LIBRARY OF THE UNIVERSITY OF CALIFORNIA. RECEIVED BY EXCHANGE Class BIOLOGY LIBRARY THE HEARING OF PRIMITIVE PEOPLES AN EXPERIMENTAL, STUDY OF THE AUDITORY ACUITY AND THE UPPER LIMIT OF HEARING OF WHITES, INDIANS, FILIPINOS, AINU AND AFRICAN PIGMIES BY FBANK G. BEUNER, A.B. ASSISTANT DIRECTOR, DEPARTMENT OF CHILD STUDY AND PEDAGOGIC INVESTIGATION, PUBLIC SCHOOLS, CHICAGO. Submitted in partial Fulfillment of the Requirements for the De- gree of Doctor of Philosophy, in the Faculty of Philosophy, Columbia Uni- rersity. Reprinted From the Archives of Psychology, No. 11. NEW YORK THE SCIENCE PRESS 19O8 THE HEARING OF PRIMITIVE PEOPLES AN EXPERIMENTAL STUDY OF THE AUDITORY ACUITY AND THE UPPER LIMIT OF HEARING OF WHITES, INDIANS, FILIPINOS, AINU AND AFRICAN PIGMIES BY FRANK G. BRUNER, A.B. ASSISTANT DIRECTOR, DEPARTMENT OF CHILD STUDY AND PEDAGOGIC INVESTIGATION, PUBLIC SCHOOLS, CHICAGO. Submitted in partial Fulfillment of the Requirements for the De- gree of Doctor of Philosophy, in the Faculty of Philosophy, Columbia Uni- versity. V Reprinted From the Archives of Psychology, No. 11, NEW YORK THE SCIENCE PRESS 1908 "B7 BIOLOGY LIBRARY G CONTENTS CHAPTER P*GK INTBODUCTION 1 I. THE PEOPLES 4 PART I UPPEB LIMIT OF AUDIBILITY II. HISTOBICAL General 11 Concerning Primitive Races 19 III. THE INSTBUMENT AND ITS GBADUATION Descriptive 21 The Graduation 23 IV. DATA COLLECTED ON UPPEB LIMIT 30 Indians 35 Filipinos 41 Ainu 44 Pigmies 45 V. UPPEB LIMIT AS AFFECTED BY AGE AND SEX 46 Summary and Conclusion 51 PART II AUDITOBY ACUITY VI. HISTOBICAL Concerning Primitive Peoples 55 Quantitative Auditory Measures 57 VII. INSTBUMENT FOB MEASUBINQ AUDITOBY ACUITY 64 Instrument Illustrated 67 VIII. GBADUATION OF INSTBUMENT 72 IX. METHOD OF CONDUCTING TEST 86 X. RESULTS 89 Indians 96 Filipinos 101 Ainu 105 Pigmies 106 XI. SUMMABY AND CONCLUSION . .109 ill THE HEARING OF PRIMITIVE PEOPLES INTRODUCTION IN the pages which follow are presented some data relating to the problem of the hearing of primitive peoples. The study was made in connection with other experiments on the inferior races at the Louisiana Purchase Exposition in 1904. During the period of the Exposition, the writer, in the capacity of Assistant Superin- tendent of the Sections of Anthropometry and Psychometry, under the Division of Anthropology, in cooperation with Dr. (now Pro- fessor) E. S. Woodworth, who was his immediately superior officer, was commissioned to make a study, so far as practicable, of the mental and physical status of the alien races stationed on the Exposition grounds. In the arrangements for the tests, the entire problem of the hearing of these people was assigned to me the ways and means of testing their hearing, together with the turn and scope the particular study of hearing should take. When it came to the question of selecting the tests to be made, there was little of historical precedent to aid in making a choice. Dr. Charles Myers, in the only extended report on the hearing of primitive peoples extant, had examined three phases of hearing, namely: (1) The upper threshold of pitch, (2) the acuity for tones of medium pitch, and (3) the perception of interval. In addition to these three tests, a number of others, which might have brought out interesting and instructive results, suggested themselves to me. These related to space perception, tone memory and imagery, and clang preferences. The scope of our work, however, was subject to certain definite limitations. In all, there were stationed at the Exposition in one capacity or another, something like one thousand individuals of various races, whom it might be possible to measure. There were two of us to carry on the work. One of two alternatives, consequently, must be pursued, (a) to restrict the number of measurements which should be made of each individual or (ft) to confine the measurements to a few representative individuals and races. We chose the first in the hope that with relatively large groups some fairly definite informa- tion could be obtained. In consequence of this limitation of the number of measurements, it was thought advisable to confine the number of tests of hearing to three or four. After some considera- 2 THE HEARING OF PRIMITIVE PEOPLES tion and advisement, I selected the three 1 which Dr. Myers had made on the Papuans. So far as practicable in these tests of primtive peoples, I hoped to be able to present my conclusions in such terms that the data might, if desired at some subsequent time, be reviewed and the measurements compared. I have consequently attempted to express my data in definitely interchangeable units, that is in acoustical units which are wholly objective in character. Much of the work included in this report, for this reason, is wholly a matter of physics. It concerns itself with the graduation of the instruments employed, but, for purposes of a quantitative psychology, this phase of the problem is none the less important. In this connection, I desire to express my indebtedness of one kind or another, to those who have particularly assisted me in the undertaking and completion of the study herein reported. To Pro- fessor Cattell, first of all, I am especially indebted for suggestions relating to the problem as my particular field of research and for many valuable and kindly hints and criticisms in the treatment of 1 The instruments and methods with reference to the first two named, (1) the measure of the upper threshold for hearing and (2) the determination of the acuity for tones lying within the range of conversational speech, will be fully detailed in the proper place. With reference to the perception of interval, however, a word is necessary in this connection. The usual method of testing the perception of a musical interval is with the aid of tuning forks. I took to the Exposition two Konig tuning forks, differing from each other by four full vibrations, the one having a vibration frequency of 512, the other 516. On one of the prongs of the fork of lower pitch I placed a metallic rider, which it was possible to slide up and down and which might be fixed readily by means of an ordinary thumb screw. Thus the pitch of the fork could be raised or lowered without occasioning any wearisome delays. It seemed impossible, however, to arrange the rider in such a way as not to alter the character of the tone which followed independently of its purely pitch character. With the rider attached, the two forks possessed marked characteristics of clang tint by which they could be distinguished, wholly inde- pendently of the feeling of a pitch difference. My subjects were repeatedly warned that they should neglect the individual peeuliarities in the two tones and render a judgment based only on their recognition of a difference of pitch. It appeared that this even my most intelligent subjects were unable to do, for questionings always showed that the individual peculiarity of the tone of each fork had become fixed during the earlier moments of the test, when the differ- ence in pitch would be so marked as to be easily observed by all the subjects. Such a method of conducting the experiment as that just indicated seems neces- sary with children and intelligent adults to impress the object of the test. Since, with intelligent white subjects, it seemed impossible to secure satis- factory data, it appeared unreasonable to presume that anything could be gained from this test on the primitive peoples. This test, therefore, was abandoned altogether and the hearing tests confined to two, (1) simple acuity and (2) upper threshold of hearing. INTRODUCTION 3 the results. To Professor Woodworth, it is impossible to express the full extent of my obligations. He performed large numbers of the tests. He, chiefly, was instrumental in making the arrangements with the officials in charge of the several groups for having the natives brought to our laboratories for testing. His searching criticisms, encouragement and interest at all times during the months at the Exposition, and subsequently, in the work of graduating the instruments have been an unfailing source of inspiration. To Pro- fessor W J McGee, Chief of the Department of Anthropology at the Louisiana Purchase Exposition, I am indebted directly for my selection to carry on the work. His encouragement in the work also and his assistance and cooperation in having the peoples brought to the laboratories for measurement was of inestimable worth. To S. M. McCowan, Superintendent of the Chilocco Indian School, who had charge of the Indian Ethnological exhibit at the Exposition, I am indebted for the privilege of measuring the Indians of the School, and to Major William Haskell, U. S. A., for the privilege of measuring the Philippine Constabulary soldiery. For assistance in interpreting directions and otherwise aiding the measurements, I was under obligations to Mr. Inagaki, of Tokio, in connection with the Ainu, Reverend S. V. Verner, in connection with the Pigmies, Dr. William Newcomb, in connection with the Vancouver Indians, and to Mr. Cushman, in connection with the Cocopas. Finally, and in a more comprehensive sense, I am indebted to all those who offered themselves as subjects for measurements, the mention of whose names alone would require many pages. Though mentioned in this general way only, my feeling of thankfulness to them is no less sincere. CHAPTER I THE PEOPLES IN all, I was able to secure hearing records, which were more or less suitable for use in making various deductions indicating indi- vidual and racial differences, from about four hundred individuals. These were distributed as follows: 156 Whites; 63 Indians; 137 Fili- pinos (Christianized) ; 10 Cocopa Indians; 7 Ainu from the Island of Hokkaido, Japan; 7 Indians from Vancouver's Island; 6 so-called African Pigmies; and 4 Indians from the region of Southern Patagonia. The Whites. The Whites whom we measured were those, for the most part, who strolled through our laboratories primarily to view the exhibits, but offered themselves as subjects for our tests, willing victims to be sacrificed, as we chose, in the interest of the furtherance of scientific truth. We examined altogether about 100 of each sex but many of the records were unavailable for my pur- poses, either because the subjects were too young to be used in comparative tests or, to be sure, because they had advanced too far in years to make data concerning their hearing of value in com- parison with those of younger individuals of other races. Many of the individuals were graduates from colleges and universities, others were school teachers. A number of professional and business men and women helped make up the number. Indeed, for the most part the group was made up of intelligent people. The Indians. Except the Christianized Filipinos and Whites, the Indians constituted the most numerous group tested. However, it can not be said that the Indians measured formed a single group, for they were brought from regions as widely separated as the Vancouver Islands and Patagonia. They belonged to approximately fourteen different stocks, as may be seen by the groupings below. The greatest number of individuals belonged to the Algonkian stock. Next in order came the Shoshones, then the Sioux, Pima, Iroquoian in the order given. During the whole of the Exposition period there were kept at the Indian School for purposes of exhibition something like 77 Indians. Of this number we tested 27 males and 44 females. Of the males, 14 were full-blooded and 13 mixed-bloods. Among the females, but four were of true stock. The mixed-bloods were in all cases partly 4 THE PEOPLES STOCK OK FAMILY Algonkian Shoshones Siouan TRIBE Sac and Fox Shawnee Pottawattomie Piegan Chippeway Kickapoo Cheyenne Comanche Chemehuevi Hopi or Moki Sioux Ponca Otae STOCK OK FAMILY Piman Iroquoian Scattered Tribes belonging to as many stocks TRIBE Pima Opata Papago Oneida Cherokee Navajo Pawnee Silitz Muskoki Payallup Kwaguitl Cocopa or Seri Tehuelche white that is, partially American, Scotch, French, German, Swedish, Spanish or Irish. Not a single individual among the In- dian group, so far as we know, possessed a negro strain in his inheritance. It may well be questioned whether any group of individuals so heterogeneously conditioned as was this, might, with propriety, be lumped and treated as if representative or typical of the race. Indeed, were we dealing with traits in which tribal and stock dif- ferences are marked, as for example, anatomical features, such a procedure would be wholly unwarranted. Sensory features, how- ever, are subject to a smaller range of variation. Moreover, these Indians from the Model Indian School came largely from the Indian schools at Haskell, Chilocco, Genoa, Phoenix and Ft. Shaw. Many of the boys and girls were taken from their homes at an early age and boarded at the Indian schools where they were subjected to social habits, intellectual training and industrial occupations which are common to whites. For the most part, they conducted them- selves as do the young men and women of our cities. So far as their attitude toward society is concerned, one could not detect anything that would point directly to their immediately native origin. Hear- ing tests, moreover, look to constitutional differences rather than anything that may be directly influenced by a social veneer, hence the culture to which these boys and girls had been subjected, we might with reason aver, would affect their sensory reactions only very remotely. The membership of the Model Indian School was made up of individuals belonging to the first twenty-two tribes named above, excluding the Hopi or Moki people who were connected with a con- cession on the Pike. Obviously, the numbers are altogether too few 6 THE HEARING OF PRIMITIVE PEOPLES to bring out tribal differences. All considered, therefore, it seems best to treat the membership of the Indian School as a single cultural group, nor does it seem necessary to indicate with reference to this group such physical and mental traits as mark off the various tribes of Indians aligned under the several stocks or families. Hav- ing lived for three or four years directly under the influence and training of an American civilization, the factors which might arise from differences in home life and ancestry are for the most part obliterated. The Filipinos. The 137 Filipinos whose hearing was tested all belonged to that branch of the Philippine soldiery known as the Constabulary. They constitute the local police of the Islands, being stationed in squads of eight in the different villages and districts, to preserve order. Inasmuch as the Constabulary is a branch of the local civil service and the remuneration is considerably in excess of that received for other manual vocations, the better element of the citizenship has been attracted to its ranks. Those brought to St. Louis were men in the prime of life, none older than thirty-five years or younger than seventeen. All had attended school to a certain extent at least. None was found to be illiterate or unable to write his name, tribe, and place of residence in the Islands. Many of the men were sons in well-to-do families who had attended the Spanish and parochial high schools and colleges found in the Philippines. Rather indicative of the scholarly habits of many of the younger men was their activity in acquiring our language. It was not uncommon to observe groups of men collected in some place apart with dictionary and grammar, assiduously studying English grammatical forms and usages. In collecting the group for representation at the Exposition, it appears that the men were drafted in squads of eight from the various Constabulary regiments located in every part of the Archi- pelago, there being, in fact, eight from the Moro population of the Island of Mindanao. The tribes represented were the Tagalog, Visayan, Ilocano, Bicol, Macabebe, Pampanga, Pangasinan and other less well established tribes. It does not appear that any cri- teria of stature, strength and intelligence were used in selecting the individuals for representation. Two legs, two arms and two eyes were required. Besides, it was necessary for the recruit to under- stand enough Spanish to take the orders of the line in that tongue. Other than these, there seem to have been no prerequisites. The Cocopa or Seri Indians. The Cocopa or Seri Indians tested were all males. For one reason or another, sickness, timidity, in- dolence, the women of the tribe could not be induced to come to the THE PEOPLES 7 laboratories where the measuring was being done. It was chiefly through the instrumentality of Professor W J McGee that the Seri Indians were brought to the Exposition. 1 As the result of a careful and painstaking study of the social habits, customs and physical characteristics of this interesting group of people, Dr. McGee speaks in the following words: " Isolated to a considerable extent on Tiburon Island (in northwest Mexico) these people have successfully resisted the innovations of the white man. To-day they still culti- vate aboriginal crops by aboriginal methods. They are said to be of so low a grade of culture that they may be classed as just entering the stone age. Physically, the Seri are a gigantic people, perhaps not excelled in their physical proportions by any other known tribe. Force of circumstances has made them an agricultural people chiefly, though the Cocopa are also given to the chase." Their habitations are extremely crude and primitive. Coarse grass, branches, leaves or whatever may be most convenient are thrown upon a crude frame- work of poles for a roof, while the same sort of an improvised material serves for walls. Such a habitation serves illy the purposes of protection from either heat or storm. The Seri Indians are not as intelligent as the average of the Indian race. They are inert, unresponsive to new impressions, dull and stupid in the face of an untried problem, and succumb readily to a difficult situation. With the older members of the group, especially, our efforts to make them approach our tests intelligently had been almost baffling were it not for the very able assistance and encouragement of an intelligent native interpreter, a half-blood woman who very ably interpreted our directions to the several subjects, but even with this, in some cases, the task seemed hopeless. Auditory acuity measures, however, being extremely simple, less difficulty was experienced with respect to them. The Ainu. The Ainu, four males and three females of whom we tested, are a people of more than ordinary ethnological interest. Surrounded by peoples of yellow skin, scant beard and little body hair with a head covering of straight black hair, the Ainu are white (when free from dirt) and their bodies and faces so profusely grown with a thick coat of hair that they have been popularly described as the hairy Ainu. The hair, too, is brown and wavy rather than straight. They inhabit the Island of Hokkaido or Yezo of the Japanese group. Little is known of the people's origin or ethnie relations. Until recently they had been little disturbed by other peoples, even by the Japanese among whom they dwell. While 1 Professor McGee made some extensive explorations in the Seri country, which are reported at length in the Annual Report of the Smithsonian Insti- tution for the year 1895-96 (Part I., pp. 1-285). 