zyxwv zyxwv zyxw z zyxw Ontogeny of Vocalization in Duck and Chick Embryos GILBERT GOTTLIEB AND JOHN G. VANDENBERGH Psychology Laboratory, Laboratory of Reproductive Behavior, Division of Research, North Carolina Department of Mental Health, Raleigh, North Carolina ABSTRACT Vocalizations can be elicited from some duck embryos five and one-half days before hatching and from some chick embryos three days before hatching. All duck embryos can vocalize three and one-quarter days before hatching (at 88% of incubation period) and all chick embryos are able to vocalize one day before hatching (95% of the incubation period). Thus, the vocal apparatus of the duck embryo achieves a functional stage of development sooner that that of the chick embryo, indicating that the syringeal structures of the duck embryo begin development sooner or develop at a faster rate than those of the chick embryo. According to audiospectrographic analysis, both species are capable of uttering at least three different kinds of vocalizations prior to hatching, and these three calls are similar to the ones emitted most frequently after hatching (“distress,” “contentment,” and brooding-like calls). These vocalizations can be objectively defined by reference to the contexts in which they are emitted and the physical structure of the call notes. It was verified in situ that the constriction and vibration of the tympaniform membranes during exhalation is the basic or primary source of sound production in birds. The clavicular sac is not essential to sound production, and the syrinx, trachea, and syringeal muscles are important only for modulating and/or resonating the sounds produced by the tympaniform membranes. Based on this knowledge, a technique was devised to de-vocalize duck embryos 3 days before hatching. The technique involves immobilizing the tympaniform membranes by applying a nontoxic substance (“Collodion”) to them. Collodion forms a glue-like sheath over the membranes, and when the tympaniform membranes are thus rigidified, the embryo can not vocalize. Eighty-five percent of the embryos treated were completely mute after hatching. The long-term effectiveness of the de-vocalization procedure is not reliably known. It was possible to keep only one bird for observation beyond the first four days after hatching and she was still mute at five months of age. zyx Vocal communication plays an important role in the social behavior of birds throughout their life cycle. Even before hatching, duck embryos are able to vocalize and are responsive to the maternal and sibling calls of their own species (Gottlieb, ’68a). Early in postnatal development ducklings and chicks recognize the maternal parent and siblings of their own species on the basis of particular calls (Gottlieb, ’66). Later on, in the juvenile and adult phases of development, vocalizations continue to play an important role in individual as well as species recognition (Dilger, ’56; Lanyon, ’63; Ramsay, ’51; Stein, ’63; Thorpe, ’61), including courtship and mating behavior. The postnatal ontogeny of vocal behavior in birds has been extensively studied, mainly in terms of patterns of sound production and, to a lesser extent, signal function. Konishi (’63), for example, has studied the effects of deafening hatchlings on the vocal patterns produced in later life. Others (Lanyon, ’60; Thorpe, ’61) have J. EXP. ZOOL., 168: 307-326. studied the vocal patterns of birds reared in auditory isolation from members of their own species. In some species, the birds are dependent for certain parts of their vocal repertoire upon self-stimulative feedback or hearing the vocalizations of other birds, while in other species certain aspects of the species-typical vocal repertoire may be developed and maintained without the benefit of prior audition. Unfortunately, an unambiguous interpretation of the results of the preceding studies is handicapped by two factors relating to prenatal ontogeny. First, hearing in birds develops early in embryonic development (Gottlieb, ’68a) and, secondly, avian embryos begin vocalizing prior to the time of hatching. To assist in the goal of achieving a fuller understanding of the ontogeny of vocalization in birds, the current report presents ( 1 ) the results of a comparative study of the onset of vocalization in duck and chick embryos and (2) an analysis of the anatomicophysiological basis of sound production in avian embryos. In the context of the latter zyxw 307 308 zyxwvutsr zyxwvuts zyxwvuts zyxwv zyxwvut GILBERT GOTTLIEB AND JOHN pursuit, it proved possible to devise a simple technique for muting embryos and hatchlings. Onset of vocalization in duck and chick embryos In the first study we determined the earliest age at which duck and chick embryos are capable of vocalizing and the nature of the initial vocalizations. METHOD The data on vocalization were obtained from 101 communally incubated Peking duck embryos (Anus platyrhynchos) and 170 communally incubated White Rock chick embryos ( GatZus gallus). This sample was obtained from a much larger number of embryos studied over a three-year period. To assure the normative or representative feature of the results, embryos which were physically abnormal or malpositioned were excluded from consideration. Otherwise normal-appearing embryos which did not vocalize and did not survive for more than eight hours post-operatively were also excluded. The embryos came from groups of 20-40 eggs in which viability up to the time of testing was no less than 55% (including infertile eggs). All eggs were incubated in forced-draft incubators in which the temperature ranged from 99”-100” F with relative humidity 60-70%. The eggs were individually turned by hand four times daily during the course of incubation. Since the doors of the incubators remained open during the turning of the eggs, the eggs were lightly sprayed with warm water after the first and last daily turning to help maintain the proper level of humidity in the incubator. Particular stress was placed on the maintenance of a standard level of humidity since very high or low humidity alters the size of the air-space in the egg and such an alteration could affect the onset of embryonic vocalization. Operative procedure. The method of exposing the head of the embryo and testing for vocal ability has been fully described and illustrated in a previous publication (Gottlieb, ’68b). In brief, after the shell and outer shell membrane over the airspace are removed, the inner shell membrane is painted with lukewarm liquefied G. VANDENBERGH petroleum jelly or saline. Application of the Liquid renders the veins of the chorioallantois visible so the chorioallantois can be gently punctured without rupturing the large blood vessels. Next the amnion is pulled out so it can be’cut sufficiently to allow the embryo’s head to be drawn into the air-space. The egg is then slightly elevated and the embryo’s head is inclined laterally to allow fluid to drain from the trachea and lungs. After 30 minutes have elapsed, the embryo’s oral cavity is prodded with blunt forceps in an effort to stimulate vocalization. The embryo is prodded in this way for 60 seconds every 30 minutes for a maximum of five hours. The embryos remain in a special transparent plastic incubator (fitted with armholes) during the entire period, beginning when the egg is first opened. The age of the embryos at the beginning of testing and the sample size of each age are shown in table 1, White Rock chicks hatch