Evolution and Homologies of Primate and Modern Human Hand and Forearm Muscles, with Notes on Thumb Movements and Tool Use

Home , Common flexor tendon, Contrahens, Flexor (fish), Gorillini, Homininae

Journal of Human Evolution 63 (2012) 64e78

Contents lists available at SciVerse ScienceDirect

journal homepage: www.elsevier.com/locate/jhevol

Evolution and homologies of primate and modern human hand and forearm muscles, with notes on thumb movements and tool use

Rui Diogo a,*, Brian G. Richmond b,c, Bernard Wood b,c a Department of Anatomy, Howard University College of Medicine, 520 W St. NW, Washington, DC 20059, USA b Center for the Advanced Study of Hominid Paleobiology, Department of Anthropology, George Washington University, DC 20052, USA c Human Origins Program, National Museum of Natural History, Smithsonian Institution, Washington, DC 20560, USA article info abstract

Article history: In this paper, we explore how the results of a primate-wide higher-level phylogenetic analysis of muscle Received 21 January 2012 characters can improve our understanding of the evolution and homologies of the forearm and hand Accepted 5 April 2012 muscles of modern humans. Contrary to what is often suggested in the literature, none of the forearm Available online 27 May 2012 and hand muscle structures usually present in modern humans are autapomorphic. All are found in one or more extant non-human primate taxa. What is unique is the particular combination of muscles. Keywords: However, more muscles go to the thumb in modern humans than in almost all other primates, rein- Phylogenetic analysis forcing the hypothesis that focal thumb movements probably played an important role in human Morphology Extensor pollicis brevis evolution. What makes the modern human thumb myology special within the primate clade is not so fl Flexor pollicis longus much its intrinsic musculature but two extrinsic muscles, extensor pollicis brevis and exor pollicis Adductor pollicis accessorius longus, that are otherwise only found in hylobatids. It is likely that these two forearm muscles play different functional roles in hylobatids and modern humans. In the former, the thumb is separated from elongated digits by a deep cleft and there is no pulp-to-pulp opposition, whereas modern humans exhibit powerful thumb flexion and greater manipulative abilities, such as those involved in the manufacture and use of tools. The functional and evolutionary significance of a third peculiar structure, the intrinsic hand structure that is often called the ‘interosseous volaris primus of Henle’ (and which we suggest is referred to as the musculus adductor pollicis accessorius) is still obscure. The presence of distinct contrahentes digitorum and intermetacarpales in adult chimpanzees is likely the result of prolonged or delayed development of the hand musculature of these apes. In relation to these structures, extant chimpanzees are more neotenic than modern humans. Ó 2012 Elsevier Ltd. All rights reserved.

Introduction (1932) and Straus (1941a, b), usually investigated a wide range of non-primate species but only a few primate taxa (e.g., Brooks, 1886; An understanding of comparative myology is crucial for devel- Parsons, 1898; McMurrich, 1903a, b; Forster, 1917; Howell, 1936a, b; oping hypotheses about the functional morphology of the modern Haines, 1939, 1946, 1950; Straus, 1942a, b). This pattern contrasts human forearm and hand, and particularly their involvement in the with most of the reports published since the 1950s that have manufacture and use of tools (e.g., Day and Napier, 1961, 1963; mainly focused on primates, or even more narrowly, on hominoids Napier, 1962; Tuttle, 1969; Lewis, 1989; Marzke, 1992, 1997; or on modern humans (Hominina: see Fig. 1), and when they did Susman, 1994, 1998; Marzke et al., 1998; Susman et al., 1999; provide a non-primate comparative context it was a relatively Tocheri et al., 2008). The relatively few publications that have narrow one (e.g., Abramowitz, 1955; Day and Napier, 1963; included myological information in discussions of the evolution of Dylevsky, 1967; Tuttle, 1969; Dunlap et al., 1985; Aziz and Dunlap, the primate and modern human forearm and hand can be divided 1986; Susman, 1994, 1998; Marzke et al., 1998; Susman et al., into two groups. Publications prior to the 1950s were mainly the 1999). A notable exception is Lewis' (1989) book that includes work of comparative vertebrate, tetrapod or mammalian anato- detailed information based on his own dissections of a wide range mists who, with the exception of authors such as Howell and Straus of non-primate tetrapods. These studies have generated hypotheses regarding the evolution of primate forearm and hand anatomy, among them that modern humans are derived relative to other extant primates in * Corresponding author. ﬂ ﬂ E-mail addresses: [email protected], [email protected] (R. Diogo). possessing a true exor pollicis longus, a deep head of exor pollicis

