen times in addition to the ones already mentioned, a total of twenty times in all. The time from the removal of No. 5 until No. 18 got food the first time was twelve minutes. 432 ‘fournal of Comparative Neurology and Psychology. TABLE 33. No. 13 Invrratine No. 5. Number of times | Number of times Date. No. 5 performed No. 13 saw the | Number of times Result. Time in the act. entire performance No. 13 saw in part. minutes. Aug. 25 7 4 1 F 10 PNW, P35) 5 3 boc 9 4 2 F 10 Aug 25 meee 5 5 F 10 ANIC 26) e er 7 4 F 10 J NUE PAD eae 11 9 F 10 PN Pls cack 10 10 Ss 12 Total 49 36 3 S 62 Summary of the Results of the String Experiment. After No. 15 had failed to solve the problem in his preliminary trials, he was allowed to see No. 5 pull the string. During the first tests he was confined in the observation-box. After four tests he still failed, when left alone in the cage. He was then put into the experiment cage with No. 5. The two animals were strange to each other, and No. 13, being the larger, was inclined to follow No. 5 about the cage, punishing her as opportunity offered. Because of this, he was usually near No. 5, when she pulled the string, and often frightened her away before she could get the food. After she had been removed, No. 13, repeatedly searched the food opening, and worked alternately with the three strings nearest the food open- ing. He seemed to have associated the strings with the getting of food. When No. 5 was put back into the cage, No. 13, was more atten- tive than formerly. After No. 5 had been removed, No. 13, worked more continuously at L and at the strings. He now singled out string 2 from the others. He grabbed the knob at the end of the string, in his hands; he pounded it against the board, carried it up the wire, and pounded it against the knobs attached to the other strings. Frequently, during this behavior he dropped the string and searched L for food. He had advanced one step in his learning. It was not strings that were associated with the getting of food, but it was a particular string. Haccerty, Imitation in Monkeys. 433 The only possible explanation for this centering of attention ou a particular string, was that No. 13 was imitating the act of No. 5. By repeated and varied effort, No. 13 finally repeated in exact detail the behavior he had witnessed. TABLE 34. RESULTS OF THE STRING HXPERIMENT, Its Number of animals used in imitation tests. 2.5.02 6.3. c ses sees eens orn Cases! Of: Successfull simitations..0% Secs vied ocloaa pe cas heels ae tee enor Sefeoul CasesnompartiallyesiuCCeSShul smi taltON Sprcueterte nouns enevotoh omen tenedeisteneletetvietc aco ele Cases of imitation when the imitator was confined during the activity of (HaVEXSaI Ne ODES NI clot AeA ee ee eee APRA eerie A ecm icntnG choo. bic.0. o-oo 0. Cases of imitation when the two animals were together in the cage........ 1 NOE Cases OfeimMmMediateiMTEAGL OMe. care sins cool eye sien eieokeelerene Fd ile. Saysenantens ers coomeo Al CASESHO fe SEA MIA MIM AGLOME «aire crs sus es) costes elesioiltevercaciiete e olene iva atereten iy eeu 1 EV Cases of imitation in which the imitating animal did not himself experi- encerthe resultof the: act before) PeLLOLrmine TU. rec mcrae eit el sileraeteerereele ) Cases of imitation in which the imitating animal did experience the result OlsthesachHeroresperko mmm Gl Gere teeters ciels eier hele eeeveneih oh raenen Neer eer tions al V. Generar SuMMARY OF RESULTS AND CONCLUSIONS. Cases of Imitation. (a) With Respect to the Several Hxperiments.—The seven ex periments (Chute Experiments A and B are counted as one) to which the several animals were subjected, yielded a total of sixteen cases of successful imitation, three of which were immediate, and five cases of partially successful imitation. No one of the experi- ments failed to yield at least one case. Four of the experiments yielded imitation, successful or partially successful, for every animal given the full series of tests (100). The other three gave a total of five failures. 434 fournal of Comparative Neurology and Psychology. In tabular form this appears as follows: CASES OF CASES OF FAILURE TO IMITATION. IMITATE. Chute Experiment A and B.......... 5 2 Rope: Raperinient 4 etree eee ) Paper Haxpertment, 26 2.55% tc. <= es 3 ) Sereen Hixperimiemtra.: eau: soe oe ee aban: 2 Pine siperindterit mete aa s e ee 2 0 Bitton Bexperitve nt: eyegs ecm ss ence 3 i String Experiment ........ Soar ee Bes’ 0 oma eto er ae ears 21 5 (b) With Respect to the Individual Animals.—Of the eleven animals used, all but two exhibited imitative behavior. These two were given the full series of imitation tests and are recorded as absolute failures. Of the nine animals which exhibited imitative behavior, seven were successful in each experiment in which they were used. No. 3 succeeded twice and failed twice; No. 6 sue- ceeded four times and failed once. No. 5 made the best record, solving three of the problems alone or with slight help from the experimenter and learning all the others (four) by imitation. The record of No. 2 is almost the same, but he required more aid from the experimenter in learning one of the tricks. No. 4 learned two tricks alone, failed on two, and learned three by imitation. No. 9, No. 10, and No. 11, each had one opportunity to manifest imitative behavior, and no one of them failed to do it. No. 13 had two oppor- tunities and imitated in both cases. On the basis of their ability to learn by imitation the animals may be arranged in three classes. The first includes those animals which did not manifest a failure. Here would come No. 2, No. 4, No. 5, No. 9, No. 10, No. 11 and No. 13. In the second group are the animals which succeeded in some tests and failed in others. Here are No. 3 and No. 6. Haccerty, Imitation in Monke YS. 435 The third group contains those animals which failed to manifest imitative behavior. Here are No. 1 and No. 8. The accompanying table exhibits the records of the individual animals. TABLE 35. Recorp oF InpIvipuAL ANIMALS. F | No. of experi- N | No. of experi- ments in which OES Gee Cases of failure umber. | ments learned imitation tests | Cases of imitation. to imitate. independently. were given. | INO Alene see ae 0 i! | 0 1 No.) 22: 2 3 3 0 NO, Bee il 4 | 2 2 No. 4.. 2 3 3 | 0 NOs 5 3 4 | 4 } 0 No. 6.. 1 5) | 4 1 IN@s oes 1 1 0 1 INO; Qe 0 1 1 0 WOO ooae tate ol 0) 1 1 | 0 Noll 2 esis 2 1 | 1 | 0 INO StS ae el 2, 2 | 2 0 (c.) With Respect to the Several Species—The number of cases of imitation per species is of interest. The results show that the tendency to learn by imitation is not confined to any one species or genus among those studied. The number of animals used is too small and the variation in the number of experiments to which the several animals were subjected is too great for these results to have any significance in showing the relative imitative ability of the various species. Cebusd(6 specimens) arte osc, 3. a ween tae Ree or aly Cebus lunatus (2 specimens)................. {¢ Cebus fatuellus (1 specimem) ...,)......5...% 3 Cebus capucinus (1 specimen).............. yet Cebus flavus (ijspecimen).. on eee eee ei Cebus hypoleucus (1 specimen).............. 2 Mivcneus: ((Sespecwmens’) ts: J s.8 5 chon eee any We ae 4 Macacus rhesus (2 specimens)............... 2 Macacus cynomologus (1 specimen)....... Beets oD Of the two animals which failed one was a Cebus lunatus and the other was a Cebus hypoleucus. 436 “fournal of Comparative Neurology and Psychology. TABLE 56. THE RESULTS OF THE SEVEN EXPERIMENTS. if Number of animals: used: in! TMTrabion TeSES se om eet ote le eerie enero *26 Cases of Successtul imitations ereterere aie pete te a oie ne rete re eto een= 5 okra aangerers - 16 Gases) of partially Sucecesshinl) immita tm deer eraclotte siete) eretelet tele teustetekel= per 5 CASES Of fallure sto TMA. ee eto ieee eer ane ence ar cleaner 5 ine Cases of imitation when the imitator was confined during the activity of the. dMItAtees. Sas haw ree lee eee er emeele aye eRe Sie of otehe ete oeeaeo eect n eet 8 Cases of imitation when the two animals were together in the cage........ 18 alike @ases obsmmediate imitatlones see an cure eee eet eee tele ete teed ele ot 5 Cases Of eradualy imMiaWOne ces race einer ened eter are ahaa tod Sooo ebio. 16 LV. Cases of imitation in which the imitating animal did not himself experi- ence the result of the act before performing it................+eeee eee ant! Cases of imitation in which the imitating animal did experience the result of the ack before: perrormiimne Ws. ose ete «lene orci eletat stele oie wh aPa “lobe telah ca 10 V. ; Cases of imitation where the result of the act was obtained at the place where “the: acto was, DELLOTIMICG Sse eel ooo alle claveray tebe) tel arotsl otal stot ther 16 Cases of imitation in which the act was performed at one place and the | result was optamed) at anopiers place se. pee teteke oislste eras liek fete ienat Pele tet -elle = *Counting each animal once for each experiment in which it was used. 2. Features of Imitative Behavior. (a) Relation Between Animals.—It is significant to note that imitation did not always occur between animals thoroughly accus- tomed to each other. It might be supposed that congeniality be- tween animals was a good condition for imitation, but that this is not necessarily so the results of my experiments seem to indicate. As I shall show later, familiarity tends to lessen attention, to make each animal follow its own tendencies. Strangeness and a certain amount of pugnacity seem effective in arousing attention, which is the first condition for imitation. In the String Experiment No. 5 was a total stranger to No. 13 and the latter was highly attentive Haccerty, Imitation in Monkeys. 437 to her every movement. The same is true of No. 2 and No. 4, of No. 4 and No. 11 and of No. 4 and No. 13 in the chute experiment. The cases of imitation between animals wholly congenial are less than one-half of the cases recorded. (6) Levels of Imitative Behavior—Monkeys react to the pres- ence of one another in various ways. At least four levels of reaction are well defined. The first of these is characterized by the simple arrest of attention. One animal walks across the floor of the cage or climbs a pole, and another animal looks in its direction. That monkeys manifest this sort of reaction requires no extended experi- mentation to prove. Every moving object, and much more, every moving monkey catches their attention. In my investigation the cases where animals failed to respond in this way may be grouped into two classes. The first group has to do with animals which, through being caged together, had become thoroughly aceustomed to each other’s behavior. No. 6, who had lived in a cage with No. 4, often seemed unaffected by her conduct when he was put into the experiment cage with her. He would go about the cage hunting food and pay no attention to the actions of No. 4 who might be getting food at the time. If, however, under the same circumstances, No. 2, a strange animal, was substituted for No. 4, No. 6 would become alert and apparently see everything No. 2 did. There were other cases of the same sort. The other group of cases are those in which one animal had whipped another. The whipped animal usually attended to his enemy only to avoid him. When the latter’s attention was directed toward some object in a distant part of the cage, the vanquished animal went about hunting food for himself and did not see what the other animal did. It was, of course, quite otherwise with the bully. He was usually inclined to watch his victim, unless some- thing more interesting presented itself. These cases in which the attention of a monkey was not attracted by the act of another monkey seem explainable by the circumstances under which they occurred. They serve, therefore, to emphasize more strongly the point that monkeys do tend to give attention to the acts of one another. Since such attention is the invariable ante- 438 “fournal of Comparative Neurology and Psychology. cedent of any behavior that may be called imitative it is important to note that it exists. A level of social response more advanced than mere looking is following. Tere again, it requires but little observation of monkeys to show that the tendency to follow is very strong, especially among the Cebus monkeys. More complicated than mere looking or following is behavior of this sort: One animal performs an act, gets food in a given locality and goes away. Another animal which observes this behavior goes, immediately after, to that locality, as if to get food. What the second animal does in that locality seems at this level of behavior to have no relation to the behavior of the first animal. There were numerous instances of this sort of behavior among the animals which I have studied. In the Sereen Experiment, in particular, there were clear cases. No, 5 repeatedly went to the corner of the cage where No. 4 had gotten food by lifting the screen. The same was true of No. 2, but in neither of these cases did the imitating animal repeat the behavior of No. 4 with sufficient definiteness to succeed. In Chute Experiment B, No. 11’s attention was directed to the chute but not to the end of it. When we take account of the fact that. No. 5, No. 2, and No. 11, in the instances noted, changed their behavior either in form or in strength from what it had previously been, it is fair to speak of their behavior as imitation. This is the simplest form of behavior to which I have applied the term in this paper. In such cases I have spoken of partially suc- cessful imitation. -More clearly entitled to be called imitation is that behavior in which the animal responds to an imitatee, not only by going to a definite locality, but by attacking a particular object. In his imita- tion test in Chute Experiment B, No. 13 went at once to the end of the chute, thrust his hand up the inside, grasped the string, and pulled. The same was true of No. 4, and of No. 6 in the same experiment, of No. 6 in the Rope Experiment and of No. 4 in the Button Experiment. In these cases, attention was centered on a definite object. This investigation presents a number of other cases of similar behavior. It was not always true that when a monkey Haccerty, Imitation in Monkeys. 439 attacked the right object he repeated the movement of the imitatee in detail. The impulse seemed to be to do something to the object, and the imitating animal used his hands and teeth interchangeably. As a result the behavior of the imitator was often ill adapted to secure the profitable result. Repetition of the act usually refined such behavior until it was correct. The most perfect type of imitation is exact repetition in detail of the act of the imitatee. The case of No. 13 in the Chute Experiment already cited is an example. So also is the behavior of No. 3 in the Button Experiment, and of No. 6 in the Rope Experiment. The investigation furnishes a number of other cases which are approximately as good. (c) The Stimulus to Imitative Behavior.—Some of the animals which I have studied learned to manipulate mechanisms unaided. No. 2 did this with the chute, No. 4 did it with the screen, and a number of the monkeys learned to get food by tearing the paper. In the case of the Paper Experiment and in the case of No. 2 in the Chute Experiment, the stimulus was the mechanism itself. That the mechanism was not a sufficient stimulus in many eases is evident from the large number of failures to learn unaided which the investigation furnishes. In the Chute Experiments eight different animals were given the preliminary trials and of these six showed no interest in the’ end of the chute, most of them not even going to it. This, of course, does not prove that they might not have learned how to get food ‘if the trials had been indefinitely prolonged, nor is it necessary to prove this latter thesis in order to interpret the behavior of the monkeys as imitation. What these preliminary tests do establish is the improbability that a sudden change of behavior should occur in the sixth trial with 70% of the animals used. For the stimulus to this sudden change we must look to something other than ithe mechanism itself. It may not be out of place at this point, to say a word in reply to a criticism often made upon the use of animals kept in a zodlogical garden. The criticism is, that such animals have had innumerable pportunities to learn to do acts about which the experimenter can- 440 “‘fournal of Comparative Neurology and Psychology. not know, and hence he cannot tell what causes his animals to act as they do. This criticism does not hold against this investigation for every animal was given abundant opportunity to manifest his random activities and to exhibit his stock of tricks. That the situ- ations were unfamiliar is evidenced by the animals’ repeated failures to learn. That this criticism is less important than it has been made to seem is evidenced by two facts which come out in this study. First, of the two animals which made the best records in the investigation, No. 5 and No. 2, one had been in the garden several years, the other had never been in the garden until June, 1908, when he was shipped there from Cambridge. He had been bought from a dealer and was presumably fresh from the forest. The other fact is that not one of the Park monkeys learned to work the chute unaided, whereas No. 2 did. The additional stimulus in the imitation tests was an animal working at the mechanism and food coming from the mechanism. The relative value of these two elements in the imitation-stimulus, this investigation does not show. That in certain cases the presence of the animal was necessary, there is sufficient evidence. The be- havior of No. 6 in the Screen Experiment is a case in point. No. 6 had seen the screen lifted in the Paper Experiment. Immediately thereafter, he had torn the paper and obtained food. He had done this repeatedly and thus had learned that there was food behind the screen. Yet throughout his entire preliminary trials he failed to lift the sereen. It was only after he had seen No. 4 get food by lifting the sereen that he did the act himself. The case of No. 5 in the Button Experiment illustrates the same thing. She had had a great deal of experience with the slide door. Over and over she had served as the imitatee in the Plug Experi- ment and had eaten more than a dozen bananas which she had gotten after opening the door. Yet she was helpless to get the food when the door was opened by the button. She learned to push the button by watching No. 2 push it. On the other hand, there is evidence to show that in certain cases the behavior of the animal unaccompanied by any profitable result is not sufficient to produce imitation. In general, the monkeys did Haccerty, Imitation in Monkeys. 441 not display much tendency to repeat the mere acts of other monkeys. That they did not imitate in this way may have been due to the conditions of the experiments. Where opportunity was given for imitation, food was given as a reward. It often happened that when the attention of the imitator was only slight it would be greatly accentuated when the imitatee began to get food. No. 10 and No. 11 were kept in the same cage. No. 10 whipped No. 11 and treated him with indifference. Yet when she saw him get food in the Paper Experiment, she at once showed an accentuation of the objec- tive marks of attention. In the Rope Experiment, No. 2 was in- different to the behavier of No. 3 until he saw No. 8 with food and his attention was not drawn to the food door until he saw No. 3 get food there. His interest in No. 8 steadily inereased until he got food for himself. The same comment may be made upon the behavior of No. 8 when watching No. 2 in the Paper Experiment. In general, No. 4 lorded it over No. 6 and No. 5 when in the living cages, but she invariably became attentive to them when she saw them getting food in the experiment cage. . Thus the facts would indicate that not only the act of the animal, but also the profitable result of that act was « necessary factor in producing imitation. By further experimentation I hope to dis- cover the relative importance of these two elements. GERTY, Imitation in Monkeys. 443 ~ J Hac a, trap door; 0, Wick ale Old cage (see text, p. 355), Chute Experiment A. device to hold door shut; c, chute; d, string; e, iron for monkey to grasp. (Drawn by B. Spencer Greenfield.) 444 “fournal of Comparative Neurology and Psychology. Fic. 2. New cage (see text, p. 351). a, 0, c¢, d, front frame; e, f, g, h, back frame; i, 7, k, | and m, n, 0, p, end frames; q, brace across front of cage; «, bolts holding frames together; A, B, C, D, boards covering back of cage; H, F, G, boards covering end of cage; # and G, door; H, I, J, boards covering top of cage; Z, floor; S, slide door in large door; h’, door hinges; ww, wing nuts; X, end of brace where animals frequently perched. (Drawn by B. Spencer Greenfield.) IDE. BY Haccerty, Imitation in Monkeys. 445 ® «S game” " om am New cage showing chute, a, with one side open; 0b. trap door; 6, ‘ing; d, wire spring handle; e, rungs. Page 358. 446 ‘Yournal of Comparative Neurology and Psychology. Vic. 4. No. 2 getting food in Chute Experiment B, characteristic position. Page 358. Haccerty, Imitation in Monkeys. 447 Fic. 5. No. 6 getting food in the Rope Experiment. Page 385. 448 “fournal of Com parative Neurology and Psychology. ee mo et tn | ‘ f a ' Fic. 6. No. 6 tearing the paper in the Paper Experiment. Page 391. Haccerty, Imitation in Monkeys. 449 Fic. 7. No. 4, to the right, pushing up the screen, a, in the Screen Experi- ment. Page 400. Aiea a; + + a ite ears Fic. 8. No. 5 pulling the plug in the Plug Experiment; a, slide door; bo, string; ¢, plug. Page 411. Haccerty, Imitation in Monkeys. 451 Fic. 9. No. 5 getting food after pulling plug (Fic. 8); a, slide door; b string; c, plug. Page 412. > 452 ‘fournal of Comparative Neurology and Psychology. it ar Bit a ee Fig. 10. No. 4 pushing the button, 6, in the Button Experiment; a, slide door; ¢, string. Page 419. Haccerty, Imitation in Monkeys. 453 Fig. 11. New cage adjusted for the string experiment; ZL, opening where food came into the cage; 7, 2, 3, 4. 5. 7, strings; 2t, where string 2 entered the cage; 2k, the knob at the end of string 2. Page 425. 454 ‘fournal of Comparative Neurology and Psychology. | i | | ' Tic. 12. New cage adjusted for the String Experiment; a, chute; JD, bottom of chute on a level with opening into the cage; ¢, trap door; d, lever; 2, string 2; f, feeder. Page 426. +e) Monkeys. 1711 zon Haccerty, /[muitat ASTER t. Page 432. xXperimen No. 13 getting food in the String E EG: a3: The Journal of Comparative Neurology and Psychology VoLumE XIX _ NOVEMBER, 190 NuMBER > THE MORPHOLOGY OF THE FOREBRAIN VESICLE IN VERTEBRATES.? BY J. B. JOHNSTON. University of Minnesota. WITH FORTY-FIVE FIGURES. CONTENTS. PAGE TiO CULE TT OM ey cnsgers ©. ata eile sok Versio overiaasie. st e).c, eile’ ets ot.stiel' c \eveve tat clei velit fojetecslonsueveyeteverere 458 epee) CSCIITH LV [ATi bye cp encrepe ores ore) soared one. avesroreSrriedacn er sre ene teia Suemtohersto nme rerernesere 462 ME) REV CIOSTOMTC Se Ke ete see. Soenle arate sie cs ves dene ei eted olen sseyeue enero, aie toleel umeeten 462 DP AS elachiamsy 027 see ee bids ecahe, Gk oo coe. af aite ca sly laces Brats toh venAcr nel operates 464 INOteSHOnshneademorpnologye eatin etriecieletcncaerersneten tenets 464 Development, Of; the? LOEW RAT iis secsie 12) siete ors clcleiel ote ehauene 479 a. The optic vesicle and the primitive optic groove 479 bs Lhe tlooriok thevdiencephalomy yet ci-s-tcre ete cists oie 482 Ge RO0k OLmneRdienceplialonancs eee ciclo cie crteiotee errs 487 8. Ganoids and teleosts .......... FIRE OOO BOA DD ORR O OIG KO.Cre 488 AL. INO OMBNTS): A ooadogucodaucooodGocdcudUde GoD” a fata ro joie afore nee 489 aS AVCHELLCS PROMO WIL OS in. cic andes e's crakeyomi titeue wie. fie pene seer esaion ore ewes 495 Gem VET TTT A Steet te otecs icvsre toes ois ares oe tele ch oges Vokes ol aro¥e oted ana kerenctoneperer 495 Neurological studies from the Institute of Anatomy, University of Minne- sota, No. 6. THH JOURNAL OF COMPARATIVE NEUROLOGY AND PsSYCHOLOGY.—VOoL. XIX, No. 5. 458 “fournal of Comparative Neurology and Psychology. I]. DiS@ussiOi: oc stere arose cles e eal era oor eee Ito Der eeh iene ee 504 1.) Dhesanterior end of the head vangdsprainet- eee eee eee 504 2. The homology of the saccus) vasculosus 2244. 4-42 sere eee 5OG 3. Segmentation of the neural tube in front of the cerebellum. 507 4. Boundary between diencephalon and telencephalon ........ 508 5. The. ventricles.and)sthestelag. atcmecusiccctearcisye terion iene none 5138 6. Dorsal and ventral zones in the diencephalon and_ telen- ols) 0) 0YFH (0) OR peeneen eee ACOA Oiaewta 9 cio cana oma HOO oo cu pas. 516 (eo sLalliumiot thertelencephulomimc seein rei teceieterrene 518 SP Divisions mand enomenclaburen .- eerie erst teint bien reneiew rene 525 Summary, and CONCIUSIOMS® airec cele ote olor eet tele aro) el ol ol ote eee Eee 582 Proposeds LEVaASIOMMOL UME ySINAG eee vele ce elelele etal ele telnet nents aad 533 In all classes of vertebrates the forebrain vesicle of the early embryo gives rise to two secondary brain segments, the diencephalon and telencephalon. The morphological features of the diencephalon are remarkably constant throughout the vertebrate series, although this segment is the most peculiar and irregular in form. There is always a membraneous roof, tela chorioidea and one or two epiphyses. The dorsal border of the lateral wall presents a paired enlargement known as the ganglion habenule, to which comes from the telen- cephalon a tract of fibers which often appears as a gross feature— the stria medullaris. The massive lateral walls bound a third ventricle which is narrow from side to side. In these lateral walls develop a considerable number of special centers of which the lateral genic- ulate bodies are recognizable microscopically in fishes. The ventral wall of the diencephalon is depressed and variously shaped according to the degree of development of the olfactory and gustatory organs. From in front the optic tracts enter through the base and run up in the lateral walls. Behind the chiasma is found always in the embryo and usually in the adult (fishes, amphibians, reptiles) a transverse groove or depression known as the recessus postopticus. Behind this is a wider depression which in mammals is provided with thick walls, the tuber cinereum, but in fishes and amphibians is a broad bilobed structure with thinner walls, coneave within, known as the lobi inferiores. Between these and continuing backward and downward from them is a thin walled sac, the saccus vasculosus. Jounston, Forebrain Vesicle in Vertebrates. 459 This forms the neural part of the hypophysis im mammals and pro- jects more or less directly downward from the tuber cinereum. The connection of the sac with the tuber cinereum is the infundibulum and its cavity may be called the infundibular cavity or recess. The posterior part of the ventral wall forms a second bilobed structure, the corpora mammillaria. The telencephalon shows greater changes of form and size in different vertebrates than any other segment of the brain. At the same time the essential morphological relations are completely pre- served throughout the series from cyclostomes to man. It is part of the purpose of this paper to make this fact more clear and ex- plicit with regard to certain features of the forebrain, but those features upon which there are no differences of opinion may first be sketched. The telencephalon is very deeply bilobed—bifureated—in front. Each forward prolongation receives an olfactory nerve and is known as the olfactory bulb. This may be closely applied to the body of the telencephalon as in cyclostomes and amphibians, or it may be removed by a longer or shorter distance as in most other classes. In the latter case there is an obvious olfactory tract. The body of the telencephalon is always bilobed, the lateral halves being joined by thinner portions which are usually membraneous or thickened only by nerve fibers. Only in selachians is there a considerable thicken- ing of the median part of the roof by gray matter. The ventricle which continues forward from that of the diencephalon is always bifureated, each lateral division extending into the olfactory bulb (except in some adult mammals where it becomes obliterated during development). A membraneous roof continuous with the tela cho- rioidea of the third ventricle extends over the whole length of the telencephalon except the olfactory tracts and bulbs. In all classes the lateral halves of the telencephalon are connected at the rostral end by the lamina terminalis, a membrane formed by the closing of the anterior neuropore of the embryo. This lamina is always thickened by transverse fibers of two or more systems and sometimes divided into two bundles, the commissura anterior. At the lower border of the lamina terminalis is a prominent depression which lies just in front of the optic chiasma and is therefore called 460 ‘fournal of Comparative Neurology and Psychology. the recessus preopticus. In the walls of the lateral ventricles are found the secondary olfactory centers and the corpora striata. To these in higher forms are added more complex cortical structures which in mammals become the predominant part of the telencephalon. The above mentioned general features of the forebrain are univer- sally recognized and require no further comment. Other features remain less well understood or their interpretation is in dispute so that they present problems for solution. These problems may be stated here, and will be discussed after the description of certain features in the embryonic and adult brain which contribute to their solution. 1. The anterior end of the brain. It is generally recognized that the lamina terminalis represents the seam of closure of the anterior neuropore, but the location in the adult brain of the lower border of the neuropore is placed by different authors at various points between the anterior commissure and the infundibulum. The determination of the anterior end of the brain is of fundamental importance for fixing the segmental order of the parts of the forebrain along the brain axis. Thus, if the anterior end of the brain is at the infundi- bulum, the optic portion of the hypothalamus.together with all the telencephalon would be dorsal structures, while if the anterior end lies at the preoptic recess the optic part of the hypothalamus must be considered as belonging to the ventral portion of the brain. In the latter case the olfactory centers would belong to the first segment, in the former they would fall behind the optic chiasma around the dorsal margin. 2. The boundary between the diencephalon and _. telencephalon. This problem is closely bound up with the first one. Although His placed the anterior end of the base of the adult brain behind the infundibulum and included the pars optica hypothalami in the telen- cephalon, most authors have referred the pars optica to the dien- cephalon and this is the obvious meaning of the tables of the BNA. In the roof of the brain of true fishes and amphibians and in the brains of embryos of most vertebrates occurs a well marked trans- verse fold of the membraneous roof, the velum transversum. This is considered as marking the boundary between di- and telencephalon in Jounston, Forebrain Vesicle in Vertebrates. 461 the roof. In cyclostomes this is only slightly marked and has been clearly described recently by Sterzi. In mammals the velum has not been recognized hitherto and therefore the boundary line in ques- tion is wholly uncertain in mammals. 3. The problem of the membraneous roof and its relations to the massive walls and the nervous pallium. This question naturally presented itself to the earlier neurologists because of the obvious great differences between the mammalian or human brain in which the ventricles were apparently wholly covered by massive nervous structures and that of lower forms, especially teleostean fishes, in which a relatively great expanse of membraneous roof covers a broad ventricle in the telencephalon. A discussion of this problem natur- ally involves consideration of the next. 4. The problem of the median and lateral ventricles in the telen- cephalon and the interventricular foramina. Does the telencephalon possess a median ventricle? One interpretation of the mammalian brain would assign the whole median ventricle in front of the pos- terior commissure to the diencephalon (third ventricle). The walls bounding this ventricle then belong to the diencephalon and the boundary between di- and telencephalon is assumed to be the inter- ventricular foramina. * If, however, the pars optica hypothalami and lamina terminalis belong to the telencephalon, as was held by His, they must carry with them into the telencephalon the adjacent front part of the IIId ventricle. In the brain of a fish or even of an amphi- bian there appears at first sight to be a large median ventricle belong- ing to the telencephalon and it is more or less diffieult to fix the location of the foramina leading into the lateral ventricles. In the teleostean and ganoid brain only the olfactory tracts and bulbs appear to possess distinct cavities which could be called lateral ventricles and the whole of the olfactory centers and corpora striata are found in the lateral walls of the median ventricle. There is, therefore, need to determine whether a part of the median ventricle belongs to the telencephalon and to fix for the different classes of vertebrates the homologue of the interventricular foramina. 