on Day 20 (21st day of incubation) and Peking ducklings hatch on Day 27 (28th day of incubation). zyx zyxwv TABLE 1 Onset of evoked vocalizations i n duck and chick embryos Developmental age 1 No. in group Duck embryos Day 21,O hours 10 Day 21,14 hours 23 Day 23,O hours 21 Day 23,18hours 11 Day 24,7hours 20 Day 25,17 hours 16 Chick embryos Day 16,Ohours 21 Day 17,Ohours 32 Day 18,O hours 40 Day 18,12hours 50 Day 19,Ohours 27 vocalized % 0 39 48 100 100 100 0 3 35 66 100 IIndicates age rfr 2 hours when embryo’s head first exposed. Ducklings hatch on Day 27 and chicks hatch on Day 20. The vocalizations of the embryos were recorded on magnetic tape, using a dynamic, directional microphone with a flat frequency response up to 20,000 Hz and a “Nagra” (Kudelski) tape recorder. The audiospectrographic analysis of the calls was made on a Kay Electric Co. “Sonagraph” by means of a narrow band filter and a magnified scale. While the Sona- zyxwv 4zyxwvut 309 VOCALIZATION IN AVIAN EMBRYOS graph adequately depicts the structure (form characteristics) of each call note, the frequency analysis provided by the Sonagraph is not very exact and it has an upper limit of 8000 Hz. An accurate analysis of the frequency components of each call (up to 20,000 Hz) was obtained by using a Bruel and Kjaer Type 2107 Frequency Analyzer and B & K Type 2305 Level Recorder. 100 DAY zyxwvu zyxw ,/I zyxwv zyxwvuts zyxw C H I C K S ( H A T C H DAY 20) o D U C K S (HATCH DAY 27) 80 N 60 DAY 18 RESULTS As indicated in table 1, some duck embryos are capable of vocalizing as early as five and one-half days before hatching, and some chick embryos are able to vocalize three days before hatching. All of the duck embryos were capable of vocalizing about three and one-quarter days before hatching (Day 23, 18 hours) and all of the chick embryos were able to vocalize one day before hatching (Day 19). In the normal course of incubation, the duck embryo’s bill begins to penetrate the air-space on Day 24 and the chick embryo’s beak begins to move into the air-space on Day 19. Thus, under normal conditions, Peking duck embryos usually begin vocalizing on Day 24 ( 3 days before hatching) and domestic chick embryos usually start vocalizing sometime on Day 19 ( 1 day before hatching). However, as pointed out by Kuo and Shen (’37, p. 5 5 ) , the chick embryo actually begins pulmonary respiration before its beak penetrates the air-space, so, in some cases, the embryos can begin vocalizing before they enter the air-space. This eventuality is probably dependent upon the creation of an air pocket when the amniotic fluid begins to evaporate around the oral cavity, shortly before the embryo’s beak begins to protrude into the air-space. Thus, under normal conditions it is possible that duck embryos may begin vocalizing as early as Day 23 and some chick embryos may begin vocalizing as early as Day 18. From the comparative point of view, it is most interesting that the duck embryo is capable of vocalizing relatively earlier than the chick embryo. As shown in figure 1, when the evoked vocalization data are plotted according to percent of the incubation period, the duck embryos began vocalizing at 80% of their incubation period while the chicks did not vocalize before 21 0 ] , , I l6I1?, DAY 21 0 80 70 O h OF 100 90 INCUBATION PERIOD Fig. 1 Onset of vocal ability in domestic chick and duck embryos according to elapsed proportion of incubation period. 85% of their incubation period. Further, all the duck embryos were capable of vocalizing when 88% of the embryonic development had been reached, while all the chicks could not vocalize until the 95% point of their incubation period. (While only 3% of the chick embryos could vocalize at 85% of incubation, 100% of the duck embryos were capable of vocalizing at 88% of incubation.) As the proportion of embryos capable of vocalizing increased, the elapsed time between exposure of the head and onset of zyxwvuts zyxwvu TABLE 2 Elapsed time between exposure of the embryos’ head to air and initial vocalization i n those embryos which vocalized Developmental age No. vocalized NO. in group Duck embryos Day 21,14 hours 9/23 Day 23,O hours 10/21 Day 23,18 hours 11/11 Day 24,7 hours 20/20 Day 25,17 hours 16/16 Chick embryos 1/32 Day 17,O hours Day 18,O hours 14/40 Day 18,12 hours 33/50 Day 19,O hours 27/27 eg$::e &Fu& 105 90 30 30 0 2 zyxw 150 60 60 30 1 Embryos were tested for first time 30 minutes after exposure of head and they were re-tested at 30minute intervals thereafter. 2 All embryos vocalized before shell removed. zyxwvutsr zyxwvu zyxwv zyxw zyxwvu zyxwvuts zyxwvuts 310 GILBERT GOTTLIEB AND JOHN G. VANDENBERGH vocalization decreased. These data are shown in table 2. Thus, the older the embryo, the more readily it vocalized when exposed to air. Nature of embTyonic vocalizations. The first evoked vocalizations from the youngest embryos were discrete, single, low intensity notes of a rather uniform character. The utterance of multiple notes occurred with increasing frequency in the older groups of embryos (see Sonagrams in figs. 2, 3). There was considerable individual varia- tion in the age at which the multiple notes were first produced. However, almost all duck embryos could produce multiple notes by Day 24 and the same was true for chick embryos on Day 19. Whereas the first evoked vocalizations had a rasping or hoarse quality most likely caused by residual fluid in the air passage from the lungs to the glottis, this quality was usually absent in the multiple notes produced by the older embryos (Day 24 duck embryos and Day 19 chick embryos). The relatively CHICK EMBRYO DUCK EMBRYO I 5-3: Late Day 25 - 2-3: Late Day IS zyxwvutsrqponm d zyxwvutsrqpo c 3-4:Lato Day25 3- T 2 - 4 : D a y 20 c L 3 ,2-5:Day I R k 20 110 2!0 SEC. Fig. 2 Sonagrams depicting vocal repertoire of chick embryo at various stages and conditions prior to hatching. (See text for discussion of each vocalization.) I, 1!0 2!0 SEC. Fig. 3 Sonagrams depicting vocal repertoire of duck embryo at various stages and conditions prior to hatching. (See text for discussion of each vocalization. ) zyxwvu 311 VOCALIZATION I N AVIAN EMBRYOS earlier onset of the ability to produce single notes in the duck embryos also held true for multiple notes. In general, the duck embryos could be characterized as being more vocal or more vocally responsive than the chick embryos. For example, all of the duck embryos tested on Day 25, 17 hours vocalized while the egg was being handled prior to removal of the shell over the air-space. In chick embryos of the same relative age (Day 19, 0 hours), only 59% vocalized when prodded for the first time 30 minutes after the egg was opened. A n important ontogenetic question concerns the classification of the initial vocalizations of duck and chick embryos, especially the relationship they may bear to calls in the vocal repertoire of hatchlings. Considering the situation in which the first vocalizations were elicited, it would seem likely that these initial calls might resemble “distress” or “alarm” notes. Such contextually defined calls have been spectrographically identified by Collias and Joos (’53) in their study of chick hatchlings. While the distress notes of chick hatchlings are composed of sharply descending frequencies, “pleasure” or “contentment” notes are composed mainly of ascending frequencies. As shown in the Sonagrams in figure 2, the chick embryo is capable of producing at least three basically different kinds of call