LEMURIFORMES nomenclature of the forearm and hand muscles of primate and Lemur (Lemuridae) STREPSIRRHINI non-primate tetrapod taxa has been the subject of some confusion Propithecus (Indriidae) and controversy (see Diogo, 2007, 2009; Diogo and Abdala, 2007, Loris (Lorisidae) 2010; Diogo et al., 2009; Diogo and Wood, 2011, 2012). For example, Nycticebus (Lorisidae) statements that some non-hominoid primates have a separate 1,2 LORISIFORMES flexor pollicis longus (e.g., Marzke, 1997; Shrewsbury et al., 2003) Tarsius (Tarsiiformes; Tarsiidae) refer to older publications that often used the name ‘flexor pollicis ’ fl PRIMATES Pithecia (Pitheciidae) longus for the tendon of the exor digitorum profundus that fl Aotus (Aotidae) inserts onto digit 1 and not for a separate exor pollicis longus that has a distinct belly and tendon (see below). 3,4,5 Callithrix (Callitrichinae) Saimiri (Saimiriinae) In recent papers (Diogo and Abdala, 2007; Diogo et al., 2009; CEBIDAE Abdala and Diogo, 2010; Diogo and Wood, 2011) and monographs HAPLORRHINI CEBIDAE+AOTIDAE (Diogo, 2007; Diogo and Abdala, 2010; Diogo and Wood, 2012), PLATYRRHINI Diogo and colleagues have reported the results of their long-term Colobus (Colobinae) dissection-based study of the comparative anatomy, homologies Cercopithecus (Cercopithecini) and evolution of the pectoral and forelimb muscles of all major 6 Macaca (Macanina) groups of non-primate vertebrates and representatives of all of the Papio (Papionina) major primate higher taxa. Diogo and Wood (2011, 2012) combined ANTHROPOIDEA PAPIONINI data from these dissections with carefully validated information 7,8,9 CERCOPITHECINAE from the literature to undertake the first comprehensive parsimony CERCOPITHECIDAE and Bayesian cladistic analyses of the order Primates based on CATARRHINI Hylobates (Hylobatidae) 10,11, myological data (Fig. 1) and for a range of outgroups (tree-shrews, 12,13 14,15, Pongo (Ponginae) 16,17 18,19, dermopterans and rodents). A goal of the project was to establish HOMINOIDEA Gorilla (Gorillini) 20 the homologies of the pectoral and forelimb muscles of vertebrates HOMINIDAE Pan (Panina) and to provide the comparative context for more detailed evolu- HOMININAE Homo (Hominina) 21,22,23, tionary and taxon-based analyses. HOMININI 24,25 The present paper uses data from these dissections to test Figure 1. Single most parsimonious primate tree (L 301, CI 58, RI 73) obtained from the hypotheses about the evolution of the hand musculature of modern cladistic analysis of 166 characters of the head, neck, pectoral and forelimb muscula- humans. Because we have dissected members of all of the major ture (Diogo and Wood, 2011, 2012), showing the unambiguous transitions (homo- primate groups and the major non-primate vertebrate taxa, we can plasic: regular font; non-homoplasic: bold) of the forearm and hand musculature that better understand the descriptions of authors that use different occurred in the branches leading to modern humans: 1) Opponens pollicis is a distinct muscle (this muscle became secondarily undifferentiated in Callithrix); 2) Opponens nomenclatures for these muscles. This has allowed us to clarify digiti minimi is a distinct muscle (also found in a few other mammals, e.g., Rattus); 3) much of the nomenclatural confusion involved in the descriptions Flexor carpi radialis inserts onto the metacarpals II and III (also found in a few other of the forearm and hand muscles of primates (see above). We also mammals, e.g., Tupaia); 4) There are more than two contrahentes digitorum (reverted use these data to critically examine the proposition that modern in some haplorrhines, e.g., modern humans); 5) Adductor pollicis has slightly differ- fl entiated transverse and oblique heads; 6) Supinator has ulnar and humeral heads humans are derived in having a exor pollicis longus, deep head of (independently acquired in Lorisiformes); 7) Adductor pollicis further differentiated flexor pollicis brevis, a ‘volar interosseous’ or ‘interosseous volaris into distinct transverse and oblique heads; 8) Opponens pollicis reaches the distal primus’ of Henle, and an extensor pollicis brevis. portion of metacarpal I (feature independently acquired in Lorisiformes); 9) Opponens digiti minimi is slightly differentiated into superficial and deep bundles; 10) Flexor Materials and methods digitorum superficialis originates from the radius; 11) Flexor digitorum superficialis originates from the ulna; 12) Epitrochleoanconeus is not a distinct muscle (this muscle became undifferentiated in Lorisiformes and in hominoids, and then became We dissected representatives of each major extant non- secondarily differentiated in Pan); 13) Pronator teres often (but not usually, i.e., in hominoid primate clade (Strepsirrhini, Tarsiiformes, New World < 50% of the cases) originates from the ulna; 14) Flexor digitorum profundus is not monkeys and Old World monkeys) and of each of the five main originated from the medial epicondyle of the humerus or from the common flexor tendon (feature also found in Macaca); 15) Tendon of flexor digitorum profundus to groups of living hominoids (i.e., hylobatids, orangutans, gorillas, digit 1 is vestigial or absent (this feature was independently acquired in Colobus and it chimpanzees, and modern humans) (Fig. 1). With a few exceptions, was secondarily reverted - i.e., the tendon is not atrophied - in modern humans); 16) we use the same taxonomic nomenclature as Diogo and Wood Pronator teres is usually (50% of the cases) originated from the ulna; 17) Contra- (2011). Data included in the analysis come from four strepsirrhine hentes digitorum are missing (this feature was secondarily reverted in Pan, i.e., alone genera, two from the infraorder Lemuriformes (Lemur, family among the Hominidae, Pan has distinct contrahentes digitorum); 18) Palmaris longus is absent in >5% of the cases; 19) Adductor pollicis accessorius (or ‘interosseous volaris Lemuridae; Propithecus, family Indriidae) and two from the infra- primus of Henle’ of modern human anatomy) is often (i.e., in <50% of the cases) order Lorisiformes (Loris and Nycticebus, family Lorisidae); the present; 20) Extensor indicis usually inserts onto digit 2 only; 21) Flexor pollicis longus single extant genus of the infraorder Tarsiiformes, Tarsius; repre- ‘ is a distinct muscle (independently acquired in hylobatids); 22) Reversion of Tendon of sentatives of three of the four extant platyrrhine families: Saimiri flexor digitorum profundus to digit 1 is vestigial or absent’ (see above); 23) Reversion of ‘Flexor carpi radialis originates from the radius’ (the muscle originates from the and Callithrix (Cebidae, subfamilies Saimiriinae and Callithrichinae, radius in gorillas, chimpanzees and orangutans, but it is not clear which is the usual respectively), Pithecia (Pitheciidae), and Aotus (Aotidae; the other condition for hylobatids, and, thus, whether a radial origin constitutes a synapomorphy platyrrhine family being the Atelidae); the two extant subfamilies of hominoids or of hominids); 24) Adductor pollicis accessorius (or ‘interosseous of Old World monkeys (family Cercopithecidae) represented by ’ > volaris primus of Henle of modern human anatomy) is usually (i.e., in 50% of the Colobus (Colobinae), Papio, Macaca and Cercopithecus (Cercopithe- cases) present; 25) Extensor pollicis brevis is a distinct muscle (independently acquired in hylobatids) [for more details, see text]. cinae; the two former genera represent the tribe Papionini, while the latter genus represents the other extant tribe of the subfamily, the Cercopithecini); and five extant hominoid genera: Hylobates brevis, a volar interosseous of Henle, and an extensor pollicis brevis. (Hylobatidae), Pongo (Hominidae, Ponginae), Gorilla (Hominidae, However, with a few noteworthy exceptions, the authors of most of Homininae, Gorillini), Pan (Hominidae, Homininae, Panini, Panina), these publications relied on reports in the literature for their and Homo (Hominidae, Homininae, Hominini, Hominina) (Fig. 1). comparative information. This is problematic because the We use the traditional classification that recognizes a single extant 66 R. Diogo et al. / Journal of Human Evolution 63 (2012) 64e78 hylobatid genus (Hylobates; including species such as Hylobates silenus: VU MS1, 1 (fresh; adult male). Nycticebus coucang: SDZ syndactylus, Hylobates lar, Hylobates gabriellae and Hylobates hoo- NC41235, 1 (fresh; adult female); SDZ NC43129, 1 (fresh; adult lock, among others: for more details see Diogo and Wood, 2011). female). Nycticebus pygmaeus: VU NP1, 1 (fresh; adult female); VU Apart from the primates dissected for this study, we have dissected NP2, 1 (fresh; adult male); SDZ NP40684, 1 (fresh; adult female); specimens from all of the major groups of vertebrates. A list of the SDZ NP51791, 1 (fresh; adult female). Pan troglodytes:PFA1016,1 dissected non-primate vertebrate specimens is given in Diogo and (fresh; adult female); PFA 1009, 1 (fresh; adult female); PFA 1051, 1 Abdala (2010). (fresh; infant female); PFA 1077, 1 (fresh; infant female); PFA UNC The nomenclature for the forearm and hand muscles follows (uncatalogued), 1 (fresh; infant male); HU PT1, 1 (formalin; infant that of Diogo and Abdala (2010) (see Table 1). We addressed the male); GWUANT PT1, 1 (formalin; adult female); GWUANT PT2, 1 problem of inconsistent usage of some specific taxonomic names in (formalin; adult female); VU PT1, 1 (fresh; adult male). Papio anubis older systematic and anatomical studies, particularly those prior to GWUANT PA1, 1 (fresh; adult female). Pithecia pithecia: VU PP1, 1 Osman Hill’s studies in the 1950s (e.g., Hill, 1953, 1955, 1957, 1959, (fresh; adult male); GWUANT PP1, 1 (fresh; adult female). Pongo 1960, 1962, 1966, 1970, 1974; see References) by carefully reviewing pygmaeus: HU PP1, 1 (formalin; neonate male); GWUANT PP1, 1 all of the names and synonyms used in those studies (see Diogo and (formalin; adult male). Propithecus verreauxi: GWUANT PV1, 1 Wood, 2011). Discussions of the evolutionary changes occurring in (fresh; adult female); GWUANT PV2, 1 (fresh; infant female). Saimiri each of the major primate branches are based on the phylogenetic sciureus: GWUANT SC1, 1 (fresh; adult female). Tarsius syrichta: results of Diogo and Wood (2011) (Fig. 1). The primate specimens CMNH M-3135, 1 (alcohol; adult female). were mainly dissected by RD, and were obtained from the following institutions: the Primate Foundation of Arizona (PFA), the Depart- Results ment of Anatomy (GWUANA) and the Department of Anthropology (GWUANT) of the George Washington University, the Department For each group of muscles (ventral forearm muscles, dorsal of Anatomy of Howard University (HU), the Department of forearm muscles and intrinsic hand muscles), we discuss the results Anatomy of Valladolid University (VU), the Cincinnati Museum of working from the more inclusive synapomorphic features (e.g., Natural History (CMNH), the San Diego Zoo (SDZ) and the Canadian Primates, Haplorhini, Anthropoidea, etc.) to those apomorphies Museum of Nature (CMN). For each taxon, we provide the Linnean that are only found within H. sapiens (Fig. 1). Detailed tables that binomial, its source, its unique identifier, the number of specimens describe and use photographs to illustrate each muscle of each of from that source and the state of the specimens. Regarding the the taxa included in the cladogram of Fig. 1 are included in Diogo sample size used in the cladistic study of Diogo and Wood (2011) and Wood (2012). Detailed descriptions of the phylogenetic char- and in the discussions of the present paper, two points should be acters used and of the synapomorphies obtained in our phyloge- stressed. First, it is difficult to find primate, and particularly ape, netic analyses are given there and in Diogo and Wood (2011). specimens in circumstances where careful dissection can take place. During this project, we made a considerable effort to estab- Ventral forearm musculature lish connections with the major museums and zoos in the United States and beyond. This effort resulted in us being able to dissect, Among the synapomorphic features of the ventral forearm for example, two fresh gorillas, one fresh and one formalin muscles, the most inclusive is the insertion of the flexor carpi embalmed Pongo, and six fresh and three formalin embalmed radialis onto metacarpals II and III (e.g., Fig. 2), which was acquired chimpanzees. The second point is that the sample size used in that in the Haplorhini (Fig. 1, feature 3). In most non-primate eutherian cladistic study and in the discussions provided here refers to the mammals, the flexor carpi radialis inserts onto metacarpal III (e.g., specimens dissected by us and the total number of specimens in rats), onto metacarpal II (as is usually the case in e.g., Lemur and reported in the numerous publications that were analysed in the Propithecus), or, in a few cases (e.g., in flying lemurs), onto other extensive review of the literature undertaken by Diogo and Wood structures, but it does not attach onto both metacarpals II and III, as (2011) and particularly Diogo and Wood (2012). Thus, when we it usually does in haplorrhines (and in a few non-primate code and discuss each character we take into account all of the mammals, e.g., Tupaia). Two of the synapomorphies of hominoids information available, and in numerous cases the total sample size concern the flexor digitorum superficialis. In non-hominoid is substantial when compared with cladistic studies of other animal primates, this muscle originates exclusively from the arm (medial species that were based on muscles (see, e.g., Diogo, 2007). For epicondyle of humerus, common flexor tendon, and/or capsule of example, for character. 118 (the presence/absence of the palmaris the elbow joint), but in hominoids the muscle usually also origi- longus) Diogo and Wood (2011) took into account information nates from the ulna and the radius (Fig. 1, features 10, 11). Another obtained from dissections of more than 20 hylobatid, 19 orangutan, synapomorphy of hominoids is the loss of the epitrochleoanconeus, 25 gorilla, and 39 chimpanzee specimens. For this character, the a small muscle that connects the medial epicondyle of the humerus sample size just for apes was >103 specimens. to the olecranon process of the ulna. This muscle is usually present (presumably due to a secondary reversion) in chimpanzees (Fig. 1, Primate specimens dissected feature 12). Plesiomorphically in primates the pronator teres has a bony origin from the humerus only. In hylobatids, this muscle Aotus nancymaae: GWUANT AN1, 1 (fresh; adult female). Calli- originates from the humerus and often, but not usually (i.e., <50% thrix jacchus: GWUANT CJ1, 1 (fresh; adult male). Cercopithecus of the cases), from the ulna (Fig. 1, feature 13). In hominids, the diana: GWUANT CD1, 1 (fresh; adult female). Colobus guereza: muscle originates from the humerus and usually (i.e., 50% of the GWUANT CG1, 1 (fresh; adult male). Gorilla gorilla: CMS GG1, 1 cases) also from the ulna (Fig. 1, feature 16). (fresh; adult male); VU GG1, 1 (fresh; adult female). Homo sapiens: Plesiomorphically, in primates the flexor digitorum profundus GWUANA HS1-16, 16 (formalin); HU D43, 1 (formalin); HU D45 originates from the forearm (ulna, radius, and/or interosseous (formalin). Hylobates gabriellae: VU HG1, 1 (fresh; infant male). membrane) and arm (medial epicondyle and/or common flexor Hylobates lar: HU HL1, 1 (formalin; adult male). Lemur catta: tendon). In hominids (and also in Macaca as a homoplasy), this GWUANT LC1, 1 (fresh; adult male). Loris tardigradus: SDZ LT53090, muscle does not originate from the medial epicondyle nor from the 1 (fresh; adult male). Macaca fascicularis: VU MF1, 1 (fresh; adult common flexor tendon (Fig. 1, feature 14). In our dissections, the male). Macaca mulatta: HU MM1, 1 (formalin; adult male). Macaca flexor pollicis longus was present in modern humans and R. Diogo et al. / Journal of Human Evolution 63 (2012) 64e78 67

Table 1 Scheme illustrating the authors’ hypotheses regarding the homologies of the arm and hand muscles of adults of representative primate taxa; the ﬂexor brevis profundus 2 (which corresponds to the ‘deep head of the ﬂexor pollicis brevis’ of human anatomy) is listed here (and counted) as a distinct muscle, following the works done on numerous other mammals. Muscle names Lemur Tarsius Aotus Macaca Hylobates Pongo Gorilla Pan Homo

(49 mus.: 19 (51-55 mus.: 19 (41 mus.: 19 (46 mus.: 19 forearm; (46 mus.: 19 (38 mus.: 18 (38 mus.: 18 (45 mus.: 19 forearm; (41 mus.: 20

forearm; 30 hand) forearm; 32-36 hand) forearm; 22 hand) 27 hand) forearm; 27 hand) forearm; 20 hand) forearm; 20 hand) 26 hand) forearm; 21 hand)

Pronator quadratus P P P P P P P P P

Flexor dig. profundus P P P P P P P P P

Flexor pollicis longus ------P ------P

Flexor dig. superficialis P P P P P P P P P

Palmaris longus P P P P P P P P P

Flexor carpi ulnaris P P P P P P P P P

APPEND.: VEN. FOREARM Epitrochleoanconeus P P P P ------P ---

Flexor carpi radialis P P P P P P P P P

Pronator teres P P P P P P P P P

Palmaris brevis P P P P P P P P P

Lumbricales 1-4 P P P P P P P P P

Contrahentes dig. P (2mus.) P (8mus.) P (3mus.) P (3mus.) P (3mus.) ------P (2mus.) ---

Adductor pollicis P P P P P P P P P

Adductor pollicis access. ------P

Flexor brevis prof. 2 P P --- P P P P P P

Fbp. / Int. pal. Fbp. 3-9 (7 mus.) Fbp. 3-9 (7 mus.) Int. pal.1-3 (fbp 4,7,9) Fbp.3-9 (7 mus.) Int. pal.1-3 (fbp 4,7,9) Int. pal.1-3 (fbp 4,7,9) Int. pal.1-3 (fbp 4,7,9) Fbp.3-9 (7 mus.) Int. pal.1-3 (fbp 4,7,9)

Int. dor. 1-4 ------P --- P P P --- P

Flexor pollicis brevis P P P P P P P P P

Opponens pollicis P P P P P P P P P

Flexor digiti mi. brevis P P P P P P P P P APPEND.: VEN. AND DORSAL HAND Opponens digiti mi. P P P P P P P P P

Abductor pollicis brevis P P P P P P P P P

Abductor digiti mi. P P P P P P P P P

Intermetacarpales 1-4 P P --- P ------P ---

Int. accessorii (4 mus.) P ? ------P ------

Extensor carpi ra. longus P P P P P P P P P

Extensor carpi ra. brevis P P P P P P P P P

Brachioradialis P P P P P P P P P

Supinator P P P P P P P P P

Extensor carpi ulnaris P P P P P P P P P

Anconeus P P P P --- P P P P

Extensor digitorum P P P P P P P P P

Extensor digiti mi. P P P P P P P P P

APPEND.: DORSAL FOREARM Extensor indicis P P P P P P P P P

Extensor pollicis longus P P P P P P P P P

Abductor pollicis longus P P P P P P P P P

Extensor pollicis brevis ------P ------P

The nomenclature of the muscles follows that of Diogo and Abdala (2010) and Diogo and Wood (2011, 2012). Data from evidence provided by our own dissections and comparisons and by a

review of the literature. The black arrows indicate the hypotheses that are most strongly supported by the evidence available; the grey arrows indicate alternative hypotheses that are supported by

some of the data, but overall they are not as strongly supported by the evidence available as are the hypotheses indicated by black arrows. APPEND. = appendicular; P = muscle present; VEN.

