Morphology of the tongue of the emu ( Dromaius novaehollandiae ) . I . Gross anatomical features and topography

Despite numerous papers addressing the topic, the gross morphology of the ratite tongue and more specifically that of the emu, has been superficially or poorly described. This paper presents the first definitive macroscopic description of the emu tongue and reviews, consolidates and compares the scattered information on the gross morphology of the ratite tongue available in the literature. Twenty-three heads obtained from birds at slaughter were used for this study. Specimens were fixed in 10 % neutral buffered formalin, rinsed and the gross anatomy described. The emu tongue is divided into a body and a root. The body is triangular, dorsoventrally flattened, pigmented and displays caudally directed lingual papillae on both the lateral and caudal margins. The root, a more conspicuous structure in comparison to other ratites, is triangular, with a raised bulbous component folding over the rostral part of the laryngeal fissure. Following the general trend in ratites, the emu tongue is greatly reduced in comparison to the bill length and is specifically adapted for swallowing during the cranioinertial method of feeding employed by palaeognaths. This study revealed that it is not only the shape of the tongue that differs between ratites, as previously reported, but also its colour, appearance of its margins and root, and its length in comparison to the bill, and the shape of the paraglossum.


IntroductIon
The gross morphological features of the avian tongue have been described in numerous species (see McLelland 1979 for a review of the earlier lit erature) and the structural adaptations of this organ linked to diet and mode of feeding (Gardner 1926(Gardner , 1927)).Many of these studies, particularly the earlier works, presented comparative information on the macroscopic features of the tongue with a view to providing taxonomic data (Lucas 1896(Lucas , 1897;;Gard ner 1926Gard ner , 1927;;Harrison 1964).This information was subsequently used to classify the tongue of birds into various categories.Gardner (1926Gardner ( , 1927)), for example, recognised eight categories based on the function and adaptations of this organ.Harrison (1964), on the other hand, proposed the classifica tion of avian tongues into five functional groups, namely tongues specialized for collecting food, eat ing, swallowing, taste and touch, and nest building.
Echoing the suggestion by Gardner (1926Gardner ( , 1927) ) that microscopic data would enhance the under standing of macroscopic features, recent studies have generally supplied more comprehensive infor mation on the structure of the avian tongue by using both light and electron microscopy in addition to macroscopic descriptions (Kobayashi, Kumakura, Yoshimura, Inatomi & Asami 1998;Jackowiak & Godynicki 2005;Jackowiak & Ludwig 2008;Tivane 2008).
Due to their commercial importance, the tongue and associated hyobranchial apparatus of domestic poul try have been described in detail (see Calhoun 1954 for a review of the earlier literature; Hodges 1974;McLelland 1975;Nickel, Schummer & Seiferle 1977;Homberger & Meyers 1989).
During the past 180 years numerous publications on the ratite tongue have appeared in the form of sketches, descriptions and comparisons (Meckel 1829;Cuvier 1836;Gadow 1879;Owen 1879;Py craft 1900;Göppert 1903;Duerden 1912;Farag gi ana 1933;Roach 1952;Feder 1972;McCann 1973;Cho, Brown & Anderson 1984;Fowler 1991;Bonga Tomlinson 2000;Gussekloo & Bout 2005;Porches cu 2007;Crole & Soley 2008;Jackowiak & Ludwig 2008;Tivane 2008).Many of these studies, how ever, provide incomplete and sometimes mislead ing information on the macroscopic features of this organ.This situation is exacerbated by the fact that some descriptions are based on limited numbers of specimens ranging from embryos to fully mature birds, resulting in conflicting information that is diffi cult to interpret.The most comprehensive studies of a ratite tongue are those of Jackowiak & Ludwig (2008) and Tivane (2008) on the ostrich, although the former authors neglected to reference any of the earlier literature on this topic.
To date there have only been four reports on the gross morphology of the emu tongue.The most complete gross morphological description is that of Faraggiana (1933) who studied a single excised specimen of the tongue and laryngeal mound.Crole & Soley (2008) described the basic features of the emu tongue.In a study of feeding in palaeognathous birds , Bonga Tomlinson (2000) depicts the outline of the emu tongue in relation to the hyobranchial apparatus and surrounding mandibular rami, and briefly describes the presence of lingual papillae.Cho et al. (1984) simply note that "the emu tongue has a serrated edge".This paper presents the first definitive macroscopic description of the emu tongue and reviews, consoli dates and compares the scattered information on the gross morphology of the ratite tongue available in the literature.This study not only contributes to a better understanding of the upper digestive tract of the emu but also provides data that can be used for more meaningful future comparative studies of the ratite tongue.

