at SciVerse ScienceDirect

Journal of Human Evolution 63 (2012) 64e78

Contents lists available

Journal of Human Evolution

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

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

Rui Diogo a,*, Brian G. Richmond b,c, Bernard Wood b,c

aDepartment of Anatomy, Howard University College of Medicine, 520 W St. NW, Washington, DC 20059, USAbCenter for the Advanced Study of Hominid Paleobiology, Department of Anthropology, George Washington University, DC 20052, USAcHuman Origins Program, National Museum of Natural History, Smithsonian Institution, Washington, DC 20560, USA

a r t i c l e i n f o

Article history:Received 21 January 2012Accepted 5 April 2012Available online 27 May 2012

Keywords:Phylogenetic analysisMorphologyExtensor pollicis brevisFlexor pollicis longusAdductor pollicis accessorius

* Corresponding author.E-mail addresses: rui.diogo@howard.edu, Rui_Diog

0047-2484/$ e see front matter � 2012 Elsevier Ltd.doi:10.1016/j.jhevol.2012.04.001

a b s t r a c t

In this paper, we explore how the results of a primate-wide higher-level phylogenetic analysis of musclecharacters can improve our understanding of the evolution and homologies of the forearm and handmuscles of modern humans. Contrary to what is often suggested in the literature, none of the forearmand hand muscle structures usually present in modern humans are autapomorphic. All are found in oneor more extant non-human primate taxa. What is unique is the particular combination of muscles.However, more muscles go to the thumb in modern humans than in almost all other primates, rein-forcing the hypothesis that focal thumb movements probably played an important role in humanevolution. What makes the modern human thumb myology special within the primate clade is not somuch its intrinsic musculature but two extrinsic muscles, extensor pollicis brevis and flexor pollicislongus, that are otherwise only found in hylobatids. It is likely that these two forearm muscles playdifferent functional roles in hylobatids and modern humans. In the former, the thumb is separated fromelongated digits by a deep cleft and there is no pulp-to-pulp opposition, whereas modern humansexhibit powerful thumb flexion and greater manipulative abilities, such as those involved in themanufacture 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 wesuggest is referred to as the musculus adductor pollicis accessorius) is still obscure. The presence ofdistinct contrahentes digitorum and intermetacarpales in adult chimpanzees is likely the result of pro-longed or delayed development of the hand musculature of these apes. In relation to these structures,extant chimpanzees are more neotenic than modern humans.

� 2012 Elsevier Ltd. All rights reserved.

Introduction

An understanding of comparative myology is crucial for devel-oping hypotheses about the functional morphology of the modernhuman forearm and hand, and particularly their involvement in themanufacture and use of tools (e.g., Day and Napier, 1961, 1963;Napier, 1962; Tuttle, 1969; Lewis, 1989; Marzke, 1992, 1997;Susman, 1994, 1998; Marzke et al., 1998; Susman et al., 1999;Tocheri et al., 2008). The relatively few publications that haveincluded myological information in discussions of the evolution ofthe primate and modern human forearm and hand can be dividedinto two groups. Publications prior to the 1950s were mainly thework of comparative vertebrate, tetrapod or mammalian anato-mists who, with the exception of authors such as Howell and Straus

o@hotmail.com (R. Diogo).

All rights reserved.

(1932) and Straus (1941a, b), usually investigated a wide range ofnon-primate species but only a few primate taxa (e.g., Brooks, 1886;Parsons, 1898; McMurrich, 1903a, b; Forster, 1917; Howell, 1936a, b;Haines, 1939, 1946, 1950; Straus, 1942a, b). This pattern contrastswith most of the reports published since the 1950s that havemainly focused on primates, or even more narrowly, on hominoidsor on modern humans (Hominina: see Fig. 1), and when they didprovide a non-primate comparative context it was a relativelynarrow one (e.g., Abramowitz, 1955; Day and Napier, 1963;Dylevsky, 1967; Tuttle, 1969; Dunlap et al., 1985; Aziz and Dunlap,1986; Susman, 1994, 1998; Marzke et al., 1998; Susman et al.,1999). A notable exception is Lewis' (1989) book that includesdetailed information based on his own dissections of a wide rangeof non-primate tetrapods.

These studies have generated hypotheses regarding the evolu-tion of primate forearm and hand anatomy, among them thatmodern humans are derived relative to other extant primates inpossessing a true flexor pollicis longus, a deep head of flexor pollicis

STREPSIRRHINI

LEMURIFORMES

LORISIFORMES

HAPLORRHINI

ANTHROPOIDEA

PLATYRRHINI

CEBIDAE+AOTIDAECEBIDAE

CATARRHINI

HOMINOIDEA

HOMINIDAE

HOMININAEHOMININI

Tarsius (Tarsiiformes; Tarsiidae)

Pithecia (Pitheciidae)Aotus (Aotidae)

Hylobates (Hylobatidae)Pongo (Ponginae)

Gorilla (Gorillini)Pan (Panina)Homo (Hominina)

Lemur (Lemuridae)Propithecus (Indriidae)Loris (Lorisidae)

Nycticebus (Lorisidae)

PRIMATES

Saimiri (Saimiriinae)Callithrix (Callitrichinae)

CERCOPITHECIDAE

CERCOPITHECINAE

PAPIONINI

Colobus (Colobinae)Cercopithecus (Cercopithecini)

Papio (Papionina)

Macaca (Macanina)

1,2

3,4,5

6

7,8,9

10,11,12,13 14,15,

16,1718,19,20

21,22,23,24,25

Figure 1. Single most parsimonious primate tree (L 301, CI 58, RI 73) obtained from thecladistic analysis of 166 characters of the head, neck, pectoral and forelimb muscula-ture (Diogo and Wood, 2011, 2012), showing the unambiguous transitions (homo-plasic: regular font; non-homoplasic: bold) of the forearm and hand musculature thatoccurred in the branches leading to modern humans: 1) Opponens pollicis is a distinctmuscle (this muscle became secondarily undifferentiated in Callithrix); 2) Opponensdigiti minimi is a distinct muscle (also found in a few other mammals, e.g., Rattus); 3)Flexor carpi radialis inserts onto the metacarpals II and III (also found in a few othermammals, e.g., Tupaia); 4) There are more than two contrahentes digitorum (revertedin some haplorrhines, e.g., modern humans); 5) Adductor pollicis has slightly differ-entiated transverse and oblique heads; 6) Supinator has ulnar and humeral heads(independently acquired in Lorisiformes); 7) Adductor pollicis further differentiatedinto distinct transverse and oblique heads; 8) Opponens pollicis reaches the distalportion of metacarpal I (feature independently acquired in Lorisiformes); 9) Opponensdigiti minimi is slightly differentiated into superficial and deep bundles; 10) Flexordigitorum superficialis originates from the radius; 11) Flexor digitorum superficialisoriginates from the ulna; 12) Epitrochleoanconeus is not a distinct muscle (this musclebecame undifferentiated in Lorisiformes and in hominoids, and then becamesecondarily differentiated in Pan); 13) Pronator teres often (but not usually, i.e., in<50% of the cases) originates from the ulna; 14) Flexor digitorum profundus is notoriginated from the medial epicondyle of the humerus or from the common flexortendon (feature also found in Macaca); 15) Tendon of flexor digitorum profundus todigit 1 is vestigial or absent (this feature was independently acquired in Colobus and itwas secondarily reverted - i.e., the tendon is not atrophied - in modern humans); 16)Pronator teres is usually (�50% of the cases) originated from the ulna; 17) Contra-hentes digitorum are missing (this feature was secondarily reverted in Pan, i.e., aloneamong the Hominidae, Pan has distinct contrahentes digitorum); 18) Palmaris longusis absent in >5% of the cases; 19) Adductor pollicis accessorius (or ‘interosseous volarisprimus of Henle’ of modern human anatomy) is often (i.e., in <50% of the cases)present; 20) Extensor indicis usually inserts onto digit 2 only; 21) Flexor pollicis longusis a distinct muscle (independently acquired in hylobatids); 22) Reversion of ‘Tendon offlexor digitorum profundus to digit 1 is vestigial or absent’ (see above); 23) Reversionof ‘Flexor carpi radialis originates from the radius’ (the muscle originates from theradius in gorillas, chimpanzees and orangutans, but it is not clear which is the usualcondition for hylobatids, and, thus, whether a radial origin constitutes a synapomorphyof hominoids or of hominids); 24) Adductor pollicis accessorius (or ‘interosseousvolaris primus of Henle’ of modern human anatomy) is usually (i.e., in >50% of thecases) present; 25) Extensor pollicis brevis is a distinct muscle (independentlyacquired in hylobatids) [for more details, see text].