= ventral; --- = muscle absent; access. = accessorius; dig. = digitorum; dor. = dorsales, fbp. = flexores breves profundi; mi. = minimi; mus. = muscles; pal. = palmares; pre. =

present in; prof. = profundus; ra. = radialis; int. = interossei. 68 R. Diogo et al. / Journal of Human Evolution 63 (2012) 64e78

Figure 3. Hylobates gabriellae (VU HG1, infant male): ventral view of the left ﬂexor digitorum profundus (to digits 2e5) and ﬂexor pollicis longus (to digit 1); note that the tendon of the former muscle to digit 2 also receives a small contribution from the tendon of the latter muscle to digit 1.

established that there is no fleshy or tendinous connection between these two muscles in the photographed specimen). This connection reflects the ancestral, and probably embryonic, condition where these two muscles form a single structure (Diogo and Wood, 2011). In the 18 modern humans dissected by us, we did not see as strong a tendinous distal connection between the two muscles as we found in the gibbon infant, but according to Lindburg and Comstock Figure 2. Hylobates lar (HU HL1, adult male): ventral view of the left forearm (1979) 31% of modern humans display some type of connection musculature. Note the distinct muscle flexor pollicis longus. In this and the following between these two muscles (see below). In the great apes, the fi gures, PRO, MED, LAT and DIS mean proximal, medial, lateral and distal, respectively. tendon of the flexor digitorum profundus to digit 1 is often either a very thin, vestigial structure or it is absent (Fig. 5). According to the results of our cladistic analysis, it is more parsimonious to infer hylobatids (Table 1, Fig. 1, feature 21; Figs. 2e4), confirming that this tendon became reduced in the last common ancestor observations by previous researchers (e.g., Deniker, 1885; (LCA) of hominids and became fully developed again in the Hom- Hartmann, 1886; Kohlbrügge, 1890-1892; Hepburn, 1892; Keith, inina (two steps) than to conclude that it became reduced inde- 1894b; Chapman, 1900; McMurrich, 1903a, b; Sonntag, 1924b; pendently in orangutans, gorillas and chimpanzees (three steps) Howell, 1936a, b; Straus, 1942a; Jouffroy and Lessertisseur, 1960; (Fig. 1, features 15 and 21). However, in this case there are reasons Tuttle, 1969; Jouffroy, 1971; Van Horn, 1972; Lorenz, 1974; Marzke, to suggest that the cladistically less parsimonious hypothesis may 1992, 1997; Susman, 1994, 1998; Stout, 2000; Tocheri et al., 2008). be the most likely evolutionary hypothesis (see Discussion below). This is a potential example of parallelism in that from a similar Apart from the presence of a distinct flexor pollicis longus, ancestral configuration (i.e., the flexor digitorum profundus going modern humans also show a reversion of the character state ‘Flexor to the distal phalanges of digits 1e5), there is an independent carpi radialis originates from the radius’ (Fig. 1, feature 23). Unlike acquisition of a similar derived feature (differentiation of the belly the condition in modern humans (i.e., the only bony origin is from of the flexor digitorum profundus going to digit 1, to form a sepa- the medial epicondyle of the humerus), in great apes the flexor rate flexor pollicis longus muscle). The belly of the flexor pollicis carpi radialis muscle also originates from the radius. However, it is longus is distinct from the flexor digitorum profundus in all of the not clear which is the usual condition for hylobatids, and, thus, modern humans and the two hylobatids dissected by us (Fig. 2), but whether a radial origin constitutes a synapomorphy of hominoids in the H. gabriellae (VU HG1) male infant there is a distal tendinous or of hominids. Hepburn (1892) and Kohlbrügge (1890-1892) noted connection between the tendon of the former muscle and the a partial origin of flexor carpi radialis from the radius in various tendon of the latter muscle that goes to digit 2 (Fig. 3; N.B., Fig. 2 hylobatid species, as we did in our H. gabriellae and H. lar speci- shows some tissue that appears to run distally from the most mens, but other authors (e.g., Michilsens et al., 2009) have reported lateral flexor digitorum profundus tendon to the flexor pollicis an exclusive origin from the humerus in species such as H. pileatus, longus tendon just proximal to the wrist, but careful dissection H. moloch and H. syndactylus. R. Diogo et al. / Journal of Human Evolution 63 (2012) 64e78 69

Figure 4. Homo sapiens. A) HU D43 (adult female): ventral view of the right hand, showing that some branches of the deep branch of the ulnar nerve run mainly distally to innervate the dorsal and palmar interossei, while other branches run mainly laterally to innervate the oblique head of the adductor pollicis and then the adductor pollicis accessorius (or ‘interosseous volaris primus of Henle’ of modern human anatomy; see Fig. 4B). Note that in this hand, as well as in all the other modern human hands dissected exhibiting an adductor pollicis accessorius, the oblique head of the adductor pollicis, the so-called ’deep head of the flexor pollicis brevis’ (flexor brevis profundus 2 sensu the present work) and the adductor pollicis accessorius are all present, so the adductor pollicis accessorius should clearly not be confused with the so-called ’deep head of the flexor pollicis brevis’, as it is sometimes done in the literature (see also Discussion). B) HU D45 (adult female): dorsal view of the right hand showing the innervation of the adductor pollicis accessorius from the branches of the deep branch of the ulnar nerve that innervate the oblique head of the adductor pollicis.

All of the non-Homininae (non-African ape and modern human) plesiomorphic state in mammals (e.g., Tupaia, Rattus and Cyn- primates, as well as the non-primate mammals dissected by us, ocephalus) and in primates (e.g., Lemur, Propithecus and Tarsius)is have a palmaris longus (e.g., Fig. 2). In gorillas, chimpanzees and a single humeral head. Lorisiformes have independently acquired modern humans, the palmaris longus is missing in at least 5% of the an ulnar head and thus they resemble anthropoids in that they have cases (Fig. 1, feature 18). The reviews of Keith (1899), Loth (1931), both humeral and ulnar heads. The other two forearm extensor Sarmiento (1994) and Gibbs (1999) indicate that a palmaris longus apomorphies are much less inclusive. One, which characterizes the is present in 64%, 15%, 36% and 31% of gorillas, respectively, and in Homininae, is the exclusive insertion of the extensor indicis onto 75%, 95%, 91% and 68% of chimpanzees, respectively, and more digit 2 (Fig. 1, feature 20). In other primates and in most other recent literature indicates that the muscle is present in 85% of mammals, this muscle usually goes to more than one digit (e.g., modern humans (e.g., Gibbs, 1999; Gibbs et al., 2000, 2002). Fig. 7; N.B., other authors use names such as ‘extensor indicis et tertius’ when the muscle goes e.g., to digits 2 and 3, but we follow Dorsal forearm musculature the nomenclature of Diogo and Wood, 2012, who pointed out that the muscle should still be designated as extensor indicis in those With respect to the dorsal forearm muscles, an inclusive syna- cases in order to stress the homology of the muscle, whether it goes pomorphy of anthropoids is the presence of both ulnar and to a single digit or not). The other forearm extensor apomorphy, humeral heads of the supinator (Fig. 1, feature 6). The which characterizes the extant Hominina (and also the 70 R. Diogo et al. / Journal of Human Evolution 63 (2012) 64e78

Figure 6. Pan troglodytes (GWUANT PT2, adult female): ventral view of the right flexores breves profundi 3e9 (pulled back) and the intermetacarpales 1e4; contrary to other hominoids, in Pan the flexores breves profundi 3, 5, 6 and 8 and the intermetacarpales 1, 2, 3 and 4 usually do not fuse in order to form the interossei dorsales 1, 2, 3 and 4 (as in other hominoids, the flexores breves profundi 4, 7 and 9 correspond to Figure 5. Gorilla gorilla (CMS GG1, adult male): ventral view of the oblique and the interossei palmares 1, 2 and 3). transverse heads of the right adductor pollicis, as well as of the adductor pollicis accessorius (or ‘interosseous volaris primus of Henle’ of modern human anatomy). as well as in non-primate taxa such as Rattus, Tupaia and Cyn- Hylobatidae, likely due to an independent acquisition; Table 1 and ocephalus, is to have only two contrahentes digitorum (usually to Fig. 1, feature 25) is the presence of a distinct extensor pollicis digits 2 and 5; Table 1). A synapomorphy of the Hominidae is the brevis (e.g., Fig. 7). The presence of this muscle in both modern absence of contrahentes digitorum (Fig. 1, feature 17), but in the humans and the Hylobatidae is another example of a parallelism in node leading to Pan there was a reversion to the plesiomorphic fi that a similar ancestral con guration (i.e., the abductor pollicis has primate condition (i.e., adult chimpanzees usually have two con- two distal tendons in most other primates; see Fig. 8 and below) trahentes digitorum, but unlike in strepsirrhines these muscles fi results in a similar, homoplasic, derived con guration (in which the usually go to digits 4 and 5: Table 1; N.B., the chimpanzees dissected abductor pollicis longus gives rise to a new distinct muscle, the by us have contrahentes, although in some cases the fleshy part of extensor pollicis brevis: see below). these muscles was markedly reduced: for more details see, e.g., Diogo and Wood, 2012, in press). Another reversion that differen- Intrinsic hand musculature tiates Pan from the other members of the Hominidae, as well as from hylobatids and New World monkeys is the presence of four inter- Regarding the intrinsic hand muscles, inclusive synapomorphies metacarpales (i.e., the flexor brevis profundi and intermetacarpales include the presence of an opponens pollicis and an opponens digiti do not fuse to form the dorsal interossei) each connecting adjacent minimi (Table 1) that were acquired in the node leading to the metacarpals (Table 1, Fig. 6; see Discussion below). Primates (Fig. 1, features 1, 2). The opponens pollicis became Apart from the presence of well-separated oblique and trans- secondarily undifferentiated in a few primates (e.g., Callithrix) and verse heads of the adductor pollicis (see above) catarrhines are one, or both, of these muscles may be present in some non-primate characterized by two synapomorphies. First, contrary to the ple- mammals such as rats. Strepsirrhines and most non-primate siomorphic primate condition in which the opponens pollicis mammals have an undivided adductor pollicis. In the node inserts onto the proximal and/or middle surfaces of the metacarpal leading to haplorrhines, this muscle became partly divided into I, in catarrhines (and homoplasically also in the Lorisiformes) the transverse and oblique heads (Fig.1, feature 5), and then in the node opponens pollicis reaches the distal margin of this bone (Fig. 1, leading to catarrhines the muscle became further divided into well- feature 8). Second, in catarrhines the opponens digiti minimi is separated transverse and oblique heads (Fig. 1, feature 7) (e.g., divided into superficial and deep bundles, while in other primates Figs. 4and 5). Another synapomorphy of haplorrhines is the pres- and most other mammals this muscle is undivided (Fig.1, feature 9). ence of more than two contrahentes digitorum (i.e., other than the According to our comparative and phylogenetic analyses, one of adductor pollicis, which derives evolutionary from the ancestral the synapomorphies of the Homininae (African great apes and contrahens to digit 1; Table 1, Fig. 1, feature 4). The plesiomorphic modern humans) is that the ‘volar interosseus of Henle’ (or ‘inter- condition for primates (e.g., Lemur, Propithecus, Loris and Nycticebus) osseous volaris primus of Henle’), which very likely derives from R. Diogo et al. / Journal of Human Evolution 63 (2012) 64e78 71