5. Subdivision of the di- and telencephalon with reference to the longitudinal zones laid down by His in other segments of the brain. 462 “fournal of Comparative Neurology and Psychology. The determination of the anterior end of the brain will fix the extent of the floor plate and roof plate of His and will show the point at which the prolongation of the suleus limitans must end. In the midbrain, hindbrain and spinal cord the suleus limitans divides the lateral walls into dorsal or alar plate and ventral or basal plate. In the di- and telencephalon this suleus can searcely be distinguished with certainty, but the fact that zones of widely different functions are separated by this suleus when present leads one to seek the equiva- lent of these zones and the position of the sulcus in the di- and telen- cephalon. Other problems concerning the internal morphology and genetic history of the telencephalon—the origin of the pallium in relation to the primary centers, the evolution of functional localization, ete.,— will be discussed in the present paper only so far as necessary in connection with the subject of nomenclature. I. Descriptive Parr. In this part will be brought together under each class of vertebrates the data which bear upon these problems. As will be seen, the new facts to be brought forward are chiefly embryological (selachians, amphibians, mammals). The summaries of facts which have been previously published will be made as brief as possible and the reader is referred to the original papers for more complete accounts. 1. Cyclostomes. In cyclostomes is found a completely bilobed telencephalon, almost as strongly marked as in any class of vertebrates. The reascn for taking especial account of this fact in the telencephalon is that the division into symmetrical lateral halves which is present in all parts of the central nervous system is more pronounced in the telencephalon and in mammals and man becomes one of the most striking features of the whole brain. The cause for the decided division of the telen- cephalon is to be found without doubt in the presence of paired olfac- tory organs and the importance of the sense of smell. In amphioxus the front end of the brain is not bilobed because there is no localized olfactory epithelium. In cyclostomes the olfactory organ is paired Jounston, Forebrain Vesicle in Vertebrates. 463 in the embryo and the mode of development of the mouth, hypophysis and olfactory pit shows that the paired definite olfactory organs are older than the circular mouth. These organs have, therefore, exerted an influence on the brain for a long time previous to the period when eyclostome characters became permanently fixed, and the result is the bilobed telencephalon. This is seen in the presence of massive lateral walls which are connected above, below and in front by rela- tively thin membranes which are thickened only by transverse fibers and nowhere contain gray matter. It is seen in the forward projec- tion of the olfactory bulbs into which the paired olfactory nerves enter. The olfactory bulbs are sessile ; there is no extended olfactory tract or peduncle. It is seen further in the lateral extensions of the ventricle to which attention has been called by Studnicka and the writer (1906). These are rather wide cavities extending into the olfactory bulbs and connected with the median ventricle by narrower openings, the interventricular foramina. Owing to the pushing back- ward of the telencephalon in cyclostomes the lateral ventricle presents a posterior prolongation similar to but not homologous with the posterior horn of the ventricle in mammals. The tela chorioidea of the IIId ventricle forms in front of the nuclei habenule a broad low dorsal sac whose roof in some species is depressed by the epiphysial bodies lying over it. In front of this is a shght transverse fold (better seen in embryos than in the adult) first described by Sterzi (1907) which this author homologizes with the velum transversum of other vertebrates. This velum is continued laterally around the brain by a constriction (groove) without and a fold (ridge) within which marks the boundary between the diencephalon and telenceph- alon. In front of the velum transversum the roof continues for some distance as a thin membrane, the Jamina supraneuroporica of Sterzi, and then is thickened by the fibers of the superior olfactory decus- sation. According to Sterzi, the point corresponding to the neuro- poric recess of other vertebrates is at the front or lower border of this decussation. The conditions in cyclostome embryos (solid nerve cord, solid con- nection of brain with ectoderm instead of open neuropore, compact- ness of structure) are not favorable for the study of the anterior 464 ‘fournal of Comparative Neurology and Psychology. end of the brain, and the discussions of v. Kupffer, Koltzoff and others do not seem to the writer to offer anything of value in the present state of our knowledge. (See note, p. 535.) It should be noted that the ridge in the brain floor which contains the optic chiasma and other decussating tracts is especially high and prominent in cyclostomes and that the pit behind the ridge, recessus postopticus, is quite as deep and sharply marked as that in front of the ridge, recessus preopticus (commonly called recessus opticus). 2. Selachians. Notes on Head Morphology. In the adult selachian brain the diencephalon presents no peculiar- ities of especial importance to our problems. The limit between the Tectum mesencephali | p De \, __— Cerebellum Nee s x > Tabane: mre _— Tub. acust. Nucleus habenule, Velum transversum | fe ~ ee Epiphysis f c OI Paraphysis Telencephalon i H ( \ ‘A 4 Com. infima >——-Sac. vase. Lobus visceralis Rec. neurop. ve Ee Rec. prop. Lobus inferior i Optie chiasma Fie. 1. The mesial surface of the right half of the brain of Mustelus. From Johnston, 1906. For the significance of the abbreviations used on all of the figures, see the list at the end of this article. diencephalon and telencephalon is clearly marked dorsally by a deep velum transversum which forms the anterior wall of the dorsal sac belonging to the diencephalon, and the posterior wall of the para- physis belonging to the telencephalon (Fig. 1). In front of the para- physis the membraneous roof is complexly infolded to form the plexus Jounston, Forebrain Vesicle in Vertebrates. 465 chorwoideus of the telencephalon. The anterior limb of the plexus is attached to the massive nervous wall which overarches the front part of the median ventricle. In the more primitive selachians such as Heptanchus and in Chimera the massive roof is smaller and the membraneous roof extends farther forward. As shown in Fig. 1, the massive roof is pierced from the dorsocephalic surface by a vas- eular canal which reaches nearly to the ventricle. This has been interpreted (Johnston, 1906) as the remnant of a deep groove which separated the lateral lobes of the telencephalon earlier in the phylo- Fic. 2. A, an outline of the brain and brain ventricles of Mustelus as “_- seen from above. B, a diagram of one side of the fore-brain to show the primitive relations of the wall and ventricle. From Johnston, 1906. geny of selachians, and the process by which the present form was reached is indicated by the accompanying diagram (Fig. 2, B). That portion of the massive roof which lies behind the vascular canal is formed by the great growth of the lateral walls and it has for its basis a fiber-decussation which is comparable in position and in a part of its fiber components to the superior olfactory decussation of Petromyzon. In selachians, then, owing to the enormous devel- opment of the olfactory centers, that part of the membraneous roof of the telencephalon which contains fibers in cyclostomes is invaded by gray matter as well. 466 “fournal of Comparative Neurology and Psychology. The form of the ventricle is shown in Fig. 2, and it is necessary only to point out that quite distinct lateral ventricles are present. that they extend into the olfactory bulbs and that they are connected with the median ventricle by the interventricular foramina. A tri- angular prolongation of the median ventricle projects a short distance in front of the interventricular foramina to meet the vascular canal mentioned above. This is the recessus neuroporicus. The solution of the problems of the anterior end of the brain and of the boundary between diencephalon and telencephalon requires the study of the development. Although all the essential facts were first acquired from the study of amphibian embryos (see below), it was thought best to extend the study to other vertebrates, and as the sela- chians present the most primitive conditions the description of certain important processes hitherto overlooked will be given first as they are seen in embryos of Squalus acanthias. For the material for this study IT am deeply indebted to Dr. H. V. Neal. When I wrote to Dr. Neal of my findings in Amblystoma and of the desirability of a control study upon selachian embryos he generously sent me from his magnificent collection of mounted preparations of Squalus embryos, the specimens representing the stages from B to M of Balfour’s notation. These preparations in- clude whole mounts and sections in all three planes prepared by vari- ous stains. These are the preparations upon which was based Neal’s very valuable paper of 1898 on the segmentation of the nervous system. My attention has been directed chiefly to stages from F (15 somites) to K and M. It has been a great pleasure to verify the clear and accurate descriptions given by Neal of these embryos. I have studied the brain, cranial ganglia, the mesectoderm derived from the neural crest, the mesodermic somites, the branchial clefts and arches, and have given especial care to the region about the forebrain and the premandibular and anterior head cavities. I have made plate recon- structions from frontal and sagittal sections of several stages and have had numerous camera drawings made to illustrate points not shown in the models. I wish to comment briefly, on the basis of these preparations, upon certain problems that have been under discussion in past years which have some bearing on the special problems being considered here. Jounston, Forebrain Vesicle in Vertebrates. 467 Upon the question of the value of the pre-otic “head cavities” as dorsal somites, Neal’s statements seem to me very careful and con- servative. With regard to the differentiation of sclerotome and myotome, the preparations seem to me to warrant a positive state- ment that a sclerotome is clearly differentiated in the second or mandibular somite as well as in the third. I am in perfect agree- ment with Neal’s interpretation of the preotic head cavities, including the anterior cavity of Platt as dorsal somites. Upon the question of a lost branchiomere between the mandibular and hyoid arches, which I have discussed elsewhere (1905) I have been surprised to find so little evidence in the preparations of Squalus. While the third somite lies over the hyomandibular cleft and is somewhat constricted or dumb-bell shaped, its two parts enclose a single cavity and the somite extends over the hyoid arch with the mesodermie core of which it is connected. The fourth somite is smaller than the third and shows that it is rudimentary. While it reaches forward to the hyoid arch, it is connected with the mesoderm of the first branchial arch. The fifth somite has practically corre- sponding relations to the second branchial cleft and the following (second) branchial arch. The sixth somite is clearly connected with the third branchial arch and the seventh somite with the fourth branchial arch. These relations, indeed, may be clearly inferred from Neal’s admirable drawings from cleared specimens, Figs. 15. 16,17. These correspond to the stages from which I have drawn the conclusions stated above. The posterior ends of somites (2), 3, 4. and 5 are connected with their respective branchial arches. I must here make reference to a series of papers on the head prob- lem coming from Prof. H. E. Ziegler’s laboratory in Jena (Klink- hardt 1905, Guthke 1906, Ziegler 1908, Brohmer 1909). These authors discuss the fundamental questions of segmentation and the relationship of the cranial nerves. The total material on which the deseriptions and conclusions are based consists of ten selachian em- bryos including four of Torpedo between the stage I-K and 20 mm. ; four of Spinax in the stages K, L, M-N, and 7.78 mm.; one of Chlamy- doselachus in the stage L-M; and one of Acanthias 22 mm. in length. 468 “fournal of Comparative Neurology and Psychology. One new idea is introduced into the subject of head morphology, namely that in the mesoderm the two structures in each segment heretofore known as the somite and the branchial arch together con- stitute the somite, and the junction of the branchial arch with the pericardial cavity is to be compared with the junction of the trunk somite with the peritoneal cavity. This conception is used in support of the view that mesomerism and branchiomerism coincide. It may be questioned whether the new conception is not more in need of sup- port than that for whose support it was called in. The material which these authors have studied is far from adequate for the study of the primitive segmentation of the head or the morphology of the cranial nerves. This is especially noticeable in their statements re- garding the anterior head cavity, which could not be studied in the material on which they worked, and the other head somites, whose development is well advanced in the earliest stage represented in their material. In determining the first segment of the head no value is attached to the first two brain vesicles, the eye, the olfactory nerve. the nervus terminalis, the rudimentary nervus thalamicus, or the epiphyses. The first segment is that of the premandibular somite and the ophthalmicus profundus (to which the name Ciliarganglion is given). This nerve is related to the brain behind the cerebellum in the embryos studied by Ziegler and his colleagues, but whether they would include all the brain in front of the cerebellum in the first or premandibular segment is not stated. Most of the processes upon which an intelligent judgment regarding the primitive segmenta- tion of the head can be based have been completed in selachian em- bryos prior to the earliest stage studied by these authors, With regard to the mesectoderm derived from the neural crest which extends down into the mandibular and other visceral arches from the cranial ganglia, I can fully confirm the statements made by Neal and illustrated in his Plate 3 (1898). To these statements one addition requires to be made which is at the same time an addi- tion to the early history of the anterior head cavity and preoral ento- derm. The details of the earliest appearance of the anterior head cavity have not been given by Miss Platt or by Neal, who was con- cerned chiefly with neural segmentation, and Dohrn’s (1904) treat- ment of the anterior head cavity is unsatisfactory. Jounston, Forebrain Vesicle in Vertebrates. 469 Hoffmann’s description (1896) is more complete, but he failed to recognize the source of the mesectoderm beneath the terminal ridge. What he describes as a median mass between the lower border of the neuropore and the “infundibulum” connecting the anterior head Fie. 38. Squalus acanthias, 17 somites, median sagittal section. The noto- chord is marked by cross lines and the undifferentiated median mass and pre- oral entoderm by dots. te ia a * EOI SOY x Re i Z: ora : Ver eperee 8 eee Beescniiaee wae Fic. 4. Detail of a part of the section drawn in Fig 3. > 100. cavities with one another is the mesectoderm derived from the termi- nal part of the neural crest to be described below. In embryos of 15 somites the notochord is continuous anteriorly with a thickened mass lying over the anterior end of the archen- teron and separated from the entoderm, Laterally the mandibular 470 “fournal of Comparative Neurology and Psychology. somite is connected with both the notochord and the median undiffer- entiated mass. The archenteron is drawn out anteriorly to a very slender pointed cavity (Figs. 3, 4, 5) which ends beneath a depres- sion in the floor of the brain usually called by authors the infundib- ulum. As this is not the infundibulum I shall for the present refer to this part of the brain as the “infundibulum,” using quotation marks. Following a series of transverse sections forward one sees Fic. 5. Squalus ac., 15 somites, medial view of a model of the right half of the head. ** part of neural tube which is still open. 25.. RN ee Wie. 6. Squalus ac., Balfour’s stage D, sagittal section of anterior end. . eine / \ \ f | ra a | ie Op.V. | | | \ & : r po ; bx ane x isi’ aa OY \ age ose a & © en ae BB Fic. 7. Squalus ac., 18 somites, frontal section through the premandibular somite and median mass. ‘The right side of the section is a little further dorsal than the left and the cavity of the somite shows only on the right. < 150. In embryos of 18 somites the cavity of the premandibular somite (som 1, v. Wijhe) is first seen in the lateral part of the undifferen- tiated mass above described (Fig. 7). The median part of this mass maintains the relations just described. In embryo of 19 somites the mesoderm and preoral entoderm show no change but a very noteworthy structure has appeared between 472 ‘fournal of Comparative Neurology and Psychology. ectoderm and brain, surrounding the anterior end of the preoral entoderm. Paired sheets or flaps of cells migrating from the ecto- derm at the lips of the neuropore extend in between the ectoderm and brain and surround the anterior part of the preoral entoderm on its under or ectal surface. These sheets of cells resemble in every way the neural crest in other parts of the embryo and should be con- sidered as the terminal part of the neural crests. In embryos of 24 somites the premandibular cavities are some what larger and especially longer. The bodies of these somites are ne t Le bel i fee Fiag. 8. Squalus ac., 22 somites, frontal section through the terminal neural crest. > 150. in contact with the mandibular somite, are still continuous with the median mass and converge forward and downward like the arms of a letter V to fuse completely with the median mass. In the lateral part of the median mass directly forward from the premandibular somites and behind the “infundibulum” are found the first indica- tions of the anterior head cavities. Many mitoses appear here, especially in the front wall of this cavity. The preoral entoderm continues forward beneath the “infundibulum.” As compared with earlier stages, before the formation of the anterior cavities the pre- oral entoderm is very slender and at one point beneath the “infun- Jounston, Forebrain Vesicle in Vertebrates. 473 dibulum” it is completely obliterated. In some embryos it is impos- sible to distinguish any preoral entoderm beneath the brain in front of the “infundibulum,” but in others there is no doubt whatever that a slender rod of cells lying in this position is the remnant of the anterior part of the preoral entoderm. This rod of cells is sur- rounded by cells (mesectoderm) derived from the neural crest above deseribed, which constitute by far the largest part of the cells lying in this position. This mesectoderm forms a considerable mass of cells filling a lens-shaped space between brain and ectoderm and between “infundibulum” and neuropore and also extends as a sheet Fic. 9. Neal’s Fig. 11 modified by addition of the terminal part of the neural crest. The figure represents the lateral yiew of a cleared embryo of Squalus of 24-25 somites and shows the extent of the mesectoderm derived from the neural crest. along the sides of the “infundibulum” beneath the optic vesicle. Fig. 9 is a copy of Neal’s Fig. 11, Pl. 3, of this stage, with the addi- tion of this mesectoderm. In one embryo of 26 somites (Fig. 10) I have found the slender rod of preoral entoderm persisting beneath the “infundibulum,” but from this time on it is impossible to recognize entoderm in this position. The “infundibulum” has become depressed until it is in contact with the ectoderm and at the same time, whether because of pressure from the “infundibulum” or not, the median part of the preoral entoderm becomes obliterated while its lateral portions form 474 “fournal of Comparative Neurology and Psychology. themselves into the walls of the anterior head cavities. The mesen- chymatous cells beneath the front end of the brain are in over- whelming proportion of ectodermal (neural crest) origin. In Dohrn’s beautiful plates illustrating his articles on the man- dibular and premandibular cavities, this terminal neural crest is clearly shown. See especially Pl. 9, Figs. 7-12. Dohrn does not distinguish between premandibular mesoderm and the mesectoderm of neural crest origin. The latter he calls the anterior part of the premandibular mesoderm (Prem. 1) after the “infundibulum” has Fic. 10. Squalus ac., 26 somites, frontal section. inf., the so-called infun- dibulum. pressed down to meet the ectoderm. In Fig. 8 the origin of this from the neuropore is strongly suggested, especially as the spot marked neurop. is at the wpper border of the lamina terminalis, some distance dorsal to the extreme point to which the preoral ento- derm or premandibular mesoderm ever reaches. In Dr, Neal’s preparations which I have studied the entoderm has a different tone from the other tissues and there is a decided difference in the form of the cells and in the size of the nuclei between the preman- dibular mesoderm and the terminal neural crest mesectoderm. In embryos of 29 and 30 somites (Figs. 11 and 12) premandibular Jounston, Forebrain Vesicle in Vertebrates. 475 and anterior head cavities are small, separate from each other, but both connected with the median mass which is crowded behind the “infundibulum.” The median band connecting the anterior cavities Fie. 11. Squalus ac., 30 somites, frontal section through premandibular and anterior head cavities. >< 150. is small and disappears at this stage. The premandibular somites retain the connection with the median mass in which cavities appear in later stages and finally fuse with the cavities of the premandib- ular somites as fully described by Neal and others. The anterior 476 “fournal of Comparative Neurology and Psychology. cavities extend forward at the sides of the “infundibular” region in the position in which they were first described by Miss Platt. Ex- cept where they are in contact with the premandibular somite the Fic. 12. Squalus ac., 29 somites, frontal section through the premandibular and anterior head cavities. These cavities are slightly more advanced in development than in the 30 somite embryo shown in Fig. 11. 150. anterior somites are almost completely ensheathed by the mesec- toderm derived from the region of the neuropore. The mesectoderm sheet derived from the thalamic nerve rudiment has now come down behind the eye and met with this which lies beneath the eye. From Jounston, Forebrain Vesicle in Vertebrates. 477 this stage onward Neal’s figures 12, 13, ete., show the distribution of neural erest mesectoderm almost correctly. The mesectoderm lying in the region between the optic vesicle and somite 1 should be continued as a thin sheet between the brain and ectoderm beneath the neuropore and the origin of all this part from the terminal part of the neural crest should be taken into account. Fie. 138. Squalus ac., 39 somites, frontal section through the anterior head cavities, showing their relation to the premandibular somite (left side) and the mesectoderm derived from the terminal neural crest. The connection between n. thalamicus and terminal neural crest shows on the right side. < 150. The relations of the anterior head cavities to somite 1 and to the mesectoderm remain essentially unchanged up to embryos of 50 somites. Figs. 12, 11, 13, 14 show these relations in frontal sec- tions of embryos with 29, 30 and 39 somites. The brain, mesoderm and mesectoderm with the cranial ganglia have been modelled in an embryo with 42 somites and Fig. 15 shows these relations as seen in the model. 478 “fournal of Comparative Neurology and Psychology. Fic. 14. Section from same series as Fig. 13, passing through the tip of the anterior head cavity on the left side and in front of this cavity on the right side. >< 150. At.m-eCc. Fic. 15. Squalus ac., 42 somites. Left view of a model of the brain, meso- derm, ganglia and mesectoderm derived from the neural crest. The auditory pit with the ectoderm bordering it are also shown. » 50. The boundaries of the mesoderm are marked in black lines. 3, 4, 5, somites of Van Wijhe’s series. Jounston, Forebrain Vesicle in Vertebrates. 479 Development of the Forebrain. a. The Optic Vesicle and the Primitive Optic Groove.—The early appearance of the optic vesicles on the open neural plate has been deseribed by Locy (1895), and the writer has elsewhere (1905, 1906, 1909) brought together the evidence that the optic vesicle is derived from the alar plate of the embryonic neural tube. The study of its relations for the present purpose begins with embryos of about 15 somites. The medial surface of the right half of a model of the head of such an embryo is shown in Fig. 5. The neural tube has a cephalic flexure and is open at the neuropore and for a short distance along the dorsal surface. The lateral wall presents two coneavities, a broad shallow one for the mid-brain and a deeper one for the forebrain. The depth of this concavity is due chiefly to the formation of the optic vesicle, which seems to involve the greater part of the lateral wall of the forebrain vesicle. In the floor of the brain the optic vesicles of the two sides are connected with one another by a transverse groove or depression which has heretofore been called the “infundibulum.” That it does not be- come the infundibulum will be shown farther on. Its relation to the optic vesicles and its later history suggest that it be called the primitive optic groove. Its relations to the preoral entoderm as above described and to the neuropore as seen in Figs. 3, 4 and 5 show that this groove is in the floor-plate of the brain some distance behind the anterior end of the entoderm. J In the embryos with 17 somites, Figs. 3 and 4, the neuropore is closed except at a single point which shows as a thin place in one section of 62/, microns thickness. The distance from the primitive optic groove to this point is greater than the distance to the lower border of the neuropore in the embryo of 15 somites. The arrange- ment of the cells in the lower lip of the closing neuropore as seen in transverse and frontal sections shows that there is a process of fusion of the lips of the neuropore from below upward and therefore the last point of the neuropore to remain open is a point in the dorsal seam of the neural tube some distance removed from the anterior border of the neural plate. This part of the seam of closure which represents the length of the neuropore is what is called the 480 “fournal of Comparative Neurology and Psychology. lamina terminalis. This is not very long in the embryo of 17 somites, but grows distinctly in length as the forebrain expands. The apparent great thickness of the lamina terminalis in Fig. 3 is of course due to the persistence of the fusion of ectoderm and brain wall at this time. The primitive optic groove is deeper and:sharper in outline than in the last stage. The relations in an embryo of 24 somites are shown in Fig. 16 representing the medial view of a model of the right half of the head. In this is seen the depression of the primitive optic groove until its wall comes into contact with the ectoderm and cuts off the Fic. 16. Squalus ac., 24 somites, medial aspect of a model of the right half of the head. > 75. front part of the preoral entoderm as before described. ‘The lens- shaped space in front of the primitive optic groove is filled chiefly by mesectoderm from the terminal part of the neural crest. The lamina terminalis is marked by a small triangular pit above, the neuroporic recess, and by a rounded shallow pit below. This pit occupies the lower portion of the neuropore space and may therefore be called the terminal pit. The optic vesicle is now stalked and the cavity of the stalk is seen near the base of the brain. It is notice- able that this cavity no longer communicates directly with the primi- tive optic groove. It seems equally closely related to the terminal pit. Between the two pits the brain wall is somewhat thickened. Jounston, Forebrain Vesicle in Vertebrates. 481 This thickening corresponds to the anterior border of the neural plate and the lower border of the neuropore. It is the terminal ridge of the early embryo with open neural plate, Figs. 3, 4, 5, 6. From this terminal thickening a ridge is seen in Fig. 16 running obliquely eaudo-laterad behind the optic vesicle and cutting it off from commu- nication with the primitive optic groove. This is the beginning of a change in the relations of the optic vesicle which the following stages show completed. An embryo of 42 somites is shown in Fig. 17 drawn from a model of the right half of the head and a corresponding model of an embryo of Balfour’s stage K is shown in Fig. 18. In these it is Fic. 17. Squalus ac., 42 somites, medial aspect of a model of the right half of the head. This is the same embryo as the one shown in Fig. 15. clear that the optic vesicle is no longer connected with the primitive optic groove, but the two vesicles are now connected with one another by the terminal pit at the lower border of the lamina terminalis. This is the pit which remains connected with the hollow optic stalk as long as that persists and is known in the adult as the recessus opticus (His) ; better called recessus prwopticus. In the latter part of Balfour’s stage K the optic tract fibers begin to appear in the chiasma. ‘This lies immediately behind the terminal pit and in front of the primitive optic groove (Figs. 22, A and B). and therefore lies in the terminal ridge. The lateral prolongation of this ridge, which has been described as running obliquely across 482 “fournal of Comparative Neurology and Psychology. the primitive optic groove (Fig. 16), furnishes a pathway for the optic tract fibers as they grow in from the retina to the optic centers in the thalamus and tectum opticum. The ridge may therefore be called the optic ridge. The primitive optic groove from this stage on is to be seen just behind the transverse ridge occupied by the optic chiasma and other decussations. Referring to the general descrip- tion at the beginning of this article it will be seen that what I have called the terminal pit in the early embryo is the same as that called by His the optic recess, and that to which I have in all my previous Fig. 18. Squalus ac., late stage K of Balfour, medial aspect of a model of the right half of the head. x 25. papers given the name of preoptic recess. The primitive optic groove I have heretofore called the postoptic recess. These terms have been used by some other authors (Mrs. Gage, Sterzi, and others), but not by all. I wish to emphasize the necessity of recognizing the two pits and applying to them clearly distinctive names, because they are both related to the optie vesicle and because the postoptic recess has heretofore been confused with the infundibulum. b. Remainder of the Floor of the Diencephalon.—As soon as the head-bend of the brain tube appears, a broad depression of the floor of the forebrain vesicle can be seen which corresponds to the future Jounston, Forebrain Vesicle in Vertebrates. 483 inferior lobes (Fig. 16). The anterior part of this is the relatively deep and sharply marked primitive optic groove. The posterior boundary is less definitely marked by a slight projection of the brain floor into the ventricle. This is the tuberculum posterius. As the broad inferior lobes involve nearly the whole floor of the diencephalon, neither the infundibulum nor the mammillary bodies are to be recognized at this time. Both are developed later within the general O) Fic. 19. Squalus ac., 60 somites, parasagittal section near the middle line. The mouth is open. Primitive inferior lobes, epiphysis and velum transversum. < 33. area of the primitive inferior lobes, as specialized portions of their walls, The mammillary bodies are indicated by a rounded eaudal pro- jection of the depressed floor of the diencephalon in embryos with about 80 somites (Figs. 20, 21, 22). The infundibular recess is not found until after the completion of the changes described in connection with the optic recesses. As in man the infundibulum is the funnel-shaped depression leading from the floor of the tuber cinereum into the neural part of the 484 “fournal of Comparative Neurology and Psychology. pituitary body, so in fishes the infundibular recess must be that somewhat funnel-shaped or trough-shaped depression in the floor of the inferior lobes which leads into the saceus vasculosus, this being the neural portion of the pituitary body in fishes. The angle of ectoderm from which the hypophysis will be devel- oped can be readily recognized by the stage when the depression of the primitive optic groove has pushed the preoral entoderm out of the way and come into contact with the ectoderm. The hypophysial Fic. 20. Squalus ac., 80 somites, median sagittal section. \ 33. Pre- and postoptic recesses, mamillary recess, epiphysis and velum transversum. ectoderm is in contact with the posterior surface of this depressed part of the brain floor. As the hypophysis pushes in (Figs. 18, 19, and following) its anterior limb remains in contact with the posterior wall of the primitive optic groove. The hypophysis grows back in contact with the rounded surface of the inferior lobes, insinuating itself between the brain floor and the median mass connecting the premandibular somites. As this mass in early stages connects the anterior head cavities also with one another, it may be said that the Jounston, Forebrain Vesicle in Vertebrates. 