Introduction. — Some structural differences in the auditory organs. — Auditory responses in mammals. — (a) The dog: Kalischer's ex- periments.— Rothman's experiments. — Further experiments by Kalischer. — Johnson's experiments. — Experiments made by the Pawlow method. — Experiments of Syzmanski on localization. — (6) The cat: response of cats to tones, noises, and to articu- late sounds. — (c) Raccoons: reactions of raccoons to tones and articulate sounds. — Tests upon mice and rats. — Incidental tests upon other mammals: (a) Monkeys. (&) Horses, (c) Bats. — Auditory response in birds. — In pigeons and parrots. — Auditory response in amphibia. — Frogs. — Auditory response in fishes. — Some investigators who report lack of auditory sensitivity in fishes. — Investigators reporting sensitivity to auditory stimuli. — Intensity of sound stimuli in water. — Summary. — The lateral line organs.
Introduction. — In the preceding chapter we saw that investigators were not wholly in accord as regards the results obtained from testing the visual responses of ani- mals. In view of the lack of any standardized method for testing the responses of animals to auditory stimuli, we shall find a still greater lack of harmony in the field of audition. The auditory work has been largely done by physiologists and zoologists for the purpose of throwing light either upon the localization of function or upon com- parative anatomy. Little attention has been given to the strictly behavior side of audition — i.e., to the influence of audition upon habit formation, the instincts released by auditory stimuli, etc. The experiments have been con- ducted in such a way that we can do little more than present the results of the different investigators. In the discus- sion of audition we shall include some experiments which have been conducted for the purpose of determining the function of the lateral line organs. The functions of the semicircular canals and vestibular portions of the ear THE AUDITORY ORGAN 371 should naturally appear in this chapter, but since the physiologists have from time immemorial included these structures in their treatises, we shall not deal with them in this sketch.^ On p. 36 we indicated some of the problems which belong in a study of audition in animals. The pres- ent summary shows how far from complete are our actual data.
Some structural differences in the auditory organs. — The mammalian ear is so well known that it is not necessary to enter into a description of its parts. Since Helmholtz we have come to look upon the organ of Corti as the structure in the inner ear w^hich is essential at least to all of our finer reactions to sound. In more recent times there has appeared the tendency to regard the tectorial membrane as a part of the essential structure. Helmholtz at first supposed that the analyzing mechanism of the inner ear w^as the arches of Corti. When C. Haase called his attention to the fact that birds do not possess the arches of Corti, Helmholtz then assumed that the basilar mem- brane fibers w^ere responsible for tonal analysis. Birds and all mammals possess a basilar membrane. In birds, however, this membrane is shorter than in mammals. And although there is no real factual justification it was as- sumed by the early zoologists that the range of pitch sensi- tivity in birds w-as very limited. In the more highly de- veloped fishes, although there is no external ear, there is a well marked internal ear consisting of an utriculus and its semicircular canals, and of a sacculus with its appended lagena.. We are particularly interested in the lagena, since it is the forerunner of the cochlea of birds and mammals. In fish (squeteague) the lagena exists as a small triangular pocket. A branch of the eighth nerve is distributed to this region, ending in a sensory patch called the papilla acustica basilaris. This area contains a single otolith, the astericus. The sacculus, while communicating freely with the lagena, does not communicate (in most forms) with the utriculus and its appended semicircular canals (structures which are in some way necessary to the maintenance of equilibrium __ ^ See the excellent summary of Wilson and Pike.
372 AUDITORY AND RELATED FUNCTIONS and muscular tonicity). The sacculus, like the lagena, contains a single sensory patch, the macula acustica sacculi, and a single large otolith, the sagitta. It is supposed by Parker and others that both the sacculus and the lagena function in the reception of sound stimuli, and that these organs play no part in the maintenance of equilibrium and muscular tonus. Parker further supposes that the sacdSS ms Fig. 62. Membranous Auditory Organ of Fish {Perca fluviatilis) Seen from Within * 1, Lagena cochleae; pi, Papilla acustica lagense; o, o, o, Otoliths of the utriculus, sacculus, and lagena; s, Sacculus; u, Utriculus; ca, Canalis superior; ce, Canalis lateralis; cp, Canalis posterior.