Figure 8. Gorilla gorilla (CMS GG1, adult male): dorsal view of the two tendons of the left abductor pollicis longus, one going to the proximo-lateral margin of the proximal phalanx of digit 1 (thus probably corresponding to the tendon of the extensor pollicis brevis of humans), the other going to the proximo-lateral margin of metacarpal I (thus probably corresponding to the tendon of the abductor pollicis longus of humans).

interosseous of Henle’ and the extensor pollicis brevis) have been considered to be uniquely, or almost uniquely, found in modern humans. Below we provide a discussion of these structures and of the other forearm and hand structures to which they are anatomically and functionally associated, and consider their functional and evolutionarily signiﬁcance in light of evidence from the fossil record and from biomechanical analyses.

Flexor pollicis longus

In modern humans, the flexor pollicis longus is innervated by the anterior interosseous nerve and it usually runs from the ventral surface of the radius and the interosseous membrane to the distal phalanx of the thumb. The different nomenclatures used to describe this muscle have resulted in some confusion. Day and Figure 7. Hylobates lar (HU HL1, adult male): dorsal view of the left forearm muscu- fl lature; note the distinct muscle extensor pollicis brevis. Napier (1963) suggested that the exor pollicis longus is present in Loris, Nycticebus, Tarsius, Aotus, Callithrix, Saimiri, Macaca, Cer- copithecus, Colobus, Papio and Homo, and this was the view that a thin deep additional slip of the adductor pollicis (Diogo and informed the cladistic studies of Groves (1986) and Shoshani et al. Wood, 2011), is often (c. < 50% of cases) present (Fig. 1, feature (1996). However, Day and Napier (1963) intended the term’s 19), whereas in other primates and other mammals this muscle is traditional usage (e.g., Barnard, 1875; Duckworth,1904, 1915; Wood nearly always, or always, absent. This synapomorphic Homininae Jones, 1920; Sonntag, 1924a, b), which is to indicate the presence of condition where the muscle is often but not usually present (i.e., a tendon of the flexor digitorum profundus to digit 1. They did not c. < 50% of cases) is seen in chimpanzees and gorillas (e.g., Fig. 5). necessarily imply the presence of a distinct flexor pollicis longus Modern humans display an autapomorphic condition within the muscle (i.e., separated from the flexor digitorum profundus, with Homininae because the muscle is found in most (50%) of the cases a tendon exclusive to the thumb). In the Loris, Nycticebus and Tar- (see Table 1; Fig. 1, feature 24; Fig. 4). sius specimens dissected by Burmeister (1846), Mivart and Murie (1865), Murie and Mivart (1872), Keith (1894a, b), Allen (1897), Discussion Woollard (1925), Miller (1943), Schultz (1984) and by us, the Aotus, Callithrix and Saimiri specimens studied by Senft (1907), Beattie Four forearm and hand muscles (flexor pollicis longus, the so- (1927), Hill (1957, 1960, 1962) and by us, the Macaca specimens called ‘deep head of the flexor pollicis brevis’, the so-called ‘volar examined by Haughton (1864, 1865), Howell and Straus (1932, 72 R. Diogo et al. / Journal of Human Evolution 63 (2012) 64e78