R. Diogo et al. / Journal of Human Evolution 63 (2012) 64e78 65

brevis, a volar interosseous of Henle, and an extensor pollicis brevis.However, with a few noteworthy exceptions, the authors of most ofthese publications relied on reports in the literature for theircomparative information. This is problematic because the

nomenclature of the forearm and hand muscles of primate andnon-primate tetrapod taxa has been the subject of some confusionand controversy (see Diogo, 2007, 2009; Diogo and Abdala, 2007,2010; Diogo et al., 2009; Diogo andWood, 2011, 2012). For example,statements that some non-hominoid primates have a separateflexor pollicis longus (e.g., Marzke, 1997; Shrewsbury et al., 2003)refer to older publications that often used the name ‘flexor pollicislongus’ for the tendon of the flexor digitorum profundus thatinserts onto digit 1 and not for a separate flexor pollicis longus thathas a distinct belly and tendon (see below).

In recent papers (Diogo and Abdala, 2007; Diogo et al., 2009;Abdala and Diogo, 2010; Diogo and Wood, 2011) and monographs(Diogo, 2007; Diogo and Abdala, 2010; Diogo and Wood, 2012),Diogo and colleagues have reported the results of their long-termdissection-based study of the comparative anatomy, homologiesand evolution of the pectoral and forelimb muscles of all majorgroups of non-primate vertebrates and representatives of all of themajor primate higher taxa. Diogo andWood (2011, 2012) combineddata from these dissections with carefully validated informationfrom the literature to undertake the first comprehensive parsimonyand Bayesian cladistic analyses of the order Primates based onmyological data (Fig. 1) and for a range of outgroups (tree-shrews,dermopterans and rodents). A goal of the project was to establishthe homologies of the pectoral and forelimb muscles of vertebratesand to provide the comparative context for more detailed evolu-tionary and taxon-based analyses.

The present paper uses data from these dissections to testhypotheses about the evolution of the handmusculature of modernhumans. Because we have dissected members of all of the majorprimate groups and the major non-primate vertebrate taxa, we canbetter understand the descriptions of authors that use differentnomenclatures for these muscles. This has allowed us to clarifymuch of the nomenclatural confusion involved in the descriptionsof the forearm and hand muscles of primates (see above). We alsouse these data to critically examine the proposition that modernhumans are derived in having a flexor pollicis longus, deep head offlexor pollicis brevis, a ‘volar interosseous’ or ‘interosseous volarisprimus’ of Henle, and an extensor pollicis brevis.

Materials and methods

We dissected representatives of each major extant non-hominoid primate clade (Strepsirrhini, Tarsiiformes, New Worldmonkeys and Old World monkeys) and of each of the five maingroups of living hominoids (i.e., hylobatids, orangutans, gorillas,chimpanzees, and modern humans) (Fig. 1). With a few exceptions,we use the same taxonomic nomenclature as Diogo and Wood(2011). Data included in the analysis come from four strepsirrhinegenera, two from the infraorder Lemuriformes (Lemur, familyLemuridae; Propithecus, family Indriidae) and two from the infra-order Lorisiformes (Loris and Nycticebus, family Lorisidae); thesingle extant genus of the infraorder Tarsiiformes, Tarsius; repre-sentatives of three of the four extant platyrrhine families: Saimiriand Callithrix (Cebidae, subfamilies Saimiriinae and Callithrichinae,respectively), Pithecia (Pitheciidae), and Aotus (Aotidae; the otherplatyrrhine family being the Atelidae); the two extant subfamiliesof Old World monkeys (family Cercopithecidae) represented byColobus (Colobinae), Papio, Macaca and Cercopithecus (Cercopithe-cinae; the two former genera represent the tribe Papionini, whilethe latter genus represents the other extant tribe of the subfamily,the Cercopithecini); and five extant hominoid genera: Hylobates(Hylobatidae), Pongo (Hominidae, Ponginae), Gorilla (Hominidae,Homininae, Gorillini), Pan (Hominidae, Homininae, Panini, Panina),and Homo (Hominidae, Homininae, Hominini, Hominina) (Fig. 1).We use the traditional classification that recognizes a single extant

R. Diogo et al. / Journal of Human Evolution 63 (2012) 64e7866

hylobatid genus (Hylobates; including species such as Hylobatessyndactylus, Hylobates lar, Hylobates gabriellae and Hylobates hoo-lock, among others: for more details see Diogo and Wood, 2011).Apart from the primates dissected for this study, we have dissectedspecimens from all of the major groups of vertebrates. A list of thedissected non-primate vertebrate specimens is given in Diogo andAbdala (2010).

The nomenclature for the forearm and hand muscles followsthat of Diogo and Abdala (2010) (see Table 1). We addressed theproblem of inconsistent usage of some specific taxonomic names inolder systematic and anatomical studies, particularly those prior toOsman Hill’s studies in the 1950s (e.g., Hill, 1953, 1955, 1957, 1959,1960, 1962, 1966, 1970, 1974; see References) by carefully reviewingall of the names and synonyms used in those studies (see Diogo andWood, 2011). Discussions of the evolutionary changes occurring ineach of the major primate branches are based on the phylogeneticresults of Diogo and Wood (2011) (Fig. 1). The primate specimensweremainly dissected by RD, andwere obtained from the followinginstitutions: the Primate Foundation of Arizona (PFA), the Depart-ment of Anatomy (GWUANA) and the Department of Anthropology(GWUANT) of the George Washington University, the Departmentof Anatomy of Howard University (HU), the Department ofAnatomy of Valladolid University (VU), the Cincinnati Museum ofNatural History (CMNH), the San Diego Zoo (SDZ) and the CanadianMuseum of Nature (CMN). For each taxon, we provide the Linneanbinomial, its source, its unique identifier, the number of specimensfrom that source and the state of the specimens. Regarding thesample size used in the cladistic study of Diogo and Wood (2011)and in the discussions of the present paper, two points should bestressed. First, it is difficult to find primate, and particularly ape,specimens in circumstances where careful dissection can takeplace. During this project, we made a considerable effort to estab-lish connections with the major museums and zoos in the UnitedStates and beyond. This effort resulted in us being able to dissect,for example, two fresh gorillas, one fresh and one formalinembalmed Pongo, and six fresh and three formalin embalmedchimpanzees. The second point is that the sample size used in thatcladistic study and in the discussions provided here refers to thespecimens dissected by us and the total number of specimensreported in the numerous publications that were analysed in theextensive review of the literature undertaken by Diogo and Wood(2011) and particularly Diogo and Wood (2012). Thus, when wecode and discuss each character we take into account all of theinformation available, and in numerous cases the total sample sizeis substantial when compared with cladistic studies of other animalspecies that were based on muscles (see, e.g., Diogo, 2007). Forexample, for character. 118 (the presence/absence of the palmarislongus) Diogo and Wood (2011) took into account informationobtained from dissections of more than 20 hylobatid, 19 orangutan,25 gorilla, and 39 chimpanzee specimens. For this character, thesample size just for apes was >103 specimens.