notes prior to hatching. The earliest evoked vocalization, that produced by prodding the oral cavity with blunt forceps, shows a preponderance of descending frequencies (fig. 2-1) which extend to even lower frequencies with age (fig. 2-2). Besides the similarity in the structure of the notes, there is also a similarity in their peak frequencies (fig. 4: C-1 and C-2). The earlier evoked vocalization (C-1) shows a fundamental peak at 4500 Hz with a strong harmonic at 9000 Hz, while the later evoked vocalization (C-2) shows a peak at 4750 Hz and a marked harmonic at 9500 Hz. The second kind of call note, that produced by the undistrubed embryo at normal incubation temperature, shows an almost symmetrical distribution of ascending and descending frequencies (figs. 2-3 and 2-4). Another difference between the evoked vocalizations (figs. 2-1 and 2-2) and the “spontaneous” ones (figs. 2-3 and 2-4) is the shorter duration of the latter notes. With age these notes show a compression of the peak frequency from 24004000 HZ (fig. 4: C-3) to 3300-4100 HZ (fig. 4: C-4). The third basic vocal utterance of the chick embryo is produced “spontaneously” only in very highly developed embryos (Day 20) shortly before hatching or, in some cases, not until after hatching on Day 20. This vocalization (fig. 2-5) is more complex than the previous ones in the sense that it contains a series of brief-ascending notes of variable pitch within a relatively restricted peak frequency range of 3600 400 Hz (fig. 4: c-5). zyx * It should be noted that all of the above three calls contain low frequency components and harmonics (fig. 4) which are not always evident in the Sonagrams presented in figure 2. The possible significance of the above calls, as well as the ones described below for the duck embryo, is taken up in the Discussion section which follows. The duck embryo, like the chick embryo, is capable of producing at least three basically different vocalizations before hatching. On Day 24 (three days before hatching), it is possible to readily elicit at least two distinct types of vocalization, one by probing the embryo’s oral cavity with blunt forceps and the other by gently folding the forefinger over the embryo’s eyes and exerting mild pressure on the forehead and bill. Probing the oral cavity evokes a series of relatively high-pitched notes which are of the ascending-descending or predominantly descending type (fig. 3-l), and which peak at 4800 Hz (fig. 5: 0-1). By late Day 25 such stimulation evokes notes of a more uniformly descending character (fig. 3-3) which peak at 4000 Hz (fig. 5: D-3). Gentle pressure over the forehead, eyes, and bill on Day 24 elicits a series of relatively low-pitched notes (fig. 3-2) which peak around 2000 Hz and show a harmonic at about 4000 Hz (fig. 5: D-2). On Day 25 the same stimulation evokes a series of low-pitched notes (fig. 3-4) which peak at 1600-1800 Hz (fig. 5 : D4). On Day 25 the vocalizations of an undisturbed embryo at normal incubator temperature (fig. 3-5) resemble those produced by exerting gentle pressure on the facial region (fig. 3-4), and the frequency components are virtually identical 312 GILBERT GOTTLIEB AND JOHN G. VANDENBERGH zy zyx zyx zyx 71 /- zyxw +ZOK -1BK -I 6K -14K -1IK -1OK -8000 - -6000 - -A000 -1000 -1800 8000 -8000 6000 -6000 - - N I c -2000 -1800 -1b00 -1400 zyxw -1600 -1400 -1100 -1000 - -800 1000 -1200 -1000 - -800 -boo -ZI c zyxwvutsrqpo zyx \ 1:: -4000 YI a Y zyxw :I ,o 4zyxwvutsrqp 50 D-3: Late Day 25 FREQUENCY (Hz) D-6: Day 26-27 zyxwvutsrqp zyxw FREQUENCY ( H r ) W Fig. 5 Frequency analysis of the duck ( D ) embryo's vocalizations depicted by Sonagrams in figure 3. D-1 through D-6 correspond to 3-1 through 3-6 in figure 3. See text for further details. CL W 314 zyxwvuts zyxw GILBERT GOTTLLEB AND JOHN G. VANDENBERGH (cf. fig. 5 : D-5 and D-4). The third kind of vocalization, one emitted most frequently by the duck embryo during the hatching process late on Day 26 or early on Day 27, is a series of double notes of the ascending-descending type (fig. 3-6) with a peak frequency between 3200-3600 HZ (fig. 5 : D-6). While these vocalizations also contain low frequency components, some of the sounds range as high as 14,000 HZ (fig. 5: D-6). DISCUSSION Whether taken with reference to day of hatching or relative to length of incubation, domestic duck embryos are capable of vocalizing earlier than domestic chick embryos. Thus, the vocal apparatus of the duck embryo achieves a functional stage of development sooner than that of the chick embryo, indicating that the syringed structures of the duck embryo begin development sooner or develop at a faster rate than those of the chick embryo. Since the structural development of the syrinx has been studied over the entire course of incubation only in chick embryos (Tymms, ’13), it is not possible to specify which of the above alternatives is correct, or whether they are both correct. [Wunderlich’s (1884, pp. 73-76) study of the embryological development of the duck syrinx is incomplete, dealing largely with the distinction between bone and cartilage, and ending with the third week of incubation.] In our analysis of the ontogeny of vocal ability in chick and duck embryos, we have tended toward a conservative classification of the vocal repertoire of both species prior to hatching. Considering the markedly different contexts in which the calls were emitted, as well as the distinct differences in the spectrographic characteristics of the calls, it seems safe to conclude that both species are capable of uttering at least three different kinds of vocalizations prior to hatching. Since the postnatal vocal repertoire of the domestic chick has been extensively studied (Baeumer, ’62; Collias and Joos, ’53; Guyomarc’h, ’66; Konishi, ’63), it is relatively easy to relate the chick embryo’s vocalizations to the most prominent (i.e. most frequently emitted) calls after hatching. The call evoked by sharply probing the oral cavity of the embryo (fig. 2-1 and 2-2) has the distinctive features of what is commonly called the “distress call” in hatchlings. In addition to the contexts in which it is emitted (any number of situations which commonly produce pain, discomfort or a low level of fear), the “distress call” of the domestic chick is uniformly composed of predominantly descending long notes, extending over a relatively wide frequency range. The other well-known vocalizations of newly hatched chicks, the “pleasure notes,” are also emitted by the embryo. In the hatchling these notes are usually brief and of variable pitch, spanning a relatively narrow frequency range as they do in the chick embryo (fig. 2-5; fig. 4 : C-5). The third kind of vocalization, one produced by the undisturbed embryo at normal incubator temperature (fig. 2-3 and 2-4), resembles the sort of call uttered by hatchlings when they are in close conact and drowsy (Guyomarc’h, ’66, pp. 148150), i.e., in brooding-like situations. Embryos emit this call most frequently during the hatching process (fig. 2-4), though it also occurs earlier on Day 19 (fig. 2-3). While there have been no published studies of the vocalizations of ducklings, based on our own knowledge of the vocal repertoire of newly hatched ducks, it is possible to classify provisionally the duck embryo’s calls in relation to those of duck hatchlings. Upon consideration of the physical attributes of the calls as well as the contexts in which they were emitted, the three main vocalizations of the