cuius is the most important organ in the reception of sound, since when the sagitta on each side is pinned down, the fish are only feebly sensitive to sound. How the sagittae function " in the reception of sound is not known with certainty; but since in the squeteague they have a specific gravity of 2.84 and that of the whole head is about 1.8, it is quite probable that when sound vibrations influence the normal fish they induce the relatively lighter parts of the head, including the macula acustica sacculi, to vibrate * Figs. 62, 63, and 64 are taken from Weidersheim, Vergleichende Anato- mie der WirtelUere (Jena: Gustav Fischer, 1906). He in turn copies them from Retzius.
THE AUDITORY ORGAN against the relatively heavier otolith; in other words, the otolith is a relatively stable body against which the audi- tory hairs of the macula acustica sacculi may strike." According to such a view the fish could react to sound only on the basis of difference in intensity — there is no differentiating or selecting mechanism. In amphibia there are both a middle ear and an internal ear. The tympanic membrane lies imbedded in the dermal surface. This mem- brane is connected with the opening in the inner ear Fig. 63. Membranous Auditory Organ of a Bird {Turdus musi- cus) Seen from Within, Eight Side 1, Lagena cochleae; pi, Papilla acustica lagense; u, Utriculus; ca, Canalis superior; ce, Canalis lateralis; cp, Canalis posterior.
(Fenestra vestibuli) by a straight slender bone (there are really two bones — stapes and columella) called the colu- mella auris. The sacculus, with its appended lagena, is far more highly developed in the frog than in the fish. We have in the amphibia, at least, rudimentary basilar and tectorial membranes. In birds and in reptiles there is still further development in these structures. We find here first a true cochlea — with a scala tympana and a scala vestibuli clearly marked. The basilar membrane is in- creased in length and has become more highly differenti- ated. Figs. 62, 63, and 64 show the progressive changes in the auditory portions of the ear as we pass from fish, through the birds^ to mammals.
374 AUDITORY AND RELATED FUNCTIONS "Ca Fig. 64. Membranous Auditory Organ of a Mammal (Rabbit) Seen from Within, Right Side mb, Membrana basilaris; aa, Ampulla superior; ac, N. acusticus; ae, Ampulla lateralis; ca, Canalis superior; cp, Canalis posterior; dej Ductus endolymphaticus; rb, Ramulus basilaris.
Auditory Responses in Mammals (a) Dog: Kalischer's experiments. — Kalischer reported in 1907 some experiments upon dogs made for the purpose of testing the relation between the temporal cortex and the reception of tone stimuli. He wished particularly to test the validity of Munk's conclusion that the center of tone lies in the temporal lobe and that the perception of high tones is taken care of by the anterior portion of the center, while that of the deeper tone depends upon the activity of the posterior portion of the center. Kalischer trained his dogs so that they would take food at one tone and refrain from taking it upon the sounding of all other tones. The food was held in the experimenter's hand or was laid on a chair near by the experimenter. The tones were first sounded upon an organ which contained nine pipes ranging from Ci to c\ Later he substituted a piano and later still a harmonium^ finding the latter better suited to his purposes. In training the dogs he sounded at first only the food tone so that he might accustom the animals to being fed at the sound. "From about the third day on," says Kalischer, " I struck now and then another tone and closed my hand over the bit of food while this tone was sounding." Then he caused the food tone to sound again, this time allowing the animal to eat. In a short time Kalischer found that the desired habit arose. In his own words, he says: ' ' From the fifth or sixth day on, even if I held the bit of meat in the open hand, many of the animals would no longer attempt to seize it when the confusion tone was sounded." Some of the animals were taught to take food at high tones (c-2048 d.v.) and others at low tones (C-6-i d.v.). As may be seen, the confusion tones at first were made quite different from the food tone. After progress in association had resulted, the confusion tones were made to lie nearer and nearer the food tone. He says that it was not very difficult to train the animals to re- spond differently to two tones when they were only a semi- tone apart. In order to control his results, Kalischer made the following checks: Some of the dogs were temporarily blinded by sewing their eyelids together. He reports that the accuracy of response was not affected. This ap- parently would rule out the possibility that the dogs were reacting to visual clues. Besides this form of control, Kalischer destroyed one cochlea in other well-trained dogs. No disturbance ensued. But when the other cochlea was destroyed the habit broke down. This shows, he thinks, that the dog was not reacting to non-auditory stimuli. It is perfectly clear that if Kalischer 's control tests were accurately made, they