485 in-pushing hypophysis remains always in front of the entoderm and mesoderm except so much of the preoral entoderm as is cut off by the primitive optic groove and aborted beneath the terminal ridge. By stage M (Fig. 22) the hypophysis begins to be constricted off from the ectoderm. It presents a shorter anterior lobe which is Fie. 21. Squalus ac. stage M. sagittal section. > 33. Pre- and post- optic, infundibular and mammillary recesses, epiphysis and velum transversum. The fibers of the optic chiasma appear in the terminal ridge. directed toward the optic chiasma and a longer posterior lobe which is directed toward the mammillary recess. Opposite the tip of the posterior lobe a special outgrowth of the brain wall represents the beginning of the saccus vasculosus. Although the term infundibulum scarcely applies to anything in the fish brain, yet the depression from which the saceus grows out is the region which corresponds 486 “fournal of Comparative Neurology and Psychology. to the infundibulum of man. ‘The relative position of all these structures is shown in Fig. 22, A and B, from a sagittal series of an embryo in stage M. The study of sections can be controlled in these young stages by the study of cleared whole embryos. Neal has given a most instruc- tive series of figures from such cleared embryos and I can attest the accuracy and faithfulness of these figures. If Figs. 7 to 11 of Neal’s Plate 3 be examined, it will be seen that the optic vesicle shifts from the “infundibulum” to a point in front of the anterior head cavity. This agrees with what I have described above. I have carefully studied these whole embryos with the Braus-Driner binocular and find that all the facts with regard to the form and position of parts in the optic region of the brain derived ‘from the study of sections and models can be seen with perfect clearness in the whole embryos. In Dohrn’s papers on the mandibular and premandibular somites (1904) the relations of neural plate and preoral entoderm discussed in this section and the last are beautifully illustrated. Plates 1. 4, 6, 7, 9 and 11 show the early stages in the formation of the primi- tive optic groove and terminal ridge and the relations of the preoral entoderm and premandibular mesoderm derived from it. As is well known, the anterior head cavity of Miss Platt is found only in the Squalidz and Dohrn does not regard it as a somite. I am forced to believe, however, that he has not analyzed the conditions in Squalus acanthias with sufficient care, and that to this is to be attrib- uted his attitude toward the anterior head cavity as well as his failure to recognize the terminal neural crest and the mesectoderm derived from it. In his figures the primitive optic groove is labelled “infun- dibulum,” but it is perfectly clear to me that it is the groove related to the optic vesicles. See Pl. 9, Fig. 9, where the groove marked Inf. is the base of the optic stalk. In Pl. 1, Fig. 15, the reference line Hnt. Zw. passes across the primitive optic groove at the front and the true infundibular recess near the deep end of the hypophy- sis. Compare Figs. 23, 24, and 25 of this paper. The terminal ridge is especially clear in Dohrn’s Pl. 11, where early stages show its form as well as the early stages of Amblystoma (see below). Jounston, Forebrain Vestcle in Vertebrates. 487 c. Roof of the Diencephalon.—In the roof of the interbrain the development of the velum transversum, dorsal sac, epiphysis and paraphysis has been well described (Minot, Sterzi) so that I have = a sims em TTI Fic. 22. Squalus ac., stage M, two sagittal sections near the median plane. 25. In addition to the features shown in earlier figures, the optic chiasma, anterior, posterior and habenular Commissures are seen. nothing new to add. I wish only to note that a careful considera- tion of Dr. Neal’s sections and of the models made from them leads me to believe that the segmental position of the optic vesicle is 488 ‘fournal of Comparative Neurology and Psychology. practically the same as that of the transverse velum. The velum does not become prominent until after the optic vesicle is well formed (stage IX), but in some specimens a slight fold representing it can be recognized in models of embryos as early as 24 somites or earlier. J am inclined to think that the velum represents an infold- ing of the brain wall which is begun early on account of the with- drawal of material from the alar plate to form the optic vesicle. It is the second neuromere whose dorsal half thus gives rise to the retina, while its ventral half becomes depressed and bulged ventro- caudally to form the primitive inferior lobes above referred to. 3. Ganoids and Teleosts. The diencephalon presents no features of especial importance. There is a greater development of the inferior lobes than in sela- chians, although the olfactory apparatus is of less importance. This is doubtless to be attributed to the much greater importance of the gustatory apparatus in ganoids and teleosts. The saccus vasculosus reaches a very great development in some of these forms and it has been shown that there is an intermingling of the epithelial sacs of the saccus with those of the hypophysis. The optic tracts form a chiasma in the floor of the brain in some forms and in others cross at some distance from the brain. The telencephalon presents certain great peculiarities. It is usu- ally somewhat more elongated than that of most selachians and re- sembles that of Chimera or Heptanchus. Also the telencephalon has no massive roof, but only a broad membraneous tela continuous with that of the diencephalon. The boundary between the di- and telencephalon in the roof is marked by a velum transversum which forms the front wall of the dorsal sac of the diencephalon. The membraneous roof of the telencephalon is much more exten- sive than in selachians or other vertebrates. In many cases the lateral walls of the telencephalon are rolled outward (laterad) so that the morphological dorsal border is directed laterad or latero-ventrad. This makes the membraneous roof in these forms exceedingly broad. The ventricle is correspondingly extensive and toward its anterior end divides into lateral ventricles which extend into the olfactory Jounston, Forebrain Vesicle in Vertebrates. 489 bulbs. These relations have been described and figured by several authors. (See Kappers 1906, 1907, Johnston, 1906, Fig. 151.) 4. Amphibians. In the adult amphibian brain the large size of the telencephalon and the form and relations of its nervous and membraneous portions are of interest for our problems. The large telencephalon has thinner walls and larger lateral ventricles than are found in sela- chians. The lateral ventricles are connected with the median ven- iG 23% Fic. 24. Fic. 28. Amblystoma punctatum, neural plate stage. Sagittal section of head end. Ectoderm dark, neural plate medium, entoderm light. 25. Fic. 24. -Amblystoma p., neuropore stage. Sagittal section. The section falls to one side of the median plane in the dorsal region and shows the mesoderm lateral to the notochord. Its cephalic limit is the same as that of the notochord. >< 25. tricles by wide but well defined interventricular foramina (Johns- ton 1906, Fig. 150, 151). The lateral ventricles extend forward into the olfactory bulbs and also have a caudal prolongation into the so-called occipital pole of the hemisphere. The tela of the diencephalon is separated from the membraneous roof of the telencephalon by a prominent velum transversum which in the adult becomes complexly folded in connection with the chorioid plexus. In front of the velum is a highly developed paraphysis. 490 ‘fournal of Comparative Neurology and Psychology. Practically the whole of the membraneous roof of the telenceph- alon is involved in the two complex structures, the chorioid plexuses which extend into the lateral ventricles and the paraphysis which projects upward between the lateral lobes. The essential fact is that the membraneous roof extends forward over the interventricular foramina to meet the lamina terminalis. The roof of the telen- cephalon near the middle line is membraneous for its whole length. The questions regarding the anterior end of the brain and the boundary between diencephalon and telencephalon have been most Fic. 25. Amblystoma p., after closure of neuropore, model of the right half of the head, viewed from the medial surface. < 28. carefully studied in embryos of Amblystoma punctatum. In these embryos the entoderm, mesoderm and notochord present essentially _ the same features and the same relations to the brain as in selachians, In particular, the premandibular somites, the median undifferentiated mass in which the notochord ends anteriorly, and the preoral ento- derm have the same disposition as in selachians. The only difference is that all the structures are more compact in amphibians and the preoral entoderm is shorter. As Figs. 23, 24, 25 show, the preoral entoderm fills the angle between the floor of the neural tube (neural plate) and the ectoderm and there is a short prolongation of the archenteric cavity into it in front of the site of the future mouth. The neural plate is bounded by neural folds which meet in front in the transverse terminal ridge. This terminal ridge marks the line Jounston, Forebrain Vesicle in Vertebrates. 491 along which ectoderm and neural plate meet and, when the neural plate rolls up into a tube, the ridge forms the lower border of the neuropore. These relations are as simple and clear in Amblystoma as in Squalus (Figs. 23, 24). Even after the neuropore has closed the arrangement of cells and nuclei in this region shows the outline of the terminal ridge. After the brain is separated from the ecto- derm the terminal ridge forms a distinct fold, convex toward the ventricle (Fig. 26), which in later stages is occupied by the fibers of the optic tracts in the chiasma (Fig. 33). No neuroporic recess is to be seen in A. punctatum in early stages following the closure ie, PAB. Fic. 26. Amblystoma p., invagination of hypophysis beginning; primitive inferior lobes. Sagittal section of head. > 25. Fie. 27. Amblystoma p., hypophysial invagination at its height ; velum transversum and epiphysis; median sagittal section reconstructed from several sections. > 25. ‘of the neuropore, but in later stages a slight pit is found which may correspond to the neuroporic recess described in other forms (Fig. 33). The early appearance of the retinal areas on the neural plate was first described by Eycleshymer (1890) and the fact that the optic vesicles are formed from the lateral parts of the neural plate has been pointed out by the writer (1905, 1906). While the neural plate is still open the retinal pits are connected with one another by a shallow groove running just behind the terminal ridge (Fig. 23, r. po). As the neural plate rolls up and the optic vesicles are evaginated this groove grows deeper (Figs. 24, 25) and by the time 492 “fournal of Comparative Neurology and Psychology. the neuropore is closed the front end of the neural tube presents a prominent depression resting against the preoral entoderm and con- ic. 28. Amblystoma p., a little more advanced than the one shown in Fig. 27. Parasagittal section through the optic ridge. < 40. FAAS ee emer no) ele Fic. 29. Amblystoma p., of about the same stage as that in Fig. 28. Model of the right half of the brain seen from the medial surface. > 40. C.S: de necting the optic vesicles (Fig. 25). This is the primitive optic groove as described in Squalus. The model of this stage shows a second angle at the front of the neural tube, separated from the Jounston, Forebrain Vesicle in Vertebrates. 493 primitive optic groove by the terminal ridge. This is a pit formed in the lower part of the neuropore and is the terminal pit (Fig. 25). From the earliest stages after the formation of the neural plate and folds, the region from which the hypophysis will be formed can be accurately located. In a median sagittal section of any stage up to the time when the hypophysis is invaginated, a slight réen- trant angle is seen between the terminal ridge and the preoral ento- derm. The ectoderm of this angle will form the hypophysis. When the neuropore closes this hypophysial ectoderm is slightly thickened and is continuous with the lower border of the thick plate formed Fic. 30. Amblystoma p., stage when the paraphysis is formed. Model of the right half of the head seen from the medial surface. The model was made from a series of sagittal sections which were oblique to the longitudinal - axis, so that the surface of the model lies in the median plane at the fore brain but passes to the left of the middle at the hindbrain. » 46. by the fusion of ectoderm and neural tube in the neuropore. When the hypophysis begins to push in it presses on the anterior surface of the preoral entoderm and as the primitive optic groove becomes depressed the brain wall presses on the preoral entoderm from above. In this way the preoral entoderm is pushed back and the hypophysis . Insinuates itself between the entoderm and the posterior wall of the primitive optic groove as in Squalus. The preoral entoderm be- comes shorter and blunter, but none of it is cut off as in Squalus. 494 ‘fournal of Comparative Neurology and Psychology. The only important difference between Amblystoma and Squalus is that anterior head cavities are not formed in Amblystoma. In later stages the preoral entoderm and median mass proliferate as mesen- chyme, so that essentially the same end is reached as in Squalus. In some embryos are found indications of a connection of the archentoderm with the hypophysis through the preoral entoderm, but this and the details of the formation of the hypophysis will be described in another paper. As development proceeds the same shifting in the relations of the optic vesicles is seen as has been described in Squalus. In the lateral wall of the forebrain vesicle a thickening is formed which runs Fic. 31. Amblystoma p., nearly the same stage as that shown in Fig. 30, median sagittal section of forebrain and midbrain. x 40. Fic. 32. Amblystoma p., later stage, median sagittal section. Note the ex- treme curvature of the brain in this and following stages. > 40. from the terminal ridge in the middle line obliquely latero-caudad across the primitive optic groove. ‘This thickening is formed in anticipation of the ingrowth of optic tract fibers and may be called the optic ridge. It separates the optic vesicles from the primitive optic groove and causes them to be connected by the terminal pit. Figs. 25, 28, 29, 30 show this in sections and models. By the time the optic ridge is formed the floor of the forebrain vesicle has become depressed to form broad primitive inferior lobes and in the caudal wall of this a mammuillary recess marks the begin- ning of the mammillary bodies (Figs. 26 and 27). This is bounded Jounston, Forebrain Vesicle in Vertebrates. 495 caudally by the tuberculum posterius. Later, when the hypophysis has reached its definitive position a saccus outgrowth from the inferior lobes appears and the region at which it is connected with the brain may be called the infundibulum. There are therefore in the floor of the forebrain vesicle four recesses formed as in selachians: terminal pit or preoptic recess, primitive optic groove or postoptic recess, infundibulum and mammillary recess, The formation of the velum transversum, epiphysis and para- physis need not be described as they are already known (Minot and others). These structures are represented in Figs. 30, 31, 32, 33. Vig. 33. Amblystoma p., all the chief features of the forebrain developed. Median sagital section. < 40. 5. Reptiles and Birds. I have to say here only that the study of whole mounts of chick embryos between 20 and 40 hours of incubation shows that the same relations exist in the region of the optic chiasma as in Squalus and Amblystoma. In early embryos the optic vesicles are connected by a primitive optic groove behind the terminal ridge. Later, the optic ridge is formed, the terminal pit becomes connected with the cavities of the optic stalks, and the optic chiasma occupies the terminal ridge. 6. Mammals. The chief peculiarity of the mammalian brain is the great size of the cerebral hemispheres. In the adult, as is well known, there is 496 ‘fournal of Comparative Neurology and Psychology. - a membraneous tela over the median ventricle and this is continued as the roof of the interventricular foramen into the wall of each hemisphere. In front of the foramina the tela meets the lamina terminalis, so that as in amphibians there is a membraneous tela in the median region of.the telencephalon for its whole length. For the morphological relations of this tela, the ventricles and the chorioid plexuses it is necessary to study the embryology. While in all lower classes, except cyclostomes, a prominent velum transversum marks the boundary between diencephalon and _ telen- cephalon, in mammals the velum has not heretofore been recognized. The large collection of pig embryos in this laboratory gives excel- Fic. 34. Pig embryo of 5 mm. Model of right half of head seen from medial surface. The optic vesicles are still connected with the primitive optic groove. The Roman numerals indicate the brain neuromeres. < 25. lent opportunity for comparison with the lower classes described above. The earliest stage available is a 5 mm, pig cut in transverse series from which a model of the brain has been made (Fig. 34). From the figure it will be seen that this brain agrees very closely with that of the Squalus embryo of about 20 somites. The cavity of the optic stalk is continuous with a groove which traverses the median line, the primitive optic groove. Behind this is the primitive inferior lobe, a ventral expansion bounded caudally by the tubereulum pos- terius. In front of the primitive optic groove is a transverse ridge whose cross section in the median plane presents the form of an arch. Jounston, Forebrain Vesicle in Vertebrates. 497 This is the terminal ridge. In front of this is a median pit, the terminal pit. Following the lamina terminalis around the front end of the brain there is found in the dorsal wall a distinct transverse fold, followed by an arched portion and a second more shallow fold. In this fold and for some distance behind it are seen in later embryos the fibers of the posterior commissure and the decussation of the tectum mesencephali. The more anterior and deeper of the two Fig. 35. Pig embryo, 6 mm, A., median sagittal section. B and C, para- sagittal sections. Neuromeres numbered in Roman. > 20. folds is the velum transversum, as the further description will make clear. In embryos of 6 and 7 mm. the same change of relations in the optic region takes place as has been described above for Squalus and Amblystoma. The terminal ridge becomes prolonged caudo- laterad as the optic ridge and in these later appear the optic chiasma and tracts. The identity of these structures in the classes of animals studied is absolutely clear. 498 ‘fournal of Comparative Neurology and Psychology. The time of development of the mammillary recesses and of the neural part of the hypophysis (sacecus vasculosus) varies somewhat in pig embryos. In most 6 mm, embryos the mammillary recess is already clearly recognizable as a caudal expansion of the primitive inferior lobe (Fig. 35, A) whose border above and behind is the tubereulum posterius. From now on the mammillary recess is always clear. Between it and the primitive optic groove the hypophysis lies against the under surface of the primitive inferior lobe. In some 6 mm. embryos there is to be seen just behind the tip of the Fic. 386. Pig embryo, 9 mm., median sagittal section. >< 20. hypophysis in sagittal section a slight, but definite, depression and thickening in the brain floor, the beginning of the evagination of the neural part of the hypophysis. This shows a variable development in embryos between 6 mm. and 9 mm., but is always clearly present in 9 mm. embryos (Fig. 36). After this time the sae grows out and enwraps the tip of the hypophysis and the relations familiar to all embryologists are established. The outgrowth of this sac is the ear- liest mark of the position of the infundibulum and this is a consid- erable distance behind the primitive optic groove, as in selachians and amphibians. In later development the floor of the primitive Jounston, Forebrain Vesicle in Vertebrates. 499 inferior lobe becomes drawn down in funnel shape and the lateral portions become thickened as the tuber cinereum. — It is one of the peculiarities of the mammalian and human brain that the infun- dibulum is drawn down close behind the chiasma and forms a deeper and narrower funnel than in lower vertebrates. There is no funda- mental difference of relations, The evidence that the fold in the roof which has been called the velum transversum is correctly identified may be seen in Figs. 34 lic. 387. Pig embryo, 12 mm. Median sagittal section, reconstructed from several sections. »< 15. to 39. The posterior commissure is a point about which there is no dispute. In the arch (neuromere) in front of it appears later the epiphysis (Fig. 38). In front of the epiphysis and in the same arch appears the commissura habenularis. This arch then is the roof of the diencephalon. It is bounded in front in all other classes of vertebrates by the velum transversum. ‘The fold to which this name has been given lies in the proper segmental position. This is further supported by its relations to the structures in front of it. Immedi- ately in front of the velum the roof is raised in a distinct arch. 500 §=fournal of Comparative Neurology and Psychology. This corresponds in position to the paraphysis of lower forms (cf. Figs. 30 to 33): Since there is no glandular development known in mammals, it may be called the paraphysal arch, a name which Minot (1901) applies to the corresponding structure in birds. In front of the paraphysal arch a membrane continues forward to meet the lamina terminalis. When the lateral cerebral vesicles are formed it is seen (Figs, 387, 38, 39) that this membrane lies over the ventricle between the interventricular foramina. Fic. 38. Pig embryo, 15 mm. Median sagittal section of the forebrain. The dotted outline of the hemisphere is reconstructed from several sections. aly As development proceeds the velum transversum becomes a fold with a sharper angle but less deep in proportion to the size of the brain. The paraphysal arch remains a distinct median pouch until the lateral vesicles are well formed. In sagittal sections to one side of the median plane the lateral ventricle appears as a dorsal cavity opening by way of the interventricular foramen in front of the paraphysal arch into the median ventricle. (Fig. 38, 39.) These simple relations persist up to the 17 mm. stage or later. By the _ Jounston, Forebrain Vesicle in Vertebrates. 501 15 to 17 mm. embryo the chorioid plexus of the lateral ventricle is forming. Its position with reference to the velum transversum is shown in a parasagittal section of a 17 mm. embryo (Fig. 40) and in three frontal sections of the brain of a 15 mm. embryo (Fig. 41). The velum transversum is not only a dorsal fold, but is continued around the lateral wall of the brain as the constriction (external fur- Fic. 39. Pig embryo, 17 mm. Median sagittal section reconstructed from several sections. The outline of the hemisphere lateral to the median plane is shown in dotted lines. 12%. row and internal ridge) between the diencephalon and the telen- cephalon. By the stage mentioned the cerebral vesicle is sufficiently expanded to push back past the boundary line. In the angle between the vesicle and the diencephalon appears the chorioid plexus pushing into the lateral ventricle. It appears as a folding of the anterior wall or limb of the velum transversum and its lateral prolongation. 502 Ffournal of Comparative Neurology and Psychology. In this way appears the chorioid fissure whose further history need y opi J not be traced. Near the median plane the plexus appears as a fold projecting into the interventricular foramen and separated from the Fic. 40. A parasagittal section from the embryo drawn in Fig. 39. Fic. 41. Pig embryo, 15 mm. Three sections from a frontal series showing the relations of the chorioid fissure and plexus to the lateral hemisphere and the thalamus. The section to the left is the most ventral, that to the right the most dorsal. velum transversum by the paraphysal arch. From this stage on the plexus grows rapidly and becomes very large and in the median region both the velum and the paraphysal arch become involved in Jounston, Forebrain Vesicle in Vertebrates. 503 the plexus and their identity is lost. If the paraphysis is to be found in adult mammals it should be looked for in the chorioid plexus in the middle line between the interventricular foramina. The mem- braneous roof extending forward from the paraphysal arch to meet the lamina terminalis is relatively long in the embryo, spanning the wide opening into the lateral ventricles. In later development these interventricular foramina grow much less rapidly than the hemi- spheres and the roof in question becomes of insignificant length. This is due in part also to the expansion of the lamina terminalis by the commissures which develop in it. The neuroporie recess can be located with certainty in the pig embryos. In young stages there appears a slight ridge in each lateral wall just rostral to the preoptic recess. This is the beginning of the corpus striatum. As the two ridges converge toward the mid- dle line they cause a thickening of the lamina terminalis above the preoptic recess (Figs. 88 and 39, between r. p. and r. n.). This is the location of the anterior commissure in later stages. Above this thickening is another pit which in 15 mm. embryos is a smooth rounded concavity in the middle line, not a transverse groove (Fig. 38). That this is the recessus neuroporicus seems clear from the fact that it is above the corpus striatum and below the interventricular foramina. As the striatum and lateral vesicles grow this pit becomes deeper and more pointed. Three of the young human embryos described in recent years seem to the writer to give clear evidence that the relations in the optic region are the same in man as in fishes, amphibians and other mammals. The embryo described by Low (1908) shows optic pits on the open neural plate. These are apparently connected with one another by a groove running across the middle line parallel with the terminal ridge. The embryo described by Broman (1896) is 3 mm. long and has the neuropore closed. It shows the terminal ridge essentially like that of Squalus or of Amblystoma. The optic vesicles are connected with the primitive optic groove caudal to the terminal ridge. Broman noticed the terminal ridge and gave to its ectal, concave surface the name ‘Fossa interocularis,” but did not speak of its significance. The embryo described by Mrs. Gage 504 “fournal.of Comparative Neurology and Psychology. (1905) is a trifle larger and considerably more advanced in develop- ment. In it the optie vesicles are connected with the pit formed in the neuropore-space. ‘This pit seems to correspond to the preoptic recess above described. From these three embryos it appears very probable that the course of development in man is the same as in the pig and lower forms. II. Dzscussion. 1. The Anterior Hnd of the Head and Brain.—Owing chiefly to the lack of certain essential facts, an extensive literature has grown up about the question of the anterior end of the brain. Since the facts which were wanting are supplied in the preceding pages, a detailed review of the discussions from v. Kupffer and His onward would be unprofitable. ‘The determination of the anterior end of the brain is a matter of direct observation. In the study of succes- sive stages in the early development of a selachian, an amphibian and a mammal the essential facts are found to be perfectly clear, and the three forms agree in the form changes of the brain and in the relations of the brain to the ectoderm, entoderm and mesoderm. The anterior boundary of the neural plate is formed by a trans- verse ridge, the terminal ridge, which is continuous with the neural folds bounding the neural plate laterally. This terminal ridge is clearly seen in successive stages and is readily followed up to the time when the optic chiasma is formed in it. The optic chiasma therefore occupies the anterior border of the floor plate of the brain. This is a matter of fact, not of interpretation. Behind the terminal ridge lies a transverse groove which laterally becomes continuous with the optic pits or vesicles. This is to be seen from the earliest stages when the optic pits are recognizable in the neural tube or, indeed, on the open neural plate. This groove is the depression which His (1892, 1893) called the recessus infun- dibuli and which v. Kupffer (1893) called the sinus postopticus. Later workers have followed His, although his own numerous figures (1892) show that there are two distinct depressions in the region between the chiasma and the mammillary recess. It is clear also (His, 1892, Fig. 2 and p. 162) that the opening into the saccus Jounston, Forebrain Vesicle in Vertebrates. 505 vasculosus represents the infundibulum and the depression which he ealls recessus infundibuli must be something else. ‘This depression I have called the recessus postopticus (1902, 1906). Because of its relation to the optic pits I have called this in the early-embryo the primitive optic groove. This primitive optic groove forms a ventral projection of the brain floor which outwardly might be considered as a transverse ridge and it was to this external ridge that His gave the name Basilarleiste. To the groove within he gave the name recessus infundibuli s. basil- aris. The Basilarleiste of His is not a thickening of the brain floor but a fold which appears within as the primitive optic groove. It is important to make this clear because of the incomparable value of His’s work in matters of general morphology. The Basilarleiste in the embryos studied by His meets the front end of the notochord and of the entoderm (Seessel’s sac). Upon this relation of the brain to the notochord and entoderm His based the conclusion that the Basil- arleiste or the recessus infundibuli formed the anterior end of the brain floor. “‘In der vorderen Endflache endigt die Gehirnaxe und es bedarf einiger Vertsindigung dariber, wohin dies Ende zu ver- legen sei. In Anschlusse an y. Baer und andere habe ich selber in friiheren Arbeiten dies Ende in das Infundibulum oder richtiger ausgedriickt, in die Mitte der Basilarleiste verlegt. Andere, wie neuerdings Keibel, lassen die Gehirnaxe im Chiasma opticum aus- laufen. Die Discussion dartiber, wer mit seiner Behauptung im Recht sei, hat nur dann einen Sinn, wenn man zuvor festgestellt hat, was unter Gehirnaxe fiir eine Linie zu verstehen sei. Ich selber habe darunter stets die Mittellinie des Hirnbodens verstanden. Das heisst die Linie, welche, wenigstens auf friiheren Stufen lings der Chorda, als der anerkannten K6rperaxe verlauft. Diese basilare Axe endigt unzweifelhaft in der Basilarleiste. Versteht man dagegen unter Gehirnaxe eine Linie, welche der Mitte der Réhrenlichtung folgt, so wird diese mittlere Axe in einer Ebene liegen, welche die Grund- und die Fliigelplatte des Gehirns von einander scheidet, und ihr Endpunkt trifft die vordere Endfliche im Recessus opticus, bez. dicht vor dem Ort des Chiasma opticum. Wollen wir zur basilaren und zur mittleren Axe noch eine dritte dorsale Lingsaxe oder Lings- 506 ‘fournal of Comparative Neurology and Psychology. linie annehmen, so haben wir deren Ende am oberen Rande der Lamina terminalis zu suchen, vor der Stelle, wo die Fissura chorioidea ihren Anfang nimmt.”’ Now it must be noticed that more recent work (Platt 1891, Hoff- mann 1896, Neal 1898, and others) has shown that the entoderm does not end anteriorly in contact with the Basilarleiste, but extends forward beneath the terminal ridge. His did not study sufficiently early stages to see this. Early stages show clearly that the basal axis of the brain ends not in the Basilarleiste but in the terminal ridge in which later the optic chiasma appears. Furthermore, this is equally true of selachians, amphibians, birds and mammals. It is altogether probable that the same is true of petromyzonts also, for the depression called by Koltzoff “infundibulum” is doubtless the same as the primitive optic groove of other forms. In all vertebrates studied by the writer the entoderm extends forward beneath the transverse ridge which afterward becomes the optic chiasma. The definition of the anterior end of the head previously given (Johnston, 1905) may now be simplified to read: in all vertebrates the anterior end of the head is the point at which the brain plate meets the gen- eral ectoderm at the same time that it comes into contact with the anterior end of the entoderm. This point is marked in the adult by the optic chiasma. It has been shown in this paper that the depression in front of the optic chiasma which has been known to His and other authors as the recessus opticus, is related to the optic vesicles only secondarily and is primarily a pit in the basal part of the neuropore (lamina termi- nalis). 