8 THE HEARING OF PRIMITIVE PEOPLES natively the Ainu are hunters and farmers, those at the Exposition had been under the influence of American missionaries, chief of whom is the famous Mr. Bachelor whose influence with the native Ainu has been remarkable. Through his influence it was that the people consented to leave their native land for the journey to this far off country. One of the Ainu young men had attended Mr. Bachelor's mission school, another had been a servant to another missionary. The father of the household, an old patriarch, had also been converted to the Christian faith, but had never quite sur- rendered all of his natw instincts and superstitions. He was a farmer and bear hunter and still clung to the superstitions attaching to the erection of a bear head at the door of the dwelling to ward off evil spirits and omens. The Ainu are short and stocky, sluggish in movement, deliberate in action, excessively timid in the face of a novel situation, and, taken all in all, very immature in their mental conceptions and aptitudes. However, they were willing and patient in the tests to a degree to cast reproach upon many of our white subjects. The Ainu were brought to America by Professor Frederick Starr of the University of Chicago. We were much indebted also to Mr. Y. Inagaki, a Japanese student, familiar with the Ainu language, who interpreted our directions to the Ainu subjects. In fact, in many ways, Mr. Inagaki 's kindly interest and assistance alone made the tests on these people at all possible. Especially was this true be- cause of their excessive timidity. Vancouver Indians. The Vancouver Indians belonged to two tribes, the Kwaguitl and the Nutken. There were present at the Exposition, two members of the first named tribe and five of the latter. Like the Ainu just described, some of these people were interrelated. At least four of the group of seven tested were closely so, though we were unable to discover in all cases the exact character and extent of the consanguinity. There were Atleo, an old man of perhaps 65 years, his two daughters, Ellen, aged 35, and Anna, aged 30, or thereabouts, and a nephew, Jack Curley, aged 28. From a sci- entific point of view, of course, this was unfortunate. The physical proportions of the Vancouver Indians are rather less than those of the individual of the Algonkian stock. The Vancouver Indian is shorter and slighter of build, but on the whole, stronger and more hardy. He is certainly more active as well as more alert than the Algonkian. In their native haunts, the men are fisherman, often going miles to sea in open boats in the search of whale and seal, which are captured by skillful rowing and spearing. The Vancouver women are especially noted for their beautiful THE PEOPLES 9 blanketry, skill in weaving and dyeing. Some of the men also carve skillfully in ivory and wood. In common with most Indian tribes of the Northwest, the Vancouvers have elaborate ceremonial rites, family legends bound up with the family totem, and certain fiducial customs and habits, which are the sacred possession of the household to which they are attached. The totem and fiducial customs at- tached thereto pass down as a family coat-of-arms, as it were, by which the tribe is distinguished. In point of intelligence, the Vancouver Indians surpassed any of the Indians we tested save only those boys and girls at the Indian schools who had for years been moulded by the influences and habits of whites. We, therefore, found these people easy to handle and instruct. The Pigmies. The group of people popularly known as the Pigmies whom we tested were made up of individuals from three tribes; three Batwas, two Batsubas and one Cheri Cheri; all were males. Their ages were uncertain. They were, however, boys, almost if not fully grown, though I think none was older than 25 years. It is claimed that no Pigmies had before crossed the Atlantic, and naturally they were of peculiar interest. Only two or three of the natives had ever before left Congo territory. The Pigmies were brought to the St. Louis Exposition by Rev. S. V. Verner, a missionary who had spent some years on the African coast, and had familiarized himself with the Pigmy language and social customs. Mr. Verner related that it required some energetic persuasion to induce these people to accompany him to St. Louis. The Pigmy tribes observed by Mr. Verner in the Congo lived a parasitic existence, following the large Kaffir tribes and feasting on their bounty or refuse. It is related that companies of Pigmies and dogs, intermingled, station themselves at reasonable distances from Kaffir feasts, spying with envious eyes the feasting banqueters. No sooner is an unwholesome piece of flesh cast aside by the Kaffirs than there ensues a scramble of Pigmies and dogs indiscriminately for the rejected prize. Whether all Pigmies stand so low in the scale of social culture we are unable to say, but it is held to be applicable to the group which we tested. In physical appearance, the Pigmy presents no sign that might lead one to class him as of degenerate stock. Although not exceed- ing an American boy of twelve years in stature, his bodily propor- tions are good. Still, he is not robust nor capable of great endur- ance or extraordinary feats of strength. This inferiority, however, probably has its basis in habit, rather than in any innate physical incapacity. Active in play and frolic, and with a keen sense of 10 THE HEARING OF PRIMITIVE PEOPLES humor, the Pigmy is a thorough optimist. He really enjoys life, indeed, takes everything with such a degree of levity, that it was only with considerable effort that we were enabled to have him approach our tests with anything like the seriousness they demanded. Withal, the Pigmy is stupid and dense and apprehends meanings slowly and often incompletely. The hearing tests being very simple in character, however, were understood with a fair degree of appre- hension, and I have reason to believe that the data are reasonably representative of the group measured. The Patagonians Tehuelche Indians. I was successful in test- ing only four men of the group of Indians from southern Patagonia ; their ages being respectively, 18, 24, 35 and 55 years. At home, the individuals on exhibition at the Exposition had been employed as herdsmen on the Patagonian prairies. They had learned the use of money and, furthermore the habit of rendering no service, no matter how trivial, without a money consideration. This necessitated a bribe of money before any measurements were possible. Like many other of the Indian tribes, the Tehuelche are sullen and uncommunicative. Their cultural habits are primitive. Their habitation is a tent made of the skins of the llama or guanaco, sewed together so as to form a considerable sheet. This is then stretched across poles, with the edges spiked to the ground. Within this tent, the family cooks, sleeps and lounges. As there is no vent for the escape of smoke, and the floors are never scoured, filth and grime abound everywhere. The Tehuelche are horsemen and skilful in the use of the bolo, a triple thong loaded at the end with stone weights which is thrown great distances with unerring accuracy. They are a large people, both men and women being tall and robustly built. With respect to the four individuals measured by us, no unusual difficulty was ex- perienced in instructing them in the ways of the tests, but I question whether in grade of intelligence, they did not exceed the average of the Patagonian Indian. PART I THE UPPER LIMIT OF AUDIBILITY CHAPTER II HISTORICAL THAT considerable individual differences exist amongst persons with regard to the faintest tone that can just be sensed is commonly recognized, but differences in the range of hearing are not so readily apparent. Helmholtz first called attention to the fact that the chirp \ of the cricket is sometimes wholly inaudible to people whose hearing j is otherwise uneffected who have experienced absolutely no diminu- / tion in hearing. Even after the fact was known that such individual differences really exist, the experimental determination of the range of such variations long awaited some device which would produce and accurately evaluate the sonorous stimuli. Experiments with visual sensations present no such difficulties as are encountered by an investigator who works with sound. Among visual stimuli, qualitative characteristics are overt. They stand out in such a way as to be little confused by an individual with a normally functioning visual organ. Differences in color such as those between red and green, yellow and blue, or, indeed, yellow and red, are readily perceived, and at the same time differentiate certain qualitative effects to which there is no analogue in the field of audi- tion. When one is affected with the sensation red, it has been de- termined once for all time, that the stimulus arises from a disturb- ance in the ether amounting to approximately 450 billions of vibra- tions to the second, and that the sensation blue corresponds to a dis- turbance of approximately 790 billions. But, among the higher orders of pitch, differences in tones as great as an octave are scarcely observed, even when they follow each other in immediate succession. Indeed, among comparatively low pitch values, the perception of tone differences is relatively uncertain. In truth, pitch differences stand out as variations in degree only, while color differences naively are differences in kind. Coming to the question of accurate tone analysis, the physical difficulties are still more involved. What the prism has been able to do for the physicist in assisting him in establishing the wave-length 11 12 TEE HEARING OF PRIMITIVE PEOPLES of any ray of light whatsoever, the resonators of Helmholtz and Konig do only very unsatisfactorily, in helping to fix the components of any tonal compound or establishing the pitch of a given unknown sound. The latter, indeed, must still be accomplished by compli- cated registering devices. Again, the nature of auditory stimuli is such as to make every unfamiliar tone in nature an almost wholly unknown quantity, which it becomes necessary to establish, empiri- cally, always anew. And with a shrill tone, the empirical method alone suffices to fix the pitch even roughly. Unfortunately, the importance of knowing the exact character of the stimuli employed has not always been appreciated by investi- gators in the field of hearing. Especially is this to be regretted in the reports of investigations on the limits or the range of audition. It thus happens that on account of a diversity of statement and lack of precision in the definition of the tones employed, it is wholly impossible to compare the data of different investigators. But they serve to emphasize the futility of any research in the field of hearing unless the physics of the problem involved has first been clearly worked out. Some figures relating to the upper threshold of hear- ing, given out by different investigators, will serve to illustrate what I have just indicated. Blake and Appunn 1 who are among the foremost investigators of the upper limit of audibility, think the human ear to be sensative to tones of 50,000 or 60,000 double vibrations to the second. Preyer 2 placed the extreme upper limit at 40,000 vibrations; Konig, 3 with the use of short sounding rods and a modification of the Galton whistle, got results substantially in agreement with those of Preyer. With a Galton whistle blown by a constant air blast, Zwaardemaker 4 believed he could produce audible tones whose vibration rate ex- ceeded 33,000 to the second. But all these data were called in question by Melde, 5 who pointed out that previous investigations were valueless because of instrumental errors. Melde 's objective ex- periments with different makes of instruments led him to believe that no ear is sensitive to tones above 24,000 double vibrations to the second. Melde 's investigations were repeated and elaborated by Schwendt, 6 who reached the conclusion that with the Galton whistle, 1 Annal. d. Phys. u. Chem. 64: 409. 1898. 2 "Die Grenzen der Tonwahrnehmung " (English trans.), Proc. Mus. Assn., 1876, pp. 1-32. 9 Annal. d. Phys. u. Chem. 69: 626-66, 721-38. 1899. * Arch. f. Ohrenhk. 35: 30. 1893; Ztschr. f. Psychol. 7: 10. 1894. Pfluger's Arch. 71: 441. 1898; Annal. d. Phys. u. Chem. 67: 781-793. 1899. 'Pfliiger's Arch. 75: 346-64; 76: 189-91. 1899. UPPER LIMIT OF AUDIBILITY HISTORICAL 13 audible tones of greater vibration frequency than 22,000 can not be produced. However, with an instrument modeled after the type of the steam whistle (Edelmann's), Schwendt 7 found that where greater intensities of air blasts might be employed, a tone of the value of 49,000 double vibrations might still be heard. With the same whistle, Edelmann 8 put the upper limit at 50,000 double vibra- tions. By a singular method Stumpf and Meyer, 9 in which some Konig forks and an Appunn Galton whistle were graduated by a method of difference tones, believed a greater vibration rate than 20,000 inaudible. Still more recently, after some extended objective experimentation with a Galton whistle of the Hawksley pattern, Dr. Charles Myers 10 concluded, at least so far as the Hawksley whistle goes, that tones above 24,000 double vibrations to the second can not be made of such an intensity as to be audible, nor indeed measured by any known means. It is scarcely necessary to enter into a detailed discussion of the factors responsible for such wide discrepancies in results as these shown between instrument makers on the one hand and scientific investigators on the other. A review of the literature relating to this work, notwithstanding, forces the conviction, that the variations are due to instrumental differences almost wholly being complicated by the failure to allow for the physical and physiological factors involved in the tests conducted. Much of the early work suffers from inefficient methods of arriving at the vibration frequencies of the tones that have been employed in making the physiological meas- urements. A great deal of the later work on the upper limit of hear- ing is of the same character. If we add to this difficulty the fact that the physiological value of tones for the ear has been ignored, it is not improbable that most of the differences would be accounted for. In keeping with this conviction are the recent investigations of Wien and Zwaardemaker. Wien 11 and Zwaardemaker and Quix 12 have shown that the ear is most sensitive to tones whose pitches are of the middle values, those lying roughly within the range covered by conversational speech (400 D. V. to 4000 D. V.). Both above and below these values, the ear's sensitivity is found to diminish rapidly. Indeed, as early 1 Verhandl. d. Naturforsch. Gesellsch., Basel, 12: (1900) ; Arch. f. Ohrenhk. 48: 1. 1900. *Annal. d. Phys. 2: 469. 1900; Ztschr. f. Ohrenhk. 36: 330. 1900. 'Annal. d. Phys. u. Chem. N. F. 61: 760-79. 1897. 10 "The Pitch of Galton Whistles," J. of Physiol. 23: 417. 1902. "Pfliiger's Arch. 97: 1. 1903. "Arch. f. (Anat. u.) Physiol. (Suppl.) , 1902, 367-98; also Ztschr. f. Psychol., u. s. w. 33: 407. 1904. 14 THE HEARING OF PRIMITIVE PEOPLES as 1878, Helmholtz (Tonempfindungen) pointed out that the ear's range of sensitivity could not be safely divorced from the factor of the intensity of the tonal elements, but the importance of this fact, as related to the upper and lower thresholds of audition, has until recently received little thought. Scripture and Smith 13 in review- ing the factors involved in the wide variations discovered by those who have investigated the ear's range of sensibility again called attention to the importance of the intensity factor and suggested that perhaps the differences in recorded experimental results might be explained on this basis. Some experiments with the Galton whistle, in which various degrees of wind pressure were employed convinced them that if the intensity of the stimuli could be made sufficiently great, the ear would be found to be sensitive to tones whose vibration frequency exceeded 50,000 or even 55,000 D. V. It must be confessed, however, that their empirical evidence for such a conclusion is not particularly convincing. In the light of recent experimental results, it would not be over- confident to believe that such wide discrepancies as the figures from different observers show would largely disappear were it possible to reckon and allow for the two factors just indicated the physical, concerned with graduation and intensity, and the physiological, con- cerned with the ear's relative sensitivity. Both of these factors are so interrelated in the historical data to be considered, that it is im- practical to attempt to separate them. 14 This condition comes about partially at least because no uniform type of instrument has been employed for measuring the upper threshold of hearing. Appunn, 15 Preyer/ 6 and Koenig 17 used small tuning forks in which the instru- ments themselves possess certain physical limitations confining the possible intensities of the tones to very narrow limits. Koenig 's "rods" possess the same deficiency. It appears that all of these in- struments uniformly were assigned tonal values altogether too high. 13 See " Highest Audible Tones," Studies from the Psychological Laboratory of Tale University, 1894, p. 105. 14 To Schwendt particularly do we owe a method for evaluating the vibra- tion frequencies of tones, which is wholly independent of the experimenter's auditory sensitivity. As Schwendt remarks, it is a method that may be employed equally well by a person wholly devoid of hearing, and consequently eliminates so far as such a thing is possible, the element of the personal equa- tion. The method consists of an adaptation of the Kundt dust figures to tubes of small bore and sound waves of extremely small extension. See Annal. d. Phys. u. Chem. N. F. 61: 760-69. 1897. a Annal. d. Physik u. Chem. 64: 409. 1898. "Pfliiger's Arch. 71: 441. 1898; Wiedemann's Annal. 51: 683. 1894; 52: 238. 1894. 1T Pfluger*8 Arch. 75: 346. 1899. UPPER LIMIT OF AUDIBILITY HISTORICAL 15 Melde, 16 who put to experimental test the pitch values assigned to the tuning forks used in Preyer and Appunn 'a experiments, entirely dis- credited them. His results were later confirmed by Schwendt. 17 The Appunn fork marked (g 8 ) 50,000 (double vibrations) was found to have a vibration rate of only 13,157 D. V. The pitches of the remaining forks of the series were overstated to about the same degree. What has just been said of the tuning forks applied with equal force to the sounding rods of Konig. His highest pitched rod, ac- cording to the optographic measurements of Melde, gave a tone of (f 7 ) 21,845 D. V., although Konig had assigned to the same rod a value of 40,000 D. V. Much of the error, no doubt, arose on account of the method used for assigning the pitch values to the different forks and rods. Preyer, Konig and Appunn believed that forks could be graduated with sufficient accuracy for ordinary scientific purposes, with the unaided ear. According to Preyer, 18 practised musicians can distinguish with certainty a difference in pitch amounting to one-half a vibration between the limits of "C" and * ' c 2 . " Were it possible for practised observers to discriminate tonal differences as accurately for all parts of the hearing scale, the gradu- ation of instruments for measuring the upper limit of hearing might well be made by some such subjective method as these investigators employed. But the perception of interval is extremely deficient for tones above "c 3 ," as the experiments of Konig 19 with sounding rods and those of Preyer 19 with tuning forks have pointed out. With tones in the sixth octave an interval of a fifth is scarcely observable, while for tones in the seventh octave and above an interval of an octave passes unnoticed even by musical ears. As Schwendt has well remarked, it is not improbable that the production of audible tones of a pitch exceeding 25,000 vibrations to the second, with a tuning fork or rod is a physical impossibility owing to the extreme weakness of tones coming from such sources. Where the forks are made as small as those must be to give a tone above 25,000, the energy given out is much diminished. Schwendt further pointed out that, even were their production possible, no known means exists for determining the vibration rate. The methods in use for de- termining the vibration frequencies of tones are wholly inapplicable to tones of so small energy value as these. Indeed, Schwendt found the resonance method inapplicable for the forks and rods of a vibra- tion frequency exceeding 15,000. 18 "On the Limits of the Perception of Tone" (trans.), Proc. Musical Assn. 1896-7, pp. 1-32. 19 Helmholtz, " Tonempfindungen " (4th ed.), p. 147. IQ THE HEARING OF PRIMITIVE PEOPLES By far the most common device for measuring the upper limit of hearing has been some form of the Galton Whistle. This type of threshold-whistle, 20 devised by Sir Francis Galton, is constructed on the model of the closed organ pipe, with vibrating lip and resonance cavity. In a closed organ pipe, Helmholtz and also Lord Rayleigh 21 found that theoretically, at least, the vibration frequency is a definite function of the length of the resonance cavity. Knowing the velocity of sound in air at the temperature prevailing "V a " the length of the resonance cavity "L," the vibration frequency "N" in double vibrations may be computed directly from the formula *-. 