were adequate; but from experiments which Johnson has carried out (p. 379) the whole matter is left in doubt. It is very probable that Kalischer himself did not give the customary clues to the dogs after the cochlea had been destroyed, since he was not expecting them to react. Nevertheless, he himself was convinced that his dogs had been trained to react on the basis of the difference in pitch between the two tones. He then began 376 AUDITOKY AND RELATED FUNCTIONS his operative work. The first operation consisted of ex- tirpation of one temporal lobe from an animal whose cochlea on the same side had previously been destroyed. According to Munk, the eighth nerve makes a per- fect crossing and this operation should render the ani- mal completely deaf, Kalischer reported that his animal reacted to tones as before, no matter whether extirpation was partial or complete. Before proceeding to extirpate the opposite temporal lobe, Kalischer allowed four or five weeks for recovery, during which time he continued train- ing the animals. When the second temporal lobe was re- moved from some animals, the visual area was also injured. After this operation the dogs no longer reacted to spoken commands^ nor did they show by pricking up the ears or by movements of the head any sensitivity to loud noises. Later they began to resume such movements at very loud noises and at very loud commands, but did not learn to react differently to them. Before the operation the least whistle or call had been enough to bring forth the appropriate re- sponses. Even the movements of the head and ears at noises finally disappeared after the destruction of the posterior corpora quadrigemina with a needle. The animals' reactions to tone, however, were much less affected. Tests were re- sumed three days after the operation. Kalischer says that some disturbance of tonal habits followed, but he asserts that the animals undoubtedly reacted properly to the food tones and kept still on the sounding of the confusion tones. He ascribes the disturbance wholly to the shock of the operation. After two weeks the animals began to ex- hibit the old habits. They snapped in accustomed fashion at the food tone and shrank back at the confusion tones. They reacted also accurately to chords and discords. It was possible even in the most mutilated of these dogs to retrain them to react to a new food tone and to inhibit reaction to the former one. He concludes, accordinglj^, that the reception of noises and that of tones are mediated by different end organs and that different centers are involved for each. The center for noise he locates in the temporal lobe, the afferent pathway to which passes through the posterior corpora quadrigemina. The end organ for noise reception is indeterminate. On the other hand, he considers that the cochlea contains the end organ for tone and that the center for tone is infra-cortical and even below the posterior corpora quadrigemina, since the only known audi- tory pathway to the cortex passes through the quad- rigeminal bodies. Kalischer concludes on the behavior side that we must attribute to the dog an exceedingly fine sensitivity to absolute pitch.
Rothman's experiments. — Later Rothman, not satisfied with Kalischer 's behavior work, decided to make the same type of experiment. His method of training the dogs was similar to that of Kalischer except that he allowed an assistant to feed the dogs upon the sounding of the food tone. After some of the dogs had been trained he made the following operation. Both posterior corpora quadrigemina were removed on four dogs. Three had been previously trained. After the operation he tried to train two of the three dogs for about a month to react positively to c-1024 d.v. and negatively to the other c's on the organ, but without success. The fourth animal previously trained to react only to c-1024 d.v., and which had learned the problem faultlessly, was again subjected to the training process four weeks after the operation. Three weeks were spent in training to the same tone, without success. Then for twenty-three days he tried to teach this animal to react to the words " Mach schon, " with the same outcome. From Rothman 's brief account one gathers that the animals reacted when noises or tones were made, but did not react differently to them. Post-mortem showed total destruction of the posterior corpora quadrigemina in all animals. Ex- tirpation of both temporal lobes in five animals produced lack of sensitiveness to both tone and noise when the entire area described by Munk was removed. If the removal was not complete some trace of the reactions remained. A sixth dog, however, not previously trained, having been deprived of both temporal lobes and of one convolution of the gyrus sylviacus was successfully trained in three weeks to respond to one of the tones. From these and other 378 AUDITORY AND RELATED FUNCTIONS results which we cannot enter into, Rothman concludes that the dog's auditory center lies in the temporal region, but extends over a wider area than that defined by Munk. According to him the pathway from the end organ passes through the posterior corpora quadrigemina and the in- terior geniculate bodies.