2. The Homology of the Saccus Vasculosus—Here I wish only to point out the homology of the saccus vasculosus of lower verte- brates with the neural part of the pituitary body in man. The saccus vasculosus is an evagination from the floor of the diencephalon which is more or less branched, is lined by ependymal cells and sensory cells, and is supphed by nerve fibers ending in its epithelial lining. In all lower forms it comes into close relations with the hypophysis. In many cases the subdivisions of the two structures become inter- mingled or interlaced. In man the neural part of the pituitary body Jounston, Forebrain Vesicle in Vertebrates. 507 has the same relations but the cavity becomes obliterated. During early development the evagination appears in identical manner and relations in all vertebrates and the writer can see no ground for doubting the complete homology of the structure in all vertebrates. This homology was understood by His but seems not to be univer- sally accepted. Edinger (1908, p, 203) publishes a schematic figure of a sagittal section of the vertebrate brain in which he shows an infundibulum: in contact with the hypophysis and behind it a wholly separate evagination of the brain floor which he calls the saccus vasculosus. ‘This diagram stands in contradiction to the drawings from actual specimens in the same book (Fig. 167, Chimera; Fig. 175, 176, Varanus; Fig. 178, Ammoccetes; Fig. 181, Siredon; Fig. 219, Hexanchus). In all of these there is only one evagination of the brain floor between the optic chiasma and the mammillary bodies and it comes into relation with the hypophysis. The writer does not know of any vertebrate in which the condition shown in Edinger’s diagram is found. 3. Segmentation of the Neural Tube in Front of the Cerebellum.— In the hindbrain the neuromeres are generally recognized as brain segments corresponding to the segments of the organs in the head. In front of the cerebellum there is no such unanimity of opinion. The writer has discussed this subject at length (1905) and has found nothing in the studies here reported to change any of the conclusions there expressed. On the contrary, the conclusions there based on indi- rect evidence from other authors are confirmed by direct observation. The segments in the mes-, di- and telencephalon are clearly indicated in Fig. 42, representing parasagittal sections of the brain of a pig of 7mm. The optic vesicle is here seen somewhat out of line with the other neuromeres but no one would doubt that it represents one brain segment. In front of it is the first segment, from which the telencephalon is formed. Opposite the optic vesicles in the median region is the velum transversum. Behind the optic vesicle are clearly seen in the figure three segments. In connection with the first of these (neuromere iii) appears later the epiphysis. The other two (iv and v) obviously enter into the mesencephalon. In selach- ians these two segments have connected with them respectively the 508 “fournal of Comparative Neurology and Psychology. thalamic nerve of Miss Platt which probably forms the ciliary gan- glion, and the part of the neural crest which forms the ophthalmic division of the trigeminus. The terminal part of the neural crest in close relation with the neuropore presumably gives rise to the ganglion of the nervus terminalis in selachians. If this be true, every neuromere of the embryonic brain has connected with it in one class of vertebrates or another some sensory nerve or sense organ (includ- ing the optic vesicle and epiphysis. The five brain segments are equally clearly to be seen in Figs. 34 and 35. 4. Boundary between Diencephalon and Telencephalon.—The pos- fod Fic. 42. Pig embryo, 7 mm. Two parasagittal sections to show the seg- ments of the forebrain and mid-brain. Compare figs. 34, 35 and 18. terior boundary of the diencephalon has never been in dispute. It is the constriction between the forebrain and midbrain vesicles and is later marked dorsally by the posterior commissure and ventrally by the tubereulum posterius. When the forebrain vesicle becomes divided into diencephalon and telencephalon the exact location of the boundary between them has not been entirely clear. In all verte- brates in which a definite velum transversum is recognizable this is considered as the mark of the boundary. The existence of a paraphysis and lateral plexus chorioideus in front of the velum and of a dorsal sac and one or two epiphyses behind it is now so thor- Jounston, Forebrain Vesicle in Vertebrates. 509 oughly understood as to need no further comment (Gaupp 1898, Minot 1901, Johnston 1905, 1906). The velum transversum has been described in eyclostome embryos by Sterzi (1908) and in mammals in the foregoing pages, so that the boundary line sought for is now clear in the brain roof in all classes of vertebrates. From the velum transversum a groove or constriction continues around the sides of the brain. Owing to the early evagination of the optic vesicles this constriction in the dorsal half of the brain oceupies the space left vacant, so to speak, by the withdrawal of the retinal tissue. Ventrally the groove is to be thought of as lying in front of the neuromere to which the optic vesicle belongs. The diencephalon consists in its dorsal half of but one neuromere after the withdrawal of the optic vesicle ; in its ventral half it includes two neuromeres, the more posterior of which is narrow while the more anterior one forms the depression of the brain floor which I have called the primitive inferior lobe. The boundary between the diencephalon and the telencephalon in the brain floor has been in dispute because of the obseurity which has existed over the optic recesses and the anterior end of the brain. His placed the boundary behind the infundibulum and assigned the pars optica hypothalami to the telencephalon. He was led to this by his conviction that the telencephalon consisted of a complete brain ring or segment and by his belief that the end of the brain axis was in the Basilarleiste or infundibular recess. As shown above, the optic chiasma is formed in the terminal ridge and therefore occupies the extreme anterior border of the floor plate of the neural tube. If the telencephalon is a complete transverse segment of the brain, as His always insisted, there is no alternative but to include the optic chiasma within it. The primitive optic groove which bounds the optic chiasma behind belongs to the same neuromere with the optic vesicles and therefore is included in the diencephalon. The telen- cephalon can include no more than the optie chiasma and the associ- ated decussations in the brain floor which lie in the terminal ridge. The boundary between the diencephalon and telencephalon is marked by the velum transversum above and by the primitive optic groove or postoptic recess below. In adult mammals, in which both these land- 510 6“fournal of Comparative Neurology and Psychology. marks have disappeared, the boundary ean be defined by a line drawn just behind the interventricular foramen and meeting the posterior surtace of the chiasma ridge (Fig. 44). The external groove which separates the diencephalon and telen- cephalon is usually well marked and in the brains of amphibians, Fic, 48. Sketches to illustrate the boundary line between the diencephalon and the telencephalon. The brains of a selachian (A) and an amphibian (B) are outlined as seen from the medial surface and the boundary set by His is indicated by a dotted line, that fixed in this paper by a heavy con- tinuous line. reptiles and mammals increases in depth and prominence with the enlargement of the cerebral hemispheres. The description of the early development has shown that the lateral chorioid plexus is formed in mammals immediately in front of the velum transversum and of the groove which continues from the velum transversum around the lateral wall of the brain. From this it results that in the Jounston, Forebrain Vesicle in Vertebrates. gi adult the chorioid fissure is found at the bottom of the very deep groove between the hemispheres and the brain stem and that the boundary between diencephalon and telencephalon runs just along the posterior (thalamic) border of the fissure. These relations are clearly set forth by G. Elliot Smith in a recent paper (1908) on the forebrain of Lepidosiren, which agrees in essentials with that of mammals. Fic. 44. Sketch of the human brain for comparison with Fig. 43. When the internal structure of the brain is taken into account it is seen that the boundary line indicated by the development separates centers of different significance. Before it lie the primary and sec- ondary olfactory centers, behind it in the nucleus habenule and inferior lobes (tuber cinereum) lie the tertiary olfactory centers with reflex functions. This is not true of the boundary line laid down by His which placed the region of the infundibulum (pars optica hypothalami)y in the telencephalon. His was apparently not followed 512 “fournal of Comparative Neurology and Psychology. in this by the Basle nomenclature commission and the tables of neurological terms adopted by the commission contradict His’s explanatory notes in that the tables place the pars optica hypothalami in the diencephalon while His states that it belongs to the telen- cephalon. (See His, 1895, pp. 161, 162.) This is perhaps because anatomists generally have felt the incongruity of assigning the tuber cinereum, infundibulum and hypophysis to the telencephalon. This objection fails when only the chiasma and the fiber decussations adjacent to it are included in the telencephalon. The usage adopted by the BNA goes to the other extreme and involves at least as bad consequences. The BNA ineludes the lamina terminalis in the pars optica hypothalami, and implies that the lamina terminalis is the front wall of the diencephalon. The discus- sion of this usage, which is widely followed by anatomists, will come best in the next section, but here it may be pointed out that it implies the inclusion in the diencephalon of various structures which cer- tainly can not be so interpreted. a. The lamina terminalis contains the anterior commissure, and according to the researches of G. Elliot Smith the corpus callosum and hippocampal commissure develop in it also. These commissures would then all fall in the anterior wall of the diencephalon. This is obviously impractical and confusing and would lead to endless diffi- culties in fixing an arbitrary boundary. b. The gray matter in the wall of the preoptic recess constitutes generally in vertebrates an important secondary olfactory center which, unless there are strong reasons for assigning it to the dien- cephalon, should be placed with the other secondary olfactory centers in the telencephalon. All the facts of development and general mor- phology, however, favor the retention of this center in the telen- cephalon. c. In many fish-like vertebrates the larger part of the telencephalon (corpus striatum and olfactory lobe) lies lateral to the lamina terminalis and forms the wall of the median ventricle. These struc- tures in fishes would be included in the diencephalon and there would be endless confusion as to the boundary line in various classes. No such confusion and no practical difficulties in the description of the Jounston, Forebrain Vesicle in Vertebrates. 513 adult brain arise from the recognition of the boundary suggested above which is clearly marked in the development. 5. The Ventricles and the Tela.—An essential part of the question of the boundary between diencephalon and telencephalon is the prob- lem of the median ventricle; specifically, does any part of the median ventricle belong to the telencephalon? The view held by His was that the anterior part of the median ventricle belonged, with the pars optica hypothalami, to the telencephalon. The view which makes the lamina terminalis the anterior wall of the diencephalon assigns the whole of the median ventricle in front of the aqueduct to the diencephalon. It is obvious that the writer must agree with His in recognizing a median ventricle in the telencephalon, although a shorter part of the median ventricle is included than was included by His. The above diagrams (Figs. 48 and 44) show the boundary line of His and that adopted by the writer and it is clear that the short part of the median ventricle between this line and the lamina terminalis belongs to the telencephalon and makes communication with the lateral ventricle through the interventricular foramina. The view which regards the lamina terminalis as the anterior wall of the diencephalon and of its ventricle denies the existence of any median portion in the telencephalon. This means one of two things: either the diencephalon is the terminal segment of the brain and the telencephalon hes lateral to it as two hemispheres, or the dien- cephalon is terminal and the telencephalon consists of ultra-terminal hemispheres. Neither of these is true. Aside from the fact that the latter view involves a contradiction in terms, it cannot be considered, because all the evidence shows that the hemispheres are lateral struc- tures. (a) In the ontogeny of all vertebrates the hemispheres arise as evaginations or expansions of the lateral brain wall behind the lamina terminalis; (6) the lateral ventricles thus formed remain always as lateral prolongations of the ventricle and the median veu- tricle always extends forward beyond the interventricular foramina ; (c) when the hemispheres by their great growth push forward beyond the lamina terminalis, as they do in most vertebrates, they are still connected with the lateral wall of the brain stem and in the middle line the lamina terminalis is always the most anterior structure of 514 ‘fournal of Comparative Neurology and Psychology. the brain in the adult as in the embryo. It is no more true to say that the telencephalon is ultra-terminal than to say that it is post- optic or post-velar. The occipital lobe extends as far behind the velum transversum as the frontal lobe extends in front of the lamina terminalis. The whole hemisphere is a great expansion of a part of the lateral wall of the brain between the lamina terminalis in front and the optic vesicles and primitive optic grooves behind. ‘The point in dispute is whether that portion of the preoptic brain segment which is not carried out in the hemispheres belongs in the dien- cephalon or the telencephalon. The first step in answering this question is to see clearly that in the early embryo the lateral hemispheres and the median portion exist together undifferentiated as a simple ring or segment in front of the optic vesicles. This segment is bounded from the earliest stages, even before the neural tube is closed, by the sharply marked primitive optic groove and the optic vesicles. It is only some time after the formation of the optic vesicles that the dorsal part of this simple segment bulges out at either side to form the lateral hemi- spheres. If the segment is simple at the start, is there any ground for separating the ventral part and adding it to the diencephalon which hes behind the primitive optic groove? The only thing to give support to this view is the connection of the hollow optic stalk with the preoptic recess. Since the optic vesicles have always been referred to the diencephalon, their close relation to the lamina termi- nalis through the preoptic recess suggests the inclusion of the lamina terminalis in the diencephalon. Now, however, it is shown that the optic vesicles are primarily connected with the post-optic recess and are only secondarily related to the preoptic recess. In view of this there remains no ground for separating the median and lateral structures which develop from this primitive first seg- ment. The embryological facts leave only one course open, namely, to consider the lateral hemispheres as the dorsal portion, the region of the optic chiasma as the ventral portion of one segment. Finally, it is impossible to harmonize the relations of the velum transversum in lower vertebrates and in all embryos with the view that the lamina terminalis bounds the diencephalon. That the velum JoHNSTON, Forebrain Vesicle in Vertebrates. 515 transversum marks this boundary dorsally is universally agreed. It stands at some distance from the lamina terminalis, however, and if the boundary line is to follow the latter it must run along the root of the brain and then around its front wall, an obvious absurdity. A reeent writer on the development of the forebrain (Fanny Fuchs, 1908) states that it is only for practical convenience in describing early stages that the term telencephalon should be used at all. Whatever is left after the development of the hemispheres (she recognizes tacitly that there is something left) should be reckoned with the diencephalon. She states that in Rana the telencephalon has no roof because the di- telencephalic groove meets the upper end of the lamina terminalis. Since she has not studied early stages, has not recognized the velum transversum and gives no figures to show what she includes in the lamina terminalis, her conclusions on this point can have little value. Such figures as Fraulein Fuchs cives show a long median ventricle extending far forward beyond the interventricular foramina. The roof of this, as the writer’s own preparations show, is the same as the roof of the similar median ventricle in all other vertebrates, namely a membrane reaching from the velum transversum to the upper border of the lamina terminalis. Fraulein Fuchs has simply included this roof in what she calls the lamina terminalis. This author has studied only the obvious features in the later stages of the development of the forebrain and these give no sufficient ground for any conclusions regarding the morphological value of the telencephalon. There may be quoted here the conclusion of His in his paper on the general morphology of the brain (1892, p. 383): “dass eine solche allgemeine Morphologie nur dann endgiiltig zu gewinnen ist, wenn wir auf die allerersten Entwicklungsstufen zuriickgreifen.” The tela chorioidea of the third ventricle and lateral ventricles requires some comment. It must be noted first that in all verte- brates, embryo and adult, a membraneous tela extends over both diencephalon and telencephalon from the habenular commissure to the dorsal border of the lamina terminalis. In all vertebrates except adults of higher forms there is an obvious narrow place in the nervous brain wall between the diencephalon and telencephalon, and the tela 516 “fournal of Comparative Neurology and Psychology. is widest here. ‘This seems to the writer to be due to the with- drawal of nervous material from the dorsal part of the lateral brain walls to form the optic vesicles. This withdrawal of retina-sub- stance leaves a gap which is filled by membraneous tela only. The tela in the median region of the telencephalon is perfectly evident in any brain, embryonic or adult, so far as the writer is acquainted. From the posterior part of this tela next to the velum transversum arises the paraphysis or the rudimentary paraphysal arch, and forward from that the tela stretches across the median ventricle be- tween the interventricular foramina. It is equally evident that at all stages of development these foramina are roofed by lateral pro- longations of the tela. This is true of all forms with the apparent exception of cyclostomes, teleosts and ganoids. In cyclostomes this is probably due to the compression of the forebrain by the oral funnel and olfactory organ. In ganoids and teleosts the interven- tricular foramina have been widened beyond recognition by the eversion of the lateral walls. In amphibians and higher forms the prolongation of the tela over the interventricular foramina to become the roof of the lateral ventricles has great importance for the forma- tion of the lateral plexuses. The beginning of these has been de- scribed and the only further comment which the writer wishes to make is that the complexity and mystery which the text-books throw around the relations of the velum interpositum and the lateral plexuses should be brushed aside for the sake of the student, who finds the subject difficult enough without artificial stumbling blocks being put in his way. The student should be told simply that the median tela extends laterally as the roof of the lateral ventricle and this becomes infolded to form the lateral plexus. This process con- tinues around the side wall just in front of the junction of the hemisphere and thalamus. 6. Dorsal and Ventral Zones in Diencephalon and Telencephalon. —The suleus limitans of His marks the boundary between alar plate (dorsal zone) and basal plate (ventral zone). The dorsal zone throughout the central nervous system is sensory, the ventral zone motor. Both zones include gray matter and fiber tracts belonging to the correlating mechanisms, and in those segments in which the Jounston, Forebrain Vesicle in Vertebrates. 517 primary sensory or motor centers are reduced or wanting owing to reduction or absence of the peripheral organs, the correlating mechan- isms constitute the whole of the zone concerned. The writer has repeatedly (1902, 1905, 1906, 1909) emphasized the fact that the longitudinal zones constitute the most fundamental divisions of the brain and hence the sulcus limitans is the most important landmark in the brain. The two sulci converge at the anterior end of the brain to meet in the lamina terminalis and this meeting-point marks the anterior end of the central axis of the brain. The end of this axis His placed at about the middle of his lamina terminalis, namely in the recessus preopticus. The facts set forth in this paper show that the chiasma region must be taken from the lamina terminalis and added to the brain floor. Still, in the writer’s opinion, the central brain axis has its ending in the recessus preopticus (Figs. 43 and 44). The reasons for this are to be found-in the following facts: (a) the ventral zone of the brain becomes greatly reduced in volume in front of the third nerve by the absence of all motor centers; (b) it is further reduced by the distribution of fiber tracts to various parts of the diencephalon and telencephalon; (c) the sensory centers are represented in the telencephalon by the large olfactory apparatus; (d) the correlating mechanism of the dorsal zone is greatly hypertrophied in connection with the olfactory centers and in higher forms in connection with the somatie cortical centers ‘(neopallium). In other words the ventral zone at the front end of the brain is represented chiefly by the decussating fibers (optic chiasma and commissures of Gudden and Meynert) of the ventral commissural system, while the dorsal zone contains both sensory and correlating mechanisms which are very large. These facts account for the bending down of the sulci lmitantes to meet near the ventral border of the lamina terminalis. There is no evidence known to the writer tending to show that the recessus neuroporicus has any sig- nificance in this connection. It is only a convenient practical mark of the dorsal border of the lamina terminalis and the anterior end of the brain roof. In the diencephalon the location of the suleus limitans is still more difficult. The typical formation of the ventral zone extends 518 fournal of Comparative Neurology and Psychology. no farther forward than the nucleus of the III nerve, or at most the nucleus of origin of descending fibers in the medial longitudinal fasciculus. ‘The ventral commissural system is interrupted by the downgrowth of the substantia reticularis to form the inferior lobes and mammillary bodies. This downgrowth has so completely altered the relations of parts in the diencephalon that it is practically impos- sible to trace a boundary line between dorsal and ventral zones. The inferior lobes themselves doubtless represent a part of the correlat- ing substance of the dorsal zones (Johnston 1906, p. 277 and fol.). 7. Pallium of the Telencephalon.—A long discussion has been waged over the subject of the general morphology of the pallium since the discoyery by Rabl-Riickhard (1882) of the forebrain roof of teleosts. It would not be profitable to enter into the details of this discussion. The hypothesis of Rabl-Riickhard and Edinger was to the effect that lower forms possessed no true or nervous pallium, but that the membraneous pallium as seen in teleosts and other fishes has been transformed into a massive pallium by the development of nervous elements in it. The hypothesis of Ahlborn was that the anlage or beginnings of the pallium of higher forms must be found in the massive portions of the brain of lower forms, that a membraneous (ependymal) roof can never be transformed into a nervous pallium. For many years the Rabl-Riickhard-Edinger hypothesis dominated the field of forebrain morphology by sheer force of the authority of its sponsors. Studnicka made an effort to show the truth of the Ahlborn thesis, but for the time was over- borne by Edinger and his followers. Although in his first work the present writer accepted Edinger’s views, a wider study of the sub- ject led him to a treatment in 1906 much more nearly in accord with the view of Ahlborn and Studnicka. Kappers and others have added to the discussion and with the fuller knowledge of the comparative anatomical and embryological facts the general morph- ology of the membraneous and nervous portions of the forebrain may be regarded as a closed subject. Much of the discussion has been due to misunderstanding and differences in the use of terms and it will be sufficient here to define the terms applied to the parts of the forebrain and indicate briefly the differences in form in various classes of vertebrates. Jounston, Forebrain Vesicle in Vertebrates. 519 The term hemisphere is applied in the BNA to each half of the telencephalon. It would therefore include the right or left half of all that les in front of a plane passing behind the interventricular foramina and the chiasma-ridge. It is well known that this portion of the brain is not hemispherical in form in all classes. It is some- what so in cyclostomes, many selachians, amphibians, reptiles, birds and mammals. In Heptanchus, Hexanchus, Chimera the hemi- sphere is more elongated and the membraneous roof is more exten- sive. In ganoids and teleosts the width of the membraneous roof is greatly exaggerated. ‘The nervous walls are rolled outward so that the membraneous roof is attached along the lateral or even latero-ventral aspect and arches up over the ventricle. This ever- sion of the forebrain walls in teleosts has made the term hemisphere inapplicable in the descriptive sense. However, most of the organs which make up the hemisphere in other forms are present in the teleostean telencephalon and these organs hold the same fundamental morphological relations to one another and to other parts of the brain. Therefore the term hemisphere may be employed throughout the vertebrate series, although in no two classes does the telenceph- alon approach in the same degree the form of a sphere. In each hemisphere are represented nervous and membraneous portions. ‘The membraneous portions include the lamina terminalis and the tela chorioidea. The lamina terminalis is supposed to be coextensive with the anterior neuropore, but there is no neuropore in eyclostomes and teleosts and in some other vertebrates (some amphibians at least) the upper border of the neuropore is not marked in the early embryos. Where an unambiguous recessus neuropor- icus exists it is the clear mark of the dorsal border of the lamina terminalis. Where this landmark is not clear an arbitrary border for the lamina terminalis must be placed at some distance in front of the interventricular foramina. ‘The tela forms the roof from the lamina terminalis to the tela of the diencephalon, from which it is separated by the infolded velum transversum. ‘The term pars supra- neuroporica of the lamina terminalis which was used by Burckhardt (1894) and is used by Edinger for this portion of the brain roof is wholly without justification. ; 520 Fournal of Comparative Neurology and Psychology. The nervous portion of the hemisphere includes numerous struc- tures the arrangement of which will be spoken of in the next section on nomenclature. The term pallium has been loosely used by various authors for the membraneous roof of the telencephalon, the dorsal part of the nervous portion and the superficial cell layers in the nervous portion. Edinger uses it in all these senses and in the last edition of his textbook (1908, Bd. 2, p. 249) he distinctly states that the epithelial roof of the teleostean telencephalon is the pallium. “Dieses Dach der Hirnblase heisst Hirnmantel, Pallium cerebri. Dazu gehért auch der auf Fig. 220 noch rein epithelial gebliebene Abschnitt, derselbe, welcher schon bei den Selachiern und Amphibien aus eigentlicher Gehirnsubstanz besteht.” This ambiguity is very unfortunate. Since we have the convenient term tela for the mem- braneous roof of the forebrain, the term palliwm should be reserved for the cerebral cortex. The question, then, whether teleosts (or other forms) possess a pallium should be answered, not with Rabl- Riickhard by pointing to the membraneous roof, but by ascertaming whether there is present any nervous substance whose fiber connec- tions and functions warrant its being compared with the cortex of higher forms. It is still too early to define in an exact way what is meant by cerebral cortex. It is not sufficient to define it as superficial layers of cells in the telencephalon because in all classes of vertebrates and in man, superficial gray matter is found in the forebrain whose fiber connections and functions are very different from those of the true cortex. To say that the cortex consists of superficial gray in the roof or dorsal wall of the forebrain gives no means of deter- mining its extent or boundaries. Although the term cortex implies and was first used for superficial layers, it has come in recent years to signify the brain substance which constitutes certain functional mechanisms, whether superficial or not. It is necessary to define the cortex by its fiber connections and from the functional point of view. In mammals the general cortex is understood to be a collection of highly complex centers which exercise functions of correlation and control over bodily movements, ete., through lower sensory and motor centers. The sensory impressions coming to these cortical centers Jounston, Forebrain Vesicle in Vertebrates. 