1933), Patterson (1942), Jacobi (1966), Kimura and Tazai (1970), the carpus in hylobatids. This shunt, which is commonly less Jouffroy (1971) and Landsmeer (1986) and by us, the Cercopithecus substantial than the tendons it connects, is a slender flat (4e5mm specimens investigated by Hill (1966) and Lewis (1989) and by us, wide) structure that passes obliquely inferomedially from the the Colobus specimens dissected by Brooks (1886), Polak (1908), tendon of the flexor pollicis longus to the lateral edge of the Jouffroy and Lessertisseur (1960) and by us, and in the Papio common flexor digitorum profundus tendons to digits 2e5. specimens studied by MacDowell (1910), Hill (1970), Swindler and According to Tuttle, the ’flexor shunt’ probably serves different Wood (1973), Tocheri et al. (2008) and by us there is no distinct functions during the various activities undertaken by hylobatids. flexor pollicis longus going exclusively to digit 1 (Table 1). Thus, the For instance, he suggested that when the hand is used as an apparently contradictory statements in the literature about the anatomical hook during vigorous arm-swinging, the ‘shunt’ prob- presence/absence of a flexor pollicis longus in non-hominoid ably transfers some of the contractile force of flexor pollicis longus primates are mainly the result of nomenclatural pluralism and to the flexor digitorum profundus tendons and thereby helps to they should not be used as the basis of cladistic character states stabilize the wrist. In contrast, when the hand is used for fine without additional clarification. manipulation, particularly when the metacarpophalangeal joints of Our dissections confirm previous reports that in most non- digits 2e5areflexed, the ‘shunt’ is probably somewhat slack, hominoid primates the flexor digitorum profundus sends allowing the flexor pollicis longus to independently flex the distal a tendon to digit 1. However, it is not independent of the tendons phalanx of the thumb. When the thumb is abducted to grasp large going to the other digits even though, when it is present, it flexes branches, the ‘shunt’ is probably also stretched to the extent that the distal phalanx of the thumb in the same way that the flexor the distal phalanges of the thumb and the medial four digits are pollicis longus does in modern humans. In the great apes, the flexor flexed synchronously to produce secure grips. As a result, and digitorum profundus tendon to digit 1 is usually either vestigial or contrary to what occurs in Pan, Pongo and Gorilla, in hylobatids the even absent (e.g., Fig. 5). Straus (1942b) showed, using evidence pollex normally plays an active role in maintaining a suspended from both his own dissections and from data available to him in the posture and in facilitating the distinctive locomotion of these literature, that among 47 chimpanzees such a tendon was absent in hominoids. After observing living animals, Van Horn (1972) sug- 14 individuals (30%), too small/thin to have had any useful function gested that the flexor pollicis longus of hylobatids is important in 10.5 individuals (22%), and was in direct functional continuity during the arm-pull phase of climbing, in which the terminal with the radial muscle belly of the flexor digitorum profundus in 22 phalanx of the thumb is flexed as the animal lifts its weight from individuals (48%). Likewise, among 16 gorillas Straus (1942b) a previous support (i.e., this muscle is an important component of showed that this tendon was absent in five individuals (31%), the power used to grasp). Stout (2000) found that the tendon of the rudimentary in six individuals (41%), and present in four individ- flexor pollicis longus of hylobatids does not flex the distal phalanx uals (28%). Among 27 orangutans, the tendon was absent in 24 of the thumb, but instead it adducts the thumb, flexes the pollical individuals (89%), rudimentary and functionless in two individuals metacarpophalangeal joint and stabilizes the pollical interphalan- (7%) and only completely developed in one individual (4%). Thus, geal joint. Further functional studies of hylobatids are needed to there is no functioning tendon in c. 95% of orangutans, in c. 72% of test these hypotheses, particularly because even among modern gorillas and in c. 50% of chimpanzees. humans there is significant variation with respect to the presence of In hylobatids, the anatomy of the flexor pollicis longus is more a fully independent flexor pollicis longus. For instance, as explained complex. Our de novo dissections (Table 1; Figs. 2e4) con firm that above according to an extensive review undertaken by Lindburg both modern humans and hylobatids have a flexor pollicis longus and Comstock (1979), only c. 70% of modern humans have distinct from the flexor digitorum profundus, unlike the a flexor pollicis longus that is anatomically entirely independent extensively-connected muscle bellies seen in other primates from the flexor digitorum profundus. Shrewsbury et al. (2003) (corroborating the descriptions of authors such as Deniker, 1885; posited that evidence of restrictive thumb/index tendosynovitis Hartmann, 1886; Kohlbrügge, 1890-1892; Hepburn, 1892; Keith, and pain in modern humans is usually associated with connections 1894b; Chapman, 1900; McMurrich, 1903a, b; Sonntag, 1924b; between the flexor pollicis longus and the flexor digitorum pro- Howell, 1936a, b; Straus, 1942a; Jouffroy and Lessertisseur, 1960; fundus. In some cases, the condition can be relieved by removing Tuttle, 1969; Jouffroy, 1971; Van Horn, 1972; Lorenz, 1974; Marzke, the connection between these muscles and occasionally the 1992, 1997; Susman, 1994, 1998; Stout, 2000; Tocheri et al., 2008). thickened flexor pollicis longus must be removed as well. According However, a few authors (e.g., Payne, 2001) suggest that in the to Shrewsbury et al. (2003), the repetitive force used in a precision Hylobates specimens dissected by them the flexor pollicis longus grip in concert with habitual repetitive behaviours associated with blends with, and is thus not really separate from, the flexor dig- precision handling might have favoured independence of these two itorum profundus. muscles as modern humans became increasingly dependent on the In an influential paper about the evolution of tool use in the dexterity required to make and use tools. Hominina, Susman (1994:1573,Fig. 3) stated that contrary to the There are reasons to consider that the tendon of the flexor condition in modern humans, chimpanzees and other primates digitorum profundus to digit 1 became independently reduced in have only a tendon that mimics the flexor pollicis longus and lacks orangutans, gorillas and chimpanzees, while the presence of a separate muscle belly; “in lesser apes (gibbons and siamang) a separate, robust flexor pollicis longus in hylobatids and modern there is a muscle belly, but it is functionally coupled with the flexor humans may be associated with having a well-developed and/or digitorum produndus.electromyography experiments on an adult functionally more independent thumb. The first hypothesis is female gibbon (H. lar) did not elicit flexion of the thumb separate supported by the fact that the tendon of the flexor digitorum pro- from flexion of the fingers, as is the case in humans”. The apparent fundus to digit 1 is highly variable in great apes. In a few cases it is inconsistency between Susman’s interpretation and the interpre- fully developed, in others it is atrophied, in others it is ‘rudimen- tations of researchers who have described the hylobatid flexor tary’ and may no longer attach to the main body of the flexor dig- pollicis longus as distinct (e.g., Marzke, 1992) is likely due to the itorum profundus muscle, and in others it is completely missing variable presence of a connective tissue ‘shunt’ (Tuttle, 1969) that (see above and also, e.g., Shrewsbury et al., 2003, Fig. 5). Straus connects the tendons of the flexor digitorum profundus and flexor (1942b) included those cases where there is no continuity with pollicis longus (Tuttle (1969) referred to it as the ‘radial component the main body of the flexor digitorum profundus in the category of of the flexor digitorum profundus musculature’) often at the level of rudimentary, vestigial or functionless tendon to digit 1. However, it R. Diogo et al. / Journal of Human Evolution 63 (2012) 64e78 73 is important to note that, as stressed by Shrewsbury et al. (2003), Pan-Homo LCA, is retained in the Hominina, and independently lost the tendon is not necessarily functionless in all of those cases. in the evolution of Pan and Gorilla (and possibly Pongo), or 2) that Within the specimens dissected by them, the tendon had an the Pan-Homo LCA lacked the tendon insertion, and that it evolved attachment pattern that was ligament-like in all four specimens of rapidly at the base of the Hominina clade independently of its orangutan, two of the chimpanzees, and the two juvenile hama- evolution in hylobatids. In either scenario, the evolution of the dryas baboons. The tendon attached distally to the pollical distal flexor pollicis longus in hominoids involves homoplasy. In a recent phalanx and laterally to the radial and ulnar basal tubercles. Tendon paper, Almécija et al. (in press) strongly support the idea that the fibres in the hamadryas baboons had distal attachments to the plesiomorphic hominoid, hominid and hominin condition is very ungual pulp over the distal phalangeal tuberosity. However, the likely a moderately long thumb, longer than the thumb of modern proximal radial and ulnar bands of the tendon both sent branches great apes, thus indirectly supporting the idea that the tendon of to the metacarpophalangeal joint in the region of the adductor the flexor digitorum profundus to digit 1 became independently pollicis insertion, and also (in the apes) in the region of the flexor reduced in orangutans, gorillas and chimpanzees. pollicis brevis insertion. No fibres were found continuing further Some time between the late Pliocene and early Pleistocene, the proximally over the wrist to the radial aspect of the flexor dig- thumb of the members of the subtribe Hominina evolved signifi- itorum profundus tendon. According to Shrewsbury et al. (2003), cantly greater robusticity, the thumb of modern humans being more this distal tethering of the tendon to digit 1 to bone on both sides of robust than that of any other extant hominoids, including hyloba- the distal interphalangeal joint probably helps to stabilize the joint. tids (Susman, 1994). However, the hypothesis remains untested The hypothesis that the presence of a separate, robust flexor regarding whether other derived muscular features of the modern pollicis longus in hylobatids and modern humans may be associ- human hand occurred with the evolution of thumb robusticity. ated with having a well-developed and/or functionally more independent thumb is in line with the results of electromyography Deep head of the flexor pollicis brevis and the ‘volar interosseous of (Marzke et al., 1998) and hand pressure studies (Rolian et al., 2011; Henle’ Williams et al., 2012) when various modern human subjects were manufacturing and using Oldowan tools as well as other objects. In modern humans, the so-called ‘deep head of the flexor pol- The recruitment of the flexor pollicis longus and the extensor licis brevis’ usually runs from the carpal region (e.g., capitate, pollicis brevis allowed the subjects to maintain the meta- occasionally from trapezoid) to the medial side of the palmar carpophalangeal joint in extension (using the extensor pollicis surface of proximal phalanx of digit 1 and/or an adjacent sesamoid brevis, which in modern humans usually attaches onto the prox- bone (e.g., Fig. 4) and it is normally innervated by the deep branch imal phalanx of the thumb: see below), while simultaneously using of the ulnar nerve (and/or occasionally by the median nerve). The the flexor pollicis longus to flex the distal phalanx of the thumb at so-called ‘volar interosseous of Henle’ usually runs from the base of the pollical interphalangeal joint. The results from experiments metacarpal I to the ‘wing tendon’ of the extensor apparatus of digit documenting hand pressure during stone tool making underscore 1 (e.g., Susman et al., 1999). In the specimens examined by us, this the importance of hand posture on the distribution of pressure latter structure, which is normally innervated by the deep branch of across the thumb and other digits (Williams et al., 