Specimen collection
The heads of 23 adult (14-15 months) emus of ei ther sex were obtained from a local abattoir (Oryx Abattoir, Krugersdorp, Gauteng Province, South Africa) immediately after slaughter of the birds.The heads were rinsed in running tap water to remove traces of blood and then immersed in plastic buck ets containing 10 % neutral buffered formalin.They were allowed to fix for approximately 4 h while being trans ported to the laboratory, after which they were immersed in fresh fixative for a minimum period of 48 h.Care was taken to exclude air from the oro pharynx by wedging a small block of wood in the beak.

Specimen preparation and recording
The specimens were rinsed in running tap water and each preserved head was used to provide infor mation on the gross anatomical features of the tongue and its topographical relationships within the oropharyngeal cavity.This was achieved by incising the right commissure of the beak, disarticulating the quadratomandibular joint and reflecting the mandi ble laterally to openly display the roof and floor of the oropharynx (Fig. 1).The length (from the apex to the caudal edge of the caudal papillae) and width (between the tips of the last lateral papillae) (Fig. 2) of 16 tongues were measured and the lateral and caudal lingual papillae counted.The bill length was measured on the mandibular rhamphotheca from the commissure to the rostral bill tip.Relevant ana tomical features were described and recorded using a Canon 5D digital camera with a 28-135 mm lens and a Canon Macro 100 mm lens for higher magni fication photographs.
Three tongues were removed from the heads by lift ing the organ from the floor of the oropharynx and cutting through the frenulum as well as the paired ceratobranchiale and urohyale of the hyobranchial apparatus.The mucosa was stripped from the tongues to expose the intraglossal elements (Fig. 7 and 8) of the hyobranchial apparatus.All specimens were studied and described at the Department of Anatomy and Physiology, Faculty of Veterinary Sci ence, University of Pretoria, South Africa.This study was approved by the research committee of the Fac ulty of Veterinary Science (Protocol no.V040/08) and complied with the regulations regarding the ethical use of animals.The terminology used in this study was that of Nomina Anatomica Avium (Baumel, King, Breazile, Evans & Vanden Berge 1993).

reSultS topography
The tongue of the emu consists of a rostral pigment ed body and a caudal, variably pigmented root, both of which lie within the confines of the nonpigment ed regions of the roof and floor of the oropharynx (Fig. 1).The tongue body occupies the middle third of the floor of the oropharynx and is a triangular structure with the apex pointing rostrally.The tongue root (Fig. 1 and 4) extends from the caudal lingual papillae to the glottis and is flanked by, but does not extend to, the paired ceratobranchiale of the hyo branchial apparatus.In the closed gape, the caudal margin of the tongue body lies beneath and in con tact with the rostral border of the choana, whereas the triangular tongue root fits snugly into the rostral aspect of the choana.In some tongues the apex is observed, in the closed gape, to make contact with the base of the median palatine ridge which origi nates at the border of the pigmented and nonpig mented regions of the palate.