521 usually pass over chains of three neurones. The existence of neurone chains of only two links connecting the peripheral sensory surtace with the cortex is somewhat in dispute, but it is certain that such chains are relatively few in number. The general cortex provides in its structure the means for association and correlation between the areas concerned with various modes of sense impressions. This general cortex has its efferent pathway over the cortico-spinal tract and other bundles descending through the cerebrai peduncle. In lower vertebrates, in which the general cortex is not yet known, the telencephalon seems to consist of olfactory centers and corpus striatum and it is generally believed that the first cortex to appear was olfactory in function. The writer was the first to attempt a definition of the olfactory cortex (1901, p. 239). It was pointed out that the cortical center receives olfactory fibers of the third order, not of the second order. The olfactory pathway consists of: fila olfactoria—bulbus and tractus olfactorius—lobus olfactorius and its efferent fibers—cortex. This definition of the cortex has since been adopted and further developed by Kappers (Kappers and Theu- nissen 1908, Kappers 1908). However, olfactory fibers of the third order run to other centers in addition to the cortex. In the diencephalon the nucleus habenule and hypothalamus, and in the telencephalon itself, the epistriatum (nucleus amygdale in mam- mals), receive olfactory fibers of the third order (Edinger 1896, Johnston 1898, 1901, Kappers 1906, 1908). The epistriatum fulfills this definition in selachians, ganoids and perhaps teleosts when there is no other part of the forebrain that does meet the conditions. How- ever, in higher vertebrates (Kappers 1908) a part of the epistriatum becomes gradually pushed back until it finally occupies a position in immediate proximity to the pyriform lobe (nucleus amygdalz), while a true cortical formation appears in the roof of the hemi- sphere in dipnoans (Elliot Smith 1908) and all higher classes. Now the epistriatum of forms above fishes, whose history has been so beautifully traced by Kappers (1908) does not represent all of the formation to which he gives the name epistriatum in selachians. The writer has shown that the epistriatum in Petromyzon (1902) and Acipenser (1898, 1901) receives an ascending tract from the 522 “fournal of Comparative Neurology and Psychology. hypothalamus and Kappers (1906, 1908) has recognized this tract also in selachians. I have interpreted this as an ascending gusta- tory tract (1906, p. 304). The center into which the tract enters at first (Petromyzon) receives secondary olfactory fibers, but in most fishes, especially the selachians in which the olfactory appa- ratus is highly developed, receives tertiary fibers as well. The entrance of an ascending, presumably gustatory tract, into a ter- tiary olfactory center in fishes creates a condition analogous to that found in the general cortex of mammals; namely, a center serv- ing for the correlation of two sorts of sense impressions which are received over neurone chains of three links. We seem, therefore, to have in the so-called epistriatum of fishes a primitive olfactory cortex. The gray matter does not consist of superficial layers of cells, but forms part of the wall of the ventricle. This primitive epistriatum, as seen in Petromyzon and Selachians, is not all accounted for in the history of what Edinger and Kappers call the epistriatum in higher forms. The primitive epistriatum lies in the side wall (floor and roof) of the selachian forebrain (Kappers). The fiber connections which I have worked out in the greatest detail in Petromyzon and Acipenser, show that the epistri- atal formation in the side wall continues caudad to the border of the diencephalon, 7. ¢., nearly to the nucleus habenule. This is the region called by Kappers the dorsal part of the prethalamus. That this is telencephalic territory is readily shown by the fact that the velum transversum is attached to the lateral wall of the brain just in front of the ganglion habenule and behind the peculiar structure here being considered. The primitive epistriatum therefore consists of (1) a dorsal or roof portion, (2) an epistriatum in the narrow sense resting upon the striatum (palostriatum, Kappers) and (3) a caudal portion forming part of the wall of the median ventricle of the telencephalon. I have shown (1906, Chap. 18) that the caudal portion is of greatest size and importance in Petromyzon, is still of considerable impor- tance in Necturus, and in mammals is represented by a small struc- ture called by older authors the paraphysis but shown by Elhot Smith (1896) to be a nervous structure. The caudal portion de- creases in size and importance in the vertebrate series. The second JOHNSTON, Forebrain Vesicle in Vertebrates. 523 portion is the true epistriatum which Kappers has traced through the phylogenetic series up to the nucleus amygdale of mammals. The dorsal portion of the primitive epistriatum is seen in the roof in Petromyzon and typical selachians, probably in the short roof overhanging a shallow lateral ventricle in Chimera, Heptanchus and Hexanchus, and possibly in a corresponding structure in ganoids at the anterior end of the olfactory lobe. This structure has generally been wholly lost sight of in ganoids and teleosts and when it reap- pears in dipnoans and amphibians in exactly the same position and relations as in selachians it has been treated as a new structure, the pallial formation or hippocampus. It must be recognized that the ganoids and teleosts have no other significance than that of a side branch of the phyletic line. The dipnoan brain represents the next step in advance from the selachian, and in the dipnoans the pallial formation appears just where the dorsal part of the primitive epis- triatum is found in selachians. The further history of this olfac- tory pallium has been so clearly worked out by Elliot Smith and others that no further comment on it is needed. When all fishes are taken into account it is seen that all three parts of the primitive epistriatum receive olfactory fibers and ascend- ing fibers from the hypothalamus. Only the dorsal portien develops into what is universally recognized as olfactory cortex in higher forms (hippocampal formation). Now if it be shown that the ascending tract (gustatory) from the hypothalamus enters the hippo- campus we could say that throughout the whole vertebrate series the archipallium (Elliot Smith) is a correlating center for olfactory and gustatory impulses. If it should prove true that the gusta- tory center is in the hippocampal formation, all parts of the cortex can be defined as correlating centers; the archipallium for olfac- tory and gustatory impulses, the neopallium for impulses coming from the eye, ear, skin, muscles and joints. General visceral sensa- tion would be represented also in the archipallium. I have long felt that the term epistriatum is an unfortunate one. In only the smaller number of forms is it descriptive of the structure to which it is applied. The view expressed here and in 1906 is that the body which Edinger called epistriatum is a part of a more 524 fournal of Comparative Neurology and Psychology. extensive formation which in primitive forms has essentially the same structure and connections in all of its parts. This statement of fact is subject to revision if further studies show it to be incorrect. With regard to the name, however, I find that the extension of the term epistriatum to include all of this formation has led to miscon- ceptions of my meaning. This formation may be described as the visceral correlating center of the telencephalon, or as the correlating substance of the visceral sensory zone of the telencephalon. Instead of the term primitive epistriatum used above, this might be called the primitive visceral cortex. The dorsal portion of it becomes the true visceral cortex (archipallium) when it receives tertiary olfactory and gustatory fibers. Some of the factors which enter into the definition of the term cortex cerebri may be indicated as follows: a. The term is applied only to structures in the telencephalon (excludes lobi inferiores, ete.) ; b. The afferent paths of the cortex are predominantly of the third order (excludes the secondary olfactory centers; the cortex shows an uncertain grade of development in the more primitive forms) ; c. The cortex serves functions of correlation for afferent impulses of two or more kinds (olfactory, gustatory, optic, auditory, ete. ; excludes the epistriatum sensu stricto or nucleus amygdule) ; Whether such correlating centers are superficial in position is not of essential importance. The general cortex of mammals is separated from the ventricle only by fibers related to the cortex itself, i. e., by its own white matter. The question of superficial position is of much less importance in the case of the cerebral cortex than in that of the inferior olives, the medial and lateral geniculate bodies, and other centers which are separated from the ventricle by voluminous fiber bundles and gray masses which have no direct rela- tion to themselves. The point of view of the writer stands in contrast to that of Kappers who in his recent paper (1909) extends the concept of cortex to the lobus olfactorius (“‘paleeocortex”) although the centers concerned are simple secondary olfactory centers throughout the vertebrate series. His reason for this is that these centers occupy a superficial position JOHNSTON, Forebrain Vesicle in Vertebrates. 525 in mammals (e. g., cortex lobi pyriformis). Elsewhere Kappers insists upon tertiary afferent pathways as the essential eriterion of the cortex. The use of two different criteria at different times leads to confusion of thought. I would not apply the term cortex or paleocortex to these secondary olfactory centers, but would use the simple terms lobus olfactorius, lobus pyriformis, ete. T would apply the term cortex to certain functional mechanisms. The above sugges- tions toward the definition of these mechanisms are of necessity incom plete and in part hpyothetical. If such a term as paleeocortex were used it should be applied to the morphological forerunner (homologue ) of the true cortex. 8. Divisions and Nomenclature. The nomenclature of the brain adopted by the Basle commission is still the best that we have. largely because it embodied the results of the indefatigable work and keen insight of His. Before suggesting certain changes in the BNA tables to bring them into accord with the facts I wish to examine briefly the nomenclature offered by some recent authors. Edinger has shown great fertility and enterprise in the production of new names in brain anatomy. Edinger’s terms have arisen from his comparative studies of adult brains and are the expression of his effort to present large and obvious relationships in attractive form. He considers the lamina terminalis as the anterior boundary of the diencephalon, agreeing with the BNA. The narrow portion of the brain extending forward from the optic chiasma (very long in Chim- wra) he calls the prethalamus (1908, p. 194). When he comes to describe the telencephalon (p. 251) he describes the lamina terminalis as the plate which unites the two halves of the telencephalon. He treats the anterior commissure system as belonging to the telen- cephalon and even speaks of the “recessus prechiasmaticus” as one feature of the telencephalon. Here is a contradiction for which there is no remedy in Edinger’s mode of treatment. The difficulty is augmented by Edinger’s definition of the primitive basal portion of the telencephalon (hypospherium): the primary and secondary olfactory centers and the corpus striatum. Now the floor of what Edinger calls prethalamus is a secondary olfactory center which I have called the nucleus preeopticus. It receives fibers from the bulbus 526 “fournal of Comparative Neurology and Psychology. olfactorius and gives rise to some fibers of the tractus olfacto- habenularis. Considerations of practical. convenience and clearness would dictate that this secondary olfactory center be included in the telencephalon, not in the diencephalon. Edinger distinguishes a neencephalon from a_paleencephalon. His paleeencephalon includes the lower segments of the brain together with that portion of the telencephalon which he calls the hypo- spherium. The neencephalon is the same as his epispheerium and includes the tertiary olfactory centers (Elliot Smith’s archipallium) and the general cortex (Grosshirn, Elliot Smith’s neopallium). Is it true that the whole of the lower segments of the brain are to be set in contrast to that part of the telencephalon to which the name epispherium is given? Are the centers for the cochlear nerve in the medulla oblongata, the inferior olives, the nucleus dentatus in the cerebellum and the auditory centers in the inferior colliculus and metathalamus older than the tertiary olfactory centers or the general cortex? Or is it true only that our knowledge of them is older? The terms paleencephalon and neencephalon are undoubtedly useful as expressions of the functional evolution and growth in organiza- tion of the whole brain; but as descriptive terms for the topographical features of the brain they would not be useful in the lower brain segments and are decidedly misleading when applied to the forebrain alone. The terms hypospherium and epispherium seem to apply fairly well in mammals, but I see no advantage in introducing new terms which will not apply to the brains of lower vertebrates as well. In describing the minor divisions of the telencephalon Edinger is neither consistent with himself nor with the majority of authors. His description of the olfactory centers is quite confusing and con- tains several self-contradictions. The body into which the olfactory nerve enters he calls (1908, p. 252) the lobus olfactorius. Almost all recent authors have agreed to use the name bulbus olfactorius for this, while the term lobus olfactorius is given to the collection of secondary centers which make up a greater or less part of the body of the forebrain. To this posterior part Edinger proposes to give the name lobus parolfactorius. This term replaces the term area JOHNSTON, Forebrain Vesicle in Vertebrates. 527 olfactoria which Edinger used earlier (1896, p. 141). The present ‘use of lobus parolfactorius is likely to lead to confusion with the area parolfactoria Broce used by the BNA, which is only a specific part of the whole group of secondary olfactory centers. Edinger uses the term area parolfactoria in the BNA sense in Figs. 247, 275, 279, which are old figures reproduced in this edition without revision. This disregard by Edinger of the usage of the majority of other authors is responsible largely for confusion which arises in the work of younger authors or those who are not thoroughly famihar with the internal structure of the brain. For example, Fuchs (1908) calls the bulbar formation in the frog larva “lobus olfactorius” and applies the term “hemisphere” to the rest of the telencephalon. The confusion of other authors who attempt to follow Edinger’s work would be less if Edinger himself always used his terms in the same sense. On p. 260 of the same book he uses the term lobus parolfac- torius as synonymous with tuberculum olfactorium at least in rep- tiles, birds and mammals. This center, Edinger thinks, is a special center for the oral sense, an interpretation which G. Elliot Smith (1909) shows to be wholly improbable. To the secondary olfactory center which covers the lateral and ventral surface of the striatum Edinger gives the names lobus olfactorius (Figs. 230, 231, 234). area olfactoria (Figs. 247, 273, 274), cortex olfactorius (Figs. 240, 265, 280), and nucleus teenie (Fig. 239). Professor C. J. Herrick in the course of a very valuable paper on the subdivision of the brain (1908) gives expression to the cur- rent idea of the telencephalon in the following sentences. “The telencephalon is well named. It is terminal, not only in position but also in point of time, having been added relatively late in the phylogeny to the rostral end of the original neural tube. The BNA has done well to omit from it the pars optiea hypothalami which was originally tabulated as part of this region by Professor His. Origi- nally developed as primary and secondary olfactory centers, it has added successively more and more complexity during the whole course of phylogenetic history.” I quote this not for the sake of criticising Herrick’s work—for the whole spirit of his paper and most of the details of it are in perfect harmony with my own views— 528 “fournal of Comparative Neurology and Psychology. but because it brings out clearly the difference between the results of the comparative study of adult brains and the results of a com- plete genetic method in which embryology contributes its just share. The early development shows that as matter of fact the telencephalon is not added late in the phylogeny but is actually the first segment of the original neural tube in all classes of vertebrates. A certain part of this segment expands and grows in complexity with the increasing complexity of the vertebrate organism and of its mode of life. Further, it is clear that the telencephalon originally contained more than primary and secondary olfactory centers. The existence of the nervus terminalis is evidence of this; the existence of preoral entoderm and of a well developed neural crest in the telencephalic segment of the embryo is evidence of it; the existence of a well developed correlating center, the corpus striatum, in the brains of all vertebrates, is further evidence. Of all the authors who have represented the telencephalon as purely olfactory in funetion, not one has shown or attempted to show that the corpus striatum is accounted for by its relations to the olfactory centers alone. The present writer is the only one who has given facts to show the path- way of impulses both to and from the epistriatum and the striatum in lower vertebrates. In my description of the brains of Acipenser (1898, 1901) and Petromyzon (1902) I showed that olfactory tract fibers ended in the epistriatum and that fibers arising from the cells of the epistriatum ended in the striatum. From the striatum the well known basal bundle (Edinger, Van Gehuchten) passed back- ward. These results have been confirmed by Kappers (1906, 1908) but Edinger has persistently disregarded the fact that in his descrip- tions of the forebrain no fiber tracts are mentioned which would enable either the epistriatum or the striatum to carry out any fune- tions whatsoever. In the last edition of Edinger’s textbook the epistriatum is represented as an end-station for olfactory tract fibers. but no fibers are described in lower vertebrates which go from the epistriatum to any other part of the brain. The striatum, on the other hand, gives rise to the tractus strio-thalamicus, but no fibers are described which come to end in the striatum. The epistriatum receives olfactory impulses but has no way of giving out any impulses; the striatum has an efferent pathway but receives no Jounston, Forebrain Vesicle in Vertebrates. 529 impulses. Neither of these important forebrain centers is provided with the means of carrying on any function. As a further indication that the primitive forebrain has some func- tions in addition to the olfactory sense, the writer has described two ascending tracts to the forebrain. One of these, the tractus lobo- epistriaticus, is believed to carry up gustatory impulses to the epistriatum from the tertiary gustatory center in the hypothalamus (fishes and amphibia 1898, 1901, 1902, 1906). If this hypothesis is correct the epistriatum must be regarded ‘as a correlating center for smell and taste and so a forerunner of the smell-taste cortex. A second tract has been traced in Acipenser (1901) from the tectum opticum only as far forward as the optic chiasma where it enters the telencephalon. Whether it ends in the corpus striatum or in some other part of the forebrain remains to be seen. In my textbook (1906, p. 336) I have pointed out that the entrance of such a tract as this into the telencephalon constitutes evidence of the beginning of the correlating centers which in higher vertebrates we eall the neopallium. The writer has been convinced for some years that the elements or beginnings of all the chief parts of the telen- cephalon of mammals and man are to be found in the telencephalon of primitive vertebrates. Herrick’s revision of the nomenclature of the diencephalon and mesencephalon contains two new terms, ophthalmencephalon, and medithalamus. As a pedogogie term based on function, “ophthalm- encephalon”’ has my hearty approval. As a morphological sub- division of the brain it is open to the objection that the regions included—retina, chiasma, lateral geniculate bodies, pulvinar and tectum opticum—do not have sufficient morphological unity. The term medithalamus is offered by Herrick provisionally for the things left over after the ophthalmencephalon has been set apart. It thus in- cludes the central gray and a number of nuclei of diverse func- tions. The fact that it must include the medial geniculate body on the lateral surface of the diencephalon seems to the writer a fatal objection to the term medithalamus. In the diencephalon the epithalamus and hypothalamus are fairly clearly marked both functionally and morphologically. The hypo- 530 ‘fournal of Comparative Neurology and Psychology. thalamus requires new definition both toward the thalamus and toward the telencephalon. The latter is furnished in the new facts brought out in this paper; the proper boundary between thalamus and hypothalamus can be determined only after we have fuller knowledge of the internal structure. The metathalamus and thalamus each pre- sents morphological unity and cannot well be improved upon at present. The chief changes needed in the BNA at present are such as are required by the new facts regarding the telencephalon and the boundary between it-and the diencephalon. These will be indicated below. Fic. 45. 27and:3 1 2.91 Failed 8.16 Meso hee. ha ek al |) eS kG 2 4.18 Failed | 19.15 2591." 1 29).08) 4) 18.5400 sis 3 1.38 iaalal 50.00 I Ale NS Ag | 139825850 + 183 16.10 47 .00 41 15.91 | 20 | @ik5S 5 63 29.06 Halo .85 9.53 74 13°33 6 54l(0) 1.50 A oe | .23 16259) .16 ahs Le a . 28 6.33 9.85 .10 4.14 .19 8.09 8 . 70 1.26 9.00 10 Zeon 40 ons 9 . 20 1.30 6.70 25 2.06 22 4.00 10 13 1.20 1.55 .06 aio 09 WSI/ 11 .10 1715 1.86 .08 .79 09 1.50 12 .08 6.13 2.01 06 2.07 .07 4.07 13 .25 Lhe We 7A} X 1.45 els .85 1.46 14 .05 85 ROS 90 .96 47 1.46 15 13 31 61 .96 . 50 54 46 16 13 20 se | 733 30 23 37 17 08 25 =28) | mls 19 10 26 18 20 36 svAsy | ela 23 15 32 19 06 55 51 .06 29 06 53 20 31 16 20 .18 21 24 18 21 11 39 31 .13 21 12 35 22 18 21 18 .36 23 27 19 23 05 10 siti .05 08 05 10 24 10 1.29 20m .03 41 06 UC 25 10 OF ile | 22 26 45 41 12 26 27 15 .16 13 14 28 ale . 63 Ie | 37 29 stoi || aye | 30 a.46 73 | 31 .08 16 32 b.50 33 .90 34 -13 35 .10 36 .10 37 230 38 13 39 -10 Note. At ‘a’ the box was turned to face in a different direction. At ‘b’ it was turned. back to the original position. ‘X’ represents the effect of a definite emotional element accidentally introduced into the experiment. It is described in a note on p. 549. al — Yoakum, Behavior of Squirrels. 5 ie a Ss | + beer Ne pai Sy Ee ie be Ee 10 Bi 22/23) (“2F 25 (26° 27 26 (29° 980) Slo S24) 335) 34 39956 Sr 58) 39 ‘Trials = za 4 19 16 17 16 = GEA ectelatal oe RAB eSeeeae eae Paso wiN M4 ial a ee | |_| NE a ar AES ee) NS | Ea Bie a a | \ ANS Bae es nal cI i = ; FE , nga l : : VENER SSE Lal fa N 2 Ls Z a [- iad a ie t iC Li C1 aise bal BERS RD eine ete tig et my ne Ps 23) ar 26 25) 24 re] 2a 2h 7 6 5 4 3 ib i 7) 3 4 1d 6 5 4 3 2 / o Fic. 2. Constructed from Table II. A. Curves for squirrels 1 and 4. (Pre- viously untrained.) AB. Curve for squirrels 2 and 3. (Previously trained in “sawdust-box”’ and in “outside latch box.’’) 552 ‘fournal of Comparative Neurology and Psychology. to top of cage; tried to open door from top of box; climbed off box, tried door, went completely around box and worked at door again; thence to south side, and back to door and then around box again. Finally left problem-box. Climbed back on top, then off over string and to door; went back on top once more, bit at wire netting of cage enclosing problem-box; went around north side, and then back to box; nosed string; played with string, then came to south side, tried door, sat and gnawed at wood for two minutes. Stopped; ran around cage several times and went to door again; bit at nail in corner of box; then climbed on top, pulled and bit at string while there until the door came open. Did not notice that door was open and continued to tear at string and pull and tug at wire netting for one minute. Entered, time: 19-15 min. Aug. 1, 07, 12 M. Male No. 2 (trained), second trial: Sratched at north side; at back; ran on top, down to south side; looked around; drank water out of a vessel in corner of large cage; ran over top of box and scratched at door; worked very hard here for almost a minute; scratched at north side. Pulled and tugged at wire netting on top of cage near the point where string was passed through; examined floor of large cage and scratched at door of problem-box ; ran around box, listened, ran around again, listened; scratched at door and then at south corner, listened, etc. Stopped work after 61 min- utes and refused to start again. Squirrel No. 1, male (untrained), second trial: Touched door with nose, and then ran around box; ran over top of box; repeated this five or six times, trying door each time as he passed it. At intervals would run all over large cage. No scratching or working at door, simply touched it or tried to enter, as if door were open. Time: 4.18 minutes. Squirrel No. 4, female (untrained), second trial: Ran around box, then over it five or six times; tried door twice; touched string several times in running over the box; pulled string, gave two tugs; opened it and jumped from box and ran into food. Time: 2.91 minutes. Tt will be seen from these notes that the trained squirrels Nos. 2 and 3 both attempted to enter by biting and scratching at the door and by scratching sawdust, sticks and shavings away from the bottom of the problem-box on all four sides. The movements throughout were evidently those of the early learning processes. The notes also show individual differences quite plainly. Both had learned the two boxes, the sawdust and outside latch boxes. The tendency of No. 3, the female, was to bite and scratch at the door as she had done in learning the outside latch problem. In her work there was very little and, after a few slight attempts, no scratching of the sawdust at the bottom of the box. On the other hand, the male did not carry over the movements learned in his work upon Yoakum, Behavior of Squirrels. 553 the outside latch box, but did carry over scratching and gnawing movements acquired in the sawdust box. Both animals apparently found the task of breaking the habits acquired in the one or the other of the earlier experiments almost insuperable. The slightest noise or movement in the room, or the failure on their part to raise the latch at the first pull on the string, would invariably drive them to use some one or mere of these previously acquired movements. The work of the untrained squirrels, Nos. 1 and 4, was of a much different type. “Useless” movements were as much in evi- dence, and their activity was at all times as great as that of Nos. 2 and 3. However, the movements of 1 and 4 were random move- ments, “useless” in the sense of not getting the animal nearer its food, and further “useless” since they did not carry out any one par- ticular line of attack, as did the movements of squirrels Nos. 2 and 3. Pulling the string, biting at the wire, running over the box, etc., were random activities directed toward food. The scratching and steady work at the door by the trained animals constituted an activity which the most casual observer would judge to be directed toward effecting a definite mode of entrance to the food box. An interesting point in learning to attend to the string came out in the work of all four animals. At first, the string went unnoticed or was subjected to the least possible scrutiny and the first pulling was entirely accidental. At the seventh, eighth and ninth trials, the string itself was singled out and became the point of the most eager attack. When found, it was subjected to the severest kind of strains. Hemp string no longer withstood their attacks and had to be replaced by electric light cord, and finally by flexible steel wire. The associa- tion between the string and the food supply had become definite, but the movements used in pulling the string had not yet become auto- matic. After the ninth and tenth trials, the unnecessary time and energy spent on the string were gradually eliminated, until finally a single pull on the string became the cue for a rush to the door. Fre- quently in later trials, this pull was made so hastily that it failed to open the door. (c) The Maze.—The method of conducting the experiments on the maze is practically a repetition of that adopted in work upon the TABLE III. SHOWING INDIVIDUAL AND AVERAGE TIME OF SUCCESSIVE TRIALS IN THE Maze. (Two ANIMALS.) No. of No. I | No. IV | Trial | Woes | seriale | Average 1 29.00 1.25 5D 2 10.25 3.80 7.02 | 3 9.56 1.86 5.71 4 4.75 1.05 2.80 5 6.33 2.50 4.41 6 3.14 1.83 2.48 i 1222 1.86 1.54 8 1.19 .70 94 9 2.33 1.08 1.95 | 10 1.91 65 1.28 11 1.45 “38 91 12 .67 eel 74 13 85 .83 84 14 .58 61 59 15 43 66 54 16 *4 00 | 255 2.27 17 .29 | 61 “45 18 1.40 | 56 98 19 .56 | 45 | 50 20 60 | 55 | 57 21 45 45 | ieee 22 . 60 .63 | 61 | 23 45 36 | Al 28 -98 38 68 25 00 28 30 26 .86 26 56 27 .80 26 53 28 66 | 26 46 29 -41 .25 33 | | 30 .80 50 65 3l .ol .28 | 39 32 26 25 OF | 33 34 26 30 34 .40 25 32 35 20 29 24 36 22 | .28 25 37 50 | .26 38 38 40%) .30 32 39 50 .26 38 40 36 26 31 41 50 25 36 42 - 40 . 26 33 43 ail .28 D4 44 23 28 25 45 .23 BOR 25 46 .30 26 98 47 25 | .40 32 48 35 35 Binge 49 48 | BS ot wea 50 35 “98 Woe) ee | *No particular reason can be found for the high record of squirrel No. 1 in the 16th trial. All the conditions were as usual, the animal simply stopped in the maze. No errors. 33) Yoakum, Behavror of Squtrrels. S[PI oP 64 8+ 44+ ‘QZBU JANOH UOJAMIL_LOY) SULUALET UT S[BUIIUB OA} JO Sp1OdDIAI aq} WAOJ OTYdVAS Ul SUrmMOoyS—'s “OTT 94 Se ++ Ct Ze (hb oF GF BF Ze 9 © se — ph Ghe @ CEST IS: og 6% 87 _ 17 97s Ht +— Pate Pm ae ee BY TL zat e 8. a 9 + 2 Saul: sous PAE eS DSSS Sea] gi $l $ it Eat ia ale | SSBERE) BESENZNEEr —! tt Soe 3 | le A = PEEERee sae OI LI 556 ‘fournal of Comparative Neurology and Psychology. white rat. The Hampton Court maze used in the preliminary rat experiments at the Chicago laboratory was employed in this test." It was covered with wire netting. ‘The food box was so arranged that the animal could be transferred to the entrance by means of a small control page described above (p. 544). A sliding trap door was inserted in the passageway just in front of the last turn which led to the food box. This was left open during the running of the maze and was closed as soon as the squirrel had entered the food compart- ment. Such precautions were necessary in order to prevent the return of the animal after the nut had been found or when it was desired to remove him from the food box for another trial, or to return him to his cage after the trials of the day had been com- pleted. The numerical results of the record are quite similar to the curve t.14 The accompanying given by the learning process of the white ra table and curves show the records made by the two squirrels which learned the maze. The process of learning the maze is for these two squirrels similar in all essential particulars to that of the rat. The early trials were characterized by all possible errors and hesitancies. Time and time again the animal would almost make the run perfectly, only to stop at the last runway and return to the starting place. The elimination of errors was fairly gradual with certain persistent errors lasting in both animals until as late as the twentieth trial. One point was noticed which seems worthy of mention. The ani- mals, even after having completely learned the maze, were easily disturbed. The slightest movement on the part of the observer, any noise outside of the room, or a bright beam of sunlight on any part of the maze, all must be investigated before the squirrel would con- tinue. The hungriest squirrel could be stopped at any place in the runway and made to turn into a cul-de-sac or to go back to the starting point by moving the finger along the wire netting above him. If the observer happened to appear anywhere within the squir- See WATSON, J. B., Kinzsthetic and Organic Sensations: their Réle in the Reactions of the White Rat to the Maze, p. 10. 4Cf. WATSON, op. cit., appendix. TABLE IV. SHowi1na EFFectT oF DARKENING MAZE AND OF RoTATING MAZE. ‘Rial tera Wonnine Average Total Darkness 1 38 oo 45 2 26 .41 BE S 30 cao 31 4 26 1.40 83 5 30 .45 37 6. 31 as 42 i( 58 ie 55 8 45 .45 45 9 30 .95 62 10 28 a 31 11 26 .43 34 12 30 .41 35 iS 28 36 32 14 28 .30 29 Lights on 15 28 36 By 16 25 .30 oT 17 .26 Sil 28 18 26 .30 28 19 .26 .30 .28 20 .26 .40 foo Maze turned 180° 21 46 .96 71 22 1.05 .66 85 23 Ral 61 46 24 1205 .46 75 25 38 oo 35 26 .36 25 31 27 EST a5 28 28 238 29 30 30 .28 31 .41 32 .46 33 a3 34 30 35 28 36 dues 37 .26 38 3s Maze turned 270° 1 Se nod 1.91 2 30 .46 38 3 60 .30 45 4 28 .26 27 5 28 -25 DT. 6 40 .28 34 (i .30 .36 33 8 .28 Maze turned 360° | 3 “88° 133 38 . 26 32 30 .83 56 35 .33 34 ONMUIkWNe ow i) 558 ‘fournal of Comparative Neurology and Psychology. rel’s field of vision, the animal would stop a moment and look up, or try to get out, at the place nearest the observer. Timidity also often produced hesitancies and slowed the time without actually resulting in error. The actual care of the squirrel in the maze turns out to be distinctly more difficult than is the case with the white rat. The former is much more easily disturbed emotionally than the latter. Curiosity and the desire for social contact with the experimenter also often cause interruptions in the squirrel’s run through the maze. (d) Effects of Darkening the Maze and of Rotating the Maze.