duck embryo would seem to correspond to the conventional designations given to the chick embryo’s calls: distress calls (fig. 3-1 and 3-3), pleasure or contentment cheeping (fig. 3-2, 3-4, and 3-5), and the “brooding-like’’call (fig. 3-6). There are, of course, spectrographic differences between the vocalizations of duck and chick embryos, so, for purposes of classification, it is necessary to consider both the context in which the call was emitted as well as the structure of the zyxw Call. Thus, chick and duck embryos are capable of uttering at least three distinct kinds of vocalizations prior to hatching and these vocalizations would appear to correspond to ones most frequently emitted after hatching. zyxwvu z zyxwvutsr VOCALIZATION I N AVIAN EMBRYOS Organic basis of sound production in birds: embryonic de-vocalization technique The previous experiment clearly indicates that duck and chick embryos are capable of uttering diverse vocalizations prior to hatching. This ability, coupled with the embryos’ ability to hear before hatching (Gottlieb, ’68a), poses a serious methodological difficulty for those workers who wish to study the ontogeny of species-typical sound production or auditory perception in supposedly “naive” hatchlings. For example, in studies of the ontogeny of speciestypical sound production, it is usual to deafen the subjects shortly after hatching to determine the role of auditory feedback in the subsequent establishment of the birds vocal repertoire. With such a procedure, however, it is not possible to evaluate the possible influence of embryonic hearing and/or vocal experience. To help rectify this difficulty, in our study of the organic basis of sound production in embryos, we have developed a technique for de-vocalizing embryos several days before hatching. Because of our own particular research interests, we have employed the technique most extensively with duck embryos. Since the de-vocalization technique simply involves the immobilization of the internal and external tympaniform membranes, and the vibration of these membranes seems to be indispensable to sound production in all avian forms (see Discussion), the present technique should have wide applicability in Aves. Due to interspecific variation in the accessibility (location) of the tympaniform membranes, however, it is more difficult to immobilize these membranes in certain species than in others. In chick embryos, for example, the unusual arrangement and location of the tympaniform membranes (Meyers, ’17 ) make it exceedingly difficult to de-vocalize them completely with the present technique. Domestic chicks differ markedly from passerine birds (Miskimen, ’51) , owls (Miller, ’34), and most waterfowl (Ruppell, ’33) in the arrangement and location of the tympaniform membranes, the latter three groups being highly similar in these respects. After the de-vocalization technique was perfected with hatchlings, 160 Peking duck 315 embryos were treated on Day 24 and tested for vocal ability after hatching. The devocalization procedure and the results are described in the following sections. It should be mentioned that a number of procedures other than the one described were explored and none of these other procedures was successful in terms of rendering the embryo completely mute except in those instances when the embryo died. Procedures involving surgical alteration of the syrinx, trachea, or syringeal musculature merely altered the quality or intensity of vocalization and not the ability to produce sound per se. When properly executed the present procedure completely mutes the embryo. Prior to devising the technique for devocalizing embryonic ducks, a series of observations were made on the vocal apparatus of newly hatched ducklings. It may be useful to recount these observations briefly so the reader can better appreciate how we arrived at the present technique. In the following account when it is implied that the vocalizations were “normal” or “unimpaired,” this means that observers who have had experience with literally hundreds of ducklings could not perceive any marked changes post-operatively. It is entirely possible that spectrographic analysis might have shown some minute qualitative changes in the ducklings’ vocalizations after our various operations, but for the present purpose such subtle changes were not of interest since our aims were to determine the anatomical basis of sound production and to develop a technique to mute the duckling. Opening the body cavity of etherized ducklings from the anteroventral aspect and incising the clavicular sac exposed the syrinx and its attendant muscles. (Puncturing the clavicular sac does not alter the duckling’s ability to vocalize.) The definitive sexual dimorphism of the male and female syrinx occurs early in embryonic development in domestic ducks (Day 12: Wolff and Wolff, ’51) and it is readily apparent in the syringes of hatchlings shown in figure 6-1 and 6-2. To determine whether the syrinx itself or the syringeal muscles were necessary for sound produc- zyxwv 316 zyxwvu zyx GILBERT GOTTLIEB AND JOHN G . VANDENBERGH a 9 m CDI zy zy zy bi Q c.l WI VOCALIZATION IN AVIAN EMBRYOS 317 318 GILBERT GOTTLIEB AND JOHN G. VANDENBERGH tion the following steps were taken. First, a hole approximately 2 mm in diameter was burned by electrocautery into the ventral surface of the syrinx immediately posterior to the syringeo-tracheal junction. The presence of such an opening in the syrinx of either male or female ducklings did not severely impair their ability to vocalize. (The vocalizations did have a more rasping quality than before the operation, probably due to the entrance of fluid into the syrinx and bronchi.) As a second step, all muscular attachments to the syrinx, i.e., the sternotrachealis and the bronchotrachealis muscles, were then removed from another set of ducklings. This operation only served to reduce the intensity (loudness) of the vocalizations of the ducklifigs so treated. The search for the organs essential for sound production only met with success when we examined the structures posterior to the syrinx. Incising the trachea and inserting an inverted Y-shaped polyethylene tube into the top portion of each bronchus eliminated all vocalizations. The tube was constructed to duplicate the inner dimensions of the trachea and bronchi, and the opening at its upper end was made to protrude through the skin of the duckling’s neck. The bifurcated tube was only effective for several hours because it either was forced cephaled (out of the bronchi) or it became filled with fluid. In the former case, the ducklings could vocalize and, in the latter case, the ducklings died. Use of the bifurcated tube suggested that intubation of the trachea of a freshly killed duckling would permit us to observe the movement of the vocal and respiratory structures below the syrinx during artificial inspiration and expiration. Thus, a glass pipette of 2 mm 0.d. was inserted into the trachea of a dead duckling approximately 2 cm above the syrinx. Alternate blowing and sucking on the pipette produced respiratory movements and, with practice, a recognizable “peeping” sound could be produced by the abrupt but gentle withdrawal of air. The quality of the artificially produced vocalization varied with both the speed with which air was withdrawn and the tension placed on the syrinx and bronchi as the trachea was manually stretched in a cephalsd direction. Under ordinary conditions, the