4L In point of fact this theoretical formula is not wholly valid even for closed pipes of relatively large dimensions as has been experimentally demonstrated by Savart, 22 Liscovius and Wertheim 23 and others and, indeed, the formula has been shown to be wholly inapplicable to pipes of small bore. 24 Differences in the pressure of air blast employed, the ratio between the length and width of the resonance cavity, the dimensions and shape of the mouth slit, together with the materials of which the whistles are made, all have been proven to be extremely important factors in determining the pitch of Galton whistles as well as all other closed pipes of this variety. As the diameter of the resonance cavity increases, the tone deepens but, on the other hand, the pitch becomes more acute as the wind pressure becomes greater. Such considerations make im- possible any mathematical formula generally applicable to threshold pipes. It has been found, moreover, that no matter how painstaking and skilful the construction in the attempt to duplicate a threshold whistle differences are certain to result, which make it necessary to graduate objectively each whistle independently. In the experiments of Stump f and M. Meyer, 25 in which Galton whistles were used, the graduations were also made subjectively. These investigators depended on the observation of difference tones. By blowing two whistles, whose pitches differed by about 2000 double vibrations to the second, simultaneously, a difference tone resulted whose pitch these investigators believed they could place "This whistle is described in Galton's "Inquiries into Human Faculty," p. 38. n Phil. Mag. 22: 344. 1879. 22 Wiillner's " Experimental Physik," Bd. 2, p. 324. 28 Op. tit. Bd. 1, p. 886. "Vid. Myers, J. of Physiol. 28: 417. 1902. "Annal. d. Phys. u. Chem. 61: 760-79. 1897. UPPER LIMIT OF AUDIBILITY HISTORICAL 17 accurately by ear. Then, by increasing the pitch of each whistle alternately, graduations were made until the point was reached where the tones faded entirely. A similar method was pursued with the Konig tuning forks and the sounding rods. Certain serious criticisms, however, have been offered against the method. In the first place, difference tones are difficult to recog- nize at all times and one is never quite certain whether the difference tone is the difference between the vibration frequencies of the fundamentals of the two tones employed or of their octaves, or, per- haps, of the second upper partials. It depends altogether upon the relative intensities of the various components, factors indeed, of which we are never quite certain in the production of very acute tones. Helmholtz 26 observes that while the method of difference tones is possible with pitches whose vibration frequencies fall below 10,000, for values higher than this the method is extremely uncer- tain. In point of fact when Stumpf and Meyer's graduations were put to experimental test they were shown to be not at all reliable. Employing a resonance method for evaluating the pitches of the instruments used by these investigators, Melde 27 found that the graduations were not accurate to within 10,000 double vibrations for the shriller tones those of f 7 and above. There has been on the market, during recent years, an improved form of threshold whistle devised by T. L. Edelmann. 28 It is asserted that this instrument overcomes the objections put against the older Galton form. Edelmann believes that with his whistle tones of 110,000 D. V. can easily be produced and the vibration fre- quencies fixed; tones, to be sure, which are far too high for any ear to sense. The mechanism of the whistle is such as to consume con- siderably more energy in its operation than the ordinary Galton type and it likewise gives out a tone whose intensity is many times in excess of that given out by the Galton whistle. Zwaardemaker and Quix 29 discovered that the upper threshold of hearing might be raised a major third by increasing the intensity of the tone 1,000 times, and Wien thinks that it may be raised at least a full octave if the intensity can be increased 10,000 times. 30 28 " Tonempfindungen " (4th ed.), p. 203. a Uber die verschiedenen Methoden der Bestimmung der Schwingungs- zahlen sehr hohen Tone, Annal. d. Phys. u. Chem. 67: 781-93. 1899. 29 This whistle, together with the methods for its graduation, is fully described by the inventor in an article in Annal. d. Physik (N. S.) 4: 469-82. 1900. "Ztschr. f. Psychol., u. s. w., 33: 407. 1904. "Pfluger's Arch. 97: 1. 1903. 18 THE HEARING OF PRIMITIVE PEOPLES These conclusions confirm the conviction of Scripture and Smith, referred to above. 61 Exactly in the same direction point some of the experiments of Dr. Chas. Myers, in which an Edelmann whistle with different degrees of wind blast was used. His statement of the facts is so convincing that I quote his own words: "When the wind pressure was 30-35 mm. of water, a very faint quivering note was heard. The note rose gradually until a wind pressure of 140 mm. of water was reached when the tone disappeared. Increasing the pressure 100 mm. the tone was again audible, and, at a wind pressure of 800 mm. of water, the tone was at least an octave higher." With a pipe length of 1.3 mm. and mouth width of 0.75 mm., employing the Schwendt dust figures for evaluating the pitch numbers, Meyers got the following significant figures: When the wind blast, measured by water pressure was 36 mm., the vibration frequency was 5,673. With a water pressure of 109 mm. the vibration frequency was 10,942 With a water pressure of 680 mm. the vibration frequency was 23,315 With a water pressure of 800 mm. the vibration frequency was 28,332 The last figure exactly corresponds to that given, for the same ad- justments, in the chart which the makers sent with the Edelmann whistle used in the experiments to be detailed in this writing. This tone and many much higher were easily heard not only by myself, but also by a majority of my adult subjects. There is no question but that the fundamental objection to the Galton whistle used by Zwaardemaker, Stumpf, Dr. Chas. Myers and others, as well as to the tuning forks and sounding rods, lies in this, that the tones pro- duced are too feeble, not only to cause a disturbance of the lycopodium powder in the dust tubes, but also to reach the physio- logical threshold of the auditory end organs for tones of the upper pitches. So much for the literature referring specifically to the physics and physiological factors of the problem. Now, let us look for such sensory differences as have been discovered. In the light of the sources of error mentioned above, the sensory data thus far collected are almost hopelessly bemuddled. Zwaardemaker, 32 perhaps, has collected more data of individual and age differences as regards the upper limit of hearing than any one else, but his results are not even comparable among themselves. In some of his experiments, a form of the Galton whistle was em- ployed, graduated by the Stumpf-Meyer method. In still other of *Loc. tit., p. 108. "Annal. d. Phys. u. Chem. (N. F.) 61: 760-79. 1897. UPPER LIMIT OF AUDIBILITY HISTORICAL 19 his experiments, the pitch values of the different whistle lengths were fixed by comparing the tones given out with those of the Konig rods or the Konig tuning forks. The untrustworthiness of both of these methods of making graduations has been pointed out above. Zwaardemaker's "e 8 ," a tone which many of his subjects heard distinctly was found' 33 to be wholly fictitious. Regarding the upper limit of hearing of whites, the literature is not at all scant. But even were it possible to separate out the data in which the instrumental defects have been least prominent, it would still be difficult to draw comparisons. Without exception, the distrbution of cases has been omitted in the data presented. Most of the work on the upper range of hearing has been carried on for the purpose of determining age differences. It has been pretty well established that the upper limit of hearing contracts with increasing years of life. Zwaardemaker, 34 Cuperius, 34 Alderton, 35 Myers, 36 and others have contributed rather convincing data on this phase of the problem. The results obtained are so significant as to justify the inclusion of tables summarizing them: Zwawrdemaker" Cuperius 3 " 1 Ages Whistle Length Pitch Whistle Length Pitch Under 10 yrs. 1.22 mm. e 7 10-20 " 1.39 " dis T 1.08 mm. f 20-30 " 1.39 " " 1.19 " e 7 30-40 " 1.58 " " 1.31 " 40-50 " 2.23 " cis 1 1.39 " dis T 50-60 " 2.93 " h" 2.08 " cis 7 Over 60 " 3.03 " cis 6 3.02 " cis 8 Alderton reports as follows the examination of 500 individuals with the Gatonj whistle: For children up to 12 years pipe length on the average . . 1.24 mm. For adults pipe length on the average 3.03 mm. Dr. Charles Myers's 36 tests of Scotch children point in the same direction. The Upper Limit of Audibility among Primitive Races. Bo far as I have been able to discover, no measurements of this character save those of Myers and my own herein reported, have ever been 33 Myers, op. tit. "Arch. f. Ohrenhk. 33: 54. 1893; Ztschr. f. Psychol. 7: 11. 1894. 35 Arch, of Otology, 23: 71. 1894; 25: 43. 1896. 86 Report of the Cambridge Anthropological Expedition to Torres Straits, Vol. 2. 1903. "Ztschr. f. Psychol. 7: 10. 1894. 20 THE HEARING OF PRIMITIVE PEOPLES made. In their work on the upper limit of hearing of the native Papuans of the Murray Islands, Rivers and Myers 38 used the Hawksley pattern of the Galton whistle. This was an instrument with an extremely small bore, such as to make it possible to produce very high pitched tones. Dr. Myers, who reports the hearing tests, was inclined to put little faith in the graduations of his instrument, a task which was performed after he had returned to England, so the data relative to sensory differences are given in terms of the length of the cavity of the whistle. 38 His results are presented so as to show at the same time race and age differences. I summarize them in the following table: Murray Islanders Aberdeenshire Folk Whistle Whistle No. Age Length No. Age Length Children 2 5-9 yrs. 2.25 mm. Children 4 5-9 yrs. 1.97 mm. 15 10-15 " 2.07 " 18 10-15 " 1.99 " Adults 5 16-19 " 2.25 " Adults 9 20-29 " 3.00 " 20-29 " 2.22 " 12 30-39 " 3.17 " 30-39 " 2.63 " 9 40-49 " 3.53 " 40-49 " 3.19 " 10 Over 50 " 4.68 " Over 50 " 3.77 " It appears from this table that for all ages the upper threshold of hearing of the Papuans is lower than for Scottish people of cor- responding years. It is a conclusion that is significant, notwith- standing that the numbers tested for the several ages were small. More of this, however, in connection with the discussion of my own figures. 88 Dr. Myers employed a number of methods in making the graduations of tonal values corresponding to the different cavity lengths in the tests. Chief of these are the resonance method and optographic method. See J. of Physiol. 24: 417. 1902. CHAPTER III THE INSTRUMENT AND ITS GRADUATION I EMPLOYED, for measuring the upper range of audibility, the Edelmann modification of the Galton whistle. It is probably too commonly known to require description. 1 It differs from the familiar Galton form, in being modeled after the pattern of the steam whistle instead of the closed organ pipe. In the opinion of Edelmann, this improved pattern possesses some marked advantages over the old form. In the first place, the different parts are con- structed separately, allowing of finer work. The whistle cavity itself is a perfect cylinder, which makes it possible to turn it out very delicately on a lathe. Then, the whistle possesses a means whereby the width of the lip opening may be varied, so as to allow for the large quantity of air that must pass through it in producing low pitched tones, and still avoid the air puff with extremely shrill tones. In measurements with the Galton whistle this air puff, which accompanies very high tones, is extremely confusing especially to untrained subjects. Not infrequently a subject states that he is not certain whether he hears a tone or wind only. Dr. Myers 2 observes in connection with his work on the native Papuans in which the older Galton type of instrument was employed, that there was con- stant confusion between the perception of the sound and that of the air puff, which always accompanied it. While I do not think that this difficulty is entirely obviated by making the mouth width of the whistle adjustable, as in the Edelmann pattern, yet there is no ques- tion but that the Edelmann whistle is superior in this respect to the old Galton form. In my own experiments, I had never observed that any subject found difficulty in distinguishing between the ac- companying air puff and the tone and, indeed, when the whistle was so far as 25 centimeters from the ear, almost no air puff was ever audible. 1 Those not familiar with the instrument, I refer to the inventor's (Pro- fessor Edelmann's) able description and careful drawings to be found in the Annalen der Physik for 1900. 4 Folge, Bd. 2, S. 469. Those also unfamiliar with the Galton whistle may find a description of this instrument in Galton's "Inquiries into the Human Faculty," 1883, p. 275. 3 Op. tit., p. 4. 21 22 THE HEARING OF PRIMITIVE PEOPLES The manufacturers send out with each Edelmann 8 whistle a chart giving the pipe length, the mouth width, and the vibration frequency corresponding to each of some twenty different tones, ranging in pitch from that represented by a vibration rate of 6,000 to that represented by a rate of 50,000 double vibrations to the second. It is claimed that each whistle has been graduated inde- pendently and empirically at the factory, and that, for each tone, that mouth width was selected by trial and practise which would produce a note of optimum purity and strength. In the chart sent out with the whistle which fell into my hands, it appeared that the graduations for those tones lying between e 3 and e 8 had been made with a uniform mouth width of 0.75 mm. With the whistle so adjusted, those tones lying in the region of e 6 were pure, clear and free from that peculiar harshness which results when a considerable quantity of air escapes with the tones. The contrary, however, was true when the whistle cavity was diminished for the production of tones in the region of e 8 . These latter tones were decidedly harsh. This harshness was obviously due to the accompaniment of air puffs, which escaped with the tones. They stood out so prominently as to confuse even the most careful subjects, and must have proven a very distracting element to children, and especially to the primitive peoples. It therefore seemed advisable to vary the adjustment from that prescribed in the chart sent out with the instrument even if it would necessitate an entire re-gradua- tion. After careful experimentation with the assistance of Professor Woodworth, a mouth width was hit upon which give admirable results for all tones from e 5 upwards. Indeed, so free was the tone from wind blasts, that when the threshold range was passed, no audi- tory stimuli of any character were sensed as coming from the whistle. This mouth width was 0.55 mm. It chanced, however, that the resulting tone was 1 predominantly the first overtone instead of the fundamental, but, since the fundamental tone was inaudible except for vibration rates between ten and fifteen thousand, it was thought not to be a particularly disturbing factor. All the meas- urements at the Exposition, consequently, except those where the upper limit was found to be extremely low, were made with a con- stant mouth width of 0.55 mm. This, to be sure, rendered the table which accompanied the whistle entirely worthless. A wholly new set of graduations must be made to meet the new conditions. The Edelmann whistle used in my tests was kindly loaned to me by the C. H. Stoelting Co., 38 W. Randolph Street, Chicago, for the double purpose of exhibition and experimentation. THE INSTRUMENT AND ITS GRADUATION 23 To carry on such an extended series of experiments called for more time and more elaborate preparations than it was possible to insti- tute at St. Louis. I therefore allowed this work to await my return to Columbia University after the Exposition had closed. While collecting the data, the measurements were tabulated in terms of mouth width and pipe length, from which it was possible readily to transcribe them into terms of vibration frequencies, and into musical nomenclature when desirable. The graduation of the whistle was a difficult task. In the work of determining the vibra- tion frequencies corresponding to the different pipe lengths em- ployed in the actual test, several devices were tried with varying degrees of success. It was found especially difficult to devise any rotating system of sufficient speed and delicacy to register satis- factorily disturbances of a sensitive flame for such rapid vibration rates as those from 15,000 to 40,000 D. V. The optical method for registering vibrations used in physical laboratories, however, was tried for some of the lower pitches, but even here the figures were by no means wholly satisfactory, by reason of the inaccuracies in determining the speed of the rotating system, the sliding of parts, etc., etc. The most satisfactory device for making the graduations, because the most accurate and objective, proved to be the "dust figure" method of Kundt, first adapted to the use of tubes of small bore by Schwendt. 4 The Kundt dust figures are illustrated and described in all general texts on acoustics; hence, they require no elaborate description in this place. The dust figure method is the one, moreover, which is employed at the factory for standardizing the Edelmann whistles before they are sent out. The modifications which are essential to adapt this method to tones as high as those employed for testing the upper limit of audibility are significant as to detail and procedure. It is unnecessary to enter here into a description of the mode of procedure in anything like a detailed way. Something requires to be said, however, with reference to some of the difficulties which are encountered. One of the first essentials with regard to the dust figure method is that the tubes used for resonance chambers shall be of optimum dimensions. A series of tubes varying in length and bore are consequently necessary to secure satisfactory results. For evalu- ating the most acute tones, I drew out some very thin tubes to a bore of 2 mm. and a length of 26 mm. just long enough to allow for the formation of five or six half -wave lengths, that is, for five 'Pfluger's Arch. 75: 546. 1899. 24 THE HEARING OF .PRIMITIVE PEOPLES crests and as many troughs. It is quite essential that a number of these artificial wave troughs and crests be formed if the work is to be at all delicate, since, in making the actual measurements of the length of the several waves, accuracy is enhanced if the span covered by several half waves is measured with the aid of calipers and the figure thus obtained divided by their number to secure the length of a single wave. Tubes with bores of from 6.0 mm. to 10 mm. and lengths from 100 to 250 mm. were used for tones of the sixth and the lower third of the seventh octaves (c 6 -e 7 ), and still larger tubes for the tones of still lower pitch values. Just enough of the lycopodium powder, which had been previously carefully dried, to cover the bottom of the tube, was evenly distributed along its entire length. The tube was then slightly turned so as to raise the powder to one side. This facilitates the formation of dust figures, the aerial disturbance within causing the dust to fall, and while falling, to collect at the points of rarefaction within the reson- ance chamber. In making dust figures, it is necessary that the resonance tubes be kept free from extraneous vibratory influences, else the resulting dust figures will be confused and impossible to interpret. To avoid jars of all kinds, I had the tubes carefully clamped between large pieces of cork, which took up most of the disturbances transmitted to them. Then to make as much as possible of the sound energy leaving the whistle enter the tubes, the whistle mouth was brought as close to the mouth of the resonance tube as possible, and, to further facilitate the movement of air, the ends of the tubes adjacent to the source of sound were flared into a funnel form whose widest diameter was about 15 mm. When everything proceeded favorably, satisfactory dust figures generally resulted in from ten to fifteen minutes, but failures and disasters were frequent. Indeed, much patience and repetition were required to secure perfectly reliable results, frequently as many as six or seven trials being necessary to get anything like a satisfactory measurement. It was essential that each graduation be the average of as many determinations as pos- sible to eliminate chance results. My data, in every instance, are the average of five or more determinations. Let us assume that the dust figures have formed in the resonance tube. Knowing the distance between two adjacent wave crests, it is a simple matter to compute the vibration frequency of the tone that gave rise to the dust figures. The result is accomplished directly by a simple formula, THE INSTRUMENT AND ITS GRADUATION 25 where "N" represents the vibration frequency, "V a " the velocity of sound in air at the temperature prevailing at the time the meas- urement is made, and "L" represents one half wave-length, or the distance encompassed by two adjacent wave crests. The "V a " implies a temperature correction. At Centigrade, under ordinary barometric pressures for this latitude, sound has a velocity of 330.7 M. per second. To secure the corrected value of ee V a " for any given temperature, I used the well-known formula of Kayser and Kirchoff, 5 V a = 330.7 V 1 + 0.00376 t<* During the summer of 1904, within doors on the shady side of a building the temperature varied (according to the United States Weather Bureau reports 6 ), between 17 and 37 C. Our hearing tests, however, were made in a basement where the prevailing temperature was lower, and unfortunately we have no record of this during this period. Still the difference between this and the above, I think, does not amount to a figure to be significant. My corrections were made on the basis of the Weather Bureau statistics. 