Primate specimens dissected

Aotus nancymaae: GWUANT AN1, 1 (fresh; adult female). Calli-thrix jacchus: GWUANT CJ1, 1 (fresh; adult male). Cercopithecusdiana: GWUANT CD1, 1 (fresh; adult female). Colobus guereza:GWUANT CG1, 1 (fresh; adult male). Gorilla gorilla: CMS GG1, 1(fresh; adult male); VU GG1, 1 (fresh; adult female). Homo sapiens:GWUANA HS1-16, 16 (formalin); HU D43, 1 (formalin); HU D45(formalin). Hylobates gabriellae: VU HG1, 1 (fresh; infant male).Hylobates lar: HU HL1, 1 (formalin; adult male). Lemur catta:GWUANT LC1, 1 (fresh; adult male). Loris tardigradus: SDZ LT53090,1 (fresh; adult male). Macaca fascicularis: VU MF1, 1 (fresh; adultmale). Macaca mulatta: HU MM1, 1 (formalin; adult male). Macaca

silenus: VU MS1, 1 (fresh; adult male). Nycticebus coucang: SDZNC41235, 1 (fresh; adult female); SDZ NC43129, 1 (fresh; adultfemale). Nycticebus pygmaeus: VU NP1, 1 (fresh; adult female); VUNP2, 1 (fresh; adult male); SDZ NP40684, 1 (fresh; adult female);SDZ NP51791, 1 (fresh; adult female). Pan troglodytes: PFA 1016, 1(fresh; adult female); PFA 1009, 1 (fresh; adult female); PFA 1051, 1(fresh; infant female); PFA 1077, 1 (fresh; infant female); PFA UNC(uncatalogued), 1 (fresh; infant male); HU PT1, 1 (formalin; infantmale); GWUANT PT1, 1 (formalin; adult female); GWUANT PT2, 1(formalin; adult female); VU PT1, 1 (fresh; adult male). Papio anubisGWUANT PA1, 1 (fresh; adult female). Pithecia pithecia: VU PP1, 1(fresh; adult male); GWUANT PP1, 1 (fresh; adult female). Pongopygmaeus: HU PP1, 1 (formalin; neonate male); GWUANT PP1, 1(formalin; adult male). Propithecus verreauxi: GWUANT PV1, 1(fresh; adult female); GWUANT PV2,1 (fresh; infant female). Saimirisciureus: GWUANT SC1, 1 (fresh; adult female). Tarsius syrichta:CMNH M-3135, 1 (alcohol; adult female).

Results

For each group of muscles (ventral forearm muscles, dorsalforearmmuscles and intrinsic handmuscles), we discuss the resultsworking from the more inclusive synapomorphic features (e.g.,Primates, Haplorhini, Anthropoidea, etc.) to those apomorphiesthat are only found within H. sapiens (Fig. 1). Detailed tables thatdescribe and use photographs to illustrate each muscle of each ofthe taxa included in the cladogram of Fig. 1 are included in Diogoand Wood (2012). Detailed descriptions of the phylogenetic char-acters used and of the synapomorphies obtained in our phyloge-netic analyses are given there and in Diogo and Wood (2011).

Ventral forearm musculature

Among the synapomorphic features of the ventral forearmmuscles, the most inclusive is the insertion of the flexor carpiradialis onto metacarpals II and III (e.g., Fig. 2), which was acquiredin the Haplorhini (Fig. 1, feature 3). In most non-primate eutherianmammals, the flexor carpi radialis inserts onto metacarpal III (e.g.,in rats), onto metacarpal II (as is usually the case in e.g., Lemur andPropithecus), or, in a few cases (e.g., in flying lemurs), onto otherstructures, but it does not attach onto both metacarpals II and III, asit usually does in haplorrhines (and in a few non-primatemammals, e.g., Tupaia). Two of the synapomorphies of hominoidsconcern the flexor digitorum superficialis. In non-hominoidprimates, this muscle originates exclusively from the arm (medialepicondyle of humerus, common flexor tendon, and/or capsule ofthe elbow joint), but in hominoids the muscle usually also origi-nates from the ulna and the radius (Fig. 1, features 10, 11). Anothersynapomorphy of hominoids is the loss of the epitrochleoanconeus,a small muscle that connects the medial epicondyle of the humerusto the olecranon process of the ulna. This muscle is usually present(presumably due to a secondary reversion) in chimpanzees (Fig. 1,feature 12). Plesiomorphically in primates the pronator teres hasa bony origin from the humerus only. In hylobatids, this muscleoriginates from the humerus and often, but not usually (i.e., <50%of the cases), from the ulna (Fig. 1, feature 13). In hominids, themuscle originates from the humerus and usually (i.e., �50% of thecases) also from the ulna (Fig. 1, feature 16).

Plesiomorphically, in primates the flexor digitorum profundusoriginates from the forearm (ulna, radius, and/or interosseousmembrane) and arm (medial epicondyle and/or common flexortendon). In hominids (and also in Macaca as a homoplasy), thismuscle does not originate from the medial epicondyle nor from thecommon flexor tendon (Fig. 1, feature 14). In our dissections, theflexor pollicis longus was present in modern humans and

Table 1Scheme 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 numerousother mammals.

Muscle names Lemur

(49 mus.: 19

forearm; 30 hand)

Tarsius

(51-55 mus.: 19

forearm; 32-36 hand)

Aotus

(41 mus.: 19

forearm; 22 hand)

Macaca

(46 mus.: 19 forearm;

27 hand)

Hylobates

(46 mus.: 19

forearm; 27 hand)

Pongo

(38 mus.: 18

forearm; 20 hand)

Gorilla

(38 mus.: 18

forearm; 20 hand)

Pan

(45 mus.: 19 forearm;

26 hand)

Homo

(41 mus.: 20

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

Epitrochleoanconeus P P P P --- --- --- P ---

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

APP

END

.: VEN

. FO

REA

RM

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

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 ---

APP

END

.: VEN

. AN

D D

OR

SAL

HA

ND

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

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

APP

END

.: DO

RSA

L FO

REA

RM

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.

R. Diogo et al. / Journal of Human Evolution 63 (2012) 64e78 67

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

Figure 2. Hylobates lar (HU HL1, adult male): ventral view of the left forearmmusculature. Note the distinct muscle flexor pollicis longus. In this and the followingfigures, PRO, MED, LAT and DIS mean proximal, medial, lateral and distal, respectively.

R. Diogo et al. / Journal of Human Evolution 63 (2012) 64e7868

hylobatids (Table 1, Fig. 1, feature 21; Figs. 2e4), confirmingobservations by previous researchers (e.g., Deniker, 1885;Hartmann, 1886; Kohlbrügge, 1890-1892; Hepburn, 1892; Keith,1894b; Chapman, 1900; McMurrich, 1903a, b; Sonntag, 1924b;Howell, 1936a, b; Straus, 1942a; Jouffroy and Lessertisseur, 1960;Tuttle, 1969; Jouffroy, 1971; Van Horn, 1972; Lorenz, 1974; Marzke,1992, 1997; Susman, 1994, 1998; Stout, 2000; Tocheri et al., 2008).This is a potential example of parallelism in that from a similarancestral configuration (i.e., the flexor digitorum profundus goingto the distal phalanges of digits 1e5), there is an independentacquisition of a similar derived feature (differentiation of the bellyof the flexor digitorum profundus going to digit 1, to form a sepa-rate flexor pollicis longus muscle). The belly of the flexor pollicislongus is distinct from the flexor digitorum profundus in all of themodern humans and the two hylobatids dissected by us (Fig. 2), butin the H. gabriellae (VU HG1) male infant there is a distal tendinousconnection between the tendon of the former muscle and thetendon of the latter muscle that goes to digit 2 (Fig. 3; N.B., Fig. 2shows some tissue that appears to run distally from the mostlateral flexor digitorum profundus tendon to the flexor pollicislongus tendon just proximal to the wrist, but careful dissection