duckling’s posture combined with the action of the sterno- and bronchotrachealis muscles, would impart different degrees of tension to the trachea, syrinx and bronchi. Directly associated with the artificially produced sounds, the internal tympaniform membranes collapsed into the bronchial cavities at the syringeo-bronchial junction (fig. 6-4). The external tympaniform membranes contribute to constricting the bronchial lumen to a lesser degree by forming a concavity between the semi-lunaI bronchial rings on the lateral wall of the upper bronchi. Both these actions, but primarily the infolding of the internal tympaniform membranes are clearly shown in figure 6-4. When air was gently blown down the trachea, the membranes ballooned outward and the internal tympaniform membranes reduced the interbronchial foramen to a narrow slit (figure 6-3). It was not possible to produce sound by blowing air into the trachea, regardless of the tensions placed on the trachea, syrinx, and bronchi by manual manipulation. This series of observations isolated the tympaniform membranes as the primary organs for sound production in the duckling. Although marked sexual dimorphism exists in the configuration of the duckling syrinx, no sex differences have yet been detected in the quality of naturally or artificially produced notes of embryos or hatchlings. It is possible that the syrinx becomes especially important only later on in development when distinct differences in male and female vocalizations are quite evident. zyx zyxwvut MATERIALS AND METHODS zy The internal and external tympaniformmembranes of 160 Peking duck embryos were made rigid by applying non-flexible “Collodion Merck (24% alcohol - 245 grains ether per fl. oz.) to them. Collodion is non-toxic and it rapidly forms a plastic sheath over the membranes. The location of the tympaniform membranes is shown in figure 6-1. When the tympaniform membranes are made rigid, the embryo or hatchling cannot vocalize. When the (201lodion sheath is removed, the ability to zyxwvu 319 VOCALIZATION I N AVIAN EMBRYOS produce sound is restored.’ The devocalization procedure can be completed in about 15 minutes and it is accomplished as follows. First, the egg is candled and the airspace and bill of the embryo traced in pencil on the outside of the shell. On Day 24 the duck embryo’s bill begins to penetrate the air-space and it can be seen as a tent-like protrusion during candling. Next the shell and outer shell membrane over the air-space are removed by cutting with iris scissors. The opaque inner shell membrane is painted with warm saline, thus completely revealing the bill of the embryo and the large veins of the chorioallantois. The embryo’s head and chest are brought into the air-space by grasping the embryo’s bill and slowly pulling the embryo’s head through the chorioallantoic membrane and inner shell membrane. In those cases where the embryo’s bill has not completely punctured the membranes, a slit is made in the chorioallantois at the point of the embryo’s bill, taking special care to avoid the larger veins of the chorioallantois. zy zy After the embryo’s head and chest are exposed, the down on the embryo’s ventral surface is shaved or closely cropped around the midline extending from the lower neck region to the sternum (fig. 7). Next, the skin is cleansed with 70% alcohol and a topical anesthetic (“Xylocain 10% ”) is applied to the shaved area, allowing 1-2 minutes for the anesthetic to take effect. Subsequently, a longitudinal, midline incision is made in the skin parallel to the trachea beginning about 10 mm anterior to the sternum and extending to the sternum. The easiest and safest way to make this slit in the skin is to raise the skin with forceps (fig. 7-1) and then cut the skin with iris scissors rather than a scalpel so the depth of the incision can be better controlled. When the incision is properly made, little ____ 1 After the Collodion has remained on the membranes of the growing embryo for several days, it IS very difficult to remove the hardened coating without causing tissue damage. We have not yet devised a satisfactory, non-injurious technique to remove the Collodion and thereby re-vpcalize the treated embryos after hatching. It is a simple matter to apply the Collodion, test the embryo’s vocal ability, and remove the Collodion all within a relatively short time period. /- Fig. 7 in text.) mustration of main steps in embryonic de-vocalization procedure. (Description 320 GILBERT GOTTLIEB AND JOHN G. VANDENBERGH or no bleeding is evident. After the skin is parted with forceps or hemostats, the transparent membranes (clavicular sac) over the syrinx are punctured with the sharp end of the scissors, thereby allowing access to the tympaniform membranes. The tympaniform membranes lie immediately posterior to the syrinx at the proximal end of the two bronchi (see fig. 6). The syrinx cannot be seen until the clavicular sac is punctured. In the duck as in many other avian species, the internal tympaniform membranes are much more extensive than the external tympaniform membranes. In the Peking duck, the internal tympaniform membranes extend from the first to the sixth bronchial ring while the external tympaniform membranes lie between the first and second bronchial rings. In the duck, the male syrinx is relatively large and asymmetrical (fig. 6-2) while the female syrinx is smaller and symmetrical (fig. 6-1). When the bird vocalizes air is expelled from the lungs and the internal tympaniform membranes assume a concave or “deflated shape (fig. 6-4). When the bird inhales, the internal tympaniform membranes assume a convex or bulged appearance (fig. 6-3). When the tympaniform membranes are made rigid by the application of non-flexible Collodion, it is not possible for the bird to vocalize. Collodion is briskly dabbed on the tympaniform membranes with a cosmetic “eyeliner” brush (fig. 7-2). To gain more complete access to the membranes without cutting into the sternum, the trachea is stretched by pulling the embryo’s head in a cephalad direction. This manipulation usually brings the syrinx into clear view and the Collodion is applied directly below the syrinx (fig. 7-2). Collodion dries very rapidly so care must be taken not to strip it off the membranes inadvertently during the application. Two quick applications are usually sufficient to cover the membranes. When their location is known, it is not actually necessary to see the tympaniform membranes when applying the Collodion. In ducks and perhaps other species with extensive internal tympaniform membranes, it is only these membranes which require Collodion. That is, when Collodion is properly applied to the internal tympani- z form membranes of duck embryos, it i s usually not necessary to treat the external tympaniform membranes. Following application of the Collodion sheath, the incision is sutured with 000 surgical silk using a three and one-eighth circle, eyed needle as shown in figure 7-3. For those investigators who do not wish to de-vocalize embryos, this same technique can be used with hatchlings. It is to be emphasized that there are inter-specific differences in the accessibility and location of the tympaniform membranes, so the present technique for applying Collodion to the membranes may have to be modified to a greater or lesser degree with species other than the Peking duck (Anas platyrhynchos). The essential feature for complete de-vocalization is immobilization of the tympaniform membranes. T