6 In the sound-proof room of the Psychological Laboratory of Columbia University where the work of graduation was done, the temperature remained quite uniformly at 22.2 C., making the velocity of sound approximately 344 M. per second. Temperature variations can not well be neglected in making tests for the upper limit of hearing with such a device as the Edel- mann whistle, if results be sought which aim to be more than ap- proximately correct. Taking the lower and upper extremes of temperature, during the days that my hearing measures were made, and making the necessary corrections, a whistle cavity length of 1.3 mm. and mouth width of 0.55 mm. would give a vibration fre- quency of 40,840 D. V. and 42,054 D. V. respectively. This varia- tion is certainly sufficient to be quite significant, especially in view of the fact that tests on the different peoples were made during different seasonal conditions. With regard to most of the data relating to the Filipinos, the correction was especially necessary, because the season had so far advanced when these people were measured that the rooms had to be heated artificially, and the temperature, which was almost uniformly 22 C. differed markedly 6 See Wiillner's "Experimental Physik (1894), 1: 931. 8 1 was able to secure from the office of the U. S. Weather "Bureau at St. Louis temperature and humidity readings taken at each hour of the day, for a period covering that included by the taking of the hearing records. From these figures I calculated the true value of " V a " for each hearing record made. 26 THE HEARINa OF PRIMITIVE PEOPLES from that prevailing during the hot summer weather when the data on Indians and whites was collected. Conditions at the Exposition made it necessary to employ a wind blast supplied from the hand bulb, which accompanies the whistle. Some such constant pressure device as that of Whipple 7 was con- templated, but there was so much delay occasioned by the failure of the apparatus to arrive and the general equipment of the labora- tory to be provided, that many of the tests of the upper range of hearing were made before such an equipment might have been in- stalled. Consequently, to keep the conditions under which the tests were conducted as nearly uniform as possible, the hand bulb method of blowing the whistle was permanently adhered to. Edelmann holds that with a whistle of the type I used, the pitch of the tone is only slightly dependent on the wind pressure employed. 8 This statement, however, can be only partially true, and indeed is ap- plicable only to a certain range of variation about the optimum wind blast for blowing it. To investigate the influence of a variable wind blast to some extent, I improvised a wind pressure device which allowed of dif- ferences in the force of the blast. An ordinary wet spirometer, found in the Columbia University Psychological Laboratory, was weighted to the required wind pressure by loading it with slugs of iron ; then, with an ordinary foot bellows, the quantity of air in this reservoir was kept constant. As in all such experiments, a water manometer or U-tube, was inserted in the lead as close to the whistle as convenient to measure the pressure of the air blast passing into the whistle. But before allowing the air to pass through the whistle, it was made to flow through a drying device ; a bottle filled with the crystals of calcium chloride, by which the moisture was, so far as possible, removed inasmuch as moisture in the air tends to interfere seriously with the formation of the ly cop odium dust figures in the resonance chambers. The cavity length of the whistle being 1.3 mm. and mouth width 0.55 mm., with a constant wind pressure indicated by 40 mm. of water in the U-tube, the resulting tone was too faint and weak to produce satisfactory dust figures. It sounded of uncertain piteh and was by no means pure. When the wind blast showed 100 mm. of water in the U-tube, the tone was clear but observably lower than when the whistle was blown by the rubber bulb. With the wind pressure increased to 500 mm. of water, the tone came forth clear and piercing. My T "A Compressed Air Device, etc.," Amer. J. of Psychol. 14: 107. 1903. *Annalen d. Physik, 4 Folge, 11. 1900. THE INSTRUMENT AND ITS GRADUATION 27 subjective judgment was that the pitch was the same as that pro- duced by the use of the bulb. The table below shows the result, in terms of wave lengths and vibration frequencies, for a number of different wind pressures employed : Wind Pressure : Wave Lengths in mm. Pitch Values mm. of water Average of 5 determinations D. V. 40 A faint tone too weak to produce dust figures 100 16.268 21,330 500 8.33 41,217 1,000 8.724 39,460 Rubber bulb 8.428 40,840 These averages were obtained from the following individual de- terminations : Pressure 100 mm. of Water 3 crests measured 23 mm. Wave length 15.34 mm. 2 17 17.00 2 19 19.00 3 22 14.66 3 23 15.34 Average wave length 16.248 mm. 500 mm. of Water 5 crests measured 21 mm. Wave length 8.4 mm. 4 15 7.5 4 18 9.0 6 25 8.34 5 21 8.4 Average wave length 8.33 mm. 1,000 mm. of Water 4 crests measured 19 mm. Wave length 8.5 mm. 6 26 8.66 6 . 23 7.66 5 22 8.8 4 18 9.0 Average wave length 8.524 mm. The Bulb 5 crests measured 19 mm. Wave length 7.6 mm. 4 17.5 8.75 6 25 8.33 6 26 8.66 5 22 8.8 Average wave length 8.428 mm. These figures are in general in agreement with those obtained by Dr. Charles Myers, in investigating a similar problem with the 28 THE HEARING OF PRIMITIVE PEOPLES Gallon whistle. It is to be noted, however, that a wind pressure of 1,000 mm. of water, gave a tone which is actually lower than that produced by a wind pressure of 500 mm. of water, a fact which does not accord with Dr. Myers's 9 experimental conclusions. His experi- ments lead him to believe -that the pitch of the whistles increases regularly with increase of air blast. The difference in favor of the lower wind pressure found by me, I believe, however, is not signifi- cant in that I do not think it exceeds the limits of the accuracy of the method. It is to be further noted that with the bulb, the figures do not differ materially from those with a wind pressure of 500 mm. and 1,000 mm. So far as I was able to determine, the bulb gives approximately a pressure of 800 mm. of water though with a pressure of so short duration it is difficult to evaluate ac- curately the total force given out, with the means at my command. Inasmuch as in all my investigations the bulb supplied the wind blast, it is the graduations in which the same source of wind pressure was used that concern us chiefly. Indeed, for this reason, in all of the work of standardizing the instrument, from which the tables to follow were made, the rubber bulb alone was employed. When using the hand bulb in the making of the graduations, an effort was made to reproduce as nearly as possible the conditions as they obtained in taking the original hearing records. That the condi- tions were more than approximately reproduced and, moreover, that with the use of the bulb, fairly constant conditions can be maintained from day to day, is borne out by the rather uniform character of the data secured. Were one unable to give relatively constant and uniform blasts to the whistle in blowing it with the hand bulb, some marked differences in the vibration frequencies of the tones produced would result, and would show in the dust figures. It is significant that such was not found to be the case, and that no greater differences in the character of the dust disturbances were experienced when the hand bulb supplied the air than when the air came from a uniform and constant pressure source, as is shown in the foregoing tables. To illustrate this point, I shall present some typical measurements in which the hand bulb was employed. As too much space would be required to present the individual measure- ments for the graduations of the whole series of whistle lengths used in the original tests, I will content myself with two samples selected at random. In the following series, the length of the whistle bore measured 1.5 mm., the mouth width being 0.55 mm.: "The Pitches of Galton Whistles," J. of Physiol. 28: 417. 1902. THE INSTRUMENT AND ITS GRADUATION 29 2 crests measured 9.2 mm. Wave length 9.2 mm. 12 57.1 9.52 12 52.81 8.8 20 87.34 8.73 14 65.41 9.33 Average wave length 9.116 mm. Vibration frequency 37,560 D. V. In the next series, the cavity length of the whistle was 2.2 mm., the mouth width remaining the same as in the preceding experiment : 2 crests measured 16.0 mm. Wave length 10.66 mm. 4 22.3 11.15 9 44.1 9.08 6 31.2 10.04 3 15.4 10.26 Average wave length 10.45 mm. Vibration frequency 30,270 D. V. These two sets of measurements are in every way typical of all that were made and indicate about the same degree of variation between the individual determinations as was experienced on the average. The same procedure, as in these samples, was followed for every whistle length employed in the original measurements. The accompanying table gives the vibration frequencies corresponding to each whistle length as empirically determined by aid of the Kundt-Schwendt resonance method. In this table the distance between the lips of the whistle remained uniformly at 0.55 mm. The average wave-length of the several determinations has been omitted in each case since it would afford no information vital to an interpretation of the figures presented : Length of whistle Vibration Frequencies Length of whistle Vibration Frequencies cavity in mm. (Complete Vibrations) cavity in mm. (Complete Vibrations) 1.2 42,960 2.5 28,048 1.3 40,840 2.6 27,448 1.4 39,220 2.7 26,854 1.5 37,560 2.8 26,264 1.6 36,360 2.9 25,724 1.7 35,100 3.0 25,212 1.8 34,000 3.1 24,754 1.9 33,060 3.2 24,196 2.0 32,180 3.3 23,020 2.1 31,170 3.6 22,217 2.2 30,270 3.8 20,973 2.3 29,508 4.0 18,490 2.4 28,766 CHAPTER IV DATA COLLECTED ON THE UPPER LIMIT DURING the earlier months of the Exposition, while we were getting our bearings, equipping our laboratories, and installing our apparatus which was somewhat tardy in arriving, we spent our time amusing the public and, incidentally accumulating data on some few tests. For the most part we limited ourselves to a single test and measured as many individuals as we could in this one particular only. In consequence, I was enabled to secure considerable material relating to the upper threshold of hearing. Unfortunately these data had to be secured under somewhat unfavorable conditions, in that the test was always made in the presence of a crowd. The sound room had then not yet been completed and there were many distracting noises that might have tended to distort the results to some extent. On the whole, notwithstanding, I believe the data satisfactory, since a comparison of this material with some secured on whites in the sound room showed no significant differences. The individuals were tested one at a time. The Edelmann whistle was held twelve inches from the subject's ear and blown; the other ear meanwhile being closed by pressing the tip of the finger into the auditory meatus. At first the pipe was so adjusted that a tone resulted whose pitch was so low as to be easily sensed by all ears. The pitch was then gradually raised until a point was reached where the tone was no longer audible. A reading was then taken and recorded for the last audible sound. Beginning with an inaudible tone, the pitch was now lowered until it could again just be sensed and this whistle length recorded. The average of the two tabulations, if a difference existed and there usually did was taken as the measure of the upper limit of the person tested. Each ear was tested. Almost without exception, the right ear was the one first examined. In the measures on primitive peoples, the procedure was in all respects essentially the same as just outlined, except that the tests were made within a specially constructed booth. 1 Although this 1 This booth was constructed in one corner of a room of our suite, which had been closed to the public, being set apart by us for making such measure- ments as required privacy. The dimensions of the booth were approximately six by nine feet, and seven feet in height. On two sides the heavy stone and 30 DATA COLLECTED ON THE UPPER LIMIT 31 room was not completely sound-proof, it did exclude extraneous noises sufficiently well to keep the subject free from distractions. To contribute still further to this same end, the person tested was kept in the dark and his back turned to the experimenter, in order, so far as possible, to exclude visual stimuli and allow the subject's mind to concentrate wholly upon the auditory sensations presented to him. Moreover, the subject was seated in a chair, in as com- fortable a position as possible. In other words, I attempted to make the external conditions as suitable as could conveniently be done for sensing the auditory stimuli and consequently securing the most nearly normal results. By way of check, a record was made of my own hearing im- mediately following that of each individual tested. Conditions made it necessary for me to make personally the test of my own hearing. Still I do not think that the source of error arising from this method is really significant. Of course, the element of expectancy was some- thing of a factor so that there might have entered into the experi- ments auditory images which it were easily possible to confuse and mistake for the stimulus at a time when it was really inaudible. I have, however, practically no auditory imagery, at least, I have never been able to observe any in myself, so it is scarcely possible that images of sufficient intensity to be confused with even faint stimuli should have arisen under the circumstances just noted. In any event, the factors concerned with the personal test did not vary from day to day or from hour to hour, and hence are not significant on the whole. In this connection may be mentioned the need for this check, discovered during the time the tests were being made. It frequently occurred that the whistle became partially clogged with bits of rubber, or other matter coming from the inside of the rubber bulb. This clogging had the effect of lowering the pitch of the upper tones from a fifth to an octave, although otherwise no ap- parent change in the character of the tones resulted, so that had it not been for the personal check-test, the subject would have received an unfairly high rating. brick walls of the building served to exclude the most penetrating sounds. Although the remaining sides were of wood, the walls were double, the space of about four inches between the two being filled with sawdust. Sawdust to the depth of six to eight inches also covered the roof while the whole booth rested in a bath of sawdust in order to exclude any sounds which might be conducted into it through the cement floors of the building. Entrance was made through a single padded door, and the sole illumination came from an incan- descent electric bulb which was suspended over the apparatus. The arrange- ment was such as to exclude extraneous sounds entirely for all practical purposes. 32 THE HEARING OF PRIMITIVE PEOPLES It is unnecessary to name all the factors and considerations which entered vitally into the making of the tests, but there is still one which requires to be mentioned. This relates to the ear's suscepti- bility to fatigue. Observers are unanimous in their experience with respect to this point. Professor Seashore 2 has remarked in con- nection with his audiometer that to be satisfactory for testing chil- dren, owing to the presence of fatigue, a device must be employed which. will make it possible to complete the test in not to exceed two minutes. Especially with subjects who are unaccustomed to making introspective observations, it is found that tones die out and are inaudible long before threshold values are reached. Par- ticularly is this true where continuous tone devices are employed in making auditory measurements such, for example, as the tuning fork, the sounding rod, or the Galton or Edelmann whistle, blown by some constant and continued air supplying device. For this very reason such devices and means for testing the range of audi- tion are unserviceable. The Edelmann whistle blown by means of the bulb with its short, quick sound obviates all this difficulty. It is a difficulty, too, which can not be overlooked with safety, especially in dealing with children and primitive peoples. The Measurements: We shall now turn directly to the data regarding racial differences in the upper limit of hearing. In Table I. are presented the figures representing the averages for both the right and left ears. In the first column are indicated (1) the num- ber of individuals constituting each group examined; and in parallel columns; (2) the averages; (3) the average of the devia- tions from each average; and (4) the standard deviation, of each group. From these data it is possible, without difficulty, to compute directly any of the different variability coefficients desired, and in consequence, the reliability of each average, and the probability of a difference between any two groups. Owing to the fact that among the Filipinos, the Pigmies, Patagoniaus and Cocopa, there were no women tested, the data have not been separated so as to show the influence of sex, except that relating to whites. In tables II. and III. the individual records are distributed, so as to present in parallel columns a picture of the distribution of the individuals of each group examined. These tables represent the figures for both the right and the left ears. In Table IV. is pre- sented a distribution of the cases according to age, for the three most numerous groups measured. These again have been placed in parallel columns to afford a more ready comparison, and to give 2 Univ. of Iowa Studies, 2: 55. 1899. DATA COLLECTED ON THE UPPER LIMIT 33 in pictorial form the relations which the different groups sustain to each other in this respect. For the smaller groups, the ages will be presented in connection with their discussion. TABLE I UPPER LIMIT OF AUDIBILITY Racial Differences as Shown by Averages No. Right Ear Left Ear Average A. D. S. D.* Average A.D.3 S. D.* Whites. . 156 63 97 10 7 7 6 3 32,285 D. V. 31,975 " 29,916 " 32,123 " 28,846 " 28,269 " 33,223 " 30,240 " 2271 2190 1755 827 1666 1209 2071 3240 2344 2663 2180 977 1873 1413 2468 3551 33,087 D. V. 31,580 " 29,886 " 31,794 29,529 " 27,571 " 34,081 " 28,630 " 1891 2460 1737 1408 2946 819 3212 2366 2482 3028 2089 1566 3199 852 3428 2592 Indians Filipinos Cocopa. . . . A inns Vancouvers Pigmies Patagonians TABLE II RIGHT EAB UPPEB LIMIT OF HEARING Distribution of Individual Cases Vibration Frequency Whistle Length Whites Indians from School Christian Filipi- nos Cocopa Indians Ainu Van- couver Indians Pigmy Patagonian Indians 42,960 1.2 1 _ _ _ _ __ 40,840 1.3 39,220 1.4 3 1 ___ 37,560 1.5 8 1 __ 36,360 1.6 9 6 __ ^_ __ 35,100 1.7 10 6 2 1 1 34,000 1.8 15 4 2 1 __ 33,060 1.9 18 6 4 2 __ 1 32,180 2.0 26 6 11 1 1 31,170 2.1 20 12 14 4 1 1 __ 30,270 2.2 18 8 20 2 1 1 29,508 2.3 6 3 10 __ 1 1 28,766 2.4 6 5 5 2 1 28,048 2.5 7 3 10 1 1 __ 27,448 2.6 3 1 2 1 __ 26,854 2.7 3 1 5 ___ 2 1 26,264 2.8 2 1 3 25,724 2.9 1 1 5 25,212 3.0 1 - __ 24,754 3.1 __ __ 24,196 3.2 1 1 1 __ 23,020 3.4 1 22,217 3.6 1 __ 20,973 3.8 1 __ 18,490 4.0 8 A. D. Average Deviation. * S. D. Mean Square Deviation. THE HEARING OF PRIMITIVE PEOPLES TABLE III LEFT EAB Distribution of Individual Cases Vibration Frequency Whistle Length Whites Indians from School Christian Filipi- nos Cocopa Indians Ainu Van- couver Indians Pigmy Patagonian Indians 42,960 .2 40,840 .3 39,220 .4 1 1 1 37,560 .5 8 2 1 36,360 .6 3 2 35,100 .7 9 5 1 34,000 1.8 14 8 1 33,060 1.9 18 6 6 1 1 32,180 2.0 9 4 11 2 1 31,170 2.1 11 11 21 3 2 30,270 2.2 5 3 10 1 1 29,508 2.3 2 11 13 1 28,766 2.4 5 7 1 2 28,048 2.5 2 4 5 1 27,448 2.6 1 5 1 26,854 2.7 1 5 1 4 26,264 2.8 1 6 1 25,724 2.9 3 2 25,212 3.0 24,754 3.1 1 1 24,196 3.2 1 23,020 3.4 22,217 3.6 1 20,973 3.8 1 18,490 4.0 TABLE IV TABLE SHOWING THE NUMBEB OF PERSONS OF EACH AGE OF THE THREE MOST NUMEROUS GROUPS OF PEOPLE MEASURED Ages 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 12 14 12 19 Whites ............. 