— Table IV and Fig. 4, appended below, show the effects of darkening the maze and of rotating the maze 180°, 270°, and 360° on two squir- rels trained to run the maze in the light. The irregularity in the record apparently produced by the absence of the light appeared only when the light was turned out after the animals had obtained their “cue.” That it is caused by distinctly emotional changes and not by a loss of “cues” due to the darkness, follows from the fact that when all lights in the room were turned out before the squirrel was started, there was neither hesitancy nor error. Even in the cases where the lights were turned out after the animal had started, a simple hesitancy was the only error present— the rest of the trip being made as automatically as in the light. When the maze was rotated 180° and 270° respectively, the be- havior of the animal was decidedly different. Errors were a fre- quent occurrence. The animals often ran back to the starting place, hesitating and disturbed on both the forward and the backward runs. The rotation to 270° seemed provocative of the greatest confusion. The last error after the change was made appeared at the tenth trial. The rotation to 360°, after practise at the two positions just men- tioned, produced no errors. There was slight confusion, however, and the time was slower than the fastest perfect time of the earlier tests. This series of trials does not appear in the curve. These records have not added much in any exact way to our knowledge of the sensory “cues” used by the squirrel in learning and later in automatically running the maze. Those who observed the animals at any time were convinced that the sensory factors operative in the learning process and in the perfected maze habit were other 559 Yoakum, Behavior of Squirrels. ‘aTqv) 9q} UT UMOYS ST [VII] Pore JO out] oy, “FON [edatmnbs Jo yaoM oy} Aq popor[Vaed jou T ‘ON Jodamnbs Aq apBUt sp10d1 9.18 SOUT] poop sv UMBAp pue q poyAeUT Sjied sy, ‘SfodamMbs OA} JO PAODAI VSRAGAL OY} WoOAJ PouAIOJ o1B YW poylvUr oAaN aq Jo suonaog ‘ezeut Supyejoa Jo pue ozeum Surueyiep Jo Joyo Surmoyg “AL 9qVy, Woay poponaysuoj—F “IT 560 ‘fournal of Comparative Neurology and Psychology. than those contributed by the distance receptors. These latter sense organs frequently presented difficulties to the learning process rather than actually assisting it. Such cannot be considered as rigidly proved, but the general and specific behavior of the animals points to such a conclusion. Ill. Tests on TemMPprERATURE SENSE. The habit of the squirrels on cold days or whenever the tempera- ture of the room became decidedly lower than usual, of burrowing under shavings, sawdust or cotton, led to an attempt to devise means of testing the temperature sense by the use of the discrimination method. In this connection, the animal’s method of covering itself is interesting. The squirrel will draw the shavings, or cotton, up in a pile in one corner of the cage, and will then burrow into it. When finally hidden in the pile, all that is visible is a portion of © the tail; if this is drawn aside the tip of the nose and finally the | entire head become visible. The little animal thus lies curled up in its nest with the tail as the final addition to its covering. The temperature test to be described was made in the first place with a view to determining the general features of the temperature sense; in the second place, for the purpose of testing the range and fineness of this sense. ‘The tests as originally planned are not com- pleted. One of the animals was accidentally killed, and the work on the other was stopped by reason of the experimenter’s removal to a distant locality. It is probably better to look upon these results as being qualitative and preliminary rather than to consider them as being standards quantitatively determined. The apparatus was constructed as follows: Two galvanized iron boxes (A and B, Fig. 5) were made with outside dimensions of 9 x 9 x 24 inches. They were built like square-sided pipes, as shown in the drawing, the inner opening being 5 x 5 inches and running the entire length of the box. The space between the outer and inner jackets was two inches deep and entirely enclosed the central cavity with the exception of the ends. This enclosed space was supplied with three vents, an inlet (I), an outlet (O) for the water supply, and an air vent (V) to relieve the pressure when Yoakum, Behavior of Squirrels. 561 Py Hy / 562 ‘fournal of Comparative Neurology and Psychology. water was admitted to the vessel, or when the vessel was emptied. This third vent was not needed so long as the circulation of water was constant. The outlet was placed at the bottom of the encircling jacket and brought to a level with the upper part of the box. Arranged thus, it could be used to siphon all the water out, which not only prevented the possibility of rust, but also at the same time afforded the opportunity for beginning work immediately on sub- sequent days, by eliminating the necessity of heating a large body of cooled water in the jackets. In the bottom of each of these 5 x 5 inch inner passages small cubical pans were sunk flush with the floor to a depth of one inch. These pans served as food receptacles. They were placed near the end farthest from the opening used as an entrance for the animal (thus avoiding any possibility of the use of vision). The back of each box was covered by a ground glass plate, behind which a 16 candle power light (EL) was placed. The boxes were painted throughout a dull black, and every effort exerted to make them abso- lutely alike. The experimenter found this successful so far as his own discrimination was concerned. The temperature of these boxes was regulated by forcing hot and cold water of the desired temperature to circulate through them. A triple faucet was used, two vents giving cold and hot water, respect- ively, and the third so attached that the water from the first two could be mixed and any desired temperature of water obtained. The overflow ran into a sink near by. The regulation of the temperature was not wholly exact. In the first place, the temperature of the room varied slightly from day to day. The extreme limit of this variation during this series of experiments was 3 degrees. The temperature near the experimental boxes necessarily varied according as the needs of the experiment demanded high or low temperatures. Further, this surrounding air varied considerably during the experiment. It was finally decided that the temperature should be said to be that obtained by a rapidly registering Centigrade chemical thermometer when its glass surface was held against the inner surface of the uniformly heated box, five inches from the mouth or entrance. The mercury of the bulb Yoakum, Behavior of Squirrels. 563 TABLE V. Tests ON TEMPERATURE DISCRIMINATION OF SQUIRRELS. Male Squirrel Female Squirrel Series | 40°C. | 15° ©. 4oec. | 25° ©. 40°C. | 25°C. | 40°C. | 30°C. by days | Right Wrong Right |Wrong Right | Wrong Right Wrong 1 1 3 4 0 4 | Z is 5 ey 2 3 4 0 0 6 6 2 3 33 4 2 2 4 3 9 Dy 4 3 0 4 0 5 4 9 2 5 6 1 1 3 3 6 if 0 6 3 2, 4 1 5 4 4 3 a 6 yy 2 0 Gl 3 6 1 8 6 0 2 1 9 0 4 1 9 4 0 5 0 3 0 6 1 10 5 0 4 2, 5 2, 5 3 11 5 0 6 0 8 1 8 1 12 4 0 3 if 4 1 7 3 il 10 0 My 0 10 O 14 4 0) It 0 9 0 15} 2 2 3 0) 10 0) 16 2 1 2 0 We 4 0 4 0 18 3 0 3 0 19 3 0 Totals: | 225 elie 59 10 86 32 76 24 eit Ast iia | | = ——| nee Percentage. .80% | 85% | 73% | 76% 1 At this point a change in conditions was made, see p. 567. 2 Exclusive of series 15, 16, 17 and 18. did not come in contact with the side of the box. It was determined by a series of trials that when the temperature thus obtained was 40° - Centigrade, the air in the center of the opening the same distance from the entrance was from 1.5° to 1.25° lower. These conditions were kept constant and tested before and during each daily series of experiments. Flexible rubber tubing was used for making the water connec- tions from the overhead piping. This tubing permitted the raising of one box over the other, which was necessary in order to rule out the position factor. Shifting the large heavy boxes was done by means of a rope and pully (P). The temperature boxes them- selves, the space in front of them, the passageway between them, 564 fournal of Comparative Neurology and Psychology. and a considerable space behind them, were screened in by wire net- ting. Sliding wire partitions served to confine the animal in any desired portion of the enclosed space during the adjustment of the apparatus. The passageway between the boxes as they stood in position was five inches wide. The animal entered from behind and passed between the boxes to the open space in front, and there had the choice of turning into the opening of either box. It will be seen that these tests were made upon few animals. All of the animals, however, were in good condition and the experimenter feels that the results, as far as they go, are representative. The first set of tests was made upon the male squirrel. The conditions of this test were as follows: Box A, the standard (which contained the food in all the tests), was kept at a temperature of 40° + 2° C. Box B, the variable, during the formation of the dis- crimination habit was kept at the temperature of 15° + 2° C. A record of this test 1s shown in the first three columns of Table V. After discrimination had become definite, the temperature of the variable box was raised to 25° + 2° C. Columns 4 and 5 of the same table show the results of this change. The second set of tests was made upon a female. In this case, as in the one above, the standard box A was kept at a temperature of 40° + 2° C., but the variable B was kept at a temperature of 25° + 2° ©. After the association had become definite, the tempera- ture of the variable was raised to 30° + 2° C. The latter part of this test (discrimination between 40° and 30°) was not completed, but the work was carried far enough to leave no doubt that the dis- crimination of this difference was possible. Table V, columns 6, 7, 8 and 9, show the results of the whole test. A similar set of tests was made upon three white rats. Table VI shows the records of these animals and the temperatures used. (a) Discussion of results of tests upon squirrels.—The behavior of the squirrels in this test varies little from recorded descriptions of like experiments in other sensory fields. The “controls” used were such as to eliminate sensory factors other than that of tem- perature. It may be well to mention in some detail the precautions taken. Yoakum, Behavior of Squirrels. 565 Position was ruled out by reversing the direction of the turn the animal must make to enter the proper box. Our experience in this part of the control is quite similar to Professor Yerkes’ descrip- tion.?® TABLE VI. Tests ON TEMPERATURE DISCRIMINATION OF WHITE Rats. No. 1 No. 2 No. 3 = = | = = == a8 = Series 40° C 24° C | 40° C 24° C, 40° C 24° C by days Right Wrong | Right Wrong Right Wrong E wee sae ie | | | | 1 3 Cee eer ae Ss lie eS 3 2 3 3 | 3 3 4 D 3 2 4 | 4 2 4 3 4 3 2 3 3 3 3 5 Dee 2 aaines ill 6 2 4 3 6 2 | 0) | 4 0 | ih 2D al 3 2 | 5 0 4 1 8 2 Oy | 4 3 3 2, 9 6 2 | 2 4 4 2 10 Tees n | 3 1 | 4 1 il | 2 0 | 6 1 ! 5 2 12 2 3 || 5 0) | 6 1 13 9 Z 5 0 4 2 14 9- 0 | 5 0 5 0 15 10 (pil amnesia a ee 10 1 16 10 OF el) 5.08. Fel eO 9 0 17 Ira) ally xy 10 0 18 HS) geclOe Gee r eeas 10 0 -——- —« — Totals. ..-2\ 75 25 | (96 | 2 93 27 [seer Meaete 6 esl | ! ele Percentages. . .75 % ! 81 % | EGS It was found impossible to eliminate the position factor to the point where it became non-operative. This failure to rule out posi- tion was used in the latter part of the series as a semi-control test. Going twice or three times into the box in one position was sufficient to establish a preference in favor of that side. Changing the right- left relation of the boxes at this stage constituted a severe test, both of the temperature association and of the squirrel’s patience. No failures to choose correctly between the boxes are recorded at the stage where this test was tried. ~’The Dancing Mouse, The Macmillan Co., ’07, p. 91 ff. 566 Fournal of Comparative Neurology and Psychology. As has already been mentioned, all visual difference between the boxes was eliminated. The boxes were of the same size, both were painted black and closed at the back by ground glass windows. Behind these windows, 16 ¢. p. electric lights were placed in order to prevent the possibility of unequal lighting (at times these lights were interchanged). ‘The food boxes were sunk into the bottom of the passage way, and nut hulls and the kernels used for food could not be seen from the upper part of the entrance. To prevent any possibility of this, however, both food boxes were partly filled with empty nut hulls, and toward the close of a series, with whole nuts and pieces of nuts. In a control test, both hghts behind the boxes were turned off entirely, leaving both boxes in total darkness; finally, first one light and then the other was turned off to determine the firmness of the association. Smell was eliminated by rubbing the side and bottom of the five- inch passage way with nut kernels and with other aromatic sub- stances.. Food was frequently placed in both food boxes, when the discrimination had been established, to test thoroughly the absence or presence of smell associations as well as visual. In order to be certain that all the factors other than temperature were alike in both boxes, the box which was made the standard on one day (40°) was used as the variable on the succeeding day. This al- teration was not constant, 7. e., not made every other day. It was feared that even this regularity might be learned by the squirrel, so that the change of the standard temperature from one box to another was made quite frequently in the midst of a single day’s series. To prevent discrimination in the runway on any basis whatever, the sides were covered with asbestos. It was not particularly in- tended that discrimination on the basis of temperature should not take place here, but in order to eliminate secondary criteria, it seemed better to eliminate differences in temperature as well. After the association had been formed, an effort was made to de- termine how the squirrel detected the temperature of the box. This was very incompletely done, for the reason that it was desired to confuse the animal as little as possible in order that the later experi- ments might proceed immediately. Asbestos pieces were prepared Yoakum, Behavior of Squirrels. 567 to fit snugly over the front lower edge of the entrance. These were placed over both boxes and thus shut out all temperature discrimina- tion by contact, thus forcing the animal to depend upon the air in the entrance to the tunnel. The effect is shown in the records of the first set of tests, series 15, 16, 17 and 18, Table V. The squirrel was plainly confused, which he showed by sniffing the air and testing each box once or even twice before entering. He soon learned, however, to depend upon the air at the entrance of the box in deciding which box to enter. It seems quite possible that under the first set of conditions, the squirrel was discriminating partly by actual contact with the metal of the box and partly by means of the air at the entrance. (b) Discussion of results of tests upon white rats—Table VI shows the work of three white rats on temperature discrimination. The method used was slightly ditferent from that pursued with the squirrels. The rats seemed so uncertain and irregular in their early tests, that it was practically necessary to develop a position habit. Such a habit was accordingly developed and then broken. The rat learned to find the food in one box by position, then the position was changed and he was compelled to learn a new position. To the observer at least the earlier stages of the learning process seemed to be much more easily detected by this method than by the method used in the experiments upon the squirrels. IV. Concuvusion. 1. In the solution of the problems placed before the squirrel in the series of experiments, it is shown that the squirrel learns by the trial and error method. His learning curves are in the main similar to the curves representing the learning processes in the white rat. 2. The greater irregularities in the curves obtained from the squir- rel may, perhaps, be explained on the basis of emotional factors that are more prominent in these animals than in the rat. The greater sensivity to emotional disturbances seems to be due to the fact that the distance receptors play a larger role in the life of the squirrel than in the rat. Experimental proof of this seems possible. 3. Successful training in certain problems is highly prejudicial 568 ‘fournal of Comparative Neurology and Psychology. to the further training of the same animal in problems which pre- sent a large number of similar conditions. Different methods of opening the same problem box constitute such unfavorable conditions. A latch problem box and the maze, for example, do not present enough identical conditions to interfere with the learning process of the one when the other has been learned first. 4. The grey squirrel and the white rat can form associations upon the basis of a temperature sense. Actual tests show that the squirrel can discriminate between two boxes when they differ in temperature by 25°. In the case of one animal tested, a difference of 10° in the temperature of the box was found discriminable. Further tests with a more accurate apparatus would in all probability show that the discriminable difference in temperature may be much less. The experiments upon the white rats show that a difference of 16° is easily discriminated. 5. Incentives which may be used easily in further investigations with the squirrel are: hunger, disagreeable odors and tastes (bad nuts, onions, etc.), gnawing impulses, and love of exercise and greater freedom. ‘This last incentive is especially strong in the squirrel. The ability of the squirrel to detect an edible nut from a faulty one probably contains the sense, or senses, in which the keenest dis- criminative power of the squirrel is reached. TROPIC AND SHOCK REACTIONS IN PERICHATA AND LUMBRICUS. BY EK. H. HARPER. From the Zoological Laboratory of Northwestern University. Wi1TH Two FIGURES. CONTENTS. PAGH TTL OG UC CLOT tasare coe treks tersreseneyetereciece Orevereie astro tue res htaelierelie Sees ere stelcre ste allem avers 569 MovementsroG Perichetay oie. cle sis « ccexstsyelsie olaeiele sietehecone Leone cheer opens: voewets 569 Methods or Stimulant ompacres <'<'sc%e1s chee sc. clei se aysicereileie verereiwie: oiavsccuelens 572 ChemicalyStimulation or the Anterior Wmdl s2iae accsmctesielscleisis Scicleis <' s 572 INGE ATIVE S EEA CLONE UY) CS te stcysscrs: sisua evan ole. epayete: oipiate iisietere ar eterrereseleleveretererstolstte 573 Relations top thesmuscwlaburer seri. ace wane ceiver eels son anal rons chores 575 Frequency of the reaction types under different degrees of stimulation 576 VATA Uy en Olt UCR y [CS eachevaus auecelels ©. s2aiere isco Where: Sebevev neve SUG ok oreteleie me ec 576 The tropic tendency in the different reaction types ..............e. 578 The two-phased character of the reaction types ............2..s00- - 580 BU muersemMedsemprindli chee responses! a... cre cic to iele » 175. the near vicinity. In Catostomus (young’) the cells have a similar arrangement, the great majority being grouped around the anterior end of the cavity of the tectum, a few being found along the mid- dorsal line and an occasional cell placed farther laterally. In brief, in selachians the cells are much more numerous, extend the whole length of the tectum and the largest collection is in the posterior region; in ganoids and teleosts the cells are fewer, are confined to the anterior half of the tectum and are collected chiefly at the 604 “fournal of Comparative Neurology and Psychology. anterior border. In fishes I have seen none of the large cells along the course of this bundle caudal to the velum medullare anterius. In Necturus the cells are found throughout the length of the tectum, are most numerous in the posterior part and are very few near the anterior border. Except near the anterior border, every transverse section of the tectum shows from three to six of these cells and one or two of the number are situated laterally, somewhere within the medial one-half of the tectum. One cell was seen at the lateral border of the tectum. In Cryptobranchus the arrangement is similar, but the greater number of cells are in the anterior half of By S. PAu bt A Vic. 6. < 50. Acipenser, and the mammals studied including the human embryos. The course of the bundle in Seyllium eanicula, Acipenser and Nec- turus was described in my previous paper. I wish to make certain additional notes on the basis of preparations made since that paper 612 fournal of Comparative Neurology and Psychology. was published. In Seyllium canicula the course of the radix mesen- cephalica through the cerebellum was somewhat complex and I sug- gested that this accounted for Edinger’s mistaking this bundle for ¢ tractus tecto-cerebellaris. In Seyllium stellare the course of the bundle is identical with that in S. canicula except that it makes a much less complex bend or loop in passing through the cerebellum into the velum medullare anterius. The course of the bundle is so direct that there is no chance of losing or mistaking it in following it Tic. 11. Acipenser rubicundus; some cells of the nucleus tecti as seen in a transverse section. The bundle of fibers is the beginning of the radix mesen- cephalica. 175. through transverse sections. The same is true of Squalus acanthias. Also, in sagittal sections of Scyllium: stellare the whole course of the bundle is very clearly followed (Figs. 9 and 10). The way in which the mesencephalic bundle crosses the motor root bundles in these sagittal sections makes it quite impossible to make any mistake as to which root the mesencephalic bundle enters. In Acipenser the bundle was previously followed only into the tectum, not to its cells of origin. I have since traced the bundle in JoHNsTON, The Radix Mesencephalica Trigemint. 613 other preparations with perfect clearness around the lateral part of the tectum to the nucleus magnocellularis tecti. The fibers run in a broad bundle immediately outside the ependymal layer and at the anterior border of the tectum are seen in transverse sections bending toward the median line and joining the large cells (Fig. 11). In the turtle the whole course of the bundle is remarkably clear. At its exit there is some intermingling with the motor roots, which is not seen in fishes or amphibians and suggests the complex relations seen in adult mammals. In mammals the structures in the region of the trigeminal roots are more crowded than in lower vertebrates and the relations of the mesencephale root have been difficult to make out on that account. So long as the fact that the cells of origin lay in the brain did not seem to be of especial importance for the interpretation of the bundle, authors inclined to believe that it joined the sensory root. After the formulation of the neurone doctrine and the law that sensory fibers arise from peripheral ganglion cells and motor fibers from central cells, most authors saw a connection of the mesencephalic bundle with the motor root of the trigeminus. I believe that this accounts for the change of opinion from Meynert and the older authors to Kolliker, Van Gehuchten, Cajal and recent authors. The mesencephalic root is very difficult to trace to its exit in mammals, the differences of opinion are most natural, and the opportunities for seeing what one wishes to see are admirable. When I first studied this root in Seyllium I fully expected to find it join the motor root, as I had been convinced so far as mammals and birds were concerned by the work of Cajal and of Wallenberg. I was greatly surprised by what I found in Seyllium, but in all the forms studied since I have never been able to trace the bundle into the motor root or to find good reason for doubting its connection with the sensory root. Instead of burdening the paper with an extended description and with numerous figures of the various brains, I will make one or two notes on the course of the bundle in adult brains and depend upon the figures from the human embryonic and feetal brains to demonstrate the essential relations in mammals. It is necessary to give attention only to the relations of this bundle to the motor and sensory roots, since the course of the upper part of the bundle is well understood. 614 fournal of Comparative Neurology and Psychology. The first mammal in which the bundle was followed with certainty was the mole, whose brain is comparatively simple and primitive. In it the bundle runs caudad over the motor nucleus without any complex relation to it and then bends downward and forward to enter the spinal trigeminal tract and leave the brain in the sensory root. The large cells in the locus coeruleus are widely separated from br Con | === a2 | V mes aes V motZ, Poet vit : S Fie. 12. Sketches to show the course of the mesencephalic root in the common rat. 25. Camera drawing. seen that these fibers form definite bundles which go out with the motor root. They are the crossed fibers of the motor trigeminus. These are all the fibers which enter the motor root in this specimen. The mesencephalic root continues in the transverse plane from the sensory root to a point dorsal to the motor nucleus, where it turns forward to go through the isthmus to the tectum mesencephali. The connection of the mesencephalic bundle with the sensory root in this Jounston, The Radix Mesencephalica Trigemint. 623 section is unmistakable, since both lie in the plane of the transverse section and the mesencephalic fibers run directly among the sensory bundles. The motor bundles are crossing the mesencephalic bundle obliquely. In the sections caudal to this one, other fasciculi of the / Ij, ig’ / “i /) Hy Hs fl ml Vmes. Vmot nuc Fie. 20. Section next anterior to that shown in Fig. 19. An interlacing of the mesencephalic fibers with the crossed motor fibers. The two form with the sensory root a Y-shaped figure. The mesencephalic fibers descend among the sensory fibers in the stem of the Y, while the motor fibers in their oblique course forward leave the section at the point where the two bundles interlace. mesencephalic bundle are cut lengthwise and show exactly the same relations as are seen in this section. In the section next cephalad there is an interlacing of the fibers of the crossed motor root with the most cephalic fibers of the mesencephalic bundle. As this is the 624 “fournal of Comparative Neurology and Psychology. point at which all the confusion of the two kinds of fibers arises, this section is drawn in Fig. 20. The motor roots are cut obliquely, the sensory root is cut nearer to its point of exit. The crossed motor bundle crosses the mesencephalic bundle at such an angle that the two kinds of fibers can be clearly distinguished. It is not necessary to describe other sections of this series. Cephalad from this the mesencephalic bundle has its usual position. The relative position of the cell column of the mesencephale root and the motor nucleus is clear in these sections. The motor nucleus extends forward six sections from the one drawn in Fig. 20, and the cells of the mesen- cephalic root appear in the fourth section cephalad from that of Fig. 20, so that there is an overlapping of the two for three sections of this series or six sections of the complete series. The cells of the mesencephalic root le one millimeter or more dorsal to the motor nucleus and are separated from it by the bulk of the mesencephalic bundle itself and by the crossed motor bundle. Tn this embryo the mesencephalic root, coming from its cell column which lies dorsal and cephalic to and separated from the motor nucleus, runs ventro-laterad lateral and caudal to the motor bundles, which it crosses obliquely, and leaves the brain in the heart of the sensory root. Fetus of 42 em., transverse sections. The 42 em. feetus is taken up next in order to compare the transverse sections with those just described. One section from each half of the brain is shown in Figs. 21 and 22. The two sections are nearly at the same level. In the section through the left half of the brain (Fig. 21) the cerebellar portion was broken away as indicated by the dotted line; in the right hand section (Fig. 22) the reticular formation was broken as indi- cated by the dotted line, but neither of these breaks in any way affected the structures under consideration. An examination of the figures will show that the chief relations here are identical with those in the earlier foetus. Medullation is now much more general. The reticular formation contains many medullated fibers and the pons, acustic area and cerebellum show fairly numerous fibers. The sensory root is cut just at the point where a part of its fibers are turning into the spinal trigeminal tract, which is better formed on Jounston, The Radix Mesencephalica Trigemini. 625 the right (Fig. 22), since that section is somewhat further caudal than the left-hand one. Many fibers of the sensory root enter the chief sensory nucleus and the cephalic end of the substantia gela- tt NSS Lane SO NET les Vsens e ~~ ese e es N Fie. 21. Human feetus, 42 cm. roots on the left side. surface of the pons. leave the section. x 20. Transverse section through the trigeminal The short heavy line at the left below is the outer @, £, x indicate the ends of the motor bundles, where they All the fibers below and to the left from these are sensory. tinosa. In each section is seen a large bundle of sensory fibers pass- ing dorsad over the outer surface of the sensory nucleus. A part of the fibers enter that nucleus, the remainder seem to pass on to the cerebellum. 626 ‘fournal of Comparative Neurology and Psychology. The motor nucleus is indicated in Fig. 21 by a dotted line and in Fig. 22 by the line along which the section was broken. Its dorso- lateral surface is covered, as in the younger specimen, by bundles of the motor root. On each side there is a broad but loose bundle of decussating fibers and beneath this are denser bundles of fibers arising \ I1Vven a LEO" iw aes aN \N\S V mot eee SS - ‘ \ Fie. 22. Same specimen as in Fig. 21, two sections farther caudal on the right side. < 20. from the motor nucleus of the same side. The motor bundles run ventro-cephalad and are cut obliquely in the section a short distance ventral to the lower border of the motor nucleus. The mesencephalic root of both sides is clearly formed of fibers from the sensory root and only obliquely crosses the motor bundles Jounston, The Radix Mesencephalica Trigemini. 627 as those pass downward and forward beneath it. In sections caudal to those shown in these figures the same relations maintain, except that the mesencephalic bundle is more widely separated from the motor bundles in successive sections. It is seen in Fig. 21 that fibers pass out from the mesencephalic bundle to end in the sensory nucleus. The close relation of this bundle to the sensory nucleus is clearer in the next sections caudad. Here it is distinctly seen that the mesen- cephalic bundle runs through the sensory nucleus and in the case of many sensory fibers it is difficult to tell whether they accompany the bundle or end in the sensory nucleus. Also it appears that many sensory root fibers run through the lateral part of the sensory nucleus or over its lateral surface and continue parallel with the mesen- cephalic bundle to enter the cerebellum. Some of these relations are shown in Fig. 22, which is taken from the second section in this series caudal to the one shown in Fig. 21. In the sections next cephalad from those figured, the same interlacing of mesencephalic fibers with decussating motor fibers as was seen in the 27 cm. fcetus is present and gives a confused picture. This interlacing is difficult to unravel in transverse sections because of the greater number of fibers medullated as compared with the younger stage and a drawing of this would not be clear unless it were made diagrammatic by omitting many of the fibers. The study of the sections convinces me that there is only an interlacing together with a certain crowding, as compared with the younger foetus, and that the mesencephalic fibers all join the sensory root. From the horizontal sections next to be described this interlacing is drawn as accurately as possible under the camera without any schematizing, and the evidence for the con- clusion here stated will be seen in those figures. It should be borne in mind, however, that in both the series of transverse sections above described the greater part of the mesencephalic root clearly passes down to the sensory root wholly separate from the motor root, so that any doubt that might exist regarding its possible relation to the motor root would attach to only the small part of its fibers. The cells of the mesencephalic root are drawn in black in Figs. 21 and 22. It is clearly seen here that these cells are separated from the motor nucleus by the motor root bundles and by the arcuate 628 “‘fournal of Comparative Neurology and Psychology. : Wy p Mi jy / Fic. 28. Human foetus, 37 cm., frontal section through the inferior colliculus and isthmus. >» 20. Description in the text. Jounston, The Radix Mesencephalica Trigemint. 