established that there is no fleshy or tendinous connection betweenthese twomuscles in the photographed specimen). This connectionreflects the ancestral, and probably embryonic, condition wherethese two muscles form a single structure (Diogo andWood, 2011).In the 18 modern humans dissected by us, we did not see as stronga tendinous distal connection between the two muscles as wefound in the gibbon infant, but according to Lindburg and Comstock(1979) 31% of modern humans display some type of connectionbetween these two muscles (see below). In the great apes, thetendon of the flexor digitorum profundus to digit 1 is often eithera very thin, vestigial structure or it is absent (Fig. 5). According tothe results of our cladistic analysis, it is more parsimonious to inferthat this tendon became reduced in the last common ancestor(LCA) of hominids and became fully developed again in the Hom-inina (two steps) than to conclude that it became reduced inde-pendently in orangutans, gorillas and chimpanzees (three steps)(Fig. 1, features 15 and 21). However, in this case there are reasonsto suggest that the cladistically less parsimonious hypothesis maybe the most likely evolutionary hypothesis (see Discussion below).

Apart from the presence of a distinct flexor pollicis longus,modern humans also show a reversion of the character state ‘Flexorcarpi radialis originates from the radius’ (Fig. 1, feature 23). Unlikethe condition in modern humans (i.e., the only bony origin is fromthe medial epicondyle of the humerus), in great apes the flexorcarpi radialis muscle also originates from the radius. However, it isnot clear which is the usual condition for hylobatids, and, thus,whether a radial origin constitutes a synapomorphy of hominoidsor of hominids. Hepburn (1892) and Kohlbrügge (1890-1892) noteda partial origin of flexor carpi radialis from the radius in varioushylobatid species, as we did in our H. gabriellae and H. lar speci-mens, but other authors (e.g., Michilsens et al., 2009) have reportedan exclusive origin from the humerus in species such as H. pileatus,H. moloch and H. syndactylus.

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 toinnervate the dorsal and palmar interossei, while other branches run mainly laterally to innervate the oblique head of the adductor pollicis and then the adductor pollicisaccessorius (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 dissectedexhibiting 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 thepresent 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 flexorpollicis 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 adductorpollicis accessorius from the branches of the deep branch of the ulnar nerve that innervate the oblique head of the adductor pollicis.

R. Diogo et al. / Journal of Human Evolution 63 (2012) 64e78 69

All of the non-Homininae (non-African ape andmodern human)primates, as well as the non-primate mammals dissected by us,have a palmaris longus (e.g., Fig. 2). In gorillas, chimpanzees andmodern humans, the palmaris longus is missing in at least 5% of thecases (Fig. 1, feature 18). The reviews of Keith (1899), Loth (1931),Sarmiento (1994) and Gibbs (1999) indicate that a palmaris longusis present in 64%, 15%, 36% and 31% of gorillas, respectively, and in75%, 95%, 91% and 68% of chimpanzees, respectively, and morerecent literature indicates that the muscle is present in 85% ofmodern humans (e.g., Gibbs, 1999; Gibbs et al., 2000, 2002).

Dorsal forearm musculature

With respect to the dorsal forearm muscles, an inclusive syna-pomorphy of anthropoids is the presence of both ulnar andhumeral heads of the supinator (Fig. 1, feature 6). The

plesiomorphic state in mammals (e.g., Tupaia, Rattus and Cyn-ocephalus) and in primates (e.g., Lemur, Propithecus and Tarsius) isa single humeral head. Lorisiformes have independently acquiredan ulnar head and thus they resemble anthropoids in that they haveboth humeral and ulnar heads. The other two forearm extensorapomorphies are much less inclusive. One, which characterizes theHomininae, is the exclusive insertion of the extensor indicis ontodigit 2 (Fig. 1, feature 20). In other primates and in most othermammals, this muscle usually goes to more than one digit (e.g.,Fig. 7; N.B., other authors use names such as ‘extensor indicis ettertius’ when the muscle goes e.g., to digits 2 and 3, but we followthe nomenclature of Diogo and Wood, 2012, who pointed out thatthe muscle should still be designated as extensor indicis in thosecases in order to stress the homology of the muscle, whether it goesto a single digit or not). The other forearm extensor apomorphy,which characterizes the extant Hominina (and also the

Figure 5. Gorilla gorilla (CMS GG1, adult male): ventral view of the oblique andtransverse heads of the right adductor pollicis, as well as of the adductor pollicisaccessorius (or ‘interosseous volaris primus of Henle’ of modern human anatomy).

Figure 6. Pan troglodytes (GWUANT PT2, adult female): ventral view of the rightflexores breves profundi 3e9 (pulled back) and the intermetacarpales 1e4; contrary toother hominoids, in Pan the flexores breves profundi 3, 5, 6 and 8 and the inter-metacarpales 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 tothe interossei palmares 1, 2 and 3).

R. Diogo et al. / Journal of Human Evolution 63 (2012) 64e7870

Hylobatidae, likely due to an independent acquisition; Table 1 andFig. 1, feature 25) is the presence of a distinct extensor pollicisbrevis (e.g., Fig. 7). The presence of this muscle in both modernhumans and the Hylobatidae is another example of a parallelism inthat a similar ancestral configuration (i.e., the abductor pollicis hastwo distal tendons in most other primates; see Fig. 8 and below)results in a similar, homoplasic, derived configuration (inwhich theabductor pollicis longus gives rise to a new distinct muscle, theextensor pollicis brevis: see below).

Intrinsic hand musculature

Regarding the intrinsic handmuscles, inclusive synapomorphiesinclude the presence of an opponens pollicis and an opponens digitiminimi (Table 1) that were acquired in the node leading to thePrimates (Fig. 1, features 1, 2). The opponens pollicis becamesecondarily undifferentiated in a few primates (e.g., Callithrix) andone, or both, of these muscles may be present in some non-primatemammals such as rats. Strepsirrhines and most non-primatemammals have an undivided adductor pollicis. In the nodeleading to haplorrhines, this muscle became partly divided intotransverse and oblique heads (Fig.1, feature 5), and then in the nodeleading to catarrhines the muscle became further divided intowell-separated transverse and oblique heads (Fig. 1, feature 7) (e.g.,Figs. 4and 5). Another synapomorphy of haplorrhines is the pres-ence of more than two contrahentes digitorum (i.e., other than theadductor pollicis, which derives evolutionary from the ancestralcontrahens to digit 1; Table 1, Fig. 1, feature 4). The plesiomorphiccondition for primates (e.g., Lemur, Propithecus, Loris andNycticebus)

as well as in non-primate taxa such as Rattus, Tupaia and Cyn-ocephalus, is to have only two contrahentes digitorum (usually todigits 2 and 5; Table 1). A synapomorphy of the Hominidae is theabsence of contrahentes digitorum (Fig. 1, feature 17), but in thenode leading to Pan there was a reversion to the plesiomorphicprimate condition (i.e., adult chimpanzees usually have two con-trahentes digitorum, but unlike in strepsirrhines these musclesusually go to digits 4 and 5: Table 1; N.B., the chimpanzees dissectedby us have contrahentes, although in some cases the fleshy part ofthese muscles was markedly reduced: for more details see, e.g.,Diogo and Wood, 2012, in press). Another reversion that differen-tiates Pan from the othermembers of theHominidae, aswell as fromhylobatids and New World monkeys is the presence of four inter-metacarpales (i.e., the flexor brevis profundi and intermetacarpalesdo not fuse to form the dorsal interossei) each connecting adjacentmetacarpals (Table 1, Fig. 6; see Discussion below).