5 Indians (School) ... 12 Filipinos ............ 3 Average age: Whites, 23 years, 5 months; Indians, 19 years, 2 months; Filipinos, 21 years, 1 month. 6 11 1 10 16 8 8 12 16 14 4 20 14 2 12 13 1 5 11 1 5 10 2 2 Reference to Table I. shows that of Whites, the records of 156 individuals were included in this study. Owing to some rather significant changes, that occur in the range of audibility during the earlier and later years of life as has been already indicated by the studies of Zwaardemaker, 5 Alderton, and others, it was thought advisable to include in these data only the records of those in early manhood and womanhood, those individuals whose ages ranged See Alderton, Arch, of Otol. 23: 171. 1894; 25: 45. 1896; also Zwaarde- maker, Ztschr. f. Psychol. 7: 10. 1894, and Arch. f. Ohrenhk. 32: 53; 35: 299. DATA COLLECTED ON THE UPPER LIMIT 35 from sixteen to thirty years. During these years, no very signifi- cant changies have been discovered in the respect just indicated. I excluded also, not only in case of the whites but also among the other races, the record of every individual who had noticed particu- larly any diminution in his hearing acuity. Obviously those with defective hearing should not be included in any comparative meas- ure of the hearing function, since they form a distinct functional group or species. The average age of the whites was found to be 23 years and 5 months. About one third of them were older than 25 years, while one fifth only were younger than 20. The majority, therefore, of the Whites used here for comparative purposes were between the ages of twenty and twenty-five years men and women in the younger years of adulthood, at a period of life most favorable for accurate testing. On the average, the results show a slight superiority of the left over the right ear, the average for the latter being 33,087 D. V., as against 32,285 D. V. for the former. This is in accord with the observations of Preyer and Fechner, who observe that the left ear is superior to the right in all its functions, due, they believe, to the fact that human beings are left brained. In the case of my own measurements, however, much may have had to do with the order of testing the two ears. During these tests, the right ear was almost invariably first tested, so that a mental element arising from practise may have been responsible for the superiority of the left ear. If instead of the average we take the median as the measure, the difference in result is not changed. The reliability of the averages, which may be directly computed from the S. D. given in Table I., is such as to indicate that the chances are about ten to one that the true average will not vary from the one given by more than 200 vibrations, which in reality, is within the range of instrumental accuracy; that is, the true average will not be as high as 32,500 vibrations to the second or as low as 32,000. Reference to tables II. and III. will show that the distributions are fairly normal. Exclusion of the records of all individuals who had experienced some hearing defect has served to eliminate the skewness which might otherwise be looked for at the lower end of the curve. INDIANS So far as I am aware, no study of the hearing of our American Indian has ever been undertaken. There are, to be sure, in the literature relating to this interesting race of people, some general ob- 36 THE HEARING OF PRIMITIVE PEOPLES servations of a wholly unscientific character which, for the most part, attribute to the savage of the American forest and plains, a remarkable sensory acuity. The works, for example, of James Feni- more Cooper, contain statement after statement, all purporting that the Indian has ears that hear so keenly that he is able to detect sounds in the forest that are wholly inaudible to the ears of a white man most favorably gifted in this respect. And, indeed, whether from the popular literature relating to Indians, or from a pre- conceived notion that a savage ought to be superior to civilized peoples in sensory acuity, the opinion generally prevails that the ears of the Indians are very much keener than are those of the Whites. No doubt much of this conception arises from the belief in what is commonly known as the doctrine of compensation. Accord- ing to this view, if one sense or mental function is lacking, in any respect, the others are the keener to compensate for the loss. More- over, based on the olfactory sense of the dog, the visual acuity of the hawk and the superior audition of certain of the felines, there has arisen the belief that the senses degenerate under the influences of civilization and higher culture. The figures relating to the upper limit of the hearing of the vari- ous Indian tribes represented at the Model Indian School at the Ex- position, are given in tables L, II., III. and IV. 6 The hearing of 71 Indians included in this group was taken; 14 full-blooded males and 13 of mixed blood; 4 full-blooded females and 40 of mixed blood. Of this number, 8 were younger than sixteen years so that data relating to them are not included in the general average. Only 6 individuals of the group were older than 25 years (See Table IV.) while but 12 of the 63 were older than 20 years. In other words, 81 per cent, of all the Indians examined were between 16 and 20 years old. The average age of the entire group is only 19 years and 2 months, while that of the Whites with whom they are compared is 23 years and 5 months. It is therefore evident that, if the upper range of hearing progressively decreases from earliest childhood, the Indians are favored in the comparison with the Whites. For the right ear, the tone marking the upper threshold of hearing of Indians is on the average 31,975 double vibrations to the second, with an average deviation amounting to 2,190 vibrations. The average for Whites for the same ear was 310 double vibrations higher. On purely statistical grounds 7 one would be justified in For general remarks relating to these Indians more specifically consult page 4. 7 In making these computations, the formulae commonly used in statistical data for measuring the reliability of a difference were employed. See Thorn- dike, " Mental and Social Measurements," 1904, p. 139, et seq. DATA COLLECTED ON THE UPPER LIMIT 37 inferring that the chances are almost two to one (exactly 1.91 to 1) that an actual difference exists between the upper limit of hearing of Whites and Indians respectively. The difference however is too small, considering that neither group is very numerous, to point strongly in the direction of a real difference. The age factor, more- over, is of some importance in affecting the data as may be seen when we take the records of those individuals only whose ages run from 16 to 20 years the figures which encompassed the largest number of the Indians examined. We have, then, 51 Indians and 43 Whites. When this reduced group is taken, the average upper threshold value for Indians, in case of the right ear, is found to differ only slightly from that found for the whole group, namely, 32,080 vibrations. On the contrary, the average for the Whites, when thus limited to those between 16 and 20 years, is increased to 33,587 vibrations. On statistical grounds, therefore, the chances are 10 to 1 that the upper limit of hearing for Whites is 1,000 vibrations in excess of that for Indians, and 400 to 1 that a real difference exists between the upper limit of hearing of the two peoples, on the average. Turning our attention now to the left ear, the difference between the two peoples seems still greater. The left ear too probably gives a figure which more nearly represents the actual state of the organs on account of the practise which the individual had in the experiment, as pointed out above. For this ear, on the average, the Whites hear tones of a pitch amounting to 1,500 vibrations higher than do Indians. Moreover, the probability of a real difference between the two peoples is at least 500 to 1, which, indeed, is extremely high. Although the conclusion is not so positive as though a larger number of each group had been measured, yet the data point in that direction to a degree amounting almost to certainty. Not only do Whites hear tones, which are more acute, on the average, than do Indians, but a glance at the distributions found in tables II. and III. will show that the entire curve for Indians for both ears extends lower than that of Whites. The amount of variability likewise ap- pears to be greater in the case of Indians. As regards the right ear, one white woman only heard a tone higher than any of the Indians, and for the left ear, the best Indian did as well as any White. Again looking at tables II. and III. from another angle, it may be observed that the relative inferiority of the range of Indians* hearing is not due to a few extreme records, which would have a tendency to distort the figure representing the average. As re- gards the right ear, the data presented show that 44 out of 67, or 66 per cent, of the records of Indians fall below the average for 38 THE HEARING OF PRIMITIVE PEOPLES Whites, and in case of the left ear, 41 of the 65 records or 63 per cent. Cocopa Indians The Sen. Of this tribe, we were able to make COCOPA INDIANS Highest Audible Tone Name Hi . Skik Mert Age 6 14 14 El Puck 15 Jack 17 John Roy 18 Joe 20 Jerry 40 Pablo 56 Tom 70 Average 32,123 S. D. 977 Right Ear 32,180 D.V. 31,170 33,060 31,170 33,060 31,170 34,000 31,170 ( 18,000 (16,000 Left Ear 30,270 D. V. 31,170 " 34,000 " 31,170 " 31,170 " 33,060 " 34,000 " 29,508 " (21,420 " ) 31,794 1,566 measurements of ten males only. On account of the variations in age, too, the data are not very suitable for comparative purposes. The youngest member of the group was a lad of six years, while the oldest had passed the age of three score and ten. All of the essential data with reference to these people are presented in detail in the table above. The records of the two oldest men have been enclosed in parentheses to indicate that they have not been included in casting up the averages for the whole. Both Tom and Jerry heard even ordinary conversation with difficulty, so that obviously their range of hearing could not be taken as normal to the group. Excluding, then, these two records, it is seen that the average upper limit of audibility of Cocopa Indians for the right and left ears respectively is 32,123 and 31,794 vibrations (double) to the second, figures which do not differ materially from the averages obtained from the more intelligent Indians at the U. S. Government schools. However, considering that only one of the eight individuals whose hearing records contribute to the average was older than 20, it is probable that the range of hearing of Cocopas in the long run would be found to be less than that of the more intelligent Indians; but even so, the difference would likely be psychological rather than organic. Compared with the upper limit of hearing of Whites, the Cocopas fall significantly lower. Not only is this true as regards the average result, but age for age, the individual records are found to fall decidedly below the medians for Whites, as may be seen by com- paring the distributions of the groups exhibited in tables II. and Name Bob Age 40 Sex Male Tribe Kwasruitl Right 26 264 Charley Jasper 28 27 u ( Nutken 28,766 28048 Curley 21 it 26 264 Atleo Ellen 64 35 (( Female M a 30,270 29508 Anna 30 M K 28,766 Average 28 269 S. D. 1.413 DATA COLLECTED ON THE UPPER LIMIT 39 III. In case of the upper limit of hearing for the right ear, 6 of the 10 Cocopas stand lower than the median record for Whites and 6 of the 9 in case of the left ear. Vancouver Indians Kwaguitls and Nutkens.0 Vancouver Indians, we tested five males and two females, belonging to two local tribes. VANCOUVER INDIANS KWAGUITLS AND NUTKENS Highest Audible Tone Left Ear 26,854 D.V. 28,766 " 26,854 " 28,048 " 28,766 " 26,854 " 26,854 " 27,571 " 852 " With the exception of Atleo whose hearing record is the best of the group these Indians were all in the prime of life. They are Indians, moreover, who have only to a small degree been influenced by contact with Whites, inasmuch as their homes are far removed from those parts where civilization has made its march. So far as the data will permit of generalization, therefore, I believe these peoples to be fairly representative of the more intelligent Indian of North America, before his organism has become modified by its adjustment to the social conditions and habits of the invading Caucasians. In the case of both ears, the degree of variability among the records is fairly small. By reference to the table above, it is seen that the range of all the cases is encompassed by less than 4,000 vibrations to the second; or that part of the musical scale lying between gis 6 and bis 6 , i. e., within the range of less than two whole tones. The records for the left ear (exhibited in Table III.) are lower, as a whole, but the difference between the highest and lowest record is still smaller, being only about a semi-tone (1912 D. V.). It is, however, worthy of note that not a single record of the Vancouver Indians is as high as the average for Whites. In fact, the highest Vancouver Indian record for the right ear is 1,000 D. V. lower than the mean for Whites, while the highest for the left ear is 3,500 D. V. lower. The numbers are too small to apply statistical methods with satisfaction, but there certainly is no question of the evident tendency toward the inferiority of these Indians in hear- 40 THE HEARING OF PRIMITIVE PEOPLES ing range, as compared with. "Whites. The records of the two women are poorer than those of any of the men, but additional representa- tives of the women might tend to reverse this result. Patagmvian Indians The Tehuelche.We were able to examine only four men of this tribe, the data concerning whom are given below in detail. These Indians represent a grade of culture slightly lower than that of the Vancouvers just considered; a tent-living instead of a house-dwelling people, a nomadic instead of a home- building folk. PATAGONIAN INDIANS THE TEHUELCHE Highest Audible Tone Name Age Right Ear Left Ear Cosimero 24 26,854 D. V. 26,264 D. V. . Canjo 35 28,766 " 27,444 " Senchel 55 Hearing very defective. About 10,000 D. V. for both ears. An acuity defect. Boni Farci 18 35,100 D. V. 32,180 D. V. Average 30,240 " 28,630 " S. D 3,551 " 2,592 " Only three records were included in the average. The number examined is too small to draw any general conclusions. However, with the exception of Boni Farci, the upper limit of hearing of all was found very much, inferior to that of the average for Whites. I leave the data without further comment. Indians as a Whole. From the foregoing, it is evident that whether taken tribe by tribe or as a whole, on the average as well as individually, the experimental results indicate that Indians do not possess as great a range of hearing as do Whites. If, indeed, we lump the records of all the Indians irrespectively, those who have had the cultural advantages of the Whites the tribes represented in the group taken from the Indian School with the smaller groups represented by Indians closer to nature, we find that for the right ear, 60 of a total of 83, or 73 per cent, of the individuals rank below the average for Whites, and 64 of 82 Indians, or 77 per cent, stand lower than the average of Whites as regards the left ear. It is worthy of note too that the better records were made by those Indians who had attended, more or less, the Government Indian Schools. It was my impression also, formed at the time of making the measurements, that the better records were made by those indi- viduals who were, all around, more intelligent and alert. Un- fortunately, it had not occurred to me to make record of the degree of intelligence of each person when tested, although it might be nothing more than a personal opinion based upon observation only. DATA COLLECTED ON THE UPPER LIMIT 41 Roughly, this could easily have been done with a fair degree of ac- curacy, since I was associated with and saw at work each person for the better part of an hour. Had I made such an observation, it would have enabled me to determine in a loose way, at least, how far the differences in the upper limit of hearing between Whites and Indians are psychological and to what extent organic in character. A little further on, we shall have occasion to revert to this subject again. Filipinos. In all, 97 Filipinos were tested for their upper threshold of hearing. The data for four, however, were so palpably in error, that the records were rejected. We thus have 93 records which form the basis for the following study. So far as I am aware, no statement has ever been made with reference to the hearing of any of the Filipino people, in any of its aspects. It is doubtful, too, whether so good an opportunity for testing these peoples, and under circumstances so favorable, will soon be found again. If taken in the Islands, it would be no easy matter to collect into a laboratory, for testing, as many as a hundred individuals, representing, as did these, almost every section of the Philippine Archipelago. 8 We feel especially gratified with the results obtained on the Filipino peoples. For the opportunity of making measurements of the Filipinos, I am under obligations to Major Haskell, IT. S. A., under whose orders the men came to the laboratory for the testing and without which orders it is doubtful whether the tests could have been made at all. Obligations are also due to Dr. Wilson, of the Philadelphia Museum, who, in the capacity of Director of the Philippine Exhibit, used his influence in our behalf to the extent of recommending that we be given permission to make the measurements. The men were brought to the laboratories, four at a time, this number being tested in the forenoon and the same number imme- diately after dinner. It may here be stated that the hearing ex- aminations were but two of a great number made on each indi- vidual. The men were ignorant as to the object of the measurements nor was there a hint given of their purpose. Notwithstanding, they approached the tests with a great deal of interest and zeal, and seemed eager to compare their several records with those of their fellows. None of them suspected that there might be such a thing as a racial difference inasmuch as they referred to my record as one, in all respects, comparable with one of theirs. The men were taken into the sound-room, one at a time, and the test conducted with as great dispatch as possible in order to avoid the flagging of interest and the onset of fatigue. 8 See page 6 for a more detailed description of these peoples. 