^ Further experiments by Kalischer. — Kalischer in 1909 continued his report. A dog was first trained to react in one way to a high tone and in another way to a low tone. After the habit was perfected one labyrinth was destroyed, making the animal wholly deaf on that side. (Destruction of both cochlea and vestibular apparatus was made com- plete.) Training was continued for two or three weeks after the operation, which had not at all damaged the dog's habit of responding to the low tones. The second cochlea was then exposed, and the desired part of it de- stroyed. The part destroyed was removed with a needle. The first animal operated upon, which had been trained before the operation to respond to the two tones A^ and c^, showed no loss of accuracy in the habit when one cochlea was entirely destroyed and removed and the other de- stroyed as far down as the lowest turn. Only a small portion of the (one) cochlea-vestibular apparatus was left intact. The part of the organ of Corti and of the mem- brane of Reissner contained in this part of the cochlea, and also the cells of the spiral ganglion, which belong to this turn of the cochlea, were uninjured. Reaction to spoken words, etc., was perfect. Kalischer made a number of such cochlea extirpations. In general he concludes that the theory of Helmholtz and others, that the different parts of the cochlea and of the basilar membrane act selec- tively as receptors of long or short sound waves, is unten- able. He assumes further that the vestibular apparatus possesses an auditory function and that it is necessary for pitch discrimination.^ 2 Since destruction of both internal geniculate bodies destroys the possibility of training to both noise and, tone, ^ Swift has also made experiments similar to those of Kalischer. He comes to the conclusion, from his experiments, that while the Johnson's experiments. — After the experiments of Kalischer and Rothman had appeared Johnson undertook in the Hopkins laboratory to repeat their work npon blind dogs. The forks used in the first test were standard forks mounted upon resonators. Two were used as stimuli, middle c-256 d.v., and g-384 d.v. To the c the dog had to respond by going to the right and climbing up in a chair, and to the g by going to the left and climbing up on a box. In the early tests, in which two dogs were used, the experimenter was in the room. Between the tests the dogs usually crouched between the experimenter's legs or else sat upon their haunches near by. In a remarkably short time the habit arose. After the response to simple tones had been perfected clangs were tried. Again the dogs soon learned to climb up on the chair when the clang contained the c and to climb up on the box when it con- tained the g. Their responses soon became so machine-like that Johnson became suspicious of the stimulus to which the animals reacted. It became absolutely necessary to test the dogs in the absence of the experimenter. Several different methods were tried. The one proving most satis- factory is the one described on p. 85 ff. After several months of experimentation by this improved method both upon normal and upon blind dogs Johnson having failed to confirm both his own earlier work and that of Kalischer reached the conclusion that if the dog is sensitive to dif- ferences in pitch it does not appear in tests of such a char- acter. Since Johnson in his preliminary work was able to repeat both Kalischer 's and Rothman 's findings in every detail, it would appear that their dogs were reacting to the same types of accidental stimuli as were Johnson's dogs in the early tests. Johnson did not determine the nature of the secondary criteria in any detailed way. He supposes that his own dogs in the tests where positive results were obtained were using a variety of cues such as (1) breathing of experimenter, and (2) involuntary movements of the body, etc.
center for pitch cannot lie in the temporal lobe, it does lie in the cortex, but in some other region.
380 AUDITORY AND RELATED FUNCTIONS Since we know that the dog does respond to auditory stimuli, such as calls, whistles, etc., Johnson next designed a simple experiment. Two electric buzzers, differing in pitch and in their overtones, were employed. One buzzer was placed directly over the right-hand food box and the other over the left-hand food box. Here we have not only differences in pitch and timbre, intensity, etc., but differ- ence in localization as well. The problem assigned the dogs was to go to the right when the buzzer over the right compartment sounded and to go to the left when the left buzzer sounded. The dogs quickly learned this problem. When the buzzers were interchanged, however, the habit broke down. They had been reacting apparently on the basis of difference in localization. While these experiments of Johnson completely discredit the work of Kalischer, Rothman, and Swift, it would not be safe to conclude until further experimentation has been undertaken that the dog is insensitive to pitch difference.