tongue body (Corpus linguae)
The tongue body is dorsoventrally flattened (Fig. 5) with the dorsum being slightly raised in the centre and sloping towards the margins.It varies in length between 21-27 mm (average of 23.6 mm), and in width between 20-29 mm (average of 25.9 mm) (Fig. 2).The apex (Apex linguae) is rudimentary and varies in shape from a sharp point (Fig. 1), to a blunt or rounded tip.In some instances the apex is invaginated by a shallow groove forming two small er points (Fig. 2).The dorsal surface (Dorsum ling uae) is pigmented giving it an ashgrey/brown col our in formalinfixed specimens (Fig. 1 and 2).The ventral surface (Ventrum linguae) (Fig. 6) is lighter in colour than the dorsal surface with the epithelium appearing glasslike (transparent).The rostromedi al region of the ventrum is slightly concave.A con spicuous, lightcoloured, fingerlike projection ex tends along the midline from the tip of the frenulum to end bluntly caudal to the apex (Fig. 6).This line represents the rostral projection of the basihyale (see below) (Fig. 8).From the rostrolateral surfaces of the frenulum two raised bands (crura) (Fig. 6), are directed and taper towards the apex.Numerous pale doughnutshaped structures with a darker cen tre are clearly visible beneath both the dorsal and ventral surfaces of the tongue body (Fig. 2 and 3).Light microscopy confirmed that each of these struc tures constitutes a glandular unit with a central lu men/duct opening onto the lingual surface (Crole & Soley 2008), which grouped together form the lin  (McLelland 1993).In some tongues, these structures are obscured due to a darker colouration of the dorsum and only the openings, resembling pits, are visible (Fig. 4).

Margins (Margo linguae)
The three margins of the tongue body display two sets of lingual papillae (Fig. 1 and 2), the left and right lateral lingual papillae (Papillae linguae late rales) and the caudal lingual papillae (Papillae ling uae caudales).
The first lateral papillae originate on either side of and just caudal to the apex.These are the smallest of the lateral papillae and are directed laterally or caudolaterally.The rest of the papillae progressive ly point more caudolaterally and become longer and more slender.The last papillae are the longest and most caudally directed, and in some specimens ex hibit a pale tip.In some instances individual papillae emanate from the base of adjacent papillae (Fig. 2) and not directly from the lingual margin.The number of papillae present on the lateral lingual margins is variable and not necessarily equal on both sides.Although the left and right lateral margins demon strate a similar range of papillae (three to eight on the left side and five to eight on the right side), there appears to be a consistently higher number of papil lae on the right margin than compared to the left.The average number of lateral papillae on the tongues studied is 11.2.The doughnutshaped struc tures seen below the surface (Fig. 3) end abruptly just beyond the root of the lingual papillae, although in the last lateral and caudal papillae they extend to the papillae tips.
The caudal lingual papillae (Fig. 1, 2 and 4) are ru dimentary and poorly defined compared to the lat eral papillae and demarcate the caudal boundary of the tongue body.In some instances (n = 4) the cau dal papillae appear as a fused, centrally positioned structure with variable incisures and small projec tions (Fig. 4).In other specimens (n = 4) the fused component is flanked on either side by a single, more typical papilla.In a number of tongues (n = 8) the fused component displays a shallow median groove resulting in the formation of two median pa pillae which are accompanied by a variable number (zero to two) of adjacent papillae (Fig. 2).The cau dal papillae vary in number between one to four (av erage two and a half).In one specimen, a structure similar in appearance to a lingual papilla was ob served to project dorsally from the mucosa covering the left ceratobranchiale, just caudal to the last lat eral papilla.

tongue root (Radix linguae)
The tongue root (Fig. 1 and 4) is a fleshy triangular structure, which in most specimens, is nonpigment ed.The caudal extremity of the root ends as a rounded, raised bulbous structure (pigmented in some specimens) that extends into the rostral as pect of the laryngeal fissure (glottis).The mucosa of the tongue root is continuous with the rest of the

Frenulum (Frenulum linguae)
The frenulum (Fig. 5 and 6) is a fleshy nonpigment ed structure which attaches the caudal half of the tongue body to the oropharyngeal floor.It is triangu lar in shape, with the rostral attachment to the ven trum of the tongue forming the point of the triangle.The mucosa along the lateral edges is thrown into longitudinal folds.These folds are obliterated when The lingual skeleton shown in dorsal (7) and ventral (8) view The broad paraglossum (Pg) lies dorsal to the rostral projection of the basihyale (Br) within the tongue body.The body of the basihyale (Bb), the rostral parts of the paired ceratobranchiale (Cb) and the urohyale (U) are all imbedded within the frenulum (see Fig. 6).Bar = 5 mm the tongue body is lifted dorsally from the oropha ryngeal floor (Fig. 5).The rostral point of the frenu lum houses the body of the basihyale while the two lateral edges enclose the rostral parts of the paired ceratobranchiale which merge rostrally with the body of the basihyale (Fig. 6).Extending caudally from the body of the basihyale, along the midline, is the urohyale, also housed within the frenulum (Fig. 6) (see also Fig. 8).