42 THE HEARING OF PRIMITIVE PEOPLES It is seen, from the age distributions found in Table IV. (See p. 34) that 4 men were younger than 18 years and 5 older than 25. The oldest of the group was but 30. Were one to select the indi- viduals deliberately with the test of the upper threshold of hearing in view, it is difficult to see how a more favorable lot could have been secured. The mean age was 21 years and 1 month. Turning now to tables I., II. and III., it is seen that the figure representing the upper threshold of hearing of Filipinos for the right ear was found to be 29,916 vibrations to the second on the average, and for the left ear 29,886 vibrations. The small degree of variability found in the group is at once striking. The average deviation is very much smaller than that for either Whites or In- dians (See Table I.). So little variability, no doubt, may be ac- counted for by the fact that the variation in age is likewise small and, moreover, it may be noted that in mental alertness and general intelligence, these individuals were more nearly on a par than those of any other group measured, and, consequently, all mental factors concerned in the tests would become more largely equalized. When compared with Whites, the upper limit of hearing of the Filipinos is decidedly lower, not only on the average, but also as regards the general distribution of the individual cases ; a fact which stands out in the general distributions found in tables II. and III. The figures show that Whites on the average have an upper limit of hearing higher by 2,369 vibrations for the right ear and by 2,301 for the left. On the basis of mathematical probability, the chances are such as to amount almost to an absolute certainty (10,000 to 1) that, even were the numbers infinitely increased, the upper limit of Filipinos would, on the average, be lower than that of Whites. Indeed, the chances are about 4 to 1 (3.9 to 1) that the difference between the upper limit of hearing of Whites and Fili- pinos amounts to at least 2,000 vibrations. And, inasmuch as the number measured is sufficiently large to render the data susceptible to fairly accurate statistical treatment, the reliability of these figures may be accepted with a certain degree of confidence. Something of the standing of the upper limit of hearing of the Filipinos as a whole may be inferred from the distribution of the records as found in tables II. and III. Not only the mode, but the distribution as a whole is found to fall distinctly below that for Whites. Take, for example, the records of the right ear. But 19 of the whole number of 93 (about 20 per cent.) are as high as the median record for Whites ; for the left ear, only one Filipino record is as good as the median for Whites. In the data just presented, were included the records of 13 DATA COLLECTED ON THE UPPER LIMIT 43 Filipino students attending American colleges and universities, who were temporarily connected with the Exposition and whom we tested as they chanced to stroll into the laboratories. These records afford an opportunity for testing to what extent the factor of in- telligence was instrumental in effecting differences found to exist between the upper limit of Filipinos and Whites. In mentality, I take it, these Filipino students were on a par with the freshmen and sophomores in our American colleges and universities. They ranked, moreover, in point of intelligence at least equal to the Whites with whom the Filipinos are compared above, inasmuch as the Whites were for the most part artizans and tradespeople, with a few who had completed college courses. Below I give the individual records of these Filipino students, both for the right and left ears. Also I give the averages of the thirteen records and the variability. FILIPINO STUDENTS MOSTLY TAGALOGS Right Ea 31,170 D 34,000 31,170 32,180 29,508 33,060 30,270 30,270 V. Highest Audible Tone Left Ear Right Ear 33,060 D.V. 32,180 D. V. 31,170 31,170 32,180 28,766 30,270 30,270 28,766 32,180 35,100 30,270 33,060 Average 31,878 A. D. 1,323 Left Ear 32,180 D. V. 32,180 33,060 28,766 33,060 31,157 1,361 Even these intelligent and educated Filipinos, it is seen, have an upper range of hearing distinctly lower, on the average, than do Whites; the difference for the right and left ears respectively being 407 and 1,930 vibrations. Were the numbers of these intel- ligent Filipinos larger, it would be interesting to note what the prob- abilities of a real difference are between their upper limit and that of Whites. The data would incline one to believe that they would amount to practically a certainty. It requires to be noted, more- over, that with one exception, the records of the Filipino students were among the best of the Filipinos tested, so that there can be no doubt but that the psychological factor is significant in accounting for some of the differences discovered between Filipinos and Whites, though, to be sure, by no means all. Army officers, who had done service for some time, in the Philip- pines, and some teachers who had spent two or three years in the Islands, with whom I conversed, without exception informed me that they, too, had experienced defective hearing in the Philippines. But they attributed this abnormality to the action of the quinine, 44 THE HEARING OF PRIMITIVE PEOPLES which they found it necessary to take in large quantities, on account of the prevalence of fevers. Since the natives are immune to malaria and other tropical maladies, and do not use quinine or other drugs with similar medicinal properties, the explanation just noted could not account for the defective hearing of the Filipinos, which, as will hereafter be pointed out, extends not only to a diminution of the range but to an actual lack of acuity as well. It would be in- teresting to know whether this is common to all dwellers within tropics, inasmuch as the results are in accord with the data collected from the inhabitants of the Torres Straits.* Amu. Excepting the Vancouver Indians, the Ainu rank lowest as regards the upper limit of hearing of any of the peoples I examined. Owing to the fewness of the numbers, it is impossible to do more than present the results with a statement of their general tendency, and perhaps this can best be seen from the comparative tables II. and III. It may be noted that all of the Ainu records fall lower than the average for Whites for both the right and left ears. The general tendency seems to indicate an upper limit slightly lower even than that of Filipinos. THE AINU Highest Audible Tone Name Age Sex Right Ear Left Ear Yazo Osawa 23 Male 31,170 D. V. 31,170 D. V. Kutorge Hiramura 38 " 30,270 " 24,754 " Santukuno Hiramura . . 53 " 25,212 " 26,264 " Goro Bete 28 " 29,508 " 32,180 " Sangea Hiramura 56 " Less than 10,000 D. V. (both ears). Record not included in the average. Ume Osagwa 19 Female 28,048 D. V. 33,060 D. V. Shutratek Hiramura . . 33 25,212 " 26,264 " Kin Hiramura 6 " 30,270 " 30,270 " Average 28,846 " 29,529 " A. D 1,666 " 2,946 " Much of the auditory inferiority of the Ainu is undoubtedly to be accounted for on purely psychological grounds. They seemed an excessively stupid people, ranking next to the lowest of all the primitive peoples collected on the Exposition grounds. Their minds seemed unresponsive and lethargic. They apprehended meanings poorly. Things once apprehended, moreover, held their attention for a moment only when they seemed immediately to relapse into a state of mental indifference. I never could feel quite certain that they were hearing even when they said they were, inasmuch as in the region of the threshold values the number of false statements was exceedingly large. "See Report Cambridge Anthropol. Exp., Vol. 11. UN! DATA COLLECTED ON THE UPPER LIMIT 45 The Ainu have been little studied and little is known of them. So far as I am aware, no scientific measurements of them have ever before been made, so it is quite unfortunate that more of the race were not available for study. It is a race, too, which, according to travelers and missionaries, is fast dying out; hence, in the near future, will no longer be open to scientific observation. The Pigmies Batwa, Batsnba and Cheri Cheri. the so-called Pigmies proper, there were at the Exposition only six representa- tives. I present hearing records of the entire number. THE PIGMIES BATWA, BATSUBA AND CHEBI CHERI Highest Audible Tone Name Shamba Age . .. 30 Tribe Batwa Eight Ear 31,170 D. V. Left Ear 31,170 D. V. Malinga ... 16 M 37,560 " 37,560 " Bushaba 13 1C 32 180 " 39,220 " Latuna" . . . 15 Batsuba 35,100 " 35,100 " Otabenga .. . 17 Cheri Cheri 30,270 " 31,170 " Linn 1110 . . 17 33 060 " 30,270 " Average . . . . . 33,323 " 34,081 " A. D. . 2.071 " 3,212 " The records of the Pigmies are all high. Were the same relative distribution to continue for a hundred Pigmy hearing records, they would be found to possess an upper range of audition superior to that of any other people, including Whites. The general distribu- tion of their results stands higher than that for Whites. ( See tables II. and III.) Only one white male was found to possess a range of hearing higher than that of any Pigmy. How shall we account for this manifest superiority of the Pigmy's hearing? Certainly not on the basis of a superior mental attitude toward the tests, since the Pigmy ranks low in the mental scale only slightly higher than the Ainu whom we have just con- sidered. Moreover, their interest in the test can not be said to have been especially keen. It was certainly by no means the equal of the Filipinos whose upper limit was found to be especially low. The question of age was a factor, no doubt, since the Pigmies were all boys, but even this does not account for all the difference found. Perchance, something of a relation may exist between a high degree of sensitivity and extreme motility. The Pigmy is a motor indi- vidual. Like his next of kin, the negro, in his native haunts, he is perpetually on the move. His reactions to incoming stimuli likewise are direct and excessively overt. Other than this, there is certainly nothing in the Pigmy's environment or mode of living which should tend to develop and cultivate any peculiar aptitude in the way of sensory acuity, with which he seems naturally to be gifted. CHAPTER V THE UPPER LIMIT OF HEARING AS AFFECTED BY AGE AND SEX FROM the data I was able to collect at St. Louis and some gathered subsequently from children in the public schools, it was possible to select material which will throw some additional light upon the relation that obtains between range of hearing and age and sex. As regards Whites alone, the data represents tests on 385 individuals; 209 males and 176 females, ranging in age from 5 to 65 years. By consulting the tables and charts which follow (See tables V. to XII.) it will be observed that the records of the individuals have been distributed into four-year groups, the first group representing the upper limit of hearing of children between 5 and 8 years in- clusively; the second group 9 to 12 years, and so on. Beyond 49 years, the numbers tested were so few that the four-year groupings were dropped and the records lumped. Such a procedure, too, is justifiable on other grounds. With the approach of senescence is met a decrease of sensitivity, in general, in consequence of which many of the data would bear record of a decline in general sen- sibility rather than a diminution only in the one particular function of hearing piercing tones. In Table V. the data are so arranged as to contribute informa- tion toward the significance of age in influencing the upper limit of TABLE V HIGHEST AUDIBLE TONE, ACCOBDING TO AGE White Males and Females White Males and Females No. of Ages Cases Right Ear. Left Ear. Right Ear. Left Ear. Average Average Median Median 5-8 41 34,826 34,525 34,000 35,100 9-12 32 34,614 34,939 35,100 34,200 13-16 54 34,418 34,224 34,000 34,000 17-20 40 32,466 32,415 32,480 32,200 21-24 48 33,491 33,025 32,480 33,800 25-28 53 31,557 32,390 32,180 32,200 29-32 31 31,464 32,000 32,180 31,200 33-36 17 28,816 29,046 28,900 29,000 37-40 27 27,512 25,054 26,854 28,100 41-44 12 27,953 29,994 29,508 27,600 45-48 20 27,382 27,741 26,854 27,500 49+ 12 26,020 26,188 28,048 25,212 46 THE UPPER LIMIT OF HEARING 47 TABLE VI HIGHEST AUDIBLE TONE, ACCORDING TO SEX White Males White Females Ages No. of Cases Right Ear. Average Vibra- tion Frequency Left Ear. Average Vibra- tion Frequency No. of Cases Right Ear. Average Vibra- tion Frequency Left Ear. Average Vibra- tion Frequency 5- 8 18 35,180 34,926 23 34,535 34,211 9-12 17 34,500 34,740 15 34,861 35,082 13-16 31 34,207 33,893 23 34,671 34,713 17-20 24 32,103 31,546 16 32,991 32,849 21-24 26 33,069 31,358 22 33,565 33,714 25-28 19 30,834 31,480 34 31,048 32,059 29-32 22 31,105 31,761 9 32,740 32,168 33-36 10 29,316 30,005 7 28,101 28,161 37-40 15 24,142 26,498 12 25,152 27,255 41-44 6 25,316 27,084 6 30,748 31,035 45-48 11 27,435 27,894 9 27,319 27,571 49+ 10 25,424 26,244 2 25,224 25,826 hearing. In this table the sexes have not been segregated. The table is made to show differences between different periods of life, both in terms of the average of the individual records of each age group and in terms of the median or middle record. The latter gives the data in such terms that the influence of pathological, or functional, disturbances of hearing is largely eliminated. That is, extreme records which have a tendency to skew the average lose their unwarranted weight. From this table, as well as from others which will follow, something, also, may be inferred as to the rela- tive range of sensitivity of the two ears. In Table VI. the data are arranged so as to exhibit any sex differ- ences that might be found for the various age groups. Neither in this table nor Table V. has there been worked out, however, any measure of variability, inasmuch as the character of the distribu- tions may be seen in tables VII. to XII. Tables VII. to XII. present the original data in such form that they may be worked over by anyone who cares to review the ques- tion; and, indeed, they show more convincingly than any series of figures representing averages, modes or medians could possibly do, the tendencies of the several groups measured. Perhaps a word of explanation is necessary to an understanding and interpretation of these tables. Taking, for example, Table VII. In the column to the left are given the lengths of the whistle cavity employed in the various hearing measurements, and in the column to the right the corresponding vibration frequencies (double) of the resulting tones. At the head of the several columns are indicated the different age groups; "5-8" indicating that the data in the column beneath are those secured from boys and girls without distinction of sex ; and so 48 THE HEARING OF PRIMITIVE PEOPLES on for the remaining columns. Again, under the "5-8" caption, the "4" indicates that four individuals were measured in which the whistle cavity had a length of 1.4 mm. or the tone possessed a vibration frequency of 39,220 D. V. (double vibrations), and so on, the "12" indicating that the upper limit of twelve children was marked by a tone of 35,100 D. V. Extended interpretation and discussion of the data contained in tables V. to XII. are uncalled for in connection with the problem of racial differences, of which this paper particularly treats. A few words, however, are perhaps not out of place. It will be observed that the number of individual measurements recorded, especially as regards the younger years, is sufficient to make the conclusions fairly definite. From tables V., VII. and XII. it stands out fairly clear that there is practically no shortening in the range of hearing before the age of sixteen years, but that- after this age the upper limit falls slowly, having sunk about three- TABLE VII UPPEB LIMIT OF HEABINQ Males and Females (White) Right Ear Whistle Ages in Years - Vibra- lions Length mm. 2 w d 1 s 3 55 3 1 1 CO l> *< f 1 18 3 (Double) Per Sec. 1.2 42,960 1.3 1 40,840 1.4 4 3 3 2 1 39,220 1.5 3 3 6 3 2 1 37,560 1.6 4 3 4 3 6 2 1 1 36,360 1.7 12 7 5 5 5 3 1 35,100 1.8 4 6 15 5 6 5 4 1 1 34,000 1.9 7 2 7 5 7 8 2 33,060 2.0 6 4 8 10 5 8 7 2 1 1 32,180 2.1 1 3 4 3 6 6 6 3 2 2 2 1 31,170 2.2 1 1 3 4 3 3 2 4 1 2 1 30,270 2.3 2 3 2 2 1 1 29,508 2.4 2 1 1 3 2 2 1 1 28,766 2.5 2 2 5 2 2 3 28,048 2.6 1 27,448 2.7 2 3 2 1 1 26,854 2.8 2 1 1 1 26,264 2.9 1 1 2 25,724 3.0 2 1 1 2 2 2 25,212 3.1 1 2 1 24,754 3.2 1 2 24,196 3.4 1 2 2 2 23,020 3.6 1 1 1 22,217 3.8 1 1 1 20,973 4.0 1 18,496 <* CO W CO 01 Average D. V? 3? CO c -Tf ^ CO f if 8 TH" CO CO *< -T CO i 1 rH 00^ IQ^ " r 50^ fourths of an octave on the average by the forty-ninth year. The data giving the median record for each age group disclose the same facts, so that it seems to matter not whether we speak in terms of the average or the median (See tables V. to XII.). It will be remembered that Alderton, Blake, Galton, Zwaarde- maker, Caperius, Myers 1 and others found that by the age of 12 to 13 years, the upper limit of hearing already exhibits considerable shortening. My data necessitate a conclusion quite at variance with that of these earlier experimenters. Nor can I suggest an ex- planation for this lack of harmony in our experimental results. A mathematical statement of the probability of a difference in the range of hearing on the average for years of life beyond seven- teen, is wholly superfluous in view of the distributions of individual records exhibited in tables VII. to XII. which makes this conclusion absolutely certain. But to explain this shortening in range with *See discussions under Chapter II., page 19, et seq. 50 THE HEARING OF PRIMITIVE PEOPLES increasing years is a more serious question. Whether it is but a symptom of a general insensitivity of the organism as the individual ages or only an atrophying of certain tissues through disuse, which were of service to man in a lower stage of his culture, or whether it may be due to other factors, are questions which await experi- mental determination. In view of the relative inferiority of the extent of hearing range among primitive peoples, the former of the two suggested explanations perhaps seems the more plausible. TABLE IX UPPEB LIMIT OF HEAEING Males Right Ear Whistle Length mm. s 2 to 2 9 S Ages in Yearn M $ 8 ! I 31! Vibra- tion Fre- quency D. V. 1.2 42,960 1.3 40,840 1.4 2 1 2 1 39,220 1.5 1 2 4 2 1 1 37,560 1.6 2 2 1 3 1 36,360 1.7 6 4 4 4 3 1 35,100 1.8 3 2 8 2 3 3 34,000 1.9 4 1 4 4 4 3 1 33,060 2.0 3 3 6 2 2 4 2 32,180 2.1 1 4 1 2 3 4 2 2 1 31,170 2.2 1 3 2 2 3 2 2 1 1 1 30,270 2.3 1 1 2 2 1 29,503 2.4 2 1 2 2 1 28,766 2.5 1 2 3 1 2 2 28,048 2.6 1 27,448 2.7 1 1 2 1 26,854 2.8 1 1 1 26,264 2.9 1 1 25,724 3.0 1 1 2 2 2 25,212 3.1 1 1 2 1 24,754 3.2 2 24,196 3.4 1 1 2 23,020 3.6 1 1 22,217 3.8 20,973 4.0 1 18,496 QO o CO OS 1C CO N * *o CO CO CO co CO CO CO S . t>. D cocococoe*e*cs ^i s^ i K] OQ > s 1 2! Ill aw i |x L Ix ir ax X X X k X X .8 . C^C^COCC^lCCOOOTHrHrHCie^ 1= X x 84 THE HEARING OF PRIMITIVE PEOPLES was not required for any subject tested, values for positions of the secondary coil nearer to the primary than this were omitted from the tabulations. It will be observed that while the secondary coil was nearer to the primary than 27 centimeters readings were made at intervals of one centimeter, but for positions of the secondary coil more distant, graduations were made only at intervals of 2 centimeters. This plan was followed since obviously the fall of electrical potential along the induction circuit is not directly as the distance. Indeed, it falls very much more rapidly than this, and while the change from graduations at intervals of one centimeter to graduations at intervals of two cen- timeters does not equate the steps, it does serve to keep the records from being too widely dispersed at the upper end of the curve. The matter of inequality in steps, however, injects no inaccuracy into the final results, so long as basically the values of each position are accu- rately determined. And, since each position in these tests receives its value in ergs of sound energy emitted or centimeter-seconds of con- densation of the sound wave, the question of the number of steps to be made in the tests acts as a matter of convenience in manipulation only. It would make but an insignificant difference in the average of any group whether graduations had been made at intervals of 1 or 2 or 5 centimeters. The excursions of the telephone plate (a), corrected as outlined above for an electrical current of 1.9 volts, and 0.5 ampere, appear in the second column of Table XIV., on page 83. In the third column are given the values of the condensations (A) computed from Wien 's formula 19 already given, namely, "These formulae call for certain measurements relative to the values of " k," " c," " N," " R," " a," p," and " d," respectively. " fc," which represents the temperature coefficient, is a constant found from the well-known formula of Keyser (Phil. Mag. 38: 256. 1894). "c" stands for the velocity of sound in air at the temperature prevailing when the graduations were made. In the' experiments under consideration its value was taken as 340.2 M. The radius of the free vibrating area of the telephone plate measured 2.24 cm. The sev- eral values of " a," the excursion of the telephone plate at its middle point, appear in the second column of the table. The distance between the ear and the center of the telephone plate " d " throughout the entire series of hearing tests was uniformly 100 cm. However, inasmuch as the ear was sealed into one end of a tube and the telephone into the other, both to a large extent air tight, a value of one centimeter was assigned to this distance. Indeed, it is improbable that the loss of energy in a sealed tube would be any considerable quantity for such a short distance of propagation. The value of " p," the temperature coefficient, is 0.00128. " N " equaled 500 D. V. THE GRADUATION OF THE INSTRUMENT 85 In the fourth column are given the values in ergs (centimeter-gram seconds) of the quantity of sonorous energy passing a square centi- meter area perpendicular to the line of the sound wave 's progression. The latter are computed from Lord Rayleigh's formula already given, namely, CHAPTER IX METHOD OF CONDUCTING TEST ALL the measurements for simple auditory acuity were made in the sound room, 1 and consequently under conditions, the most favorable for quiet that it was possible to secure at the Exposition. In order to obtain a correct estimate of an individual's hearing, it is very essential that the test be conducted under circumstances as favorable as possible. It is difficult to keep the attention focused on faint stimuli, no matter through which sense avenue they are received, but faint auditory stimuli are especially elusive. Fatigue of the auditory sense organ results after a moment of stimulation, so that even under the most favorable adjustment of external condi- tions, it is not possible to continue an auditory test longer than four or five minutes, even with a subject of far more than average intelligence. Hallucinations enter as very disturbing factors. It is a common observation with those who have made tests of the hearing of both children and adults to find that the individual vigorously asserts that he hears the sound used as a stimulus even though the source has been entirely removed. This is especially true of continuous tones, or such as follow in a rhythmical order, as the ticking of a watch, or the regular falling of water drops. Faint auditory sensations are likewise subject to decided illusory effects. Any extraneous noises are apt to be interpreted as the tone to which the subject is expectantly attending. Such noises as are produced by the scraping of the feet in walking over an earthen, brick or cement floor are especially likely to be heard as the stimulus to which the subject is directing his attention, if a metallic click is the stimulus employed. Noises from the street act in the same way. Those things which tend to be distracting elements with adult Whites are bound to be emphatically so in tests on primitive peoples and younger children, whose power of attention is weaker, and who are, consequently, more easily dis- tracted and are unable to single out one from among a number of somewhat similar stimuli which they will hold at the focus of at- tention. To overcome so far as possible the error which might creep in, on account of the elements just enumerated, a check method sug- 1 For a description of this booth the reader is referred to the foot-note on page 30. 