e s t of effectiveness of de-vocalization procedure. One hundred and sixty Peking duck embryos were subjected to the devocalization procedure on Day 24 (25th day of incubation); 95 of these hatched and were tested for vocal ability on the day after hatching. For the test of vocal ability the ducklings were placed in an “imprinting” apparatus (Gottlieb, ’65a) for 20 minutes during which time they were exposed to the maternal attraction call of their species. Under these testing conditions, all normal Peking ducklings vocalize innumerable times during the 20-minute test period. Data on hatchability, sex ratio, etc. were also tabulated in the treated embryos. For control purposes, similar data were tabulated on 29 Peking duck embryos whose head and chest were exposed on Day 24 and from which embryonic EKG and behavioral recordings were made by means of needle electrodes in their skin (Gottlieb, ’65b). zy z zyxw RESULTS As shown in table 3 , the de-vocalization procedure was 85% effective. That is, only 14 of the 95 treated hatchlings could vocalize. At least two of these vocalized essentially normally, while the other 12 were only occasionally capable of a very low intensity vocalization. In terms of viability, it can be seen in table 3 that the de-vocalization procedure is no more traumatic than opening the egg and making behavioral and physiological zyxwvu zyxwv zyxwvuts zyxwvutsrq zyxwvut 32 1 VOCALIZATION I N AVIAN EMBRYOS TABLE 3 Comparison of viability and vocal ability in duck embryos subjected to two d i f f e r e n t operative procedures three days before hatching, one involving exposure and de-vocalization and t h e other involving exposure and behavioral recording 160 duck embryos subjected to devocalization procedure 29 duck embryos not submitted to de-vocalization Drocedure of the de-vocalization procedure is not known with any degree of reliability. The one bird which has been kept for observation is five months old at this time (December, 1967), and she is still mute. zyxwvu DISCUSSION The success of the present technique in muting embryos and hatchlings sheds some I % definitive light on the anatomico-physiologMales 50 ? ical mechanism of sound production in Females 50 ? birds. While it has been known for over 200 In air-space 43 55 Hatched 59 69 years (Herissant, 1753) that the tympanid 58 ? form membranes play a critical role in 63 ? 0 avian sound production, it has never been in air-space 77 94 experimentally demonstrated that these not in air-space 45 38 Mute after h a t c h i n g 85 0 elastic membranes are the fundamental source of sound production. That is, to our Note: Of the embryos i n the de-vocalization and control groups which hatched, the median age at the knowledge, it has not been previously demtime of the operation was 24 days 17 hours and 24 onstrated that if the tympaniform memdays. 19 hours respectively. The mkdian age of those which didn’t hatch was 24 days, 12 hours (de-vocal- branes are made rigid in situ, such a ized group) and 24 days, five hours (control group) procedure absolutely prohibits sound proat the time of the operation. If the embryo’s bill was in the air-space at the time of the operation, it was duction in the living bird. Since other almore likely to hatch i n both the de-vocalized ( p < 0.001) and control ( p < 0.01) groups. Thus, the terations of the vocal apparatus (slitting older the embryo at the time of the operation the more likely its bill had penetrated into the air-space, the trachea, severing the vocal muscles, and the greater likelihood it would hatch. There destroying the clavicular sac, puncturing were no statistically reliable differences in hatch the syrinx) failed to eliminate sound prorate between the de-vocalized and control embryos whether the embryos were in the air-space or not at duction, it therefore seems safe to conclude the time of the operation. that the tympaniform membranes are the recordings 011 Day 24. Also, there were no primary or basic source of sound in the significant differences in survival between vocal apparatus of birds. Since we are in the sexes. The likelihood of the duck em- the curious position of claiming to provide bryo hatching after removing the shell and definitive evidence for a conclusion which membranes at the large end of the egg, was reached 200 years ago, it may help to whether submitted to the de-vocalization put our belated contribution in the approtechnique or not, was considerably en- priate perspective by briefly reviewing the hanced if the embryo’s bill had even barely experimental evidence on the role of the penetrated the air-space (table 3). Specifi- tympaniform membranes in avian sound cally, 77% of the embryos hatched that production. were in the air-space at the time of the If a clear distinction is made between operation, whereas only 45% hatched of anatomical and physiological (functional) those not in the air-space at the time of de- evidence, the physiological side of sound vocalization ( p < 0.001). This functional production has actually never been demoncriterion (air-space penetration) is related strated in birds. The main researchers into age, but it is more important than age terested in this problem (Herissant, 1753; per se. However, when the embryo’s bill Cuvier, 1800; Yarrell, 1833; J. Muller, has penetrated the air-space, the embryos 1847; Griitzner, 1879; Wunderlich, 1884; are much more likely to vocalize during the Riippell, ’33) built up a plausible and more operation itself and this may not be desir- or less accurate account of the physioable for certain kinds of experiments. 2 Our source of duck eggs is the C & R Duck Farm, Due to an extraneous difficulty2 it has Westhampton, Long Island, New York. Since there not been possible to keep a large number of have been several recent outbreaks of duck viral enteritis i n the Long Island afea, we have been asked by the muted ducklings for observation be- the U.S.Department of Agrlculture to take appropriate and It has been necessary to sacrxfice all yond the first four days after hatching. precautions of our Peking ducklings within a few days after Consequently, the long-term effectiveness hatching. zyxwv zyxwvu z 322 zyxwvut zyxwvutsr GILBERT GOTTLIEB AND JOHN G. VANDENBERGH logical side of avian sound production based on ( a ) anatomical considerations and ( b ) a functional analogy based on the mechanism of sound production in certain musical instruments. Following the lead of Herissant and Cuvier, all of the above investigators agreed that the vocal apparatus of birds is analogous to the mouthpiece of a wind instrument. In the case of instruments like the trumpet or trombone (Cuvier), the lips of the player are said to act like the tympaniform membranes, while with instruments such as the oboe (Herissant), the reeds inside the mouthpiece serve the same function as lips (or tympaniform membranes). The main difference between Herissant’s and Cuvier’s model was Herissant’s belief that the clavicular sac was also essential to sound production since it acts as a counterforce to the air streaming through the narrow opening formed by constriction of the tympaniform membranes during exhalation. As we shall see, this mistaken idea about the importance of the clavicular sac was recently revived by Ruppell (’33). The only published experiments that were performed in the eighteenth and nineteenth centuries which in some way related to the functional side of sound production were the following. 