Edgeworth (1935) and Huber (1930a,b, 1931) divided the hyoid muscles into two main groups: dorso-medial and ventral (Table 2). The plesiomorphic configuration for sarcopterygians is a single ventral hyoid muscle, the interhyoideus, and two dorso-medial hyoid muscles, the adductor arcus palatini and the adductor operculi (note that the ‘adductor hyomandibulae Y’ and ‘levator operculi’ of Latimeria are not homologues of the adductor hyomandibulae and levator operculi of actinopterygians such as teleosts: Diogo, 2007, 2008) (Table 2). The depressor mandibulae, levator hyoidei, branchiohyoideus and cervicomandibularis of extant dipnoans, amphibians and reptiles seem to develop from the anlage that gives rise to the adductor arcus palatini in other osteichthyans (Diogo, 2007, 2008) (Table 2 and Figs 3, 5, 6). The adductor operculi is not present as an independent muscle in extant dipnoans, amphibians and reptiles; at least in dipnoans it is most likely fused with the ventral hyoid muscle interhyoideus (Diogo, 2007, 2008) (Table 2 and Fig. 3). The number of hyoid muscles found in extant mammals, and particularly in therians (placentals + marsupials), is much greater than that found in extant non-mammalian tetrapods (Table 2). Also, in non-mammalian vertebrates the hyoid muscles are mainly restricted to the region of the second branchial arch and occasionally to the mandibular and/or neck regions (Figs 3, 5, 6), whereas in extant mammals these muscles extend more anteriorly, covering much of the anterior region of the head (Figs 8, 12, 14). With the exception of styloideus, stylohyoideus, digastricus posterior, jugulohyoideus and stapedius, all the mammalian hyoid muscles listed in Table 2 are usually designated as facial muscles because they attach to freely movable skin and are associated with the display of facial expressions (e.g. Ruge, 1885, 1897, 1910; Boas & Paulli, 1908; Lightoller, 1928a,b, 1934, 1940a,b, 1942; Huber, 1930a,b, 1931; Edgeworth, 1935; Andrew, 1963; Gasser, 1967; Jouffroy & Saban, 1971; Saban, 1971; Seiler, 1971a,b,c,d,e, 1974a,b, 1975, 1979, 1980; Minkoff et al. 1979; Preuschoft, 2000; Schmidt & Cohn, 2001; Burrows & Smith, 2003; Burrows et al. 2006; Burrows, 2008). Some researchers have suggested that the mammalian facial muscles derive exclusively from the interhyoideus of non-mammalian tetrapods (e.g. Huber, 1930a,b, 1931), but our dissections and comparisons support authors such as Lightoller (1942) and Jouffroy & Saban (1971), who claim that at least some of these muscles (e.g. platysma cervicale, platysma myoides, mandibulo-auricularis) correspond to part of the dorso-medial hyoid musculature (e.g. cervicomandibularis) of other tetrapods (Table 2 and Figs 6, 8, 12, 14). The evolution and homologies of the mammalian facial muscles have been, and continue to be, controversial. In light of the overall analysis of the data obtained by our dissections and comparisons and by a review of the literature, it can be said that some of the hypotheses proposed and shown in Table 2 (black arrows) are in fact well supported by the data that are now available. For instance, the data available on topology, functional morphology, development and innervation strongly suggest that the platysma cervicale, platysma myoides, occipitalis, auricularis posterior and some of the extrinsic muscles of the ear (e.g. antitragicus, helicis and/or transversus and obliquus auriculae) of mammals have a common phylogenetic and ontogenetic origin (e.g. Boas & Paulli, 1908; Huber, 1930a,b, 1931; Gasser, 1967; Jouffroy & Saban, 1971; Saban, 1971; this work) (Figs 8, 12, 14 and Table 2). These same lines of evidence also suggest that the interhyoideus profundus, sphincter colli superficialis, sphincter colli profundus, naso-labialis, levator labii superioris, levator labii superioris alaeque nasi, buccinatorius, dilatator nasi, maxillo-naso-labialis, nasalis, depressor septi nasi, levator anguli oris