2012), the ulnar nerve, may also originate from the carpal bones adjacent supporting the suggestions of Marzke et al. (1998) suggestion that to metacarpal I and its insertion onto digit 1 may extend onto the the differentiation of these two muscles needed to have occurred proximal phalanx and/or sesamoid bone of the thumb (e.g., Fig. 4). for early members of the human lineage to be able to habitually Susman (1994; Fig. 3) suggested that the ’deep head of the flexor make effective stone tools. pollicis brevis’ and the ’interosseous volaris primus of Henle’ (as Evidence from the fossil record also lends some support to the described above) are derived in modern humans. However, our hypothesis that a well-developed tendon to digit 1, whether as part comparative investigation suggests that the ‘deep head of the flexor of flexor digitorum profundus or as a separate flexor pollicis longus, pollicis brevis’ is present in most primates, while the volar inter- was the primitive condition for the Hominoidea and probably the osseous of Henle is seen in <50% of a few primates including Pan-Homo LCA. While some (Shrewsbury et al., 2003; Marzke and gorillas and chimpanzees (e.g., Fig. 5). Lewis (1989) embraced the Shrewsbury, 2006; Tocheri et al., 2008) have noted that the flexor hypothesis of Forster (1917) who suggested that the plesiomorphic pollicis longus (or flexor digitorum profundus to digit 1) does not condition for eutherian mammals is to have ten flexores breves insert onto the distal phalanx’s volar pit, or ‘proximal volar fossa’ profundi, each inserting onto the lateral and medial sides of each (Shrewsbury et al., 2003), it is clear that the tendon inserts on the digit, and four intermetacarpales (see, e.g., Fig. 6) connecting the ridge that is, when present, just distal to the fossa (Susman, 1998; adjacent metacarpal bones. In primates, the plesiomorphic condi- Shrewsbury et al., 2003). The presence of an attachment (ridge) to tion is the same, but two of the 14 muscles have differentiated, each the distal pollical phalanx is seen in all of the fossil hominoid taxa forming two muscles: the flexor pollicis brevis and the opponens known to date, including Proconsul (Begun et al., 1994) and Oreo- pollicis from flexor brevis profundus 1, and the flexor digiti minimi pithecus (Moyà-Solà et al., 1999). Similarly, all fossil pollical thumb brevis and the opponens digiti minimi from flexor brevis profundus distal phalanges attributed to (pre-modern) Homo show evidence 10 (Fig. 1, Table 1). In hominoids other than chimpanzees, as well as of a volar ridge, with an accompanying proximal volar fossa in New World monkeys, the flexores breves profundi 3, 5, 6 and 8 (Shrewsbury et al., 2003) that marks the attachment of a tendon usually fuse with the intermetacarpales 1, 2, 3 and 4 to form the (Susman, 1998). The attachment site is also present on the distal dorsal interossei 1, 2, 3 and 4, respectively, whereas the palmar pollical phalanges of both early and potential fossil taxa of the interossei 1, 2 and 3 are derived directly from the flexores breves subtribe Hominina, including Orrorin tugenensis (Gommery and profundi 4, 7 and 9, respectively. The model of Lewis (1989), model, Senut, 2006; Almécija et al., 2010), Ardipithecus ramidus (Lovejoy which explains why the dorsal interossei muscles are bipennate et al., 2009), and. Australopithecus afarensis (Ward et al., in press). whereas the palmar interossei are unipennate, is supported by the Whether Orrorin is a member of the subtribe Hominina (Senut developmental studies of authors such as Cihák (1972), which point et al., 2001; Richmond and Jungers, 2008) or not (e.g., Wood and out that at least some flexores breves profundi effectively become Harrison, 2011), the clear indication of a tendon insertion on the fused ontogenetically with the most dorsal intrinsic muscles of the distal pollical phalanx at almost six million years ago suggests hand (the intermetacarpales), forming the dorsal interossei of adult that either 1) the presence of tendon to digit 1 was present in the modern humans. Chimpanzees have undergone a secondary 74 R. Diogo et al. / Journal of Human Evolution 63 (2012) 64e78 reversion to the plesiomorphic state, probably because their flex- designated this structure either a component of the adductor pol- ores breves profundi do not fuse with the intermetacarpales during licis or they have referred to it as the ‘interosseous volaris primus of ontogeny to form the dorsal interossei (Table 1; Fig. 6). As explained Henle’. Linscheid et al. (1991) have designated the adductor pollicis above, chimpanzees also display a secondary reversion of a syna- accessorius sensu the present paper as the ‘accessory adductor pomorphy of the Hominidae in that adult chimpanzees have two muscle’ or the ‘adductor pollicis accessory’, but their study is contrahentes digitorum in addition to the the adductor pollicis, one somewhat confusing because despite using this name they clearly going to digit 4 and the other to digit 5 (in other adult hominids consider that this muscle corresponds to the flexor brevis pro- there is usually none). Because the studies of Cihák (1972) suggest fundus 2, and not to part of the adductor pollicis, of other mammals that in at least modern humans the contrahentes are lost (i.e., (see, e.g., their Fig. 8). According to the authors, the adductor pol- ‘reabsorbed’) during ontogeny, the presence of metacarpales and of licis accessorius sensu the present paper is present in all eight contrahentes in chimpanzees is very likely due to a prolonged or modern human hands dissected by them, inserting onto the ulnar delayed development of the hand musculature of these apes, i.e., in wing tendon of the thumb; the origin is from metacarpal 1 in four this particular case extant chimpanzees are seemingly more hands, from metacarpals 1 and 2 in two hands, from metacarpal 2 neotenic than modern humans (Diogo and Wood, in press). This is in one hand, and from the trapezium in one hand. Contrary to the in line with other recent studies that have pointed out that hypothesis defended in the present paper, they thus consider that although in the literature it is often stated that modern humans are the ‘deep head of the flexor pollicis brevis’ of modern human in general more neotenic than other primates, the empirical data anatomy corresponds to part of the oblique head of the adductor collected in the last decades reveal that both paedomorphic and pollicis, and not to the flexor brevis profundus 2, of other mammals. peramorphic processes have been involved in the mosaic evolution Moreover, they use the name ‘adductor pollicis, accessory oblique of humans and of other hominoids (see, e.g., Bufill et al., 2011; and head’ to designate an extra head of the adductor pollicis that is references therein). present in six of the eight modern human hands dissected by them The main controversy concerning the evolution and taxonomic and that originates from metacarpal 2. distribution of the so-called ‘deep head of the flexor pollicis brevis’ In summary, our comparative and phylogenetic analyses indicate and of the volar interosseous of Henle mainly concerns the identity that the flexor brevis profundus 2 has been present and has basically of one of the flexores breves profundi (i.e., number 2, which ple- kept the same anatomical configuration from the LCA of eutherian siomorphically in mammals goes to the ulnar side of digit 1). mammals to H. sapiens, where it forms the so-called ‘deep head of Although the so-called ‘volar interosseous of Henle’ is usually not the flexor pollicis brevis’. The names ‘deep head of the flexor pollicis shown in modern human anatomical atlases, it is present in the brevis’ and ‘superficial head of the flexor pollicis brevis’ are thus majority of modern humans. For example, Abramowitz (1955), inappropriate and in our opinion should be abandoned, because the Lewis (1989), Susman et al. (1999), Henkel-Kopleck and Schmidt former corresponds directly to the flexor brevis profundus 2 of other (2000) and Morrison and Hill (2011) report the presence of volar mammals while the latter derives from the flexor brevis profundus 1 interosseous of Henle in 100%, 92%, 86%, 69%, and 91%, respectively, (the opponens pollicis being also derived from the flexor brevis of the individuals examined. This structure was also found in the 18 profundis 1: see above and Table 1). We propose that the ‘deep head modern human individuals examined in this study (GWUANA HS1- of the flexor pollicis brevis’ and ’superficial head of the flexor pollicis 6, HU D43 and HU D45; Fig. 4). brevis’ are therefore designated as flexor brevis profundus 2 and as As the alternative name ‘interosseous volaris primus’ indicates, flexor pollicis brevis, respectively (see, e.g., Fig. 4; Table 1). As we some authors consider that this structure, (e.g., Figs. 4and 5; see have argued above, the ‘volar interosseous of Henle’ of modern below) corresponds to a vestigial flexor brevis profundus 2 (i.e., it is humans does not correspond to the flexor brevis profundus 2 of a fourth, and most radial, palmar interosseus). However, our other mammals (i.e., it is not a true palmar interosseus), but instead comparative and phylogenetic analyses strongly suggest that the it is very likely derived from the adductor pollicis (Fig. 9; Table 1: ‘interosseous volaris primus’ corresponds instead to a structure adductor pollicis accessorius; see below). that is derived from a thin, deep additional slip of the adductor pollicis (Diogo and Wood, 2011) that we suggest should instead be Extensor pollicis brevis named the adductor pollicis accessorius (see below and Table 1). In fact, the ‘interosseous volaris primus’ often blends with the In modern humans, the extensor pollicis brevis, which is adductor pollicis distally (see Diogo and Wood, 2011, 2012) and, at innervated by the posterior interosseous nerve, usually runs from least in modern humans, it is innervated by the branches of the the dorsal surface of the radius and the adjacent interosseous deep branch of the ulnar nerve that innervate the adductor pollicis membrane to the base of the proximal phalanx of digit 1. Within the (and not by the branches of this deep branch that innervate the 200 adult modern human upper limbs examined by Kaneff (1959, dorsal and ventral interossei: Fig. 4; Bello-Hellegouarch et al., 1968, 1969, 1979, 1980a, b), the extensor pollicis brevis is missing in submitted). Moreover, contrary to the statements of some authors just over 1% of cases and in just over 5% of cases this muscle is (e.g., Susman, 1994), our analyses revealed that the so-called ‘deep reduced to a tendon connected to the tendon of the abductor pol- head of the flexor pollicis brevis’ of modern human anatomy is licis longus or to ligaments adjacent to the insertion of this tendon. present in the vast majority of primates and that it corresponds to In most of the cases he examined, Kaneff reported that the fleshy the flexor brevis profundus 2 of other mammals (Table 1; Diogo and belly of the extensor pollicis brevis blends with the belly of the Wood, 2011, 2012). For example, the structure designated as the abductor pollicis longus. According to the ontogenetic studies of flexor brevis profundus 2 in rats is the structure that is designated Lewis (1910), the abductor pollicis longus and extensor pollicis as ‘deep head of the flexor pollicis brevis’ of modern humans. Such brevis usually only become separate entities late in modern human a structure is also found in the vast majority of non-human development. In our comparative sample, the only other group primates, except for the New World monkeys in which this struc- apart from modern humans with a distinct extensor pollicis brevis ture is either missing or, more likely, has fused with the flexor is the hylobatids. A separate extensor pollicis brevis was found in brevis profundus 1 component of the ‘superficial head of the flexor the hylobatids dissected by Bischoff (1870), Kohlbrügge (1890- pollicis brevis’ of these monkeys. As Diogo and Wood (2011) argue, 1892), Duckworth (1904) and Michilsens et al. (2009) and by us most authors have found a flexor brevis profundus 2 in the non- (e.g., Fig. 7). The exception is Deniker (1885), who did not find human and non-platyrrhine primates, but many have erroneously a distinct extensor pollicis brevis in the gibbon foetus dissected by R. Diogo et al. / Journal of Human Evolution 63 (2012) 64e78 75 C D A DP B DP DP DP