lingual skeleton
The lingual skeleton consists of the paraglossum and the rostral projection of the basihyale (Fig. 7 and 8), both of which are imbedded in the tongue body.The paraglossum is a broad, thin, teardrop shaped cartilaginous plate imbedded within the lin gual parenchyma.The rostral tip is pointed while the base varies from gently rounded, to scalloped.
The paraglossum is situated dorsal to the rostral projection of the basihyale, to which it is attached by loose connective tissue.The basihyale runs almost the full length of the paraglossum, ending near its rostral tip.The edges of the paraglossum do not ex tend to the apex or lingual margins, nor into any of the lingual papillae.

dIScuSSIon
There is no definitive information in the literature on the topography of the emu tongue within the oro pharynx.The sketch by Faraggiana (1933) shows the tongue in relation only to the laryngeal mound whereas Bonga Tomlinson (2000) simply depicts the outline of the emu tongue body in relation to the hyobranchial apparatus and mandibular rami.From the specimens examined in the current study it was observed that the apex of the tongue did not extend further than half the distance from the commissure to the rostral bill tip.This contrasts with the position ing of the tongue body indicated by Bonga Tomlinson (2000), which shows it to occupy a far more rostral position relative to the surrounding structures.How ever, despite differences in the appearance of the various ratite tongues, the topographical relation ships of this organ in the emu are generally similar to those illustrated in the ostrich (Göppert 1903 The general shape of the tongue in birds usually mimics that of the bill (Bradley 1915;Harrison 1964;Koch 1973;Hodges 1974;Nickel et al. 1977) or the palate (McLelland 1979).However, in comparison to other bird families, the ratite tongue is greatly re duced in length relative to the bill (Faraggiana 1933;Ziswiler & Farner 1972;McLelland 1979 1).Tongue structure in birds is highly variable and closely related to feeding (McLelland 1979), with the ratite tongue being described as a rudimen tary or vestigial organ adapted for rapid swallowing of large food items (Gadow 1879;Pycraft 1900;McLelland 1979;Bonga Tomlinson 2000).Two spe cific adaptations of the avian tongue for swallowing have been recognized, namely, the occurrence of caudally directed lingual papillae (Harrison 1964;McLelland 1979;King & McLelland 1984) and/or a reduction in tongue size (McLelland 1979).The emu tongue body displays both of the above mentioned adaptations, as does that of the cassowary (P.John ston, personal communication 2008).Two reasons for tongue reduction in ratites can be advanced.In birds that swallow food whole (Harrison 1964;McLel land 1979) the tongue is unnecessary and therefore rudimentary (Harrison 1964;King & McLelland 1984) as well as nonprotrusable (King & McLelland 1984).
It is also suggested that because of the cranioiner tial feeding method employed by ratites, a longer tongue extending to the bill tip would be injured due to the rapid bill closure involved in this feeding tech nique (Bonga Tomlinson 2000).
There are surprisingly few accounts documenting the general appearance of the emu tongue, with both Fowler (1991) and Sales (2006Sales ( , 2007) ) simply quoting the observation of Cho et al. (1984) that "the tongue of the emu has a serrated edge".The fringed appearance of the emu tongue body is also illustrated by Bonga Tomlinson (2000).The most comprehensive description of the general shape of the emu tongue is that of Faraggiana (1933) who described the basic features noted in this study.How ever, as this author was limited to a single spec i men, some differences were apparent.In addition to the rounded apex described by Faraggiana (1933), pointed or split apices were observed in the present study, whereas the tongue body appeared broader than that depicted in the earlier work.
It is clear from previous studies that the shape of the tongue body differs between ratites (Cho et al. 1984).These differences in tongue shape are compared in Table 1 and indicate that the tongues of the emu and cassowary (P.Johnston, personal communica Short and/or blunt 3,4,6,8,13,14,17,18 Caudal "lingual pocket" Present study, 21 Personal observation tion 2008) share similar gross morphological fea tures.It should be noted, however, that it is not only tongue shape that differs between ratites.The ap pearance of the tongue body margins, tongue root, the prevalence of pigmentation, tongue size relative to the length of the bill, the occurrence of special features (for example, the lingual pocket in the os trich), and the shape and composition of the para glossum all define differences in ratite tongue struc ture and appearance (see Table 1).
It is also noteworthy that in birds with an omnivo rous diet the tongue conforms to a generalised pat tern described as triangular with a pointed apex, with the chief adaptive feature being that of caudally pointing spines (papillae) on the caudal margin (Gardner 1927).This statement would certainly be true for the emu, which also enjoys a varied diet (Davies 1978).