Apart from the presence of well-separated oblique and trans-verse heads of the adductor pollicis (see above) catarrhines arecharacterized by two synapomorphies. First, contrary to the ple-siomorphic primate condition in which the opponens pollicisinserts onto the proximal and/or middle surfaces of the metacarpalI, in catarrhines (and homoplasically also in the Lorisiformes) theopponens pollicis reaches the distal margin of this bone (Fig. 1,feature 8). Second, in catarrhines the opponens digiti minimi isdivided into superficial and deep bundles, while in other primatesandmost othermammals thismuscle is undivided (Fig.1, feature 9).

According to our comparative and phylogenetic analyses, one ofthe synapomorphies of the Homininae (African great apes andmodern humans) is that the ‘volar interosseus of Henle’ (or ‘inter-osseous volaris primus of Henle’), which very likely derives from

Figure 7. Hylobates lar (HU HL1, adult male): dorsal view of the left forearm muscu-lature; note the distinct muscle extensor pollicis brevis.

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

R. Diogo et al. / Journal of Human Evolution 63 (2012) 64e78 71

a thin deep additional slip of the adductor pollicis (Diogo andWood, 2011), is often (c. < 50% of cases) present (Fig. 1, feature19), whereas in other primates and other mammals this muscle isnearly always, or always, absent. This synapomorphic Homininaecondition where the muscle is often but not usually present (i.e.,c. < 50% of cases) is seen in chimpanzees and gorillas (e.g., Fig. 5).Modern humans display an autapomorphic condition within theHomininae because themuscle is found inmost (�50%) of the cases(see Table 1; Fig. 1, feature 24; Fig. 4).

Discussion

Four forearm and hand muscles (flexor pollicis longus, the so-called ‘deep head of the flexor pollicis brevis’, the so-called ‘volar

interosseous of Henle’ and the extensor pollicis brevis) have beenconsidered to be uniquely, or almost uniquely, found in modernhumans. Below we provide a discussion of these structures and ofthe other forearm and hand structures to which they are anatom-ically and functionally associated, and consider their functional andevolutionarily significance in light of evidence from the fossilrecord and from biomechanical analyses.

Flexor pollicis longus

In modern humans, the flexor pollicis longus is innervated bythe anterior interosseous nerve and it usually runs from the ventralsurface of the radius and the interosseous membrane to the distalphalanx of the thumb. The different nomenclatures used todescribe this muscle have resulted in some confusion. Day andNapier (1963) suggested that the flexor pollicis longus is presentin Loris, Nycticebus, Tarsius, Aotus, Callithrix, Saimiri, Macaca, Cer-copithecus, Colobus, Papio and Homo, and this was the view thatinformed the cladistic studies of Groves (1986) and Shoshani et al.(1996). However, Day and Napier (1963) intended the term’straditional usage (e.g., Barnard,1875; Duckworth,1904,1915;WoodJones, 1920; Sonntag, 1924a, b), which is to indicate the presence ofa tendon of the flexor digitorum profundus to digit 1. They did notnecessarily imply the presence of a distinct flexor pollicis longusmuscle (i.e., separated from the flexor digitorum profundus, witha tendon exclusive to the thumb). In the Loris, Nycticebus and Tar-sius specimens dissected by Burmeister (1846), Mivart and Murie(1865), Murie and Mivart (1872), Keith (1894a, b), Allen (1897),Woollard (1925), Miller (1943), Schultz (1984) and by us, the Aotus,Callithrix and Saimiri specimens studied by Senft (1907), Beattie(1927), Hill (1957, 1960, 1962) and by us, the Macaca specimensexamined by Haughton (1864, 1865), Howell and Straus (1932,

R. Diogo et al. / Journal of Human Evolution 63 (2012) 64e7872

1933), Patterson (1942), Jacobi (1966), Kimura and Tazai (1970),Jouffroy (1971) and Landsmeer (1986) and by us, the Cercopithecusspecimens investigated by Hill (1966) and Lewis (1989) and by us,the Colobus specimens dissected by Brooks (1886), Polak (1908),Jouffroy and Lessertisseur (1960) and by us, and in the Papiospecimens studied by MacDowell (1910), Hill (1970), Swindler andWood (1973), Tocheri et al. (2008) and by us there is no distinctflexor pollicis longus going exclusively to digit 1 (Table 1). Thus, theapparently contradictory statements in the literature about thepresence/absence of a flexor pollicis longus in non-hominoidprimates are mainly the result of nomenclatural pluralism andthey should not be used as the basis of cladistic character stateswithout additional clarification.

Our dissections confirm previous reports that in most non-hominoid primates the flexor digitorum profundus sendsa tendon to digit 1. However, it is not independent of the tendonsgoing to the other digits even though, when it is present, it flexesthe distal phalanx of the thumb in the same way that the flexorpollicis longus does in modern humans. In the great apes, the flexordigitorum profundus tendon to digit 1 is usually either vestigial oreven absent (e.g., Fig. 5). Straus (1942b) showed, using evidencefrom both his own dissections and from data available to him in theliterature, that among 47 chimpanzees such a tendonwas absent in14 individuals (30%), too small/thin to have had any useful functionin 10.5 individuals (22%), and was in direct functional continuitywith the radial muscle belly of the flexor digitorum profundus in 22individuals (48%). Likewise, among 16 gorillas Straus (1942b)showed that this tendon was absent in five individuals (31%),rudimentary in six individuals (41%), and present in four individ-uals (28%). Among 27 orangutans, the tendon was absent in 24individuals (89%), rudimentary and functionless in two individuals(7%) and only completely developed in one individual (4%). Thus,there is no functioning tendon in c. 95% of orangutans, in c. 72% ofgorillas and in c. 50% of chimpanzees.

In hylobatids, the anatomy of the flexor pollicis longus is morecomplex. Our de novo dissections (Table 1; Figs. 2e4) confirm thatboth modern humans and hylobatids have a flexor pollicis longusdistinct from the flexor digitorum profundus, unlike theextensively-connected muscle bellies seen in other primates(corroborating the descriptions of authors such as Deniker, 1885;Hartmann, 1886; Kohlbrügge, 1890-1892; Hepburn, 1892; Keith,1894b; Chapman, 1900; McMurrich, 1903a, b; Sonntag, 1924b;Howell, 1936a, b; Straus, 1942a; Jouffroy and Lessertisseur, 1960;Tuttle, 1969; Jouffroy, 1971; Van Horn, 1972; Lorenz, 1974; Marzke,1992, 1997; Susman, 1994, 1998; Stout, 2000; Tocheri et al., 2008).However, a few authors (e.g., Payne, 2001) suggest that in theHylobates specimens dissected by them the flexor pollicis longusblends with, and is thus not really separate from, the flexor dig-itorum profundus.