PP PP PP PP

MI MI MI MI OH OH OH OH

AB PPI PPI

Figure 9. Simplified scheme showing our hypothesis about the evolution of the adductor pollicis accessorius (or ‘interosseous volaris primus of Henle’ of modern human anatomy): A) First stage of evolution, with a mainly undivided oblique head of the adductor pollicis (OH); B) Second, hypothetical stage of evolution, with the oblique head plus an additional bundle (AB) that has an origin separated from the main body of this head but still a common insertion together with that main body; C) Third, hypothetical stage of evolution, in which the insertion of the additional bundle is also somewhat separated from the main body of the adductor pollicis, forming an almost completely differentiated PPI; D) Fourth, hypothetical stage of evolution, showing a PPI that is even more lateral (radial) and thus even more separated from the main body of the oblique head of the adductor pollicis, resembling a PPI configuration that is commonly found in modern humans (DP, distal phalanx of digit 1; MI, metacarpal I; PP, proximal phalanx of digit 1). him. Contrary to the condition in modern humans, in hylobatids the going to the phalanges of the thumb (flexor pollicis longus to the extensor pollicis brevis usually does not extend to the proximal distal and extensor pollicis brevis to the proximal), have more fore- phalanx of the thumb, inserting instead onto the base of metacarpal arm muscles (i.e., 20) than any other primate studied by us (Table 1). I and/or the adjacent sesamoid or carpal bones. In gorillas (e.g., Hylobatids also have a flexor pollicis longus and an extensor pollicis Hepburn, 1892; Straus, 1941a, b; Raven, 1950; Preuschoft, 1965; brevis, but as stated above they normally lack an anconeus, so they Sarmiento, 1994), a tendon of the abductor pollicis longus often have 19 forearm muscles (Table 1). With respect to the hand muscles, inserts onto the proximal phalanx of the thumb and the term phylogenetically plesiomorphic primates such as strepsirrhines and ‘extensor pollicis brevis’ has been used to describe this (N.B., in Tarsius usually have more than 30 muscles, but modern humans other great apes, including those dissected by us, the abductor usually have only 21 muscles (Table 1). The hand muscles that pollicis longus usually inserts onto metacarpal I and/or onto carpal modern humans have lost relative to phylogenetically plesiomorphic bones: see Diogo and Wood, 2011). However, as Huxley (1864), primates (i.e., contrahentes, intermetacarpales and interossei acces- Macalister (1873), Bischoff (1880), Deniker (1885), Tuttle (1970), sorii) attach to digits 2e5(Table 1). The hand muscles that are Kaneff (1979, 1980a, b), and Aziz and Dunlap (1986) have stressed, conserved as separate structures in modern humans are those that in gorillas the tendon going to the proximal phalanx of the thumb is insert onto the thumb. Modern humans usually also have an addi- the result of a bifurcation of the tendon of the abductor pollicis, the tional pollical structure that may also be present in gorillas and other branch of this tendon usually going to the metacarpal I chimpanzees, which is often designated as ‘musculus interosseous (Fig. 8). Our dissections corroborate this. Thus, contrary to the volaris primus of Henle’ but should be designated as musculus condition in modern humans and hylobatids, in gorillas and in adductor pollicis accessorius (Table 1). other primates there is no separate extensor pollicis brevis muscle Contrary to what is sometimes suggested in the literature, none (i.e., a distinct belly separated from abductor pollicis longus). It of the muscles and muscle bundles in the modern human forearm should moreover be noted that in the literature reviews carried out and hand are unique to modern humans. Nonetheless, it is probably by Straus (1941a, b) and Sarmiento (1994), for gorillas they found not a coincidence that the three derived structures found in that in just over half of the cases the abductor pollicis longus only modern humans and in only one or two other primate genera (i.e., extends as far as the proximal phalanx of the thumb. adductor pollicis accessorius is also found in some Pan and Gorilla; In summary, modern humans are peculiar because they have flexor pollicis longus and extensor pollicis brevis also found in a distinct extensor pollicis brevis that attaches to the proximal Hylobates) all involve the thumb. In fact, this is consistent with the phalanx of the thumb, hylobatids are peculiar because they have hypothesis that movements of the thumb played an important role a distinct extensor pollicis brevis that usually attaches to the base of in human evolution. Marzke et al. (1998) suggested that the flexor metacarpal I and/or to adjacent bones, and gorillas are peculiar pollicis longus and extensor pollicis brevis may have co-evolved to because in about half of the cases one of the two tendons of the enable the members of the subtribe Hominina to maintain the abductor pollicis longus attaches to the proximal phalanx of the metacarpophalangeal joint in extension while flexing the distal thumb. According to our dissections and review of the literature, in phalanx of the thumb as occurs in stone tool making and particu- chimpanzees, orangutans and in non-hominoid primates the larly usage. However, the presence of the flexor tendon attachment abductor pollicis longus usually has two tendons, but with the site on distal phalanges of all known fossil taxa of the subtribe exception of one of the 20 chimpanzees reviewed by Keith (1899), Hominina suggests that some of this musculature was present early these do not extend to the proximal phalanx of the thumb. in human evolution and indeed it may be the primitive condition for the Pan-Homo LCA. Conclusions The functional significance of the adductor pollicis accessorius is more obscure, because this structure usually connects the ulnar side Our comparative and phylogenetic analyses indicate that the of the base of metacarpal I to the wing tendon of the extensor forearm muscles in primates usually number between 18 and 19 apparatus of digit 1. It is not very substantial and it is difficult to see (Table 1). Two of the 19 muscles predicted to be plesiomorphically (e.g., Morrison and Hill, 2011) how it could confer significant present in primates may be missing in some groups (e.g., the epi- biomechanical advantage across the joints of the thumb. Detailed trochleoanconeus is usually absent in hominoids except Pan and the comparative morphological and functional analyses, ideally anconeus is usually absent in Hylobates). Modern humans, because including electromyographic studies, should be carried out in they usually lack the epitrochleoanconeus but have two muscles modern humans and also in gorillas and chimpanzees in order to 76 R. Diogo et al. / Journal of Human Evolution 63 (2012) 64e78 better understand the functional and evolutionary significance of this Brooks, H.S.J., 1886. On the morphology of the intrinsic muscles of the little finger, fl enigmatic structure. In addition, further studies of more specimens of with some observations on the ulnar head of the short exor of the thumb. J. Anat. Physiol. 20, 644e661. other primate taxa are needed to check if this structure may be Bufill, E., Agusti, J., Blesa, R., 2011. Human neoteny revisited: the case of synaptic present in at least some specimens of non-hominin taxa as well. plasticity. Am. J. Human Biol. 23, 729e739. The adoption and consistent use of the names proposed in this Burmeister, H., 1846. Beiträge zur näheren Kenntniss der Gattung Tarsius. Georg Reimer, Berlin. paper is not a mere nomenclatural detail: it is a return to an old, and Chapman, H.C., 1900. Observations upon the anatomy of Hylobates leuciscus and Darwinian, tradition in which it is important to understand the Chiromys Madagascariensis. Proc. Acad. Nat. Sci. Phila 52, 414e423. phylogeny and homology of each and every anatomical structure. It Cihák, R., 1972. Ontogenesis of the skeleton and intrinsic muscles of the human hand and foot. Ergeb. Anat. Entwicklungsgesch 46, 5e194. is also a way to stress that modern humans usually have 11 muscles Day, M.H., Napier, J., 1961. The two heads of the flexor pollicis brevis. J. Anat. 95, (abductor pollicis longus, extensor pollicis brevis, extensor pollicis 123e130. longus, flexor pollicis longus, abductor pollicis brevis, flexor pollicis Day, M.H., Napier, J., 1963. The functional significance of the deep head of flexor pollicis brevis in primates. Folia Primatol. 1, 122e134. brevis, opponens pollicis, adductor pollicis, dorsal interosseus 1, Deniker, J., 1885. Recherches anatomiques et embryologiques sur les singes adductor pollicis accessorius, and flexor brevis profundis) attached anthropoides, foetus de gorille et de gibbon. Arch. Zool. Exp. Gén 3, 1e265. to metacarpal I and the pollical phalanges, and not nine as it often Diogo, R., 2007. On the Origin and Evolution of Higher-Clades: Osteology, Myology, stated in atlases and textbooks. We hope that this paper will attract Phylogeny and Macroevolution of Bony Fishes and the Rise of Tetrapods. Science Publishers, Enfield. the attention of the authors of human anatomy atlases and text- Diogo, R., 2009. The head musculature of the Philippine colugo (Dermoptera: books and will thus result in the inclusion, in these atlases and Cynocephalus volans), with a comparison to tree-shrews, primates and other e textbooks, of information about the musculus adductor pollicis mammals. J. Morphol. 270, 14 51. Diogo, R., Abdala, V., 2007. Comparative anatomy, homologies and evolution of the accessorius and the musculus flexor brevis profundus 2. pectoral muscles of bony fish and tetrapods: a new insight. J. Morphol. 268, 504e517. Diogo, R., Abdala, V., 2010. Muscles of Vertebrates e Comparative Anatomy, Acknowledgements Evolution, Homologies and Development. Science Publishers, Enfield. Diogo, R., Wood, B., 2011. Soft-tissue anatomy of the primates: phylogenetic analyses based on the muscles of the head, neck, pectoral region and upper limb, We thank R. Walsh and F. Slaby (Department of Anatomy, with notes on the evolution of these muscles. J. Anat. 219, 273e359. George Washington University), R. Bernstein and S. McFarlin Diogo, R., Wood, B., 2012. Comparative Anatomy and Phylogeny of Primate Muscles and Human Evolution. Taylor and Francis, Oxford. (Department of Anthropology, George Washington University), Diogo, R., Wood, B. Violation of Dollo’s law: evidence of muscle reversions in primate N. Rybczynski (Canadian Museum of Nature), H. Mays (Cincinnati phylogeny and their implications for the understanding of the ontogeny, Museum of Natural History), A. Aziz (Department of Anatomy, evolution and anatomical variations of modern humans. Evolution, in press. fi Howard University), F. Pastor (Department of Anatomy, University Diogo, R., Abdala, V., Aziz, M.A., Lonergan, N., Wood, B., 2009. From sh to modern humans - comparative anatomy, homologies and evolution of the pectoral and of Valladolid), A. Gorow, H. Fitch-Snyder and B. Rideout (San Diego forelimb musculature. J. Anat. 214, 694e716. Zoo) and J. Fritz and J. Murphy (Primate Foundation of Arizona) for Duckworth, W.L.H., 1904. Studies from the Anthropological Laboratory, the kindly providing the non-primate and primate mammalian speci- Anatomy School, Cambridge. C. J. Clay & Sons, London. Duckworth, W.L.H., 1915. Morphology and Anthropology, second ed. Cambridge mens dissected during this project. RD was supported by a George University Press, Cambridge. Washington University (GW) Presidential Merit Fellowship and by Dunlap, S.S., Thorington, R.W., Aziz, M.A., 1985. Forelimb anatomy of New World a Howard University start-up package, BW by the GW University monkeys: myology and the interpretation of primitive anthropoid models. Am. J. Phys. Anthropol. 68, 499e517. Professorship in Human Origins, the GW Provost and the GW Dylevsky, I., 1967. Contribution to the ontogenesis of the flexor digitorum super- Selective Excellence Program, and BGR by the National Science ficialis and the flexor digitorum profundus in man. Folia Morphol. 15, 330e335. Foundation (BCS-0725122). Forster, A., 1917. Die mm. contrahentes und interossei manus in der Säugetierreihe und beim Menschen. Arch. Anat. Physiol. Anat. Abt. 1916, 101e378. Gibbs, S., 1999. Comparative soft tissue morphology of the extant Hominoidea, including man. Ph.D. Dissertation, University of Liverpool. References Gibbs, S., Collard, M., Wood, B.A., 2000. Soft-tissue characters in higher primate phylogenetics. Proc. Natl. Acad. Sci. 97, 11130e11132. Abdala, V., Diogo, R., 2010. Comparative anatomy, homologies and evolution of the Gibbs, S., Collard, M., Wood, B.A., 2002. Soft-tissue anatomy of the extant homi- pectoral and forelimb musculature of tetrapods with special attention to extant noids: a review and phylogenetic analysis. J. Anat. 200, 3e49. limbed amphibians and reptiles. J. Anat. 217, 536e573. Gommery, D., Senut, B., 2006. The terminal thumb phalanx of Orrorin tugenensis Abramowitz, I., 1955. On the existence of a palmar interosseous muscle in the (Upper Miocene of Kenya). Geobios 39, 372e384. thumb with particular reference to the Bantu-speaking Negro. S. Afr. J. Sci. 51, Groves, C.P., 1986. Systematics of the great apes. In: Swindler, D.R., Erwin, J. (Eds.), 270e276. 1986. Comparative Primate Biology: Systematics, Evolution and Anatomy, vol. 1. Allen, H., 1897. Observations on Tarsius fuscus. Proc. Acad. Nat. Sci. Phila 49, 35e55. A.R. Liss, New York, pp. 187e217. Almécija, S., Moya-Sola, S., Alba, D.M., 2010. Early origin for human-like precision Haines, R.W., 1939. A revision of the extensor muscles of the forearm in tetrapods. grasping: a comparative study of pollical distal phalanges in fossil hominins. J. Anat. 73, 211e233. Plos One 5, e11727. Haines, R.W., 1946. A revision of the movements of the forearm in tetrapods. J. Anat. Almécija, S., Alba, D. M., Moyá-Solá, S. The thumb of Miocene apes: New insights 80, 1e11. from Castell de Barberá (Catalonia, Spain). Am. J. Phys. Anthropol., in press. Haines, R.W., 1950. The flexor muscles of the forearm and hand in lizards and Aziz, M.A., Dunlap, S.S., 1986. The human extensor digitorum profundus muscle mammals. J. Anat. 84, 13e29. with comments on the evolution of the primate hand. Primates 27, 293e319. Hartmann, R., 1886. Anthropoid Apes. Keegan, London. Barnard, W.S., 1875. Observations on the membral musculation of Simia satyrus Haughton, S., 1864. Notes on animal mechanics, 2, on the muscles of some of the (Orang) and the comparative myology of man and the apes. Proc. Am. Assoc. smaller monkeys of the genera Cercopithecus and Macaca. Proc. R. Ir. Acad. 8, Adv. Sci. 24, 112e144. 467e471. Beattie, J., 1927. The anatomy of the common marmoset (Hapale jacchus Kuhl). Proc. Haughton, S., 1865. Notes on animal mechanics, 7, on the muscular anatomy of the Zool. Soc. Lond 1927, 593e718. Macacus nemestrinus. Proc. R. Irish Acad. 9, 277e287. Begun, D.R., Teaford, M.F., Walker, A., 1994. Comparative functional anatomy of Henkel-Kopleck, A., Schmidt, H.M., 2000. Das Ligamentum metacarpale pollicis. Proconsul phalanges from Kaswanga primate site, Rusinga Island, Kenya. J. Hum. Handchir. Mikrochir. Plast. Chir. 32, 1e8. Evol. 26, 89e166. Hepburn, D., 1892. The comparative anatomy of the muscles and nerves of the Bello-Hellegouarch, G., Aziz, M. A., Ferrero, E. M., Kern, M., Francis, N., Diogo, R. superior and inferior extremities of the anthropoid apes: I - myology of the 'Pollical palmar interosseous muscle' (musculus adductor pollicis accessorius): superior extremity. J. Anat. Physiol. 26, 149e186. attachments, innervation, variations, phylogeny, review of the literature, and Hill, W.C.O., 1953. Primates e Comparative Anatomy and Taxonomy, I, Strepsirhini. implications for human evolution and medicine. Anat. Rec., Submitted. University Press, Edinburgh. Bischoff, T.L.W., 1870. Beitrage zur Anatomie des Hylobates leuciscus and zueiner Hill, W.C.O., 1955. Primates e Comparative Anatomy and Taxonomy, II, Haplorhini: vergleichenden Anatomie der Muskeln der Affen und des Menschen. Abh. Bayer Tarsioidea. University Press, Edinburgh. Akad. Wiss Miinchen Math. Phys. Kl 10, 197e297. Hill, W.C.O., 1957. Primates e Comparative Anatomy and Taxonomy, III, Pithecoidea, Bischoff, T.L.W., 1880. Beitrage zur Anatomie des Gorilla. Abh. Bayer Akad. Wiss Platyrrhini (Families Hapalidae and Callimiconidae). University Press, Miinchen Math. Phys. Kl 13, 1e48. Edinburgh. R. Diogo et al. / Journal of Human Evolution 63 (2012) 64e78 77