The lateral lingual papillae of the emu tongue show a lack of bilateral symmetry which involves differ ences in both number and shape, with a greater number of papillae usually being observed on the right margin.Faraggiana (1933) also noted that the number of papillae was not the same on each side of the tongue body whereas Bonga Tomlinson (2000) provides a definitive number of five lingual papillae on the lateral margins.In contrast, as noted in this study, the number of papillae displays a nor mal variation between specimens of 3-8 on the left and 5-8 on the right margins.
The caudal lingual papillae of the emu tongue are ru dimentary compared to other bird species and even though identifiable, are often not welldeveloped.The sketch by Bonga Tomlinson (2000) neglects to depict the caudal lingual papillae in this species.In comparison to the other ratites, the emu appears to be the only member which possesses structures recognisable as caudal lingual papillae (Table 1).However, in the ostrich and greater rhea (Table 1) the caudolateral aspect of the tongue body displays papillaelike extensions.Whether these structures represent true caudal lingual papillae remains un determined.
The function of the lingual papillae is reportedly to assist in the aboral transport of food (McLelland 1979;King & McLelland 1984).In the emu the lin gual papillae may be instrumental in removing smaller food particles from the roof of the orophar ynx in a similar fashion to that proposed by Bonga Tomlinson (2000) for palaeognathous birds (see below).
Some confusion exists in the literature regarding the naming of the caudal extremity of the tongue body (the tongue base) and the tongue root (Moore & Elliott 1946) with both terms being used inter changeably (McLelland 1975).In domestic poultry the tongue is clearly defined into a free rostral tip (apex), a body and a caudal root (McLelland 1993).Descriptions of the tongue using this terminology exist for a number of species (see, for example, Faraggiana 1933;Bailey et al. 1997;Jackowiak & Godynicki 2005;Jackowiak & Ludwig 2008).Based on the work of Lillie (1908) and Bradley (1915) it is generally accepted that the border between the tongue body and root is the row of caudal lingual papillae (Botezat 1910;Moore & Elliott 1946;Gentle 1971;Nickel et al. 1977;Bailey et al. 1997).This border coincides with the boundary between the oral and pharyngeal cavities as described for Anas by Zweers et al. 1977(cited by McLelland 1993).Some authors appear to use the term 'tongue base' synonymously with 'tongue root' (Nickel et al. 1977;Gussekloo & Bout 2005).In some studies the cau dal aspect of the tongue body has been termed the tongue base (Warner, McFarland & Wilson 1967;McLelland 1975;Bhattacharyya 1980;Bonga Tom linson 2000) or even the tongue root (Koch 1973;McLelland 1979;McLelland 1990;Kobayashi et al. 1998) whereas in other publications the term tongue base is used but not defined (Bacha & Bacha 2000;Calhoun 1954).Alternative terminology used for the tongue root includes the posterior part of the tongue (Gentle 1971), the sensory area (Bhattacharyya 1980) and the preglottal part of the tongue (Hom berger & Meyers 1989;Liman, Bayram & Koçak 2001).
The importance of clarity in correctly identifying and naming the various components of the tongue has been pointed out by Moore & Elliott (1946), particu larly in regard to the location of taste buds.Failure to recognise the caudal aspect of the tongue (the tongue root) as part of the tongue could lead to invalid conclusions about the presence of taste buds in this organ, as they are reportedly concentrated in this region (Moore & Elliott 1946;Gentle 1971;Nick el et al. 1977;Bacha & Bacha 2000;AlMansour & Jarrar 2004).
A clearly defined triangular structure represents the tongue root in the emu and is positioned between the caudal margin of the tongue body and the laryn geal entrance.This structure seems to be unique to the emu as in other ratites the tongue root is repre sented by a featureless stretch of mucosa (Table 1).
The structure of the tongue root in kiwi species is unclear (McCann 1973).The extension of the tongue root into the rostral aspect of the laryngeal entrance (Faraggiana 1933; present study) represents an in teresting modification not observed or illustrated in other ratites (ostrich and greater rhea) (Göppert 1903;Faraggiana 1933;Gussekloo & Bout 2005;Por chescu 2007;Jackowiak & Ludwig 2008;Tivane 2008).The positioning of the tongue root would also appear to assist in sealing the rostral part of the larynx when the glottis is closed, almost assuming the role of an epiglottis, which is not present in birds (Kaupp 1918;Calhoun 1954;King & McLelland 1984;Nickel et al. 1977).This argument regarding the role of the tongue root functioning as an epiglot tis in the emu has been proposed by Gadow (1879) but disputed by Faraggiana (1933).The tongue root of the emu also appears to play a special role in as sisting to close off of the rostral aspect of the choa na in the closed gape.The choana of most birds is divided into a rostral slitlike part (pars rostralis) and a caudal triangular part (pars caudalis) (King 1993) with the tongue commonly closing off the rostral part of the choana (McLelland 1975(McLelland , 1979).In the emu, the triangular choana (Fig. 1) is not divided into ros tral and caudal parts and therefore the tongue body plays no part in closing off the choana in the closed gape.Instead, the tongue root partially closes off the rostral aspect of the choana in this species.