Experiments made by the Pawlow method. — The re- sults of tests made with the Pawlow method are fully as positive as those described by Kalischer; e.g., Selionyi states that tones which differ from each other by not more than a quarter of a tone can be reacted to differently by the dog; furthermore, that those which differ from one another only slightly in timbre are received as distinct stimuli. A single tone of a familiar chord of three tones of the same timbre and intensity produces a less intense response than do two tones of the same chord. Further- more, positive results were obtained with respect to the analysis of chords, etc. It is extremely improbable, in the light of Johnson's work, that these responses are really responses to tone. From reading the description of this work we cannot gather that the results were controlled in Selionyi 's work any better than in Kalischer 's.
Experiments of Syzmanski on localization. — In con- nection with the above experiments of Johnson on the localization of sound stimuli (electric bells) it is extremely interesting to note that Syzmanski failed both with cats and dogs to establish such a habit. His experiments were quite EXPERIMENTS UPON LOCALIZATION 381 similar. Three fairly large boxes were placed in the form of an isosceles triangle. The two boxes on the ends of the base of the triangle were made large enough to contain food boxes. The food boxes were so constructed that only one could be entered. The other contained an equal amount of food which could not be reached (to equalize smell). A small electric bell was placed first on one box and then on the other. (The two food boxes could be interchanged.) The animal was released from the box at the apex of the triangle. Upon release the animal could go to the box from behind which the bell sounded and get food. The trials, as is usual in such cases, were given irregularly. Apparently only one trial per day was given each animal. Only about 21 days were used in the regular test and then, since a position habit developed, Syzmanski gave up the attempt to establish the habit. While from casual observa- tion he states that the animals can localize sounds, he thinks that the reason they failed to do so in the above test was due to the narrowness of the room (this being unfavorable to localization).
(h) Cat: the response of cats to tones, noises, and to articulate sounds. — Numerous incidental observations at- test the fact that the cat is apparently keen-eared. Such observations, however, cannot be relied upon until tests experimentally controlled have been undertaken. We are lacking totally in exact information concerning auditory sensitivity. Rough tests, such as are about to be described, have been made. The absence of even such a simple con- trol as being out of sight of the animal when the stimulus was given robs them of much of the value they might other- wise have. Shepherd tested two cats, P and M, with two tones on a harmonica. To the note A-3 (food tone) cat P had to rear up on the cage with the fore-feet and look up at the top of the cage, M, being an inactive cat, merely had to look up at the top of the cage. To the note A-1 (non- food tone) both cats were to inhibit reaction. Forty-five trials were necessary to perfect the habit in one animal — P. The other cat, M, never learned entirely to inhibit reaction to the non-food tone, although there were signs 382 AUDITORY AND RELATED FUNCTIONS of definite improvement. The stimuli were then changed to A-2 (food tone) and to A-1 (non-food tone). P per- fected this habit in about 20 trials. The stimulus was then changed to piano tones — F-1 (food tone) and F-2 (non-food tone). Forty trials in all were necessary to perfect the habit. Selionyi states that the cat can be taught to come from another room to be fed on the sound- ing of the C of a set of tuning pipes and to inhibit response to the other c's in the same set.
A rough test has also been made on the sensitivity of the cat to the difference in intensity between two noises. The noise-producing apparatus consisted of a wooden box 18 X 11 X 10 inches and a slat 13 x 4 x 5/12 inches, fastened to the top of the box by a leather hinge. By raising the free end of the slat and letting it go a noise could be made, the intensity of which varied wdth the height of the fall. To give sounds of different degrees of intensity, two sticks, one 2-^ inches in length, and the other 4^ inches, were separately used and placed perpendicular to the box under the free end of the slat. By pressing on the slat near the hinge and suddenly removing the stick two sounds could be produced which varied in intensity. Shepherd states that one cat learned in 40 trials to respond to the louder noise and to inhibit response to the fainter.