Hill, W.C.O., 1959. The anatomy of Callimico goeldii (Thomas): a primitive American Marzke, M.W., Shrewsbury, M.M., 2006. The Oreopithecus thumb: pitfalls in primate. Trans. Am. Phil. Soc. New Ser. 49, 1e116. reconstructing muscle and ligament attachments from fossil bones. J. Hum. Hill, W.C.O., 1960. Primates e Comparative Anatomy and Taxonomy, IV, Cebidae, Evol. 51, 213e215. Part A. University Press, Edinburgh. Marzke, M.W., Toth, N., Schick, K., Reece, S., Steinberg, B., Hunt, K., Linscheid, R.L., Hill, W.C.O., 1962. Primates e Comparative Anatomy and Taxonomy, V, Cebidae, Part An, K.N., 1998. EMG study of hand muscle recruitment during hard hammer B. University Press, Edinburgh. percussion manufacture of Oldowan tools. Am. J. Phys. Anthropol. 105, Hill, W.C.O., 1966. Primates e Comparative Anatomy and Taxonomy, VI, Catarrhini: 315e332. Cercopithecoidea, Cercopithecinae. Interscience Publishers Inc, New York. McMurrich, J.P., 1903a. The phylogeny of the forearm flexors. Am. J. Anat. 2, Hill, W.C.O., 1970. Primates e Comparative Anatomy and Taxonomy, VIII, Cyn- 177e209. opithecinae: Papio, Mandrillus, Theropithecus. University Press, Edinburgh. McMurrich, J.P., 1903b. The phylogeny of the palmar musculature. Am. J. Anat. 2, Hill, W.C.O., 1974. Primates e Comparative Anatomy and Taxonomy, VII, Cyn- 463e500. opithecinae: Cercocebus, Macaca, Cynopithecus. University Press, Edinburgh. Michilsens, F., Vereecke, E.E., D’Août, K., Aerts, P., 2009. Functional anatomy of Howell, A.B., 1936a. Phylogeny of the distal musculature of the pectoral appendage. the gibbon forelimb: adaptations to a brachiating lifestyle. J. Anat. 215, J. Morphol. 60, 287e315. 335e354. Howell, A.B., 1936b. The phylogenetic arrangement of the muscular system. Anat. Miller, R.A., 1943. Functional and morphological adaptations in the forelimbs of the Rec. 66, 295e316. slow lemurs. Am. J. Anat. 73, 153e183. Howell, A.B., Straus, W.L., 1932. The brachial flexor muscles in primates. Proc. U.S. Mivart, S.G., Murie, J., 1865. Observations on the anatomy of Nycticebus tardigradus. Natl. Mus. 80, 1e31. Proc. Zool. Soc. Lond 1865, 240e256. Howell, A.B., Straus, W.L., 1933. The muscular system. In: Hartman, C.G., Straus, W.L. Morrison, P.E., Hill, R.V., 2011. And then there were four: anatomical observations of (Eds.), The Anatomy of the Rhesus Monkey. Williams and Wilkins Co, Baltimore, the pollical palmar interosseous muscle in humans. Clin. Anat. 24, 978e983. pp. 89e175. Moyà-Solà, S., Köhler, M., Rook, L., 1999. Evidence of hominid-like precision grip Huxley, T.H., 1864. The structure and classification of the Mammalia. Med. Times capability in the hand of the Miocene ape Oreopithecus. Proc. Natl. Acad. Sci. 96, Gaz. 1864, 398e468. 313e317. Jacobi, U., 1966. Die Muskulatur des Unterarmes and der Hand bei Macaca mulatta. Murie, J., Mivart, S.G., 1872. On the anatomy of the Lemuroidea. Trans. Zool. Soc. Z. Morphol. Anthropol. 58, 48e73. Lond 7, 1e113. Jouffroy, F.K., 1971. Musculature des membres. In: Grassé, P.P. (Ed.), Traité de Zoo- Napier, J.R., 1962. The evolution of the hand. Sci. Am. 204, 2e9. logie, XVI: 3 (Mammifères). Masson et Cie, Paris, pp. 1e475. Parsons, F.G., 1898. The muscles of mammals, with special relation to human Jouffroy, F.K., Lessertisseur, J., 1960. Les spécialisations anatomiques de la main chez myology: a course of lectures delivered at the Royal College of Surgeons of les singes à progression suspendue. Mammalia 24, 93e151. England e lecture II, the muscles of the shoulder and forelimb. J. Anat. Physiol. Kaneff, A., 1959. Über die evolution des m. abductor pollicis longus und m. extensor 32, 721e752. pollicis brevis. Mateil. Morphol. Inst. Bulg. Akad. Wiss 3, 175e196. Patterson, E.L., 1942. The myology of Rhinopithecus roxellanae and Cynopithecus Kaneff, A., 1968. Zur differenzierung des m. abductor pollicis biventer beim Men- niger. Proc. Zool. Soc. Lond 112, 31e104. schen. Gegenbaurs Morphol. Jahrb. 112, 289e303. Payne, R.C., 2001. Musculoskeletal adaptations for climbing in hominoids and their Kaneff, A., 1969. Umbildung der dorsalen Daumenmuskeln beim Menschen. Verh. role as exaptations for the acquisition of bipedalism. Ph.D. Dissertation, Anat. Ges 63, 625e636. University of Liverpool. Kaneff, A., 1979. Évolution morphologique des musculi extensores digitorum et Polak, C., 1908. Die Anatomie des genus Colobus. Verhandl. Akad. Wet. Amst 14, abductor pollicis longus chez l’homme. I. Introduction, méthodologie, m. 1e247. extensor digitorum. Gegenbaurs Morphol. Jahrb. 125, 818e873. Preuschoft, H., 1965. Muskeln und gelenk der vorderextremitat des gorillas. Morph. Kaneff, A., 1980a. Évolution morphologique des musculi extensores digitorum et Jb 107, 99e183. abductor pollicis longus chez l’homme. II. Évolution morphologique des m. Raven, H.C., 1950. Regional anatomy of the Gorilla. In: Gregory, W.K. (Ed.), The extensor digiti minimi, abductor pollicis longus, extensor pollicis brevis et Anatomy of the Gorilla. Columbia University Press, New York, pp. 15e188. extensor pollicis longus chez l’homme. Gegenbaurs Morphol. Jahrb. 126, Richmond, B.G., Jungers, W.L., 2008. Orrorin tugenensis femoral morphology and the 594e630. evolution of hominin bipedalism. Science 319, 1662e1665. Kaneff, A., 1980b. Évolution morphologique des musculi extensores digitorum et Rolian, C., Lieberman, D.E., Zermeno, J.P., 2011. Hand biomechanics during simulated abductor pollicis longus chez l’ homme. III. Évolution morphologique du m. stone tool use. J. Hum. Evol. 61, 26e41. extensor indicis chez l’homme, conclusion générale sur l ’évolution morpho- Sarmiento, E.E., 1994. Terrestrial traits in the hands and feet of gorillas. Am. Mus. logique des musculi extensores digitorum et abductor pollicis longus chez Novit. 3091, 1e56. l’homme. Gegenbaurs Morphol. Jahrb. 126, 774e815. Schultz, M., 1984. Osteology and myology of the upper extremity of Tarsius. In: Keith, A., 1894a. The myology of the Catarrhini: a study in evolution. Ph.D. Disser- Niemitz, C. (Ed.), Biology of Tarsiers. Gustav Fischer Verlag, Stuttgart, tation, University of Aberdeen. pp. 143e165. Keith, A., 1894b. Notes on a theory to account for the various arrangements of the Senft, M., 1907. Myologie der Vorderextremitäten von Hapale jacchus, Cebus mac- flexor profundus digitorum in the hand and foot of primates. J. Anat. Physiol. 28, rocephalus, und Ateles ater. Ph.D. Dissertation, Bern University. 335e339. Senut, B., Pickford, M., Gommery, D., Mein, P., Cheboi, K., Coppens, Y., 2001. First Keith, A., 1899. On the chimpanzees and their relationship to the gorilla. Proc. Zool. hominid from the Miocene (Lukeino Formation, Kenya). C.R. Acad. Sci., Paris, Soc. Lond 1899, 296e312. Série IIa 332, 137e144. Kimura, K., Tazai, S., 1970. On the musculature of the forelimb of the crab-eating Shoshani, J., Groves, C.P., Simons, E.L., Gunnell, G.F., 1996. Primate phylogeny: monkey. Primates 11, 145e170. morphological vs molecular results. Mol. Phylogenet. Evol. 5, 102e154. Kohlbrügge, J.H.F., 1890. Versuch einer Anatomie des Genus Hylobates. In: Shrewsbury, M.M., Marzke, M.M., Linscheid, R.L., Reece, S.P., 2003. Comparative Weber, M. (Ed.), 1890. Zoologische Ergebnisse Einer Reise in Niederländisch morphology of the pollical distal phalanx. Am. J. Phys. Anthropol. 121, 30e47. Ost-Indien, vol. 1. Verlag von EJ Brill, Leiden, pp. 211e354, pp. 138e208 (Vol. 2). Sonntag, C.F., 1924a. On the anatomy, physiology, and pathology of the orang-outan. Landsmeer, J.M., 1986. A comparison of fingers and hand in Varanus, opossum and Proc. Zool. Soc. Lond 24, 349e450. primates. Acta Morphol. Neerl. Scand. 24, 193e221. Sonntag, C.F., 1924b. The Morphology and Evolution of the Apes and Man. John Bale Lewis, W.H., 1910. The development of the muscular system. In: Keibel, F., Mall, F.P. Sons and Danielsson Ltd, London. (Eds.), 1910. Manual of Human Embryology, vol. 1. Lippin-Cott, Philadelphia, Stout, K., 2000. Grip types and associated morphology of the hylobatid thumb and pp. 454e522. index finger. M.A. thesis, Arizona State University. Lewis, O.J., 1989. Functional Morphology of the Evolving Hand and Foot. Clarendon Straus, W.L., 1941a. The phylogeny of the human forearm extensors. Hum. Biol. 13, Press, Oxford. 23e50. Lindburg, R.M., Comstock, B.E., 1979. Anomalous tendon slips from the flexor pol- Straus, W.L., 1941b. The phylogeny of the human forearm extensors (concluded). licis longus to the digitorum profundus. J. Hand Surg. 4, 79e83. Hum. Biol. 13, 203e238. Linscheid, R.L., An, K.-N., Gross, R.M., 1991. Quantitative analysis of the intrinsic Straus, W.L., 1942a. The homologies of the forearm flexors: urodeles, lizards, muscles of the hand. Clin. Anat. 4, 265e284. mammals. Am. J. Anat. 70, 281e316. Lorenz, R., 1974. On the thumb of the Hylobatidae. In: Rumbaugh, D.M. (Ed.), 1974. Straus, W.L., 1942b. Rudimentary digits in primates. Q. Rev. Biol. 17, 228e243. Gibbon and Siamang, vol. 3. Karger, Basel, pp. 157e175. Susman, R.L., 1994. Fossil evidence for early hominid tool use. Science 265, Loth, E., 1931. Anthropologie des Parties Molles (Muscles, Intestins, Vaisseaux, Nerfs 1570e1573. Peripheriques). Mianowski-Masson et Cie, Paris. Susman, R.L., 1998. Hand function and tool behavior in early hominids. J. Hum. Evol. Lovejoy, C.O., Simpson, S.W., White, T.D., Asfaw, B., Suwa, G., 2009. Careful climbing 35, 23e46. in the Miocene: the forelimbs of Ardipithecus ramidus and humans are primi- Susman, R.L., Nyati, L., Jassal, M.S., 1999. Observations on the pollical palmar tive. Science 326, 1e8. interosseus muscle (of Henle). Anat. Rec. 254, 159e165. Macalister, A., 1873. The muscular anatomy of the gorilla. Proc. R. Irish Acad. Ser. 2, Swindler, D.R., Wood, C.D., 1973. An Atlas of Primate Gross Anatomy: Baboon, 501e506. Chimpanzee and Men. University of Washington Press, Seattle. MacDowell, E.C., 1910. Notes on the myology of Anthropopithecus niger and Papio- Tocheri, M.W., Orr, C.M., Jacofsky, M.C., Marzke, M.W., 2008. The evolutionary thoth ibeanus. Am. J. Anat. 10, 431e460. history of the hominin hand since the last common ancestor of Pan and Homo. Marzke, M.W.,1992. Evolutionary development of the human thumb. Hand Clin. 8,1e8. J. Anat. 212, 544e562. Marzke, M.W., 1997. Precision grips, hand morphology, and tools. Am. J. Phys. Tuttle, R.H., 1969. Quantitative and functional studies on the hands of the Anthropol. 102, 91e110. Anthropoidea, I, the Hominoidea. J. Morphol. 128, 309e363. 78 R. Diogo et al. / Journal of Human Evolution 63 (2012) 64e78

Tuttle, R.H., 1970. Postural, propulsive and prehensile capabilities in the cheiridia of Williams, E.M., Gordon, A.D., Richmond, B.G., 2012. Manual pressure distribution chimpanzees and other great apes. In: Bourne, G.H. (Ed.), 1970. The Chim- during Oldowan stone tool production. J. Hum. Evol. 62, 520e532. panzee, vol. 2. Karger, Basel, pp. 167e263. Wood, B., Harrison, T., 2011. The evolutionary context of the ﬁrst hominins. Nature Van Horn, R.N., 1972. Structural adaptations to climbing in the gibbon hand. Am. 470, 347e352. Anthropol. New Ser. 74, 326e334. Wood Jones, F., 1920. The Principles of Anatomy as Seen in the Hand. J. & A. Ward, C.V., Kimbel, W.H., Harmon, E.H., Johanson, D.C. New postcranial fossils of Churchill, London. Australopithecus afarensis from Hadar, Ethiopia (1990e2007). J. Hum. Evol, in Woollard, H.H., 1925. The anatomy of Tarsius spectrum. Proc. Zool. Soc. Lond 70, press. 1071e1184.