Little mention is made in the literature of the frenu lum in birds.A possible reason for this may be its general lack of remarkable features, serving simply to attach the tongue to the oropharyngeal floor (McLelland 1979).In the emu, the frenulum is a rel atively large structure which houses part of the hyo branchial apparatus.The lateral margins are longi tudinally folded which would seem to indicate that the tongue is capable of a certain degree of move ment.This observation lends further support to the role played by the tongue of palaeognaths in cranio inertial feeding and in drinking.During swallowing in palaeognaths the tongue is lifted and contacts the palate before moving caudally, thereby scraping any food caudal to the tongue into the proximal oesophagus (Bonga Tomlinson 2000).Palaeognaths transport food from their bill tips to the oesophageal entrance via the cranioinertial feeding method (Bonga Tomlinson 2000), also described as the 'catch and throw' method by Gussekloo & Bout (2005).The transport of food into or close to the oesophageal entrance is facilitated by a large gape and marked depression of the tongue.Tongue de pression enlarges the 'buccal cavity' (oropharyn geal cavity), which assists in moving food to the caudal oropharynx, while retraction of the tongue assists in the final transport of fluid to the oesoph a gus during drinking (Gussekloo & Bout 2005).There fore, despite the emu tongue showing such relatively reduced dimensions and rigidity, it possess a surpris ingly large range of movements in both the rostro caudal (though unable to protrude) and dorsoventral planes by virtue of the relatively large, folded frenu lum and the association of the hyobranchial appara tus with the tongue body and frenulum.
The lingual skeleton of the emu is formed by the me dian, unpaired paraglossum and the rostral projec tion of the basihyale of the hyobranchial apparatus.
The paraglossum is related dorsally to the rostral projection of the basihyale as also described by Bonga Tomlinson (2000) in the emu and the greater rhea.However, the findings of this study contrasted with those of Bonga Tomlinson (2000) in that the rostral projection of the basihyale extended further rostrally, ventral to the paraglossum, than that de picted by the author.
The paraglossum of the emu was teardropshaped with a pointed rostral tip and a rounded base al though it is depicted by Parker (1866) in Dromaius irroratus as inverted heartshaped and by Bonga Tomlinson (2000) in Dromaius novaehollandiae as arrowheadshaped.In ratites the paraglossum re mains cartilaginous and does not ossify in older birds (Bonga Tomlinson 2000), a situation also ap parent in the emu.The shape of the paraglossum differs be tween the ratites.The paraglossum of the emu (Dromaius irroratus and novaehollandiae), rhea (Rhea americana) and cassowary (Casuarius ben netii) are all basically arrowheadshaped, although individual differences are apparent, particularly re garding the form of the base (Parker 1866;Bonga Tomlinson 2000; present study).The paraglossum of the kiwi (Apteryx australis) (Parker 1891) is also a single structure but is much narrower than that of the emu, rhea and cassowary and has a split, elon gated base.The ostrich paraglossum is divided into two narrow paraglossalia which flank the rostral pro jection of the basihyale and are located ventrola teral to it (Bonga Tomlinson 2000;Tivane 2008).This arrangement differs radically from that of the emu, where the rostral projection of the basihyale lies ventral to the paraglossum, and the other ratites and has lead to some authors not recognising or misinterpreting the narrow, paired structure (Meckel 1829;Parker 1866;Webb 1957;Jackowiak & Lud wig 2008) present in the ostrich tongue.
The tongue of birds is a rigid organ due to the pres ence of the paraglossum (Koch 1973) and, except in parrots, the absence of intrinsic musculature (Zis wiler & Farner 1972;Koch 1973;Nickel et al. 1977;McLelland 1990).The rigidity afforded by the para glossum in palaeognathous birds is needed for the swallowing phase in order to push the food into the oesophagus.The rostral projection and body of the basihyale, situated ventrally in the tongue body, con nects the hyobranchial apparatus with the tongue, and due to its close association, retracts the tongue during swallowing.The great mobility of the hyobran chial apparatus in birds, attributed to the fact that it does not articulate with the skull (McLeod 1939), is the main contributor to the movement of the tongue (King & McLelland 1984;Bonga Tomlinson 2000).