629 fibers of the acustic area. In these figures the cells are removed from the motor nucleus by about the diameter of that nucleus itself. The cells shown in Fig. 22 are the nearest cells to the motor nucleus. Feetus of 87 em., horizontal or frontal sections. From this series nine sections are drawn (Figs. 23 to 32) showing the course of the mesencephalic root from the mid-brain to its exit in the trigeminus. For the purpose of demonstrating whether this bundle is connected with the motor or sensory root it will be best to follow it from above downward in the description. In sections through the decussation IV vent Fic. 24. The sixth section of this series (12 of the whole series presumably) ventral to that drawn in Fig. 23. > 20. Description in text. of the trochlearis the mesencephalic bundle is seen passing forward deep in the corpora quadrigemina, or rather along their base, and reaching nearly to the bundles of the posterior commissure. Cells are seen at intervals along the bundle, twenty or more being counted in a single section on one side. A drawing of such a section has not been made because this part of the course of the bundle is familiar to every one. Fig. 23 shows a section of the mesencephalic bundle beneath the posterior colliculus at the point where it begins to turn ventrally. Far forward appears the root of the trochlearis, to the left is the lateral lemniscus, and between them is the brachium con- 630 ‘fournal of Comparative Neurology and Psychology. junctivum whose fibers are lghtly medullated. In Fig. 24 the greater part of the fibers have turned ventrad and are ent across in the section. Owing to the fact that the fibers of this bundle are cut across in the frontal series, their course will not appeal so quickly to the eye in the following figures as in those taken from the trans- verse series, but the evidence of the relations of the fibers will be much more complete from an examination of both series of figures. In Fig. 24 the mesencephalic root appears as several small bundles of fibers obliquely placed in the section. The fibers may now be grouped into three main areas as an aid in tracing their further course. These areas are indicated roughly by the position of the IV vent Fic. 25. The third section of this series ventral to the last. >< 20. Descrip- tion in text. letters a, b and ¢. The fibers adjacent to b and ¢ form one fairly continuous flattened bundle which is wider cephalo-caudad. Adjacent to a are numerous oblique fibers, some of which are Jonger than others. These belong to the more ventrally placed bundles in the mesencephalon and are just now joining the root-bundle. Between a and the bundle b-c in this figure are some intermediate fibers which in Fig. 25 have joined } orc. In Fig. 26 the motor nucleus appears and is outlined by a dotted line and by the motor bundles which partly embrace its caudal surface. At least a part of the broad bundle running diagonally past the motor nucleus consists of decus- sating motor fibers from the other side. For convenience the motor bundles will be designated by the last letters of the alphabet and these decussating fibers are labeled z, while the homolateral bundles Jounston, The Radix Mesencephalica Trigemint. 631 y and x are seen arising from the nucleus. It should be noticed now that the mesencephalic bundle b-c hes caudo-lateral to the motor nucleus and is distinct from the motor bundles. It will make the understanding of the following figures more easy if the reader holds in mind that from here on the mesencephalic root is destined to run latero-ventrad and slightly caudad to its point of exit, while the motor root bundles run ventro-cephalad and laterad, so that at this point the bundle b-c shows its nearest approach to the motor bundles. It is this bundle b-c whose separate course has been clearly shown in the transverse sections. The bundle a in this figure is to be found (i ee f : Coan VN lV vent. Fic. 26. The second section of this series ventral to the last. 20. Descrip- tion in text. among the motor bundles, this being the point at which the inter- lacing already described takes place. The fibers will be recognized as a row of short lines set at an angle to the direction of the motor fibers. It is in reality a very thin flat bundle which lies chiefly between the decussating and homolateral motor fibers. This section is the one in which these fibers are most completely intermingled with the motor fibers. To render the relations more clear, I have drawn the same section at a higher magnification so as to be able to show the exact position of the fibers of bundle a more clearly (Fig. 27). Every fiber is drawn under the camera with great care. I am unable to find a single fiber of the mesencephalic bundle which seems to turn into the motor bundles. 632 ‘fournal of Comparative Neurology and Psychology. Above and below the section just described the relations of bundle a are entirely clear. Proceeding with the series, Fig. 28 shows the full extent of the motor nucleus and all of the motor bundles formed. The nucleus is pear-shaped with the small end caudad. A few decus- sating motor fibers are seen passing over this small part of the nucleus to join the bundle z which receives also fibers from this part of the nucleus. The bundle y is also a mixed bundle of decussating and homolateral fibers. From the large cephalic part of the nucleus are being formed the remaining motor bundles which are marked LAE mone GAs KZ LZ Fic. 27. Same as in Fig. 26. > 78. Description in text. v, w and a. Between the bundles w and y are seen the curving fibers of the bundle a of the mesencephalic root, just extricating themselves from the motor bundle a. Lateral to the bundle a are seen the fibers of the bundle b-c, now loosely scattered. The letters a, b and ¢ stand at the angles of a triangle which includes the fibers of the mesen- cephalic root and nearly all the fibers appearing within this triangle belong to that root. All these fibers are clearly traced in the inter- mediate sections. It is probable in this and the following sections that a few of the fibers in the a-b-c area end in the sensory nucleus, Jounston, The Radix Mesencephalica Trigemini. 633 for as Fig 28 clearly shows, the mesencephalic root is running directly through the sensory nucleus. There now appears in this section the most dorsal portion of the main sensory root of the trigeminus, just at the point where it is giving fibers to the sensory nucleus and where the most of its fibers are turning caudad as the spinal trigeminal tract. There also appear in the figure the root of the vestibularis and of the motor facialis. In Fig. 29 appear besides the sensory and motor roots and nuclei of the trigeminus, the motor facialis, the cochlearis, fibers of the Fic. 28. The fourth section of this series ventral to the last. 20. Descrip- tion in text. trapezoid body, ete. The motor bundles of the trigeminus are col- lected opposite the cephalic end of the motor nucleus, except the bundle z, which still receives decussating fibers and is closely related to bundle a of the mesencephalic root. The fibers of the b-c area have now collected into two fairly distinct groups of small bundles. The bundles ¢ are evidently joining the sensory root and are inter- mingled with the fibers which are passing caudad in the spinal tri- geminal tract. The other relations in this figure will be clear from the labelling. 634 “fournal of Comparative Neurology and Psychology. The following figures require little comment in addition to the description given beneath each. The sensory and motor roots now run on to their places of exit, the motor in front of the sensory. Of the motor root bundles, the one which is chiefly composed of decus- sating fibers, bundle z, lags behind the others, remains close to the Z~ Nmot.cr. SY, . & KE y Fic. 29. The third section of this series ventral to the last. x 20. The heavy lines to the left mark the outer surface of the brain. Description in the text. sensory root and leaves the brain as a separate rootlet in this fcetus. Of the mesencephalic root the bundle ¢ joined the sensory root in Fig. 29, the bundle 6 enters the sensory root in Fig. 30, while the bundle a appears as three small compact bundles, the last of which is seen joining the sensory root in Fig. 31. The relations of the Jounston, The Radix Mesencephalica Trigemini. 635 spinal trigeminal tract, the cochlearis, the facialis and the trapezoid body are drawn for the sake of orientation. Fic. 30. The fifth section of this series ventral to the last. x 20. Descrip- tion in the text. The horizontal sections of the 37 cm. foetus fully support the description of the mesencephalic root in the transverse sections of 636 fournal of Comparative Neurology and Psychology. the 27 and 42 em. stages. The general conclusion may be stated thus: in the human embryo of 15.5 mm. the mesencephalic root of the trigeminus is connected with the sensory root and is clearly not connected with the motor root or nucleus, but is widely separated Jee, BA Fic. 31. The second section of this series ventral to the last. » 20. The motor and sensory roots of the trigeminus, the cochlearis and motor facialis are shown. The outer surface of the pons is toward the left. A broken line separates the motor and sensory roots. The fibers of the motor bundles have a much darker and more opaque stain than those of the sensory bundles. Fic. 32. The sixth section of this series ventral to the last. » 20. The motor and sensory roots of the trigeminus are shown near their point of exit. In these last two sections there is nothing except their deeper stain to indicate that bundles y and ¢ are motor bundles. They are readily traced continuously through the series as described in the text. The broken line separates the motor and sensory roots. from them; in the later foetus of 27, 37 and 42 cm. stages the mesen- cephalic root is gradually crowded closer to the motor nucleus and root and where it passes over the dorso-caudal surface of the nucleus a small part of its fibers are interlaced with the decussating motor fibers and with a few of the homolateral fibers; by far the greater Jounston, The Radix Mescenphalica Trigemini. 637 part of the root remains wholly distinct from the motor root and nucleus; all the fibers of this root mingle intimately in the sensory root of the trigeminus with the fibers which end in the chief sensory nucleus and with those which go to form the spinal trigeminal tract. Discussion AND CONCLUSIONS. The important considerations regarding this mesencephalic root are (1) the character and significance of its cells of origin; (2) the course, position and connections of the bundle in the brain and (3) the peripheral distribution of its fibers. That the fibers arise from cells in the brain in all classes of verte- brates there is now, I believe, no reasonable doubt. I have shown in this paper that in fishes and amphibians the great majority of these cells lie in or near the mid-dorsal line in the tectum mesencephali. Since the same root bundle of the trigeminus arises from these cells and from those in the locus ceeruleus in mammals, I know no reason to doubt, and I know of no author who doubts, that the cells in the two situations belong to the same category. We are dealing with the same set of cells throughout all vertebrates and those in mammals must be regarded as having migrated farther from the place of origin, at the mid-dorsal line. The facts are all in favor of the supposition that these cells have been derived from the neural crest just as the giant ganglion cells in the spinal cord have been. The cells located in the locus cceruleus in mammals will require some additional ex- planation such as the supposition that they have migrated probably from the mesencephalic segment along the course of the nerve bundle. T see no possibility of denying that these cells arise in the extreme dorsal region of the brain in all vertebrates and I know of no other structures with which they can be compared than the giant cells in the spinal cord. The reason for comparing these with one another is that both send their processes out in sensory nerves. The similarity of the mesencephalic cells in size, form and struc- ture to the spinal ganglion cells and to the dorsal cells of the spinal cord is another strong argument. It is important to notice that in both classes of cells bipolar and unipolar examples are found which give rise to coarse peripheral processes (dendrites) and slender cen- 638 Fournal of Comparative Neurology and Psychology. tral processes (axones). ‘I'he ground for this comparison is as perfect as can be desired so far as the animals thus far studied are concerned. I have already pointed out that the bundle has essentially the same course and position in the brains of all classes and I need only repeat that it is always situated in the dorsal zone of the brain. ‘There is not known in any part of the brain or spinal cord of any vertebrate a root bundle of any motor nerve which runs longitudinally in the dorsal half of the brain wall. The trochlearis is the only motor nerve which enters the dorsal zone of the brain in any part of its course, and that only to decussate. On the other hand, all the sensory nerves, their roots and longitudinal bundles in all vertebrates are strictly confined to the dorsal half of the neural tube. This is not a demonstration that this bundle is sensory in function, but it shows that the presupposition is that it should be sensory. We expect all primary bundles in the dorsal half of the brain wall to be sensory ; the burden of proof rests with thosé who would consider this bundle to be motor. The connections of the bundle in the brain require further study. Since all the evidence goes to show that the large processes of the cells are peripheral sensory fibers comparable to the peripheral pro- cesses of the spinal ganglion cells or the dorsal cells of the spinal cord, we should look for connections in the brain similar to those set up by the central processes of the spinal ganglion cells. I have shown (1900) that the giant cells in the cord of fishes are bipolar and that their axones run in the dorsal tracts, so that these neurones resemble embryonic spinal ganglion cells in everything but the posi- tion of their cell bodies. The cells in the mesencephalon of the toad are unipolar, bipolar or multipolar and possess true axones which enter the substance of the tectum. The same is probably true of the bipolar cells deseribed by Merkel and Krause and of those which I have seen in the rabbit. I consider it probable also that one of the several processes which Kdlliker saw on these cells may have been a true axone. The same may be said of the two cells figured by Van Gehuchten. Each has a slender ascending process which is probably the true axone. My line of reasoning here is that the existence of Jounston, The Radix Mesecenphalica Trigemim. 639 true axones with central distribution is what we should expect in view of the fact that the peripheral process goes into the sensory nerve and is therefore the dendrite, and in view of the disposition of the processes of the giant or dorsal cells of the spinal cord which is well understood. I am satisfied that such central axones entering the substance of the tectum are not uncommon in fishes and amphi- bians. If the bipolar (and multipolar) cells in mammals are inter- preted in the most simple and direct manner, they must be placed in the same category. But the majority of cells in mammals seem to have only a single process which sends collaterals into the motor nucleus of the trigeminus. Are these collaterals to be regarded as the central axone and the point of their origin as the T-division of the single process of the ganglion cell? I should hold this hypothesis in doubt until we have a very thorough knowledge of these neurones. It involves the supposition that these neurones begin as bipolar cells (already known), change into unipolar cells as do the spinal ganglion cells, and that the single process grows to an enormous length in ease of those neurones whose cell bodies lie in the tectum. On the other hand, Van Gehuchten figures the collaterals from the large processes at the level of the motor nucleus of the trigeminus in the trout, while from his figures and from my sections of Acipenser and selachians I believe that the cell bodies in the tectum bear true axones. The only other case in the vertebrate nervous system which comes to my mind in which axones or axonic collaterals are given off from the afferent or dendritic process of a neurone in addition to an axone arising from the cell body, is the case of the giant cells in the spinal cord of fishes (l. ¢., p. 376). These neurones possess ascending axones arising from the cell body and descending axones arising from the dendrite. (For other conditions occurring see the paper referred to.) The motor collaterals in the case of the trout (Van Gehuchten) and mouse (Cajal) would seem to correspond closely to the descend- ing axone arising from the dendrite of the giant cells in the cord of teleosts. This suggests the hypothesis that what may be called a descending or accessory axone has grown in importance in higher forms, while the true axone ending in the tectum has been reduced and possibly is absent from most of the cells in mammals. On this 640 ‘fournal of Comparative Neurology and Psychology. hypothesis the motor collaterals would be analogous (not homologous) to the motor collaterals in the spinal cord, and serve for direct reflexes between the sensory surfaces about the mouth and the muscles con- trolled by the trigeminus. ‘This is perhaps the most important function of the mesencephalic root bundle and is sufficient to account for the growing predominance of the motor collaterals. I have gone thus far with these speculations in order to show that very interesting problems lie here for other workers—to determine more completely the morphology of these neurones in all classes of vertebrates, the disposition of their central processes, their origin and development and especially the history of the several processes, the axones and the motor collaterals. The size of the mesencephalic root and its constancy in all classes of vertebrates are sufficient proot of its importance and of the value of further studies along the lines indicated. The peripheral course of the fibers is the one point of crucial importance in the question at issue. In showing that the mesen- cephalic root bundle in selachians, ganoids, urodeles, anura, reptiles, insectivores, rodents, ungulates, carnivores, and man leaves the brain in the sensory root, I believe that I have established the strongest probability that the bundle is sensory in function. In mammals and man the motor root runs over the surface of the trigeminal ganglion without interchange of fibers. Im many lower forms the motor root is almost as distinct. I do not know of any case in vertebrates in which motor fibers leave the brain in a sensory root and join the motor rami peripherally. Exact and conclusive proof of the sensory character of this bundle would be obtained by one or a combination of the following opera- tions: (a) Destruction of the cells of origin or cutting the mesen- cephalic root at any point central to the Gasserian ganglion without injury to the latter, followed by examination of the peripheral trunks of the trigeminus by the Marchi method to determine the distribution of the degenerated fibers. (b) Violent tearing out of one of the peripheral sensory rami in each of several animals with later study of the cells of origin to determine which ramus it is whose rupture causes destruction of the cells by retrograde degeneration. I have not found time or favorable conditions to attempt these operations. Jounston, The Radix Mesencephalica Trigemini. 641 One clinical case has come to my notice in the hterature which furnishes almost as clear results as could be expected from the second form of operation. This is the case of facial atrophy reported by Mendel (1888). The patient had suffered, twenty-five years before her death, from interstitial neuritis of the left trigeminus. Micro- scople examination of the trigeminus showed the end products of the neuritis in the root and all the rami, but the maxillary division was very much more seriously affected than any other part of the nerve. In some sections of the maxillary nerve there appeared only small islands of normal nerve fibers among the thick connective tissue septa. The cells of the trigeminal ganglion appeared entirely normal. In the brain the only abnormal changes found were the reduction in the size of the mesencephalic bundle of the trigeminus and reduction in the number of the cells in the substantia ferruginea on the left side as compared with the right. The facial nerve showed no change either centrally or peripherally. The motor nucleus of the tri- geminus was normal. The vesicular cells in the mesencephalon from which part of the mesencephalic bundle arises appeared normal. Mendel interpreted his results as evidence that the mesencephalic bundle had specific trophic functions. The simplest interpretation seems to me to be that the destruction of nerve fibers in the peripheral rami had resulted in retrograde degeneration of part of the fibers of the mesencephalic bundle and atrophy of their cells of origin. The Gasserian ganglion appeared normal, the motor nucleus appeared normal. Either there had been no degeneration in either of those or the atrophy of certain cells had been so complete in the course of twenty-five years that there were no results apparent to Mendel. The Gasserian ganglion, however, had not offered a block to the degeneration of fibers of the mesencephalic bundle passing through it, and the atrophy of cells in the comparatively small locus cceruleus could be detected by counting. The division of the nerve which was much the most deeply affected is wholly sensory. The obvious con- clusion is that the mesencephalic bundle is sensory in function and this harmonizes with all the considerations urged in the present paper. Although I began the study of this bundle four and a half years 642 “fournal of Comparative Neurology and Psychology. ago, fully expecting to find it a motor bundle as described by Cajal, Wallenberg and others, I have been unable to find any good evidence for this interpretation. I have pointed out (1905, 1906) that the existence of a general sensory (cutaneous) column extending up into the tectum mesen- cephali is a strong support for the theory of primitive functional divisions of the nervous system with which all readers of recent literature have become familiar. This general interpretation rests upon the demonstration of true axones of the ganglion cells ending in the tectum and is more secure than before so far as the primitive brain is concerned. The hypothesis suggested above implies that the tectum mesencephali in mammals has largely lost its primary general sensory function. ABBREVIATIONS USED IN ALL THE FIGURES. a,b,c, bun@les of the radix mesencephalica. Aq., Aqueduct. - br. conj., brachium conjunctivum. cblm., cerebellum. Cc. p., commissura posterior. c. r., corpus restiforme. dec. vel., decussatio veli. > ep., epiphysis. lem. lat., lemniscus lateralis. loc. ceer., locus coeruleus. nue. coch., nucleus cochlearis. nuc. hab., nucleus habenule. s., Sensory nucleus of trigeminus. s. 1., sulcus limitans. sg. or sub. gel., substantia gelatinosa. ter. gust., tertiary gustatory tract. tr. sp. V, tractus spinalis trigemini. V cebllr, cerebellar fibers of trigeminus. V g, ganglion trigemini. V mes, radix mesencephalica trigemini. V mot, motor root of trigeminus. V mot. cr., decussating fibers of motor trigeminus. V mot. nuc., motor nucleus of trigeminus. V sens, sensory root of trigeminus. VIII c, nervus cochlearis. VIII vest., nervus vestibularis. Ven, fourth ventricle. v, Ww, £. y, 2, bundles of the motor root of the trigeminus. Jounston, The Radix Mesencephalica Trigemint. 643 LIST OF PAPERS CITED. BECHTEREW. 1887. Ueber die Trigeminuswurzeln. Newrolog. Centralb., 1887. 1899. Die Leitungsbahnen im Gehirn und Riickenmark. Leipzig, 1899. BREGMAN. 1892. Ueber experimentelle aufsteigende Degeneration motorischer und sensibler Hirnnerven. Jahrbicher f. Psychiatrie, Vol. 11. CAJAL. 1896. Beitrag zum Studium der Medulla Oblongata, des Kleinhirns und des Ursprungs der Gehirnnerven. Leipzig, 1896. GOLGI. . 1893. Intorno all’ origine del quarto neryo cerebrale. Atti della reale Accad. dei Lincei, Ser. V, Vol. II. HELD. 1893. Beitrige zur feineren Anatomie des Kleinhirns und des Hirn- stammes. Archiv f. Anat. und Entwickl., 1893. HARRISON. 1901. Ueber die Histogenese des peripheren Nervensystems bei Salmo salar. Archiv f. mikrosk. Anat. und Entwickl., Vol. 57. JOHNSTON. 1900. The Giant Ganglion Cells of Catostomus and Coregonus. Journal of Comparative Neurology, Vol. 10. 1905. The Radix mesencephalica trigemini. The ganglion isthmi. Anat. Anzg., Vol. 27. 1906. The Nervous System of Vertebrates. P. Blakiston’s Sons & Oo., Philadelphia. KOLLIKER. 1896. Gewebelehre. 6te. Auflage, Vol. 2. LUGARO. 1894. Sull’ origine di aleuni nervi encefalici. Archivio di ottalmologia, Vol. 2. MENDEL. 1888. Zur Lehre von der Hemiatrophia facialis. Neurolog. Centralb., 1888. MEYNERT. 1871. The Brain of Mammals. In Stricker’s Handbuch der Lehre von den Geweben. American Translation. Wm. Wood & Co., New York, 1872. PONIATOWSKY. 1892. Ueber die Trigeminuswurzel im Gehirne des Menschen nebst einigen vergleichend-anatomischen Bemerkungen. Jahrbiicher f. Psychi- atrie, Vol. 11. RAMON. 1904. Origen del nervio masticador en las aves, reptiles y batracios. Trabajos del lab. d. invest. biol. de la Univ. d. Madrid, Vol. 3. 644 “fournal of Comparative Neurology and Psychology. SABIN. 1901. An Atlas of the Medulla and Midbrain. Friedenwald Company, Bal- timore, 1901. SARGENT. 1904. The Optic Reflex Apparatus of Vertebrates for Short-circuit Trans- mission of Motor Reflexes through Reissner’s Fiber ; its, Morph- ology, Ontogeny, Phylogeny and Function. Bul. Mus. Comp. Zool., Vol. 45. TELLO. 1909. Contribucion al conocimento del encefalo de Jos teleosteos. Los nucleos bulbares. Trabajos d. Lab. d. invest. biol. de la Univ. d. Madrid, Vol. T. VAN GEHUCHTEN. 1895. De Vorigine du pathétique et de la racine supérieure du trigumeau. Bull. de VAcad. Royale des Sciences de Belgique. 38 Serie, Vol. 29. 1895 a. Les cellules de Rohon dans la moelle epiniére et la moelle allongée de la truite (trutta fario). Bruxelles, 1895. 1897. Contribution a l’étude des cellules dorsales (Hinterzellen) de la moelle epiniére des vertebres inferieurs. Bruxelles, 1897. W ALLENBERG. 1904. Neue Untersuchungen iiber den Hirnstamm der Taube. Anat. Anzeiger, Vol. 25. 1904, Nachtrag zu meinem Artikel tiber die cerebrale Trigeminuswurzel der Vogel. Anat. Anzeiger, Vol. 25. A NEW ASSOCIATION FIBER TRACT IN THE CEREBRUM. Wirn Remarks oN THE Fiser Tract Dissection Meruop or STUDYING THE BRAIN. BY EK. J. CURRAN, Assistant in Anatomy, Medical School of Harvard University. WITH THREE PLATES. In April and May, 1908, while as an undergraduate pursuing the elective course in advanced anatomy in the Harvard Medical School under Professor Dwight, I was repeatedly and agreeably surprised with the ease and accuracy with which many of the fiber tracts of the brain, the nuclei, and the deep origins of some of the nerves could be dissected by methods slightiy modified from those employed in study- ing other parts of the body. Through the kindness of Professor Dwight and Dr..Warren, I was allowed to use a large number of brains for dissection, in the course of which new and impressive pictures of well-known structures were continually occurring. With such novelty and distinctness did these stand out that one not familiar with the original works of the early anatomists would almost be con- vineed that a new method of attacking the difficult problems of brain anatomy had been discovered. The attempt to follow fiber tracts by dissection seems to be the most natural method for investigation of the larger structures in the brain, and one that would be first thought of. So it is not surprising to learn that before the advent of the micro- scope, among the different methods of brain study attempted by our predecessors, fiber-tract dissection had a place, and many of the larger structures were displayed with considerable skill. With the introduction of the microscope, however, this method became neglected, and one cannot see in the text-books of to-day a single THE JOURNAL OF COMPARATIVE NEUROLOGY AND PSYCHOLOGY.—VOL. XIX, No; 6. ‘ 646 ‘fournal of Comparative Neurology and Psychology. example of a fiber tract dissection of any merit. Prof. J. B. John- ston,’ University of Minnesota, published in November, 1908, an excellent paper on brain dissection in which is mapped out an elaborate course which he follows in teaching and in which anatomy is taught with reference to the function. In this paper the following of the tracts of known function is emphasized and dissection of some of the large bundles in the cerebrum is described. Recently Dr. Jamie- son*® published two beautiful dissections, and made on appeal for this kind of work as an adjunet to the section studies now so much used. I learned from this publication that he had been dissecting brains for the past four years. Professor Hoeve also has been successfully doing fiber tract dissection." As to the reliability of the method, Dr. Jamieson says that he has never been obliged to abandon a tract as an artifact; and this has also been my experience. So reliable do I consider it that I have been using it as an adjunct in research work, and in the following pages I shall endeavor to convince the reader that not only may an adequate idea of well known structures be acquired, but that by the use of this method new tracts may be discovered which can finally be verified by the microscope, and more easily thus verified on account of their having first been dissected. Microscopie sections of such can be cut in the direction of the fibers, and this does not entail the work of following a large series of transverse or, as is more usually the case, oblique sections, which in an adult human brain must be so large and numerous that the danger of losing the continuity is always very great. I have used the dissection method in class this year, and am convinced that it makes the work of the teacher and that of the student much easier and, as compared with the section studies, renders demonstration of the essentials of brain anatomy perfectly simple. Mernops. To prepare the brain for dissection, any of the hardening fluids in general use will give good results; but I prefer a 10 per cent formalin solution injected through the carotid artery while the brain is an situ. *Anat. Record, Nov., 1908. “Jour. of Anat. and Phys., April, 1909. °Anat. Record, April, 1909, p. 247. Curran, 4 New Assoctation Fiber Tract. 647 If we use a stronger solution the fibers will be firmer for dissection, but the nuclei will become bleached, and there will be no definite line of demarcation to distinguish the gray from the white matter. The brain should remain im situ for a day or two; this is a sufficient length of time to allow of hardening to such a degree as will enable it to retain its shape. It should then be removed and placed carefully in a 10 per cent solution of formalin for three weeks or a month. I have also obtained good results when the brain was taken out shortly after death and preserved in the above solution without carotid injec- tion. In case of early removal, the fluid has time to penetrate -to the center before softening takes place. The only instrument necessary is a blunt pair of forceps which, when closed, are smooth and even at the edges where they meet. When such forceps are closed, they act as a blunt dissector. As well as separating fibers with the forceps thus arranged, it is sometimes necessary to lift bundles out of their places and to pull on them in order to ascertain the direction of their fibers. For this purpose the interlocking ridges should be fine and transverse in direction, to permit of a gentle but firm grip on small bundles of fibers without the danger of tearing or breaking them. Forceps with a few large sharp teeth at the end should not be used. When museum prepara- tions are being made, a very sharp knife will be of use to cut off ragged ends and to trim the dissection. 'Too much trimming some- times produces a dissection which, though finely finished, is less instructive than one which has not been trimmed at all. The latter shows the direction of the fibers, and any work that is to be done on it should not be such as would obscure this; otherwise the full value of the work would be lost. In developing skill in this method, it is necessary to have an abundance of material at one’s disposal. Almost every viewpoint revealed by dissection will be new, and the dissector will often be unwilling to proceed further with the dissection in hand, because in trying to show a deep set of structures he will have displayed a new and attractive view of another set, not originally intended, which perhaps lie more superficially than or close to the structures first intended to be shown. , Although it may be desirable to save this 648 ‘fournal of Comparative Neurology and Psychology. specimen, it must be destroyed in further proceeding to make the one first attempted. Consequently the best plan in such a case is to begin anew on another brain if there is plenty of material available. The fibers of the brain are so easily removed that one has to be con- tinually on guard against proceeding too quickly and attempting to show too much at once. It is better to show a little in each dissec- tion and to do a greater number to cover the ground. For this reason I recommend beginning at the cortex, studying the superficial fibers first, and making a series of dissections, each deeper than the preceding one, until the whole brain has been gone through. It is well to dwell upon some of the difficulties to be met with, and to say something of the methods which I use to overcome or diminish them. By careful observation during the dissection of over 200 brains, human and others, I was able to gather certain facts, and abstract with tolerable accuracy certain laws regarding the structure of the white matter of the brain, which have been of great use to me in my subsequent dissections. The chief difficulties are caused by inter- crossings with other tracts. These intercrossings take place at various angles, and on this depends the degree of difficulty. The nearer to a right-angled crossing it is, the more difficult it will be to dissect. If the fibers are almost parallel—or merely interlocking at an acute angle—the tract can easily be followed, but if the angle at which they cross each other is much greater than this, it will be impos- sible to dissect them unless one tract is markedly larger than the other, or its fibers are in isolated bundles. When one tract is much smaller than the other, it becomes lost at the intercrossing and only the larger one can be followed. The intercrossing and intermixing of fibers from different systems is, for obvious reasons, greater near the cortex than is to be found deeper, consequently the superficial dissections and the terminals in the cortex of deep dissections are less certainly made out than deeper tracts themselves. I have also observed that fibers arising in adjacent convolutions or adjacent nuclei and going to a distance to be distributed to adjacent areas soon gather together and continue in a compact bundle, which is in some cases thoroughly isolated from the surrounding structures until near the points of distribution, when they spread out and proceed to their Curran, 4 New Association Fiber Tract. 