86 METHOD OF CONDUCTING TEST 87 gested by Professor Cattell was introduced. This consisted in hav- ing the subject interpret and give back what he received. The test in consequence is a little more than a measure of pure sensation if a measure of pure sensation is ever possible. Instead of presenting the ''makes and breaks" in rapid succession without reference to number or manner, as is ordinarily done with the telegraph key, or the "make and break" mechanism of the Seashore audiometer, I gave the stimuli in groups or rhythms; that is, a series might be presented in which two clicks in rapid succession would be fol- lowed by a pause of two seconds, the two clicks repeated and so on. Or, the series might be given in singles or in groups of three clicks followed by a rest of a couple of seconds. Graphically, the series might be represented as follows : --,,,--,,,--,,,--,,, etc., in which the dashes represent the order of the stimuli and the commas, pauses or silence. A group of three clicks would be repre- sented by the following scheme: ---,,,,---,,,,---,,,,---,,,, etc., and a rhythm of single clicks by the following : -,,,,-,,,,- 5 > > i - 1 j > j - 5 > 9 9 etc. I pursued no set order of presenting the series, so that it might not be possible for the subject to anticipate the answer which he would be expected to give. The number of clicks to a group, I found, had to be limited to three. I discovered that many of the primitive peoples were unable to count the stimuli following each other as rapidly as did these with any degree of facility in groups larger than two or three. Children, too, I afterward learned, can catch a rhythm of three but experience difficulty with one of four or five. Such a method of presenting the stimuli, as has just been indi- cated, in a manner supplies the deficiencies arising from the neces- sity of employing non-significant stimuli in testing hearing. Indeed, to a considerable degree, the test is one of which the elements are significant ; that character being attached to them by the count- ing operation. The test, moreover, has other advantages. It does away with the necessity of depending upon the cooperation of the subject a necessity present in almost all other methods for measur- ing hearing. It presents something tangible, as it were, for the attention to attach itself to. The subject, when the ordinary methods are followed, finds that as the sensation becomes increas- ingly more faint, it is impossible for him to organize his mental processes to the extent of arriving at a certain conviction as to whether he actually hears the tone or not. Where the method of reproduction is followed, this element of uncertainty does not, in any way, enter into the results. In all cases, I chose as the threshold that point where the subject 88 THE HEARING OF PRIMITIVE PEOPLES failed to indicate accurately the character of the stimuli presented. In the neighborhood of the threshold values, I made it a point to vary the character of the stimulation frequently without changing its intensity. This was done to make certain that the subject's responses were not pure guesses. It was found, however, that in case of almost all persons, there is a subconscious evaluation given to such estimates which are wholly valid. When the subject stated that he was uncertain as to the character of the sensation, whether the ticks came in twos, threes or ones, he was told to express a judgment and it so happened that frequently the subject judged correctly every variation made with a given intensity although he asserted that his estimations were in every case pure guesses. CHAPTER X RESULTS THE nature of the data as regards auditory acuity is such as to make them somewhat difficult to present intelligibly. These data relating specifically to auditory acuity are given both in terms of the condensation of the sound wave leaving the telephone and the actual energy in ergs (centimeter-gram-seconds) exerted by the sound wave over a square centimeter of surface at the same point. Since the energy required to produce a tone of threshold value, in an ear of ordinary sensitivity, amounts to less than a ten-millionth part of an erg, it is at once clear that the data we are working with are compre- hensible, popularly, only in mathematical terms. TJie data to be presented, however, are to be employed for comparative purposes only, so that the absolute values are, relatively, of insignificant mo- ment. To know that one group of people possesses a hearing sensi- tivity two, or ten, or one hundred times as great as another, does not call for an understanding of the absolute unit used, if only the unit remains constant from one series of measurements to another, and besides, possesses the additive character. The figures representing the condensation as well as the actual quantity of sound energy being dispersed fulfill both requirements, and hence no inaccuracy will result if the reader ignores the fractional factor altogether, and thinks of the results in terms of the first part of the figure only, e. g., in terms of "1.4" instead of "1.4X lO" 9 ." And in the eyes of many readers such a method of interpreting the complex data which follow, will contribute both to their clearness and ease of compre- hension. 1 In Table XV. are presented these figures in detail, indicating the race; number of cases included in the average; the average result, stated in terms of atmospheres of pressure, and ergs (centimeter- gram-seconds) ; the average deviation from this average ; and the standard deviation. From the latter, it is possible to compute di- rectly any other measures of the variability that may be desired. The data are given for both the right and the left ear. A word, perhaps, is necessary concerning the make-up of the respective groups. For a test of this kind, as was noted in connection with the 1 For details relating to the physical graduation of the instrument em- ployed in the acuity test, the reader is referred to Chapter VIII. 89 90 TEE HEARING OF PRIMITIVE PEOPLES data on the upper threshold of hearing, already presented, it was thought best to reject those individual hearing records where there was definite reason to believe that the case was one of pathological hearing. In pursuance of this, therefore, I rejected in this series of measurements the data from every person who consciously had ex- perienced difficulty in hearing. Obviously, so long as the rejections follow some consistent plan, no inaccuracy need result and I do not TABLE XV AUDITORY ACUITY In Terms of the Condensation of the Sound Wave Leaving the Instrument and also in Terms of the Energy in Ergs Leaving the Same Right Ear L< ;ft Ear No. of Cases Average A = 1(T 9 A. D. S. D. Average E=lQ-' r ergs Average A = 10~ 9 A. D. S.D. Average J? = 10- 7 ergs Whites 151 56 385 502 7.9 7.2 506 509 13 2 Indians (School)... Filipinos Coco pa In- dians 64 137 10 7.5 24.2 7.3 3.63 14.57 286 5.18 18.30 3.49 14.2 147.8 16.01 8.5 26.6 9.0 4.09 14.46 5.30 5.79 17.81 6.34 18.2 178.6 30.2 Vancouver Indians Patagonian Indians Ainu . 7 3 8 10.01 12.3 18.1 5.01 8.43 1164 5.69 8.97 1638 38.1 38.1 82.7 10.03 17.0 17.1 5.95 8.59 8.71 6.66 9.57 1136 36.5 73.0 73 4 Pigmies 5 10.3 3.55 4.44 26.8 7.5 1.24 1.32 14.2 A = condensation of sound wave at the instrument. E = energy in ergs of sound wave leaving the instrument. A. D. = average deviation. S. D. = standard or the mean square deviation. TABLE XVI AUDITORY ACUITY In obtaining the averages given in this table, the lowest 25 per cent, and the highest 25 per cent, of each group were omitted, leaving only the middle half. Right Ear Left Ear - No. of Cases Average A=10- A.D. Average E=vy- 1 ergs Average A=10- 9 A. D. Average E=10- 7 ergs Whites. Indians Filipinos 78 33 81 3.9 6.1 16.5 6.922 1.40 5.29 3.84 9.61 68.73 4.9 5.0 18.8 1.33 1.03 6.02 6.06 6.31 89.21 A = the condensation of the sound wave at the instrument. E = the energy of the sound wave in ergs per square centimeter area leaving the instrument. RESULTS 91 think that the rejection of the records of some twenty individuals, for the reason indicated, influenced the character of the results otherwise than to raise the average for each group to an appre- ciable extent. For such large groups as those of the Whites, In- dians, or the Filipinos, a sifting of the records, perhaps, would have been unnecessary, but in case of some of the smaller groups, with fewer than ten individuals, any one specially poor record would TABLE XVII RIGHT EAB Showing the Distribution of the Individuals with Respect to Auditory Acuity Intensity of Sound, A = X 10- 9 Whites Indians (from School) Filipinos Cocopa Indians Van- couver Indians Pata- gonian Indians Ainu Pigmy 1.4 1 1.5 3 1.6 4 1.8 7 1.9 6 1 2.1 5 1 2.3 9 2 2.5 6 1 3 2.8 9 3 1 3.0 13 2 2 3.3 9 2 3 1 1 3.6 9 2 3 4.0 7 4 2 2 1 4.5 6 4 2 1 5.0 6 5 2 1 5.5 7 4 2 1 6.2 6 3 4 4 1 2 7.1 5 4 5 1 8.1 2 7 6 1 9.1 4 5 8 1 10.4 6 4 7 1 11.9 5 4 7 1 1 1 14.1 3 2 9 1 15.7 1 10 18.1 2 9 1 1 21.1 1 1 8 1 25.0 2 1 8 1 1 29.9 1 7 1 36.5 6 2 43.4 5 52.5 5 64.8 6 82.3 3 99.9 2 affect the standing of the group to such an extent as to make the figure for the average wholly unrepresentative and false. Yet, in spite of corrections and eliminations, it is quite evident that all pathological cases were not excluded, as appears from a comparison of the data presented in tables XV. and XVI. In Table XVI., I 92 THE HEARING OF PRIMITIVE PEOPLES TABLE XVIII LEFT EAR Showing the Distribution of the Individuals with Respect to Auditory Acuity Intensity of Sound, A = XlO- Whites Indians (from School) Filipinos Cocopa Indians Van- couver Indians Pata- gonian Indians Ainu Pigmy 1.4 1 1.5 2 1.6 3 1.8 3 1.9 4 1 2.1 4 1 2.3 5 1 2.5 7 3 1 2.8 7 2 3.0 8 3 3.3 11 4 4 1 3.6 12 4 4.0 8 5 2 1 4.5 8 6 2 5.0 7 5 4 2 1 5.5 6 4 3 6.2 8 2 4 2 1 2 7.1 5 3 4 1 1 8.1 5 3 5 1 1 9.1 5 3 7 2 10.4 4 2 6 11.9 5 2 6 3 14.1 4 2 8 1 1 15.7 3 1 12 1 18.1 4 1 11 1 1 1 21.1 4 1 9 1 25.0 2 2 9 1 29.9 3 1 7 1 1 36.5 7 43.4 1 6 1 52.5 1 8 64.8 4 1 82.3 3 99.9 3 127.2 2 1 162.5 1 214.9 1 rejected the 25 per cent, of the records at the top and the same number at the bottom, leaving one half of all the records distributed normally about the median. Were the distribution of cases normal, such a method of procedure would affect the average result only to a very slight degree. It will be noted on the contrary, however, that the average has been shifted upward decidedly in the case of every group. In fact, this sifting has the effect of almost doubling each average, thus reducing the quantity of energy necessary to excite an auditory sensation, on the average, by one fourth. In tables XVII. and XVIII. are shown the general distributions of the individual records, as they appear in the several groups. RESULTS 93 These distributions have been made in parallel columns to facilitate the study of comparisons. The tables, likewise, show the character of the curves that the hearing records of the several groups offer, and in addition, present them in such form that group differences are directly apparent and recognizable. Such a form of distribu- tion makes unnecessary the use of graphic curves. In Table IV., p. 34, are given the distributions of the individuals of the three most numerous groups according to age. All the data relating to the smaller groups will be presented in detail in con- nection with the discussion of their several individual results. Again, making a cursory review of the data in Table IV. the fact is revealed that in case of the Whites, Indians, and Filipinos, the age lines were rather closely drawn. I accepted the record of no individual whose age exceeded 30 years or fell below 16. This was done as in the case of the data relating to the upper threshold of hearing, in order to secure, so far as practicable, homogeneous groups. The average age of the Whites selected was found to be 23 years and 5 months; of Indians, 19 years and 2 months; and of the Filipinos, 21 years and 1 month. For an acuity test I do not think this difference in average age is such as to be very significant. We shall now proceed to a consideration of the data in detail : Whites. Whites, such as those selected for these tests, indi- viduals in the prime of life, men and women who have never ex- perienced any difficulty in hearing, according to these experimental data are able to sense and interpret on the average, a stimulus pro- duced by the action of an air wave, amounting to a pressure differ- ence of 5.5 X 10~ 9 atmospheres or 7.5 X 10~ 7 ergs. This indeed is a pressure difference smaller than it is possible to secure in the most rarified vacua* It is rather difficult to compare this value with the figures which have been obtained by other observers, largely because of the dif- ferent experimental conditions under which the tests have been made. In the first place, I have been unable to find data based on the measures of more than a dozen individuals, and, indeed, since the range of individual differences in hearing acuity is as 100 to 1, within a single group, any discrepancies which exist may very plausibly be accounted for because of the paucity of numbers con- stituting the groups compared. Again, Wien 3 has shown that the ear's sensitivity for tones is a function of their pitch, and that its sensitivity for different pitched tones varies within rather extreme limits. The vibration number of the tone used is, consequently, no 2 See Rayleigh, Phil. Mag. 38 : 300. 1894. *Pfliiger's Arch. 97: 1. 1903. 94 THE HEARING OF PRIMITIVE PEOPLES doubt, also an important factor in accounting for discrepancies noted. The vibration frequency of the tone I employed I could not definitely establish. After repeated assays to fix the dominant tone by comparison with other tones of known vibration frequencies, I finally selected 500 double vibrations as being most nearly correct. But, although I think the dominant tone did not vary far from 500, there is no question but that some very pronounced over-tones present were quite effective in favoring acuity. Wien's 4 figures for a tone of approximately 400 D. V. was 1.2 X 10' 10 (E = 8.0 X 10' 11 ergs) for his own ear, when a telephone instrument furnished the sound. The figure was practically the same with a tuning fork and the Helmholtz's resonators, namely, 8.0 X 10' 10 (# = 3.6X10-' ergs). 5 Wien's results, however, make the ear to be almost one hundred times more sensitive than the experimental results of previous observers had led them to believe. For example: Toepler and Boltzmann, 6 who, according to Lord Rayleigh, were the first to make an experimental determination of this question found, with tuning forks, that the value of a sound wave's condensation at the ear, to be just audible, was 6.5 X 10~ 8 . This figure differs but slightly from Lord Rayleigh 's own conclusions from experiments .with Wolf's bottle 7 where A = 4.1 X 10' 8 for a tone of 2,730 D. V. The same writer found A = 4.6 X 10~ 8 when a tuning fork 8 of 512 D. V. was employed as the source of sound. Professor Wead, 9 also employing tuning forks found for a tone of c 2 that A = 7.1 X lO' 9 (E = 1.1 X 10' 6 ergs) was still audible. P. Ostmann 10 with tuning forks (256 D. V.) places the threshold at A = 2.1 X 10- 8 (# = 8.0 X 10- 6 ergs). These latter results are pretty much in accord with some recent experiments of Zwaarde- maker and Quix, 11 who find that tones from a tuning fork of pitch c 2 in which A = 1.5 X 10~ 8 (E = 5.4 X 10' 6 ergs) might still be heard. More recently still, 12 they secured a somewhat smaller figure A = 7.1X10- 9 (# = 1.3X10-* ergs). Lord Rayleigh 's results from his telephone experiments lead him to think his previous figures placed the sensitivity of the ear too low, since some of his 'Pftiiger's Arch. 97: 33. 1903. 6 Dissert, s. 46, 1888; also, Wied. Annal. 36: 849. 1889. 6 See Lord Rayleigh, " Theory of Sound," II., p. 433. 7 Proc. Roy. Soc. 26: 248. 1877. 9 Phil. Mag. 38: 270. 1894. Amer. J. of Sci. 36: 36. 1883. 10 Arch. f. (Anat. u.) Physiol. 1903, S. 321. "Arch. f. (Anat. u.) Physiol. 1902, S. 393. "Ztschr. f. Psych. 33: 401. 1904. RESULTS 95 subjects were able to still hear when A = 1.1 X 10' 9 (E = 2.8 X 10' 8 ergs). The first figure is about the same (A = 8.88 X 10~ 9 ) as has more recently been gotten by Webster 13 with his "phone." It was noted above that Wien's figures are from 40 to 100 times smaller than my own and those of the other investigators. Zwaar- demaker and Quix 14 attribute Wien's excessive sensitivity of the ear to the fact that the telephone receiver was held snugly against the ear and that hearing was assisted by molecular bone conduction in addition to the molar sound energy passing over the ossicles. From the character of my own data it is easy to explain differ- ences in auditory acuity as great as 20 times, such as have been obtained by different observers, where experiments have been limited to a few subjects. Among my white subjects, although not a single individual had ever observed any diminution in his hearing func- tion, the person with the best acuity required about 400 times less energy to just excite an auditory sensation than did the one who heard most poorly. And, indeed, the individual with the keenest ears heard about 90,000 times as well as did the poorest among Filipinos, although in conversing with these Filipinos it was not possible to detect any hearing deficiency. Wien 15 reported two cases of individuals who are still able to hear loud speech but whose hear- ing is from one to ten million times poorer than normal. Ostmann 16 concluded that a dimunition of hearing of one half or one-third is of slight consequence. The range of efficiency in hearing among normally hearing people is a question which, to my knowledge, has never before been investigated in this way. Ordinarily, it appears unreasonable to believe that in speech the human voice covers such wide latitudes of intensity that a person can speak 300 or indeed 1,000,000 times as loud at one time as at another and yet not be speaking appreciably loud. Our difficulty, perhaps, arises from comparing a hearing test such as the one under consideration, with the ordinary visual acuity tests in which the units are the angles subtended by light rays coming from opposite parts of the letters. These tests obviously are incomparable. But only recently von Kries 17 has shown that the minimum intensity of light necessary to excite the eye amounts to only 1.3 to 2.6 X 10' 10 ergs. To see an object, 5.6 X 10~ 10 ergs is essential, about the same quantity of energy that Wien discovered necessary for sound, and about %o the 13 Boltzmann-Festschrift, 1904, p. 874. "Ztschr. f. Psych. 33: 408. 1904. "Pfluger'sArch.97:37. 1903. "Arch, of Otology, 34: 207. 1905. "Ztschr. f. Psychol. u. s. w. 41: 393. 1907. 