1. Herissant (1753, p. 293) punctured the clavicular sac of live ducks andreported that the birds no longer vocalized. On the basis of this experiment, Herissant concluded that the clavicular sac plays an essential role in sound production, in conjunction with the tympaniform membranes. (This problematic interpretation is discussed below. ) 2. In two experiments Cuvier (1800, pp. 430-431) slit the trachea of a living blackbird and decapitated a living magpie; the blackbird vocalized more or less normally after the operation and the headless magpie continued to vocalize for ten minutes until a blood clot in the trachea caused the bird to suffocate. From these experiments Cuvier correctly localized the source of sound in the “lower larynx” (syrinx). 3. Grutzner (1879, pp. 139-141) opened the chest of a live female turkey and observed that the tympaniform membranes collapsed inward in conjunction with sound production. This experimental observation, though only correlational in nature, js the most germane one in the sense that it did show the movement of the tympaniform membranes during vocalization. Finally, in 1933, Ruppell revived Herissant’s idea on the importance of the clavicular sac operating in conjunction with the tympaniform membranes. From experiments in which he placed the excised trachea, syrinx, and bronchi of various species in a glass atmosphere chamber, Ruppell concluded that “. . . location of the syrinx in the clavicular air sac is the physiological basis and the necessary condition for the special kind of sound production in the vocal apparatus of birds” [translated from page 450 of original article (Ruppell, ’33)].3 Ruppell described the results of his experiment with excised syringes in the glass atmosphere chamber as follows. zyxw “The experiment showed that when blowing air through the bronchi i n the direction of the trachea, while a t the same time maintaining a superatmospheric pressure in the glass chamber, the drum membranes (Membrane tympaniformes internae) went into a strong oscillation and produced a strong sound, which ceased whenever the pressure in the glass chamber was reduced to the normal atmospheric pressure (1 atm.); externally this became apparent in that with continued blowing through the bronchi, the drum membranes bowed out strongly, but without sound production. In the living bird the process occurs as we have already said, when through a puncture in the wall of the clacivular air sac the air surrounding the syrinx can escape. This observation was the more surprising to me, since at that time I was unaware of the discoveries of Herissaut (sic). Later, as opportunity offered, I have induced oscillations in the syrinx of other bird types which departed not too significantly from the normal type (Sula capensis, Stercorarius skua, Anus platyrhynchos domestica 0 , Arclea cinerea). It was obviously necessary to work with large or medium-large types, since the vocal organs of the smaller birds (even of doves) did not lend themselves to this type experiment. The morphological structure of the syrinx of small birds indicates that with them also, sound production is based on the primary oscillation of the membrane tympaniformes, and follows the same acoustical p r k ciple. The interbronchial foramen is of specid significance for the oscillating process; it gives the membranes a free space in which to vibrate, and provides for the air in the clavicular sac a surface on which to operate” [translated from pp. zyxwv zyxwvu zy zyxwvut 3 This translation and subsequent ones from Ruppell‘s classic monograph were made by Crawford H. Greenewalt. The correctness of each translation quoted in the text has been verified by one of us (G.G.). zyxwvu zyxw VOCALIZATION IN AVIAN EMBRYOS 451-453 of the original German language article (Ruppell, ’33 >I. Since we found that puncturing the clavicular sac in living ducklings and embryos did not alter their ability to produce sound, it is necessary for us to conclude that the clavicular sac does not play a n essential role in avian sound production when studied i n situ. In speculating on the factors which caused Ruppell to reach an apparently erroneous conclusion about the function of the clavicular sac, two possibilities seem most evident. The first is that Ruppell’s use of excised vocal organs in a glass atmosphere chamber introduced an artifact not actually present during sound production in situ. The second possibility is that Ruppell worked exclusively with adult birds (including Anas platyrhynchos) and that the clavicular sac is essential for sound production in older birds. We tested this second possibility by puncturing the clavicular sac of a 12-week-old female Peking duck (Anas platyrhynchos) and three older (6-1 8 months) Mallard ducks (Anas platyrhynchos), including one male and two females, and we found that all of the birds could still vocalize after their clavicular sac was punctured. There were no readily discernible qualitative changes in the ducks’ vocalizations after the sac was punctured and they all vocalized while the chest wound was open. Therefore, since Ruppell himself made a clear distinction between mechanisms of sound production and mechanisms of sound modulation, we must conclude that the use of the glass atmosphere chamber led Ruppell to an erroneous conclusion regarding the role of the clavicular sac in sound production. (Why Herissant’s ducks did not vocalize after their clavicular sac was punctured remains obscure.) Further support for the essential role of the tympaniform membranes in avian sound production comes from the de-vocalization technique devised by Durant (’53) and further developed by Van Krey and Ogasawara (’64). Durant’s technique involves puncture of the semilunar membrane and cauterization of the pessulus bone. The pessulus is located at the base of the syrinx and unites the two bronchi, immediately anterior to the internal tympaniform membranes. This operation, though it reduces 3 23 vocal intensity and sometimes modifies sound production, does not result in muting guinea fowl, turkeys, or chickens (Van Krey and Ogasawara, ’64, pp. 82 and 83). Thus, these findings lend further support to the contention that the constriction and vibration of the tympaniform membranes during exhalation is the essential feature of sound production in birds, as correctly deduced by Cuvier in 1800 and reiterated by Grutzner in 1879. It would appear that the syrinx, syringed musculature, and trachea play an important role only in modulating or resonating the sound produced by the rush of air past the constricted tympaniform membrane^.^ Regarding the direction of air flow during vocalization, Miller (’34) and Miskimen (’51) obtained the same results as we did with chicks and ducklings, using a similar method on dead owls and passerine birds, respectively. Namely, sound was produced only when air was drawn or forced past the tympaniform membranes in a caudocephalad direction. Miller states (p. 208) “Gratifyingly, the sound produced was similar to a Horned Owl hoot in pitch and in many points of quality, although it lacked greatly in resonance and volume.” With reference to an English Sparrow (Passer domesticus) similarly prepared, Miskimen states (p. 494): “No sound was produced by forcing air into the trachea, but when the bulb was released and air drawn back into the bulb the bronchial tubes partly collapsed and a chirping noise was made. The internal tympaniform membrane just above the insertions of the bronchidesmus membrane on the first bronchial half-rings vibrated freely and rapidly in the inter-bronchial space.” Miskimen also obtained the same results regarding the direction of air flow and sound production with a Ring-necked Pheasant, Phasianus colchicus, and a Starling, Sturnus vulgaris. Finally, attesting to the possible greater importance of the internal tympaniform membranes relative to the external ones for sound production in many if not all avian zy zyxwvut zyxwv zyx 4 The neural control of avian sound production has yet to be completely worked out-progress in this area has been recently summarlzed by Brown (’68). If decapitated birds can vocalize, as contended by Cuvier (1800), then the proximal control of vocalization must reside in spinal neurons. 