AcKnoWledGeMentS
The FIG. 2 Dorsal view of the tongue body (Tb) showing the apex (A), lateral lingual papillae (*) and caudal lingual papillae (Cp).Tongue body length (L) was measured from the apex to the caudal papillae.The width (W) was measured between the tips of the last lateral papillae.Bar = 5 mm FIG. 4 Dorsal view of the triangular tongue root, showing the caudal extremity of the tongue root (*) folding over the laryngeal entrance (Le).In this specimen, the caudal lin gual papillae (arrows) of the tongue body (Tb) appear fused with variable incisures and small projections being apparent.The rostral parts of the paired ceratobranchi ale (Cb) are seen bordering the tongue root.Note the pitted surface of the tongue body, representing the open ings of the large underlying glands.Bar = 1 mm FIG. 5 The dorsoventrally flattened tongue body (Tb) shown in lateral profile.The folds of the frenulum (Fr) are not visi ble as the tongue body is in the raised position.Dorsum (D), ventrum (V), tongue root tip (arrows), laryngeal fis sure (Lf), choana (C).Bar = 5 mm

TABLE 1
authors thank Mr Peter Duncan for providing the emu heads; Dr Catarina Tivane for collection of the specimens; Mrs Charmaine Vermeulen for the photography; Ms Marté Smit for preparing the fig ures; the support staff of the Department of Anatomy and Physiology, Faculty of Veterinary Science, Uni versity of Pretoria; Carole Long, Secretary of the Otanewainuku Kiwi Trust, Avi Holzapfel (leader of the Kiwi Recovery Group) and Susan Cunningham from New Zealand for their assistance in acquiring literature on the kiwi; Peter Johnston from the Live Transplant Unit and Department of Anatomy, Uni versity of Auckland, New Zealand for his information on the cassowary; and the University of Pretoria for financial support.