In an influential paper about the evolution of tool use in theHominina, Susman (1994: 1573, Fig. 3) stated that contrary to thecondition in modern humans, chimpanzees and other primateshave only a tendon that mimics the flexor pollicis longus and lacksa separate muscle belly; “in lesser apes (gibbons and siamang)there is a muscle belly, but it is functionally coupled with the flexordigitorum produndus.electromyography experiments on an adultfemale gibbon (H. lar) did not elicit flexion of the thumb separatefrom flexion of the fingers, as is the case in humans”. The apparentinconsistency between Susman’s interpretation and the interpre-tations of researchers who have described the hylobatid flexorpollicis longus as distinct (e.g., Marzke, 1992) is likely due to thevariable presence of a connective tissue ‘shunt’ (Tuttle, 1969) thatconnects the tendons of the flexor digitorum profundus and flexorpollicis longus (Tuttle (1969) referred to it as the ‘radial componentof the flexor digitorum profundusmusculature’) often at the level of

the carpus in hylobatids. This shunt, which is commonly lesssubstantial than the tendons it connects, is a slender flat (4e5 mmwide) structure that passes obliquely inferomedially from thetendon of the flexor pollicis longus to the lateral edge of thecommon flexor digitorum profundus tendons to digits 2e5.According to Tuttle, the ’flexor shunt’ probably serves differentfunctions during the various activities undertaken by hylobatids.For instance, he suggested that when the hand is used as ananatomical hook during vigorous arm-swinging, the ‘shunt’ prob-ably transfers some of the contractile force of flexor pollicis longusto the flexor digitorum profundus tendons and thereby helps tostabilize the wrist. In contrast, when the hand is used for finemanipulation, particularly when the metacarpophalangeal joints ofdigits 2e5 are flexed, the ‘shunt’ is probably somewhat slack,allowing the flexor pollicis longus to independently flex the distalphalanx of the thumb. When the thumb is abducted to grasp largebranches, the ‘shunt’ is probably also stretched to the extent thatthe distal phalanges of the thumb and the medial four digits areflexed synchronously to produce secure grips. As a result, andcontrary to what occurs in Pan, Pongo and Gorilla, in hylobatids thepollex normally plays an active role in maintaining a suspendedposture and in facilitating the distinctive locomotion of thesehominoids. After observing living animals, Van Horn (1972) sug-gested that the flexor pollicis longus of hylobatids is importantduring the arm-pull phase of climbing, in which the terminalphalanx of the thumb is flexed as the animal lifts its weight froma previous support (i.e., this muscle is an important component ofthe power used to grasp). Stout (2000) found that the tendon of theflexor pollicis longus of hylobatids does not flex the distal phalanxof the thumb, but instead it adducts the thumb, flexes the pollicalmetacarpophalangeal joint and stabilizes the pollical interphalan-geal joint. Further functional studies of hylobatids are needed totest these hypotheses, particularly because even among modernhumans there is significant variationwith respect to the presence ofa fully independent flexor pollicis longus. For instance, as explainedabove according to an extensive review undertaken by Lindburgand Comstock (1979), only c. 70% of modern humans havea flexor pollicis longus that is anatomically entirely independentfrom the flexor digitorum profundus. Shrewsbury et al. (2003)posited that evidence of restrictive thumb/index tendosynovitisand pain in modern humans is usually associated with connectionsbetween the flexor pollicis longus and the flexor digitorum pro-fundus. In some cases, the condition can be relieved by removingthe connection between these muscles and occasionally thethickened flexor pollicis longusmust be removed as well. Accordingto Shrewsbury et al. (2003), the repetitive force used in a precisiongrip in concert with habitual repetitive behaviours associated withprecision handlingmight have favoured independence of these twomuscles as modern humans became increasingly dependent on thedexterity required to make and use tools.

There are reasons to consider that the tendon of the flexordigitorum profundus to digit 1 became independently reduced inorangutans, gorillas and chimpanzees, while the presence ofa separate, robust flexor pollicis longus in hylobatids and modernhumans may be associated with having a well-developed and/orfunctionally more independent thumb. The first hypothesis issupported by the fact that the tendon of the flexor digitorum pro-fundus to digit 1 is highly variable in great apes. In a few cases it isfully developed, in others it is atrophied, in others it is ‘rudimen-tary’ and may no longer attach to the main body of the flexor dig-itorum profundus muscle, and in others it is completely missing(see above and also, e.g., Shrewsbury et al., 2003, Fig. 5). Straus(1942b) included those cases where there is no continuity withthe main body of the flexor digitorum profundus in the category ofrudimentary, 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),the tendon is not necessarily functionless in all of those cases.Within the specimens dissected by them, the tendon had anattachment pattern that was ligament-like in all four specimens oforangutan, two of the chimpanzees, and the two juvenile hama-dryas baboons. The tendon attached distally to the pollical distalphalanx and laterally to the radial and ulnar basal tubercles. Tendonfibres in the hamadryas baboons had distal attachments to theungual pulp over the distal phalangeal tuberosity. However, theproximal radial and ulnar bands of the tendon both sent branchesto the metacarpophalangeal joint in the region of the adductorpollicis insertion, and also (in the apes) in the region of the flexorpollicis brevis insertion. No fibres were found continuing furtherproximally over the wrist to the radial aspect of the flexor dig-itorum profundus tendon. According to Shrewsbury et al. (2003),this distal tethering of the tendon to digit 1 to bone on both sides ofthe distal interphalangeal joint probably helps to stabilize the joint.

The hypothesis that the presence of a separate, robust flexorpollicis longus in hylobatids and modern humans may be associ-ated with having a well-developed and/or functionally moreindependent thumb is in line with the results of electromyography(Marzke et al., 1998) and hand pressure studies (Rolian et al., 2011;Williams et al., 2012) when various modern human subjects weremanufacturing and using Oldowan tools as well as other objects.The recruitment of the flexor pollicis longus and the extensorpollicis brevis allowed the subjects to maintain the meta-carpophalangeal joint in extension (using the extensor pollicisbrevis, which in modern humans usually attaches onto the prox-imal phalanx of the thumb: see below), while simultaneously usingthe flexor pollicis longus to flex the distal phalanx of the thumb atthe pollical interphalangeal joint. The results from experimentsdocumenting hand pressure during stone tool making underscorethe importance of hand posture on the distribution of pressureacross the thumb and other digits (Williams et al., 2012),supporting the suggestions of Marzke et al. (1998) suggestion thatthe differentiation of these two muscles needed to have occurredfor early members of the human lineage to be able to habituallymake effective stone tools.

Evidence from the fossil record also lends some support to thehypothesis that a well-developed tendon to digit 1, whether as partof flexor digitorum profundus or as a separate flexor pollicis longus,was the primitive condition for the Hominoidea and probably thePan-Homo LCA. While some (Shrewsbury et al., 2003; Marzke andShrewsbury, 2006; Tocheri et al., 2008) have noted that the flexorpollicis longus (or flexor digitorum profundus to digit 1) does notinsert onto the distal phalanx’s volar pit, or ‘proximal volar fossa’(Shrewsbury et al., 2003), it is clear that the tendon inserts on theridge that is, when present, just distal to the fossa (Susman, 1998;Shrewsbury et al., 2003). The presence of an attachment (ridge) tothe distal pollical phalanx is seen in all of the fossil hominoid taxaknown to date, including Proconsul (Begun et al., 1994) and Oreo-pithecus (Moyà-Solà et al., 1999). Similarly, all fossil pollical thumbdistal phalanges attributed to (pre-modern) Homo show evidenceof a volar ridge, with an accompanying proximal volar fossa(Shrewsbury et al., 2003) that marks the attachment of a tendon(Susman, 1998). The attachment site is also present on the distalpollical phalanges of both early and potential fossil taxa of thesubtribe Hominina, including Orrorin tugenensis (Gommery andSenut, 2006; Almécija et al., 2010), Ardipithecus ramidus (Lovejoyet al., 2009), and. Australopithecus afarensis (Ward et al., in press).Whether Orrorin is a member of the subtribe Hominina (Senutet al., 2001; Richmond and Jungers, 2008) or not (e.g., Wood andHarrison, 2011), the clear indication of a tendon insertion on thedistal pollical phalanx at almost six million years ago suggeststhat either 1) the presence of tendon to digit 1 was present in the

Pan-Homo LCA, is retained in the Hominina, and independently lostin the evolution of Pan and Gorilla (and possibly Pongo), or 2) thatthe Pan-Homo LCA lacked the tendon insertion, and that it evolvedrapidly at the base of the Hominina clade independently of itsevolution in hylobatids. In either scenario, the evolution of theflexor pollicis longus in hominoids involves homoplasy. In a recentpaper, Almécija et al. (in press) strongly support the idea that theplesiomorphic hominoid, hominid and hominin condition is verylikely a moderately long thumb, longer than the thumb of moderngreat apes, thus indirectly supporting the idea that the tendon ofthe flexor digitorum profundus to digit 1 became independentlyreduced in orangutans, gorillas and chimpanzees.