facialis, orbicularis oris, depressor labii inferioris, depressor anguli oris and/or mentalis of mammals derive from the interhyoideus (Table 2; see also e.g. Gasser, 1967; Jouffroy & Saban, 1971; Saban, 1971; Seiler, 1971a,b,c,d,e, 1974a,b, 1975, 1979, 1980). However, it is still not clear, for instance, if the therian mandibulo-auricularis (a muscle that is usually deep to all the other mammalian facial muscles) is phylogenetically more closely related to the other facial muscles than to deeper dorso-median muscles such as the stylohyoideus, digastricus posterior, jugulohyoideus and stapedius (e.g. Lightoller, 1934; Jouffroy & Saban, 1971; Seiler, 19711971a,b,c,d,e, 1974a,b, 1975, 1979, 1980; this work) (Table 2 and Fig. 12). Also, it is commonly accepted that muscles such as the zygomaticus major, zygomaticus minor, orbito-temporo-auricularis, frontalis, auriculo-orbitalis, temporoparietalis, auricularis anterior and auricularis superior derive from the sphincter colli profundus and/or superficialis, but Seiler (1971a,b,c,d,e, 1974a,b, 1975, 1979, 1980), based on his comparative and developmental studies, argues that at least some of these muscles may derive from the platysma cervicale and/or myoides (Table 2). However, some of Seiler's methods and interpretations are questionable. For example, in his 1980 developmental study of primates and tree-shrews, he argues that the facial muscles that are more superficial in early developmental stages are necessarily part of a ‘platysma anlage’ and thus derived phylogenetically from an ‘ancestral platysma’, whereas the majority of the other facial muscles are part of a ‘sphincter colli profundus’ anlage and thus are derived phylogenetically from a ‘primitive sphincter colli profundus’. This contrasts with Gasser's (1967) study of the ontogeny of the facial muscles of modern humans, in which various other anlages are recognized in early developmental stages. Also, it should be stressed that in adult mammals, including monotremes, at least some portions of the platysma (cervicale and/or myoides) lie deep to facial muscles such as the sphincter colli superficialis and even to facial muscles that Seiler categorizes as ‘sphincter colli profundus derivatives’ (e.g. part of the orbicularis oris and/or levator labii superioris) (e.g. Boas & Paulli, 1908; Lightoller, 1928a,b, 1934, 1940a,b, 1942; Huber, 1930a,b, 1931; Andrew, 1963; Jouffroy & Saban, 1971; Saban, 1971; Minkoff et al. 1979; Preuschoft, 2000; Schmidt & Cohn, 2001; Burrows & Smith, 2003; Burrows et al. 2006; this work) (Table 2 and Fig. 8). The majority of researchers consider that the sphincter colli of mammals derives from the interhyoideus of other tetrapods, so it is likely that the mammalian sphincter colli was plesiomorphically mainly superficial, and not deep, to the other hyoid muscles (the interhyoideus of other tetrapods is usually superficial not only to the other hyoid muscles, but to all the other muscles of the head). Monotremes are plesiomorphic mammals, and both the platypus and the echidna have a well-developed, broad sphincter colli superficialis that is superficial to most of the other facial muscles (the platypus actually lacks a sphincter colli profundus, although it has an interhyoideus profundus that seems to be derived from the deeper part of the interhyoideus; in the echidna most of the sphincter colli is superficial to the other facial muscles, but part of it passes deep to these muscles, forming a sphincter colli profundus: e.g. Huber, 1930a; Lightoller, 1942; Jouffroy & Saban, 1971; this work) (Table 2 and Fig. 8). A more detailed comparative analysis of the development and innervation of the hyoid group of muscles in vertebrates, including various key mammalian groups such as monotremes, is planned to clarify these and other controversial issues regarding the origin, homologies and evolution of the mammalian facial muscles and to test the hypotheses proposed in Table 2.