96 THE HEARING OF PRIMITIVE PEOPLES quantity necessary to excite a sensation of sound on the average according to my own experimental results. Between the intensity 2.6 X 10~ 10 ergs and that of ordinary moonlight, the difference is millions, and between the intensity of moonlight and that of sun- light, there is again a difference of at least 100,000. It is such differences with respect to the eye which have their counterpart in the field of hearing. Indeed, extremely great differences in the in- tensity of tones are not so commonly noticed as one would think. The singing of a thousand voices, though noticeably louder than that of a single singer, certainly does not appear 1,000 times as great. Under certain atmospheric conditions in the quiet country, it is possible to hear, quite distinctly, a human voice at a distance of two miles. If the loudness of the voice when at a distance of 11,000 feet be compared with that when audible only 10 feet from the speaker, some conception may be had of differences of intensity amounting to at least 10,000 to 1 and perhaps 1,000,000 to 1. In the light of such comparisons the figures showing the range of sensitivity of the normal ear are not exceptionally striking. It will be noted that not only in the case of Whites but in those of the records of all of the groups, the average deviations are ex- tremely large. Such of necessity must be the case, however, when the unit of measurement is extremely small, such as are physical units of sound. Although such fineness of measurement is not essential, it can not well be avoided. Our units are fixed beforehand and with these we are measuring physiological and psychological conditions as they are found among individuals. Part of the large average deviation, indeed, may also be explained by the dispropor- tionately large number of the cases found below the mode in the curve, which arises from the inability which we experience to mark off accurately the normal from the pathological in any functioning. In a way, I sought to eliminate some of the error arising from this source by presenting an average of the mean cases only, as appears in Table XVI. But, even under such restrictions, the average devia- tion amounts to about one third of the average result in each of the several groups. From this it appears that so far as the hearing function goes, individuals do not distribute themselves so as to con- form closely to the laws governing the normal frequency curve. Indians. (For a more detailed description of this group, see Chapter II.) The 64 Indians tested for simple auditory acuity, whose records are here presented, are the same whose records formed the basis of the study on the upper threshold of hearing for Indians, as already presented. Included in this group are the records of 14 full-blooded males and 13 mixed bloods, with 4 full-blooded females RESULTS 97 and 33 mixed bloods. Five of those tested were over 25 years; 81 per cent, were between the ages of 16 and 20 years. (See Table IV., p. 34.) It must be recalled that these Indians included under the general caption "Indian" were those only attending the Model Indian School at the Exposition, who, previous to their coming to St. Louis had been in attendance, for some considerable time, at various Indian schools throughout the United States. As has already been pointed out, in habits and culture they are to be distinguished from the Indian of the forest and plain, hence, my reason for grouping together all these Indians from the schools representing a number of different tribes. For the same reason I have chosen to consider separately those tribes representing Indians who came from their natural habitats, and who, therefore, more nearly constitute what might be called representatives of the typical Indian. For the right ear, the figure representing the condensation of a sound wave which on the average is required to just excite the organ of hearing of the Indians at the Model Indian School, is 7.5 X 10~ 9 (E = 14.2 X 10- 7 ergs) and for the left ear 8.5 X 10' 9 (E = 18.2 X 10~ 7 ergs). The figure for the right ear is 1.34 times larger than that for the corresponding ear of Whites, and for the left ear 1.18 times larger. On account of the relatively large average deviations these differences are not so significant as at first hand one might suppose. Still, the mathematical probability of a difference in favor of Whites is not unimportant. On the basis of the data, the chances are nearly 200 to 1 in favor of the superior auditory sense of Whites for the right ear, and 6 to 1 in case of the left ear. Arguing from Table XVI., where the individual records included in the average are restricted to those lying about the mean, the superi- ority of the hearing of Whites over Indians is more strikingly brought out, especially as regards the right ear. The average for Whites shows a keenness of hearing which is just about two times that for Indians. In the averages of the left ear, however, the size of the difference between the two groups is lessened, a condition, no doubt, due to the fewness of the individuals comprising both the groups under comparison. Where the individual records are so widely dispersed as are these, instead of grouping themselves rather closely about the mode, the character of the distribution, as a whole, is really more significant in the way of indicating group differences than the figures repre- senting the averages of the two groups to be compared. A compari- son of the hearing of Indians and Whites respectively can easily be made by reference to tables XVII. and XVIII., where the individual 98 THE HEARING OF PRIMITIVE PEOPLES records have been distributed according to relative position. It is clearly apparent that the modes of the distributions of the Indian hearing records if such a term as "mode" is really applicable to such a form of distribution fall rather decidedly lower than do those in the curves of Whites. Moreover, it is seen that the general distributions of the hearing records for Whites stand distinctly higher than do those for Indians for both ears, demonstrating that Whites as a whole hear better than do Indians, although many of the Indians, to be sure, possess ears that are more acute than the average acuity for Whites. Of the 64 Indians, however, the ears of 13 only, or about 20 per cent., rank as high as the median record for American or European ears as regards the right ear, and for the left, 24 Indians hear better than the median hearing record for Cau- casians, or about 38 per cent. Taken all in all, therefore, the data point rather decidedly toward a superiority of the hearing of Whites over that of Indians; such Indians at any rate as constitute the groups here considered. The numbers are rather small to indicate reliable sex differences ; 27 men and 37 women. But the average acuity of the men, for the right ear, amounted to A = 7.4 X 10~ 9 (# = 1.38 X 10~ 6 ergs) ; of the women for the same ear it was A = 7.5 X 10' 9 (E = 1.42 X 10~ 6 ergs). For the left ear, the figures for men and women respectively were A = 8.5 X 10' 9 (E = 1.83 X lO' 6 ergs) and A = 8.4 X 10' 9 (E = 1.78 X 10~ 6 ergs). If we should argue from this group alone, therefore, sex differences in hearing, among Indians, do not exist. The Cocopa (or Seri) Indians. (For a more detailed description of these people, turn to Chapter II., page 6.) Of the Cocopa Indians I was able to secure ten hearing records, all of males. Owing to the fact that the number is so small, I shall present the data relating to the hearing of the various individuals in detail : ATIDITOBY ACUITY Name A pp Rig] it Ear Le ft Ear Age (A = XlO-) (=XlO- 7 ergs) (A-xi- E= X 10- 7 ergs Shamba 30 16 13 15 17 Batwa u Batsuba 6.2 11.9 18.1 9.1 6.2 9.5 35.0 80.0 21.0 9.5 6.2 9.1 7.1 9.1 6.2 9.5 21.0 13.0 21.0 9,5 Malinera Bushaba Latuna . Average 10.31 3.55 31.0 20.8 7.54 1.24 14.8 7.6 In contrast with the records of the Ainu just considered, it is seen that the Pigmies present a rather homogeneous group so far as auditory acuity is concerned. For both ears the average deviations are small. This probably is due, at least in part, to the fact that the individuals were about of an age, and differed little tempera- mentally from one another. It will be remembered that for the upper threshold of hearing the Pigmy records were all high, and if the same relative distribution were to follow, were the number increased indefinitely, Pigmies would possess an upper threshold for hearing superior to that of any other race, not being inferior to even Whites in this respect. An equal degree of superiority was not attained by Pigmies in auditory acuity, although for the left' ear (see Table XV.) the average acuity of Pigmies is slightly higher than that of the same ear for Indians who next approach Whites in keenness of hearing ; the record for the right ear falls below that of both Whites and Indians. Little significance perhaps may be attached to an average measure, where the numbers measured are so few, but the character of the distribution of the group seems to indicate a decided inferiority for the hearing of the Pigmies as com- pared with those of both Whites and Indians. The curve of their hearing falls perceptibly lower, the average being relatively higher owing possibly to the fact that none of the Pigmies possessed any- thing in the way of an organic hearing defect, which might tend to lower the standing of the group. An explanation of the differences found between the comparative records of Whites and Pigmies in the upper threshold of hearing and for simple acuity respectively might be given on an intellectual basis. In the test for the upper 108 THE HEARING OF PRIMITIVE PEOPLES threshold of hearing, the stimuli are of longer duration. More- over, they require no interpretation, and consequently the feeling of hearing the sound, which really has only a subthreshold value, may be more easily accomplished, than the actual interpretation, which the counting of the stimuli, implied in the simple acuity test, neces- sitates. I put this forth as suggestive only. CHAPTER XI SUMMARY AND CONCLUSION IT is very difficult to compare the foregoing results with those of Myers in the same field, by reason of the differences in the method employed in collecting data. In the classic work of Myers, on Papuan hearing, several different devices for testing the hearing of the primitives were employed. 1 And to such an extent were these different measures therefore confused that it was necessary for Myers to report all the data he collected in terms of a personal fraction in which the hearing of one of the members who made up the expedition was the denominator, while that of the subject constituted the numerator. Of the 35 Islanders who were examined for auditory acuity by Myers, the hearing of seventeen only was reported and of these twelve were children, five only being adults. Of the children five could hear as far as Myers ; seven were clearly inferior ; and of the adults examined, all possessed a very low acuity. Although consequently I can speak only in general terms, Myers's conclusions do not appear to differ essentially from my own in this, that they point out clearly the obvious superiority of Whites over primitive races in the keenness of their hearing sense. With my smaller groups, as has been repeatedly stated, the num- ber examined is insufficient to do more than indicate a general tendency of the group within the region of a large probable error. Especially is this true of such peoples as the Vancouvers, the Pigmies, and the Cocopas, where it is difficult to predict with any degree of probability the character that the hearing curve of the peoples as a whole might assume, inasmuch as the records are so scattered some being fairly high; others extremely poor. But in the case of such groups as the Indians, Filipinos and Whites, the number of measurements is sufficient to give at least the general character of a complete distribution of the race as a whole. Taking the results of all the groups examined for what they are worth their standings respectively are as follows, as regards the acuity of the right ear : Whites ; Cocopas ; Indians from the School ; Pigmies ; Patagonian Indians; Vancouver Indians; Ainu, and lastly, Filipinos. For the left ear, the order is slightly changed, Whites and 1 See Report of the Cambridge Anthropological Expedition to Torres Straits, Vol. II. 109 110 THE HEARING OF PRIMITIVE PEOPLES Filipinos, however, still retaining the positions at the two extremes ; the order from the most acute people to the least acute being : Whites; Pigmies; Indians from the School; Cocopas, Vancouver Indians; Ainu; Patagonians and Filipinos. ^\ ^ R i r I i 1. i! II i <* 1 1! / * l/f " |/7 * l/4*" l/^. / 9 / -* o It will be observed that the relative positions of the three most numerous groups, namely Whites, Indians from the School, and SUMMARY AND CONCLUSION HI Filipinos remain unchanged. Indeed, they retain in respect to each other about the same relative position for both the right and left ears, and also, when the basis of comparison is that of absolute units of hearing, instead of relative position. To summarize the various comparisons which have been made in connection with the data relating to the several groups, we may show the following ratios indicating the relative keenness of the hearing sense of each group as compared with that of Whites : Right Ear Left Ear Whites Cocopas Ratio 7 to 5 Ratio 9 to 7 Whites Indians (School) Whites Pigmies Whites Patagonians Whites Vancouver Indians Whites Ainus Whites Filipinos 9 to 5 8.5 to 7 10.5 to 5 " 7.5 to 7 12 to 5 " 17.5 to 7 10 to 5 " 10 to 7 18 to 5 " 17 to 7 24 to 5 " 26.5 to 7 Preyer, Fechner, 2 Bezold and others have observed that in hear- ing tests, the left ear in general is more acute than the right. Miss Nelson, 3 on the contrary, found that in both men and women the right ear was the better. The left ear, it will be remembered, was found to be superior with respect to the tests for the upper threshold of hearing. In case of the ears of each of the larger groups, my own experiments in general confirm the observations of Miss Nelson as opposed to those of Fechner and Bezold. The acuity of the left ear not only of the three larger groups but in three of the five smaller ones, is clearly inferior to the right, the Pigmies and Ainus alone being exceptions. When making the measurements of the upper threshold, it will be recalled that it was stated that almost invariably the right ear was first tested. In consequence, I believed the superior upper limit of the left ear to be due to the effect of practise in hearing shrill tones. But this explanation will not apply to the case of the acuity test. Instead of testing invariably one particular ear first, the process was alternated the right and left ears alter- nately being first tested in successive subjects. Practise effects could not, therefore, have been operative in causing the average for acuity of one ear to be higher than the other. It is, indeed, more probable that the causal factor is organic rather than psychological. The one fact standing out most prominently as a result of these measurements is the clearly evident superiority of Whites over all other races, both in the keenness and in the range of the hearing sense. The evidence is so clear and striking as to silence effectually the contention that the hearing function, inasmuch as it is of rela- 2 Poggendorfs Annal. (4th series) 3: 500. 'Psych. Rev. (suppl.) 12: 280. 1905. 112 TEE HEARING OF PRIMITIVE PEOPLES lively less utility in the pursuits attending modern social conditions than those surrounding the life of the savage, has deteriorated and is degenerating. On the contrary, they are more nearly in keeping with the advanced positions taken by modern dynamic psychology, to the effect that not only the intellectual but sensory possibilities are to be stated in terms of the variety of motor response of which the individual is capable. Other things being equal those individuals or races possessing the greatest complexity and variety of reactions to elements in their respective environments likewise will be gifted with keener and more acute sensory mechanisms. If all discrimination of data coming to the senses must finally be stated in motor terms, as most psychologists would have us think, then those peoples whose social activities call for the greatest com- plexity of response will, of necessity, possess keener senses along those lines in which the social media call for closer discrimination. This motor aspect of a sensory function also serves, to a certain extent, to explain a rather startling auditory inferiority on the part of some of the natives of tropical lands. In these regions of warmth, where lack of thrift and indolence are fostered by nature's bounty, in its luxuriance and plenty in the way of food, in its relative im- munity from exigencies calling for protection and shelter, adaptive activities are found at their lowest ebb. Contrast these conditions with those of higher latitudes, in which the individual is in constant strife to keep himself in harmony with his surroundings. And the ear plays no insignificant role in this endless round of readjustment. Roughly, and in general, the data on hearing were found to correlate with motor versatility as regards the different races. ^Then again the more involved a test, the more probable is it that differences in the degree of intelligence of the subjects tested will be effective in modifying in an unfavorable direction the performance of the less gifted group. It has already been indicated that the test for auditory acuity which I employed was more than a simple sen- sory test, inasmuch as it required an interpretation of the stimuli presented to the ear, and for this reason it was believed that some of the differences between the acuity of the several peoples tested might be attributable to the obvious fact that striking differences in mental alertness obtained among the different races. But to what extent the mental factor was responsible for the degree of auditory inferiority in such a race as the Filipinos, it is impossible to tell with any degree of certainty from the data at hand. Only two factors have been indicated to account for differences in auditory acuity found among primitive races, and between primi- tive races and Whites. That there are many others, some perhaps SUMMARY AND CONCLUSION H3 more significant and vital than those pointed out, the writer only too well appreciates. But the field is new. And indeed the conditions surrounding the taking of the measurements herein reported were not as favorable for making an intensive study of the factors enter- ing into hearing as the importance of the problem warrants. Still this is not the phase of the study that I care to unduly emphasize. As significant as the problem of racial differences in hearing is for genetic psychology, and the writer feels this importance keenly, it was to the method employed in testing hearing, particularly, that it was desired to call attention. Psychology as a science has advanced to that point where quantitatively exact methods of research ought to be emphasized, as well for evaluating functions as for equating differences between individuals or among races. Such methods too are demanded as are possible of reinstatement, and offer data in a nomenclature more specific and determinative than normatively es- tablished units of measurement can give. If consequently the meth- ods that have been used to obtain the results herein presented succeed in accentuating the need for more exact objective methods of research in experimental psychology, the writer's purpose will have been at- tained as effectually as by the recognition of the light that has been thrown on the problem of the hearing of the inferior races. VITA Born at Streator, 111., March 28, 1874; college preparatory education, Indiana Normal School, Valparaiso; entered Junior Class, University of Nebraska, 1901; A. B. Nebraska, 1903. En- gaged as principal of public schools 1896-99; Superintendent, Dwight, 111., City Schools, 1899-1901 ; University Assistant in Edu- cation, Nebraska, 1902-03. Fellow, 1903 ; Assistant in Psychology, Columbia University, 1903-05 ; Assistant Superintendent sections of Anthropometry and Psychometry, Louisiana Purchase Exposition, 1904 ; Assistant Director, Department of Child Study and Pedagogic Investigation, Chicago Public Schools, 1905 ; Member Ameri- can Association for the Advancement of Science, and American Psychological Association. Previous publications: "Investigations Concerning Deaf Chil- dren" (in collaboration with D. P. MacMallan, Ph.D.), Report of Child Study Department, Chicago, 1906, pp. 86. "Mental and Physical Status of Truants," Report of the Chicago Parental School, 1906, pp. 9-24. "Grades of Mentality Among Public School Children," Proc. Illinois State Teachers' Assn., 1905, pp. 149-157. ry Obrair j UNIVERSITY OF CALIFORNIA LIBRARY BERKELEY Return to desk from which borrowed. This book is DUE on the last date stamped below. 29 NOV 18 1952 JUN U 1970 3 JUN 26 1971 JUN 1 1 1971 3 LD 21-100m-ll,'49(B7146sl6)476 207775 QP461 374 BIOLOGY LIBRARY 6 THE UNIVERSITY OF CALIFORNIA LIBRARY