Some time between the late Pliocene and early Pleistocene, thethumb of the members of the subtribe Hominina evolved signifi-cantly greater robusticity, the thumbofmodern humans beingmorerobust than that of any other extant hominoids, including hyloba-tids (Susman, 1994). However, the hypothesis remains untestedregarding whether other derived muscular features of the modernhuman hand occurred with the evolution of thumb robusticity.

Deep head of the flexor pollicis brevis and the ‘volar interosseous ofHenle’

In modern humans, the so-called ‘deep head of the flexor pol-licis brevis’ usually runs from the carpal region (e.g., capitate,occasionally from trapezoid) to the medial side of the palmarsurface of proximal phalanx of digit 1 and/or an adjacent sesamoidbone (e.g., Fig. 4) and it is normally innervated by the deep branchof the ulnar nerve (and/or occasionally by the median nerve). Theso-called ‘volar interosseous of Henle’ usually runs from the base ofmetacarpal I to the ‘wing tendon’ of the extensor apparatus of digit1 (e.g., Susman et al., 1999). In the specimens examined by us, thislatter structure, which is normally innervated by the deep branch ofthe ulnar nerve, may also originate from the carpal bones adjacentto metacarpal I and its insertion onto digit 1 may extend onto theproximal phalanx and/or sesamoid bone of the thumb (e.g., Fig. 4).Susman (1994; Fig. 3) suggested that the ’deep head of the flexorpollicis brevis’ and the ’interosseous volaris primus of Henle’ (asdescribed above) are derived in modern humans. However, ourcomparative investigation suggests that the ‘deep head of the flexorpollicis brevis’ is present in most primates, while the volar inter-osseous of Henle is seen in <50% of a few primates includinggorillas and chimpanzees (e.g., Fig. 5). Lewis (1989) embraced thehypothesis of Forster (1917) who suggested that the plesiomorphiccondition for eutherian mammals is to have ten flexores brevesprofundi, each inserting onto the lateral and medial sides of eachdigit, and four intermetacarpales (see, e.g., Fig. 6) connecting theadjacent metacarpal bones. In primates, the plesiomorphic condi-tion is the same, but two of the 14muscles have differentiated, eachforming two muscles: the flexor pollicis brevis and the opponenspollicis from flexor brevis profundus 1, and the flexor digiti minimibrevis and the opponens digiti minimi from flexor brevis profundus10 (Fig. 1, Table 1). In hominoids other than chimpanzees, as well asin New World monkeys, the flexores breves profundi 3, 5, 6 and 8usually fuse with the intermetacarpales 1, 2, 3 and 4 to form thedorsal interossei 1, 2, 3 and 4, respectively, whereas the palmarinterossei 1, 2 and 3 are derived directly from the flexores brevesprofundi 4, 7 and 9, respectively. The model of Lewis (1989), model,which explains why the dorsal interossei muscles are bipennatewhereas the palmar interossei are unipennate, is supported by thedevelopmental studies of authors such as Cihák (1972), which pointout that at least some flexores breves profundi effectively becomefused ontogenetically with the most dorsal intrinsic muscles of thehand (the intermetacarpales), forming the dorsal interossei of adultmodern humans. Chimpanzees have undergone a secondary

R. Diogo et al. / Journal of Human Evolution 63 (2012) 64e7874

reversion to the plesiomorphic state, probably because their flex-ores breves profundi do not fuse with the intermetacarpales duringontogeny to form the dorsal interossei (Table 1; Fig. 6). As explainedabove, chimpanzees also display a secondary reversion of a syna-pomorphy of the Hominidae in that adult chimpanzees have twocontrahentes digitorum in addition to the the adductor pollicis, onegoing to digit 4 and the other to digit 5 (in other adult hominidsthere is usually none). Because the studies of Cihák (1972) suggestthat in at least modern humans the contrahentes are lost (i.e.,‘reabsorbed’) during ontogeny, the presence of metacarpales and ofcontrahentes in chimpanzees is very likely due to a prolonged ordelayed development of the hand musculature of these apes, i.e., inthis particular case extant chimpanzees are seemingly moreneotenic than modern humans (Diogo and Wood, in press). This isin line with other recent studies that have pointed out thatalthough in the literature it is often stated that modern humans arein general more neotenic than other primates, the empirical datacollected in the last decades reveal that both paedomorphic andperamorphic processes have been involved in the mosaic evolutionof humans and of other hominoids (see, e.g., Bufill et al., 2011; andreferences therein).

The main controversy concerning the evolution and taxonomicdistribution of the so-called ‘deep head of the flexor pollicis brevis’and of the volar interosseous of Henle mainly concerns the identityof one of the flexores breves profundi (i.e., number 2, which ple-siomorphically in mammals goes to the ulnar side of digit 1).Although the so-called ‘volar interosseous of Henle’ is usually notshown in modern human anatomical atlases, it is present in themajority of modern humans. For example, Abramowitz (1955),Lewis (1989), Susman et al. (1999), Henkel-Kopleck and Schmidt(2000) and Morrison and Hill (2011) report the presence of volarinterosseous of Henle in 100%, 92%, 86%, 69%, and 91%, respectively,of the individuals examined. This structure was also found in the 18modern human individuals examined in this study (GWUANA HS1-6, HU D43 and HU D45; Fig. 4).

As the alternative name ‘interosseous volaris primus’ indicates,some authors consider that this structure, (e.g., Figs. 4and 5; seebelow) corresponds to a vestigial flexor brevis profundus 2 (i.e., it isa fourth, and most radial, palmar interosseus). However, ourcomparative and phylogenetic analyses strongly suggest that the‘interosseous volaris primus’ corresponds instead to a structurethat is derived from a thin, deep additional slip of the adductorpollicis (Diogo and Wood, 2011) that we suggest should instead benamed the adductor pollicis accessorius (see below and Table 1). Infact, the ‘interosseous volaris primus’ often blends with theadductor pollicis distally (see Diogo and Wood, 2011, 2012) and, atleast in modern humans, it is innervated by the branches of thedeep branch of the ulnar nerve that innervate the adductor pollicis(and not by the branches of this deep branch that innervate thedorsal and ventral interossei: Fig. 4; Bello-Hellegouarch et al.,submitted). Moreover, contrary to the statements of some authors(e.g., Susman, 1994), our analyses revealed that the so-called ‘deephead of the flexor pollicis brevis’ of modern human anatomy ispresent in the vast majority of primates and that it corresponds tothe flexor brevis profundus 2 of other mammals (Table 1; Diogo andWood, 2011, 2012). For example, the structure designated as theflexor brevis profundus 2 in rats is the structure that is designatedas ‘deep head of the flexor pollicis brevis’ of modern humans. Sucha structure is also found in the vast majority of non-humanprimates, except for the New World monkeys in which this struc-ture is either missing or, more likely, has fused with the flexorbrevis profundus 1 component of the ‘superficial head of the flexorpollicis brevis’ of these monkeys. As Diogo and Wood (2011) argue,most authors have found a flexor brevis profundus 2 in the non-human and non-platyrrhine primates, but many have erroneously

designated this structure either a component of the adductor pol-licis or they have referred to it as the ‘interosseous volaris primus ofHenle’. Linscheid et al. (1991) have designated the adductor pollicisaccessorius sensu the present paper as the ‘accessory adductormuscle’ or the ‘adductor pollicis accessory’, but their study issomewhat confusing because despite using this name they clearlyconsider that this muscle corresponds to the flexor brevis pro-fundus 2, and not to part of the adductor pollicis, of other mammals(see, e.g., their Fig. 8). According to the authors, the adductor pol-licis accessorius sensu the present paper is present in all eightmodern human hands dissected by them, inserting onto the ulnarwing tendon of the thumb; the origin is from metacarpal 1 in fourhands, from metacarpals 1 and 2 in two hands, from metacarpal 2in one hand, and from the trapezium in one hand. Contrary to thehypothesis defended in the present paper, they thus consider thatthe ‘deep head of the flexor pollicis brevis’ of modern humananatomy corresponds to part of the oblique head of the adductorpollicis, and not to the flexor brevis profundus 2, of other mammals.Moreover, they use the name ‘adductor pollicis, accessory obliquehead’ to designate an extra head of the adductor pollicis that ispresent in six of the eight modern human hands dissected by themand that originates from metacarpal 2.