649 different places of termination. Moreover, the more distant the areas united by these fibers, the deeper are they from the surface ; and, conversely, the more superficial the fibers in the cortex are, the closer are the areas they unite. In other words, the long associating fibers le deeply and the sort lie superficially. This is an important and a helpful law; for it enables us, when we meet a tract lying deeply, to say with tolerable certainty that it unites distant parts of the cortex or distant nuclei. A tract is more easily dissected if fol- lowed from the place where it appears as a separate bundle to its distribution in the cortex. Owing to the tendency of fibers to form into bundles in all long tracts there is a line of cleavage which, if found and followed, assists us to overcome the intercrossing diffi- eulty which appears as we approach the cortex from within. . 5 2 (COG Roel 2h eke at 1 2 Gray 10 !9 2 1 1 1 || YOT 1....) Aare oe 1 10 4 Mofale=:| 9 301.30 30 30 | cee 30 30 30 30 30) 30 30 30 19 For a third test of thirty trials YGS 2, RT 1, and RV were substituted for BV, GB, and Gray 10. Both the animals selected YGS 2 twice, RT 1 once and the food color, OS 1, twenty-seven times. 672 ‘fournal of Comparative Neurology and Psychology. It seems evident from a comparison of Table 15, with the preced- ing tables and with the following one, that light colors were most difficult for the animals to discriminate. If such is the case, the rather long time required to learn the series which matched Gray 5, is partly accounted for. Group 5 was the first one we tried, and some time was required for the animals to learn to operate the mech- anism. As this was soon learned, however, our tables show that dis- criminations were more readily made among dark than among light colors. TABLE 16. TABLE 17. VS 1 in Group 20. GBS 2 in Group 30. Color. Raccoon No. 2. Raccoon No. 3. | Color. Raccoon No. 2. | Raccoon No. 3. | Wis) tle scocdl JO Pid 247 PA) Xe), P48) AVASiS wine 3 1 it 3 Bis leer. 8 2 Z 1 || GBS 2 lige ePapf 743) 29 14 BAe se 2 ORS 2 Smeal: 5 Wi ees 2 1 2 JRAVAS) Bs 5 1 5 ROS 2 Hall 22, Gray 30.. 1 2 RVS 1 4 1 1 4 1 ih —- — — — ~ = = SS — — — || Total...| 30 30 30 30 30 ARMS So) SO) S80) std) ait) ail) axl) | | | It seemed to us that No. 3 made most of his errors when he seemed to be looking at the colors with his left eye. Since his twenty-nine correct choices in the first thirty trials of the above series showed that he could discriminate the food-color, we compelled him to walk along the row of glasses from left to right in the second series of thirty trials. His correct choices were at once reduced to fourteen. Our suspicion that his left eye was defective arose from the fact that the animal, unless prevented from doing so, invaribly went to the right end of the row of glasses, then along it to the left with his right eye thus nearest to the row. The above reduction in the number of his right choices, and our record of his earlier errors, indicate that his vision with the left eye was poor as compared with the right eye, and as compared with the vision of Raccoon No. 2. The records above do not give all the training in discrimination of colored papers which was demanded of the raccoons, for some of the control tests required many trials. The tables do show, however, that Cote anp Lone, Visual Discrimination in Raccoons. 673 nine different colors and one gray have been selected by the raccoons from colors which were equally bright for the human eye. In all, thirty-nine different colors were used, and from these the ten food- colors had to be chosen. Up to this point, however, we cannot be sure that the selections were not made by the sense of smell. The Possibility of Selection by Odor Differences. We have al- ready stated that the raccoons never swerved toward the food-glass from a distance of one or two feet. In fact, they seemed to be unable to distinguish the food-glass from the others at a distance of more than four or five inches. Thus the animal’s nose was always very close to the glasses and we were always confronted with the possibility that a keen sensitiveness to odors might account for the behavior we observed. The discrimination by smell might have come about in any one of three ways. (1) The animal might have detected the odor of food when near the glass which contained food. (2) He might, as it were, follow his own trail, and this might be done in either of two ways: (a) since the animal ate the food from his paws and also laid hold of the food-glass more than the others he might soon get an odor of food from the outside of the glass; (b) he might in a similar way detect the native odor of his paws as stronger on the food-glass than on the others, due to repeated pulling at the glass even though he did not communicate the odor of food to it. The last supposition while hardly probable is, nevertheless, possible. (3) The animal might be able to discriminate between the odors of the pig- ments of the colored papers. With closed glasses the first possibility is easy to deal with. We used bread as food and often allowed it to become so dry that it had hardly any odor for us. In addition to this we put food in all the glasses, usually the same amount that we put in the food-glass, some- times more than that glass contained. From the first day’s work with closed vessels our rule was to put food in all of them during at least half of the trials. In a large part of the experiments, we put food in all at the beginning of the day’s work. During a smaller portion of the time all the glasses contained food during the last thirty or sixty trials. In no case did this seem to modify either the animal’s behavior or his record of discriminations. 674 ‘fournal of Comparative Neurology and Psychology. Was the animal following his own trail? We first met this pos- sible difficulty by providing for each set of colored glasses six food- glasses all of the same color. After five trials with one glass we re- moved it from the room and put a second one in its place. After five or six trials more, a third food-glass was used and so on. Had the animal been selecting the food-glass by any odor attached to it (except that of its pigment) this exchange of food-glasses should have confused him. It did not do so, and in the later experiments we changed the food-glass only every tenth trial. Again, if the animal were selecting the food-color by any odor except that of the pigment, he should learn to select a food-glass from a group, all of which were covered with the same color. On July 15th No. 2 was given a set of glasses covered with OYS 1, and he pulled at all of them. At the end of half an hour he had so scratched the yellow papers that the scratched places would soon have served as distinguishing marks. The result was the same with No. 3. He was tested with the six yellow glasses on the same day. We repeated such tests with the food-color of several later groups. We also filled the glass holder with tumblers, all of which were covered with Gray 5, and gave the animals opportunity for olfactory discrimination. In the first thirty trials Raccoon No. 2 made seventy-eight pulls at the glasses, eleven of which were at the food-glass, but in six of these eleven cases the food-glass was the first one to which the animal came. He passed by one of the glasses only nine times in the seventy-eight attempts. We restate these facts in the same order in the first line of Table 18. These records show that the animals soon ceased to try to dis- tinguish one glass from another and pulled blindly at almost all of them. By going but part way down the row and then returning to the point of beginning, it was possible for the animals to pull re- peatedly at several of the glasses without coming to the one which contained food, and thus to make such records as those of the last series in each table. The failure of the animals in this experiment gives further evidence that they were not using smell in the color tests. On continuing these experiments with No. 2, though he was very hungry, the large number of selections which yielded no food seemed, to a great extent, to inhibit the impulse to pull at the glasses. CoLe anv Lone, Visual Discrimination in Raccoons. 675 We next varied the experiment by rubbing apple on the inside of the food-glass. Both the raccoons then very promptly learned to select it, but now we could often make out distinct sniffing and the animals held the nose very close to the top of the glass as they had not done hefore. This is doubtless the best evidence we have that odor was not a guide in the color tests. If the animal had to direct his nose to the top of the glasses in order to detect the odor of apple in a glass whose inner surface had been thoroughly rubbed with it, then any odor less strong TABLE 18. Gray 5 in a Group of Gray 5. Raccoon No. 2. | | . | No. of Series of | No: of Pulls No. of Pulls at | Food Glass First | Passed by With- 30 Trials each. | o Wise Food-Glass. | One Reached. out Pulling. 1 | 78 | 1 6 9 2 | 63 16 6 25 3 | 60 | 18 9 22 4 | 141 6 4 8 | TABLE 19. Gray 5 in a Group of Gray 5. Raccoon No. 3. No. of Series of | N f Pulls | No. of Pulls at Food Glass First | Passed by With- 30 Trials each. OOS Food-Glass. One Reached. out Pulling. = | : ind es! Z uy 1 | ae | 12 7 | 5 2 Hela 9 7 i 3 104 5) 5 1 4 199 2 2 0 than this could have had no effect or else it surely would have elicited the same behavior. The same reaction was elicited by the use of meat and cake as food. The only odor not excluded by these experiments is that possibly due to the different pigments of the col- ored papers. True, the papers had been kept for some months in the same drawer, and the glasses were close together in the row, yet this only makes discrimination by the pigment odor improbable, not impossible. 676 ‘fournal of Comparative Neurology and Psychology. In order to exclude this type of discrimination, we first placed the colored papers inside the glasses. We believed that this would so diminish the pigment odor, if it did exist, as to put it below the threshold of the animal. Or, if not, that it would so reduce the pigment odor that it would be completely eclipsed by the food odor when food was placed in all the glasses. In the light of the experi- ment above, in which apple was used as a stimulus, it is practically certain that, with the colored papers inside the glasses, no odor of the pigment could be sensed by the raccoons. The effect, for the human eye, of putting the paper inside the glasses was to make only a vertical strip of color visible in each glass. While the intensity of the colors seemed much reduced, the reduction seemed equal for all the papers, so that differences in brightness seemed no more pronounced than before. The test proved to be a very difficult one. We had taught the animals to select VBT 2 as a food-color when the papers were on the outside of the glasses. We continued to use it as a food-color after putting them within the tumblers. The results of the experiment are shown in Tables 20 and 21. TABLE 20. VBT 2 in Group 5.—Papers inside the Glasses. Raccoon No. 2. 7 E Color. | First Day. Second Day. Third Day. Fourth Day. - | — | Vesa al 1 ¢ 3 i ALG ab vay) 1h RS (¢ 4 2 VBI os sees Oy Gy ley) es lik es Pah) ley) al) Pal || GI a) BE Ae, ONS eee ct ts 10 4 4 TEN teh GE tay 229} 1 DRA OM Da OR aso es PS aa Ta Ne i ee aye NI Le (6) ee) euays 1 2 1 1 VD eee eae 233 UN lcm) steve 2+ a 6 4 1 4 i 2 246 (ChenY DY cocgooqas 5. 68 VSP Teo ore. 4 Ponane Di Del ae) Motaleneence 5 S000 30 30 30 30 30 | 30 30 30 30 30 | 30 30 30 30 * Our supply of OYS 1 was exhausted and on this day we used a second glass covered with Gray 5 to which the responses were 5, 5 and 4, respectively. Though only the vertical strip of color was visible, the raccoons succeeded in selecting the food-glass twenty-five and twenty-six times, respectively, on the fourth day of training. CoLeE anpD Lone, Visual Discrimination in Raccoons. 677 TABLE 21. VBT 2 in Group 5.—Papers inside the Glasses. Raccoon No. 3. Color. | First Day. | Second Day. Third Day. Fourth Day. AK 3 £3 5 6 2g) sien ono Vib? 7 QU Oki. 14 16 16 13 13 24 24 25 OMNIS Mi Neco 15) 4 1 1 6 1 1 ie aa HO Mele Se es os 6 4 66 1 3 Po | 6 Wap 5 a BF D Siac 2 2 3) 2 Gray LS carey tirade 4 O38) 2 -@ eG ee, 2 2 INOUE ooo boob 30 | 30 30 30 30 30 10 30 30 | 25* 30 30 30 * Due to error in counting trials. As a further means of excluding the pigment odor, we coated the paper-covered glasses of Group 20 with shellac, with the result shown in Table 22. TABLE 22. VS 1 in Group 20.—All Glasses Covered with Shellac. Color. Raccoon No. 2. Raccoon No. 3. BAS ate apes se. 20 cicttesecs 5 3 3 1 1 2 Beary eae | i 4 1 1 1 ] 1 2 ROSIE Boredetstsie cicvestessie'| Oo S524 2 26 25 29 30 £28 \WCLEAY Seog oe oP | 4 3 1 1 EL OSO2 rats oie kod eleven ave 9 1 1 1 RAV AS ices recteee ese aes 5 1 1 J 1 Bocas t 30 30-30" 730) 30 30 30 630 ~— 30 In the two tests recorded above the two animals should have failed to discriminate if they were selecting the food-glass by the odor of the pigment. In case of discrimination by odor the raccoons should choose cor- rectly in the dark. They were tried in the dark on Group 30 which, it will be remembered, they had learned very readily by daylight. Each animal was given thirty trials. Raccoon No. 2 selected the food-glass six times, No. 3 selected it five times. Apparently these choices were made only when the food-glass was the first one to which they came. In all other cases they pulled at every glass in order until 678 “fournal of Comparative Neurology and Psychology. the food-glass was reached. In cases of turning back before food was obtained they pulled at some of the glasses twice in each trial. Except in the first fifteen trials with Raccoon No. 2 there was food in only one glass, GBS 2. It seems fair to conclude, from these experiments, that the animals were not making their selections of the food vessel by means of the odor of the food, of the pigment, nor of their own paws. Evidently the discrimination was a matter of vision or of some sense unknown to us. Visual Discrimination. Since the possibility of discrimination by means of the sense of smell has been eliminated, we must inquire by what visual criteria the different colored papers were distinguished. Watson has pointed out that “the surfaces of the papers differ greatly owing to accidents in manufacture, dyeing, ironing, etc.,” and that there is difficulty in ‘pasting them upon surfaces so that slight differ- ences do not appear.*° Discrimination by means of these criteria must-be guarded against, and in the case of the raccoons it was done as follows. (1) Not one food-glass alone was used but a half dozen different ones of the same color. (2) As already stated (p. 674), at the end of a test, we filled the glass holder with six glasses of the same color to see whether the animals could pick out the single one in which food was placed. They simply pulled at every glass. Besides testing the animals, in this way, on several colored papers, they were also tried on gray and white. It seems apparent, there- fore, that they were not being guided in their choices by any secondary criteria which the papers may have presented. The possibilities, then, seemed limited to two. Either the animals discriminated between the several glasses (a) by means of their brightness differ- ences, or (b) by means of their differences in color. (a) As we have stated, the colors were selected so as to be of equal and very nearly equal brightness for the human eye, and as wide a range of brightnesses was used as the ninety colored papers would furnish. The value of our tests rests on the assumption that colors of equal brightness for the human eye may be somewhere near Watson, J. B. Some experiments bearing upon color vision in monkeys. Jour. Comn. Neur. and Psych., vol. 19, pp. 3-4. 1909. Cote and Lone, Visual Discrimination in Raccoons. 679 equal in brightness for the raccoon. We now have to ask whether there is any evidence to justify this assumption. After training our animals to select Gray 5 from the colored papers of Group 5 we substituted for the latter Grays 3, 4, 6, 7 and 8 of the Hering series. Raccoon No. 2 was given only sixty trials with these papers. The results appear in Table 23. TABLE 23. Gray 5 in Grays 3-8 inclusive. Gray. Raccoon No. 2. Raccoon No, 3. 3 3 1 sil 1 a 17 1 5 Omelg 2S 2 LON ee 6 yy 5 il 1 a 3 2 if 8 DP) 1 2, 1 Total 30 30 30) 30) 30) 30 It appears from this table that Raccoon No. 2 confused Gray 4 with Gray 5 in the first thirty trials and showed, in the second thirty, that he was learning to discriminate Gray 5. This animal was not hungry, as he had just been given ninety trials on Grays 5, 10, 15, 20, 25, and 30, which series presented no difficulty of discrimina- tion. ‘Table 23 shows that Raccoon No. 3 made very few mistakes. It seems evident, when these records are compared with those of the color-discrimination tables, that the brightness differences between the colors of any group were less for the raccoon than the brightness differences between any two of these consecutively numbered gray papers. This agrees with human vision, for Professor Titch- ener wrote us as follows: “We tried to arrange the four equivalent disks in a scale or order of apparent brightness, by the eye alone; and we got into great difficulties at once. I do not think that any two of us would have taken the same arrangement except by chance. We all felt pretty sure that VT2 was the lightest of the five, though it is true that suggestion (from experiments I have mentioned) may have played a part here.” 680 Fournal of Comparative Neurology and Psychology. If the raccoons were discriminating the colored papers by bright- ness differences alone these differences must have been very near their difference threshold for brightness, for they learned very slowly to make the discriminations. But the brightness difference between some pairs of these papers was greater than that between other pairs, since some of the colors were inexact matches for the gray. If, then, we were not below their difference threshold with the exact matches, but only near that threshold, the inexact matches should not have been chosen at all, or at least, fewer mistakes should have been made on inexact matches. As a fact almost an equal number was made on each class, as shown by the following table. TABLE 24. Ragone ae 2 3 Ya rc hme re fa SE allo) 3 2 3 | 2) 8 | Total average. Average number of | | errors on exact | | | | MaAtCHeS. sr. - | 9.6|20.3/15.3) 6.3 | 41|16.5| 26) 5 can 3 | 5.3 ee eco lp() |p Ab to) | 12.04 Average number of| | | errors on inexact) | | | matches........ | 10. WSS Se ls eat I Gol, | abre |) 2AS5"| LON) 455) 2655815 625) 282 5 4 | 12.76 | | | | a“ | -_— = — ~ The total averages for exact and inexact matches are practically the same. ‘This seems to be excellent evidence that we were below the animal’s difference threshold for brightness. If such was the case, they must have been discriminating by color differences alone. In this average are included all the colors, both dark and light, so that the figures apply to the whole range of papers used. When Yerkes tested the dancing mouse by means of colored papers the records showed at once that the dancer cannot distinguish green from blue, nor violet from red.?!_ As the raccoons have given no case of failure to discriminate colored papers, it would seem from this comparison that their vision is more nearly like that of human beings than like that of the dancing mouse, or the common mouse.?? Table 25 gives the average number of the raccoons’ errors on each wrong color and the number of their errors on each gray. Of course, records for Gray 5 which were made after we had used it “YERKES, R. M. The dancing mouse, pp. 147 and 149. 1907. =WAuGH, K. T. The role of vision in the mental life of the mouse. Doctor’s thesis, Harvard University. 1907. (Unpublished.) CoLE and Lone, Visual Discrimination in Raccoons. 681 The records for other grays are those obtained after the animals had been brought to avoid Gray 5, by having been trained on VBT 2 as a food-glass in the group to which Gray 5 belonged. as a food-glass do not appear in the table. TABLE 25. Raccoon No. | 2 2 3 OF Bi 3 |Total 2 | 3 HN" 83 2 3 Zz 3 | Total Av. number of se- | | lections of each | | wrong color... ..| LOMPBLGE Zale smea|| a7 44 | 17 | 30.6 |1388.5]| 2.2 | Path MPa MN abate dye ZheBr Ma aie cai PY Ae) Number of selec- | | tions of gray... 9 | 24. 15. 6 | 49 | 16/16. |135. || 0 | O 220) 3 73 ik 2 10. | | | | | If by the use of the flicker method we did not secure colors which were equally bright, for the animals, with the gray which was used as a standard, they should not have made wrong choices by selecting the gray. Or, at most, they should have made very few such errors. The first half of Table 25 shows almost as many mistakes on the gray as on any single wrong color, though some hundreds of trials which should have discouraged the tendeney to make these mistakes on gray, preceded those which are tabulated (see pp. 667 and 668). In the latter half of the table the animals seem to have profited by the fact that they had never found food in the gray. This might be because of a difference in brightness, or because, after long practice with food in colored glasses, the animals began to pay more attention to the colored papers and less to the grays. Comparison of the tables, then, reduces the question to this. If the papers were not equally bright for the animals they should have made fewer mistakes than they did. This is shown by their records on the consecutive grays. On the other hand, if the papers were equally bright for the animals and yet they were color-blind, the excellent records they made on the food-glasses finds no explanation. (b) So far, then, as evidence can be gained by the use of reflected light, we think it probable that the raccoon can be made to discrimi- nate objects by their color alone. We do not think that in their native state they are often called upon to make pure color discriminations. Samojloff and Pheophilaktowa concluded, only, that the dog can 682 “fournal of Comparative Neurology and Psychology. make color discriminations, not that he does so without training.** Our raccoons also required many trials before they made fairly good records with the first group of papers which we used. We state the above conclusion tentatively only because the use of reflected light is possibly inadequate for the solution of this prob- lem. We regard our work as preliminary, and we hope to complete the investigation by the use of methods, which had not been described until our experiments were almost completed. Summary. (1) Open feeding vessels were unsatisfactory in test- ing visual discrimination in raccoons because of an instinctive tend- ency they have to explore such vessels by touch. (2) In the cases of discrimination described in this paper the raccoons seemed not to make a selection with the eyes at a distance of more than four or five inches from the glasses. (3) So far as evidence can be gained by our use of the method of reflected light, it indicates that the raccoon is able to discriminate by differences of color. (4) Many trials were required with the first group of colors used before a high per cent of right choices was made. (5) The animals learned most quickly to select the food-glass when dark groups of colored papers were used. Memory for Visual Differences. Thirty-four days after the rac- coons had learned to select RVT 1 as a food-color, and without train- ing during this period, they were again tried with food in this glass. Raccoon No. 2 gave no evidence of having been trained on this group of papers and selected RVT 1 9, 16, 23, and 29 times respectively in each of four series of thirty trials. His learning series for RVT 1 was 14, 29, 24, and 30 correct choices in each series of thirty trials. Raccoon No. 3 selected the RVT 1 27 times out of the first *2S,amMosLorr, A., und PHEoPHILAKTOWA, A. Ueber die Farbenwahrnehmung beim Hunde. Zent. f. Physiol., Bd. 21, S. 133. “YprKes, R. M. The dancing mouse, p. 152 ff. 1907. Methods of studying color vision in animals. Science, N. S., vol. 29, p. 482. 1909. Also WATSON, J. B. Some experiments bearing upon color vision in monkeys. Jour. Comp. Neur. and Psych., vol. 19, p. 1, 1909. Cote AnD Lone, Visual Discrimination in Raccoons. 683 thirty trials and was perfect in the second thirty. He thus selected the food-glass three times more in each of the two re-learning series than in the learning series. One hundred and eight days later, and meantime without practice, he selected this color 15, 25, and 29 times in each series. ‘The utmost that can be said, therefore, is that in the case of the two animals, relearning was a little more rapidly accomplished than learning. So, also, must the fact of much more rapid learning in the later experiments be given some weight after due allowance is made for the darker colors being easier for the ani- mals to discriminate. Group 2, for example, which was not dark, was evidently learned much more quickly than it would have been without previous training. ‘The animals’ behavior is thus greatly modified by past experience, but the effect of having learned to discriminate objects by means of a specific brightness or color differ- ence, at any rate if that difference be shght, does not last more than a few days. This might be expected since in its native state the raccoon is probably not called upon either to detect or remember such slight differences as were used in these experiments. Finally, it may be remarked that the animal which gave the better “memory” record above for discrimination of colored papers was the one which gave evidence of superior motor memory for fastenings. This animal also required the greatest number of trials in both cases for the original formation of the associations. So much evidence for a mechanical law of association in animal psychology. A STATESTICAL STUDY OF THE MEDULLATED NERVE FIBERS INNERVATING THE LEGS OF THE LEOPARD FROG, RANA PIPIENS, AFTER UNILATERAL SECTION OF THE VENTRAL ROOTS. BY ELIZABETH HOPKINS DUNN. From the Anatomical Laboratory of the University of Chicago. WitH ONE FIGURE, CONTENTS. PAGE The material studied and the methods employed ....................... 685 Distribution of the nerve branches supplying the leg of the leopard frog anadevots fro oe Hy im: PAL CULAN \ yo sels couse cene al evel oteletotets ) ini primanry, branches! tomhiche. passerine 242 241 213 8 ,|| In primary ‘branches'to thigh:.-..4-%. - <2: 287 276 233 13° | ta primary branches to thighs... 4... sot 269 5’ | In primary branches to thigh:..2:......... | 289 16 |) Above branchesito'shanks.22-4.55.52s een. | 239 195 165 i eAbovelbranches tomshankeseeee rae ee 245 10| | Below branches to shank...--5......55.4.- bg? 143 106 i \eBelows branches toishamlcarsss eee eee: kOe 16s lnebranchesstorsh sank aerate 17/33 179 128 Gi si branchesitons hia eens ier ener ee 20 197 145 11) eine boranchesstorshanl< selenite ete Vere alkeyoyees =| | 4 intbranchesavonsh anlar eee eee ae renene eee | 224 | could be distinguished from the cutaneous branches containing affer- ent medullated fibers by the greater number of nerve fibers of a uniform size in the muscular branches. In the present study, after Dunn, Medullated Nerve Fibers. 715 eliminating the efferent fibers, it appears that among the afferent nerve fibers the greatest number of nerve fibers of uniform size pass to the muscles. All the findings in this study regarding the size of the medullated nerve fibers point to a fundamental uniformity underlying their dis- tribution. The probabilities seem to lie between two alternatives. Hither, first, a difference of function must be correlated with size; this might be possible among the cutaneous nerve fibers where a greater variation in size is coexistent with the need for transmission of various cutaneous sensations. Or, second, size may depend upon the amount of tissue to be innervated by the single fiber. Some pre- liminary measurements still unpublished seem to show that in the case of the medullated efferent fibers a direct relation exists between the diameter of the muscle fiber and that of the nerve fiber by which it is innervated. Herrick, 1902, in discussing the significance of the size of nerve fibers in fishes, states, page 333, “that each functional system of peripheral nerves has tolerably definite fiber characteristics, the basis for which is unknown; that these characteristics are by no means invariable, but that fibers of a given system may show considerable differences in caliber and medullation in a single animal, and that some of these differences, at least, can be correlated with the degree of functional development of the peripheral end-organ.” Johnston, 1908, has a very suggestive paper on the significance of the caliber of the parts of the neurone in which he points out the extreme differences in size in the non-medullated nerve fibers of the lamprey. A further discussion of the literature cannot be made at this time, but it is hoped that an accumulation of findings may make possible a later discussion of the caliber of the medullated nerve fiber in the leopard frog. Tt has been shown, Dunn, 1902, that the largest medullated nerve fibers passing to the leg of the frog are not found iat levels below the thigh but appear in the primary branches to the thigh. This finding discredited the theory of Schwalbe that the largest nerve fibers run the longest distance. 716 ‘fournal of Comparative Neurology and Psychology. In the course of the present investigation it was found possible to corroborate the previous finding by a study of the conditions in the unoperated leg of frog E, and to determine the condition existing among the afferent nerve fibers in the left or operated leg of frog E. Table XVII gives the measurements from frog E for the afferent fibers, and for the combined afferent and efferent fibers, and for the combined nerve fibers from frog IIB. Because of the absence of efferent nerve fibers from the operated leg of frog E the total number of nerve fibers is decreased at each level. To make possible com- parison with the unoperated leg and with frog IIB both the full number and the relative number were included, the latter in each instance standing alone below the former. The findings for frog IIB are confirmed by the measurements for the unoperated leg of frog E. The average area of the 22 largest nerve fibers at the level above the branches to the thigh is 257 square micra. That of the corresponding 16 nerve fibers at the level below the branches is 194 square micra, while the corresponding 8 nerve fibers in the branches have an average area of 276 square micra. Similarly the largest nerve fibers at the entrance to the shank are found in the branches to the shank and not in the main trunks below the branches. In the operated leg there is shown a similar distribution of the largest afferent nerve fibers at each level to the tissues of the adjacent seoment. Table XVII shows that larger medullated nerve fibers, both muscular and cutaneous, are found in the thigh than in the shank. RELATION oF THE ‘AREA OF THE AxIS CYLINDER TO THE AREA OF THE MrpuLLARY SHEATH IN Cross SECTIONS OF THE LARGEST MeEpDULLATED NERVE FIBERS IN THE OPERATED LEG oF Froe E. A swollen condition of the individual nerve fibers in the operated leg is shown by the uniformly larger size of the nerve fibers on this side when compared with corresponding nerve fibers on the unoper- ated side. As this condition prevails throughout the leg it does not vitiate the results of a comparison of the caliber of the nerve fibers at various levels of the operated leg. It does unfortunately prevent Dunn, Medullated Nerve Fibers. 717 that comparison of the nerve fibers of the operated leg with those of the unoperated leg, by which we might ascertain the relative sizes of the largest afferent and efferent nerve fibers. TABLE XVIII. Showing the ratio of the average areas in square micra of the ten largest medullated nerve fibers to the average areas of their axis cylinders at. various levels in the operated le leg of frog E Axones. Fibers. Ratio A:B. Ae: B. N. ischiadicus above branches to thigh... . 143.99 294.98 1: 2.50 N. ischiadicus below branches to thigh....| 158.37 Zola > AAO, Imvbranchesitotnighe we... , ites Tt “ oe af, aie eee ee 2 “ . . . =e 2.8 oN no oe ++ * ee Ve vee at a-® a a ie 4) f te tists ° > ese FS Write teatntifetee err ehcreacate sa ie ; 8 TA Peet 325) . . ‘3 ia He é ths fie . ‘oe ~~“. ee . * stare Bate