324 zyxwvutsr zyxwv zyxw zyxw GILBERT GOTTLIEB AND JOHN G. VANDENBERGH species, immobilization of the external tympaniform membranes (Gross, 64a) in chickens, peacocks, and guinea fowl does not mute the bird but only modifies the pitch and intensity of the sounds produced (Tudor and Woodard, '67). In the domestic chicken (Gross, '64b; Myers, 'l?), and presumably peacocks and guinea fowl, the internal tympaniform membranes are less extensive than the external ones, but the internaI membranes are located in the bronchi whereas the external membranes have a more rostra1 location. Therefore, it would appear that the internal tympaniform membranes assume greater importance than the external ones for sound production ( a ) when they are located in the bronchi and/or ( b ) when they are considerably more extensive (spatiaIly) than the external ones. While most avian species have two pairs of tympaniform membranes (internal and external), there are exceptional species such as the Trumpeter Swan ( C y g n u s cygnus) which, according to Ruppell ('33, p. 466), has only one pair of tympaniform membranes (external). Though it seems that there are always some exceptions to any generalization about the structure of the vocal apparatus in birds, there are no known exceptions to the rule that all avian species have at least one structure in their vocal apparatus which is capable of vibrating and this elastic structure provides the basis for sound production, whether it is Iocated in the bronchi, syrinx, or trachea. LITERATURE CITED Baeumer, E. 1962 Lebensart des Haushuhns, uber seine Laute und allgemeine dritter Teil Erganzungen. Zeit. Tierpsychol., 19: 394416. Brown, J. L. 1968 The control of avian vocalization by the central nervous system. In: Studies on Bird Song, R. A. Hinde, ed. Cambridge University Press, in press. Collias, N., and M. Joos 1953 Spectrographic analysis of sound signals of the domestic fowl. Behaviour, 5: 175-188. Cuvier 1800 Mdmoire sur les Instrumens de la Voix des Oiseaux. J. de Physique, de Chimie, et #Histoire Naturelle, 7: 426-451. Dilger, W. C. 1956 Hostile behavior and reproductive isolating mechanisms in the avian genera Hylocichla, Auk, 73: 313-353. Durant, A. J. 1953 Removing the vocal cords of the fowl. J. Amer. Vet. Med. Assoc., 122: 14-17. Gottlieb, G. 1965a Imprinting in relation to parental and species identification by avian neonates. J. Comp. Physiol. Psychol., 59: 345356. 1965b Prenatal auditory sensitivity in chickens and ducks. Science, 147: 1596-1598. 1966 Species identification by avian neonates : contributory effect of perinatal auditory stimulation. Anim. Behav., 14: 282-290. 1968a Prenatal behavior of birds. Quart. Rev. Biol., 43: 148-174. 1968b Prenatal development of vocal ability i n birds. In: Animal Behavior in Laboratory and Field, A. W. Stokes, ed. Freeman Press, San Francisco, pp. 157-160. Gross, W. B. 1964a Devoicing the chicken. Poultry Sci., 43: 1143-1144. 1964b Voice production by the chicken. Poultry Sci., 43: 1005-1008. Griitzner, P. 1879 Physiologie der Stimme und Sprache. In: Handbuch der Physiologie, vol. 1, pt. 2, Hermann, ed. F. C. W. Vogel, Leipzig, pp. 1-236. Guyomarc'h, J.-C. 1966 Les Cmissions sonores du Poussin domestique, Ieur place dans le compartement normal. Zeit. Tierpsychol., 23: 141160. Herissant 1753 Recherches sur les Organes de la Voix des QuadrupCdes, et de Celle des Oiseaux. Mdmoires de Mathdmatique et de Physique, tires des Registres de l'Acad6mie Royale des Sciences, 279-296. Konishi, M. 1963 The role of auditory feedback in the vocal behavior of the domestic fowl. Zeit. Tierpsychol., 20: 349-367. Kuo, 2.-Y., and T. C. Shen 1937 Ontogeny of embryonic behavior in Aves. XI. Respiration in the chick embryo. J. Comp. Psychol., 24: 49-58. Lanyon, W. E. 1960 The ontogeny of vocalization in birds. In: Animal Sounds and Communication. W. E. Lanyon and W. N. Tavolga, eds. Amer. Inst. of Biol. Sciences, Washington, pp. 321-347. 1963 Experiments on species discrim~ ination i n Myiaschus flycatchers. Amer. Mus. Novit., No. 2126: 1-16. M a e r , A. H. 1934 The vocalapparatus of North American owls. Condor, 36: 204-213. - zyxwvut zyxwvut zyxwvutsr ACKNOWLEDGMENTS Miss Evelyn Strickland, Mr. Lincoln Gray, and Mrs. Carole S. Ripley aided in the collection and tabulation of the data. Dr. Colin Beer kindly performed the Sonagraphic analysis of the calls. The experiments were partially supported by funds from U.S.P.H.S. Research grant HD-00878 from the National Institute of Child Health and Human Development. We appreciate the assistance of Dr. and Mrs. Peter N. Witt in translating certain of the foreign language literature. This paper is dedicated to professorD ~ ald K. Adams on the occasion of his retirement from Duke University. zyxw zyxwvu zyxw VOCALIZATION IN AVIAN EMBRYOS Miskimen, M. 1951 Sound production in passerine birds. Auk, 68: 493-504. Muller 1847 Uber die bisher unbekannten typischen Verschiedenheiten der Stimmorgane der Passerinen. Abhandlungen der Koniglichen Akademie der Wissenschaften (Berlin), 321391. Myers, J. A. 1917 Studies on the syrinx of GalZus domesticus. J. Morph., 29: 165-215. Ramsay, A. 0. 1951 Familial recognition in domestic birds. Auk, 68: 1-16. Riippell, W. 1933 Physiologie und Akustik der Vogelstimme. J. Ornithologie, 81: 433-542. Stein, R. C. 1963 Isolating mechanisms between populations of Traill's Flycatchers. Proc. Amer. Phil. SOC., 107: 21-50. Thorpe, W. H. 1961 Bird-Song. Cambridge Univ. Press, 143 pp. 325 Tudor, D. C., and H. Woodward 1967 A method for devoicing fowl. J. Amer. Vet. Med. Assoc., 151: 616-617. Tymms, A. 0.V. 1913 The syrinx of the common fowl, its structure and development. Proc. Roy. SOC.Victoria, 25: 286-306. Van Krey, H. P., and F. X. Ogasawara 1964 Electrosurgical devocalizing of poultry. Avian Diseases, 8: 81-85. Wolff, Et., and Em. WoIff 1951 The effects of castration on bird embryos. J. Exp. Zool., 116: 59-98. Wunderlich, L. 1884 Beitrage zur vergleichenden Anatomie und Entwickelungsgeschichte des unteren Kehlkopfes der Vogel. Nova Acta der Ksl. Leop.-Carol. Deutschen Akademie der Naturforscher (Halle), 48: 1-80. Yarrell, W. 1833 On the organs of voice in birds. Trans. Linn. SOC.(London), 16: 305421. zyxwvu zyxwvu zyxwvutsrqp zyxwv zyxwvut Added in proof: Subsequent experience with the de-vocalization technique indicates that it is much simpler to de-vocalize embryos or hatchlings than adult birds. The main problem with the older birds is obtaining access to the tympaniform membranes - such access requires a much more extensive chest incision than in the young birds. Most of the de-vocalized embryos regain their vocal ability within three months after hatching - a few individuals remain mute for an indefinitely longer period.