In summary, our comparative andphylogenetic analyses indicatethat the flexor brevis profundus 2 has beenpresent and has basicallykept the same anatomical configuration from the LCA of eutherianmammals to H. sapiens, where it forms the so-called ‘deep head ofthe flexor pollicis brevis’. The names ‘deep head of the flexor pollicisbrevis’ and ‘superficial head of the flexor pollicis brevis’ are thusinappropriate and in our opinion should be abandoned, because theformer corresponds directly to theflexor brevis profundus 2 of othermammalswhile the latter derives from theflexor brevis profundus 1(the opponens pollicis being also derived from the flexor brevisprofundis 1: see above and Table 1).We propose that the ‘deep headof theflexor pollicis brevis’ and ’superficial head of the flexor pollicisbrevis’ are therefore designated as flexor brevis profundus 2 and asflexor pollicis brevis, respectively (see, e.g., Fig. 4; Table 1). As wehave argued above, the ‘volar interosseous of Henle’ of modernhumans does not correspond to the flexor brevis profundus 2 ofother mammals (i.e., it is not a true palmar interosseus), but insteadit is very likely derived from the adductor pollicis (Fig. 9; Table 1:adductor pollicis accessorius; see below).

Extensor pollicis brevis

In modern humans, the extensor pollicis brevis, which isinnervated by the posterior interosseous nerve, usually runs fromthe dorsal surface of the radius and the adjacent interosseousmembrane to the base of the proximal phalanx of digit 1.Within the200 adult modern human upper limbs examined by Kaneff (1959,1968, 1969, 1979, 1980a, b), the extensor pollicis brevis is missing injust over 1% of cases and in just over 5% of cases this muscle isreduced to a tendon connected to the tendon of the abductor pol-licis longus or to ligaments adjacent to the insertion of this tendon.In most of the cases he examined, Kaneff reported that the fleshybelly of the extensor pollicis brevis blends with the belly of theabductor pollicis longus. According to the ontogenetic studies ofLewis (1910), the abductor pollicis longus and extensor pollicisbrevis usually only become separate entities late in modern humandevelopment. In our comparative sample, the only other groupapart from modern humans with a distinct extensor pollicis brevisis the hylobatids. A separate extensor pollicis brevis was found inthe hylobatids dissected by Bischoff (1870), Kohlbrügge (1890-1892), Duckworth (1904) and Michilsens et al. (2009) and by us(e.g., Fig. 7). The exception is Deniker (1885), who did not finda distinct extensor pollicis brevis in the gibbon foetus dissected by

DP

PP

MIOH

A DP

PP

MIOH

AB

B DP

PP

MIOH

PPI

C DP

PP

MIOH

PPI

D

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 additionalbundle (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, inwhich 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).

R. Diogo et al. / Journal of Human Evolution 63 (2012) 64e78 75

him. Contrary to the condition inmodern humans, in hylobatids theextensor pollicis brevis usually does not extend to the proximalphalanx of the thumb, inserting instead onto the base of metacarpalI and/or the adjacent sesamoid or carpal bones. In gorillas (e.g.,Hepburn, 1892; Straus, 1941a, b; Raven, 1950; Preuschoft, 1965;Sarmiento, 1994), a tendon of the abductor pollicis longus ofteninserts onto the proximal phalanx of the thumb and the term‘extensor pollicis brevis’ has been used to describe this (N.B., inother great apes, including those dissected by us, the abductorpollicis longus usually inserts onto metacarpal I and/or onto carpalbones: see Diogo and Wood, 2011). However, as Huxley (1864),Macalister (1873), Bischoff (1880), Deniker (1885), Tuttle (1970),Kaneff (1979, 1980a, b), and Aziz and Dunlap (1986) have stressed,in gorillas the tendon going to the proximal phalanx of the thumb isthe result of a bifurcation of the tendon of the abductor pollicis, theother branch of this tendon usually going to the metacarpal I(Fig. 8). Our dissections corroborate this. Thus, contrary to thecondition in modern humans and hylobatids, in gorillas and inother primates there is no separate extensor pollicis brevis muscle(i.e., a distinct belly separated from abductor pollicis longus). Itshould moreover be noted that in the literature reviews carried outby Straus (1941a, b) and Sarmiento (1994), for gorillas they foundthat in just over half of the cases the abductor pollicis longus onlyextends as far as the proximal phalanx of the thumb.

In summary, modern humans are peculiar because they havea distinct extensor pollicis brevis that attaches to the proximalphalanx of the thumb, hylobatids are peculiar because they havea distinct extensor pollicis brevis that usually attaches to the base ofmetacarpal I and/or to adjacent bones, and gorillas are peculiarbecause in about half of the cases one of the two tendons of theabductor pollicis longus attaches to the proximal phalanx of thethumb. According to our dissections and review of the literature, inchimpanzees, orangutans and in non-hominoid primates theabductor pollicis longus usually has two tendons, but with theexception of one of the 20 chimpanzees reviewed by Keith (1899),these do not extend to the proximal phalanx of the thumb.

Conclusions

Our comparative and phylogenetic analyses indicate that theforearm muscles in primates usually number between 18 and 19(Table 1). Two of the 19 muscles predicted to be plesiomorphicallypresent in primates may be missing in some groups (e.g., the epi-trochleoanconeus is usually absent in hominoids except Pan and theanconeus is usually absent in Hylobates). Modern humans, becausethey usually lack the epitrochleoanconeus but have two muscles

going to the phalanges of the thumb (flexor pollicis longus to thedistal and extensor pollicis brevis to the proximal), have more fore-armmuscles (i.e., 20) than any other primate studied by us (Table 1).Hylobatids also have a flexor pollicis longus and an extensor pollicisbrevis, but as stated above they normally lack an anconeus, so theyhave 19 forearmmuscles (Table 1).With respect to the handmuscles,phylogenetically plesiomorphic primates such as strepsirrhines andTarsius usually have more than 30 muscles, but modern humansusually have only 21 muscles (Table 1). The hand muscles thatmodern humans have lost relative to phylogenetically plesiomorphicprimates (i.e., contrahentes, intermetacarpales and interossei acces-sorii) attach to digits 2e5 (Table 1). The hand muscles that areconserved as separate structures in modern humans are those thatinsert onto the thumb. Modern humans usually also have an addi-tional pollical structure that may also be present in gorillas andchimpanzees, which is often designated as ‘musculus interosseousvolaris primus of Henle’ but should be designated as musculusadductor pollicis accessorius (Table 1).

Contrary to what is sometimes suggested in the literature, noneof the muscles and muscle bundles in the modern human forearmand hand are unique tomodern humans. Nonetheless, it is probablynot a coincidence that the three derived structures found inmodern humans and in only one or two other primate genera (i.e.,adductor pollicis accessorius is also found in some Pan and Gorilla;flexor pollicis longus and extensor pollicis brevis also found inHylobates) all involve the thumb. In fact, this is consistent with thehypothesis that movements of the thumb played an important rolein human evolution. Marzke et al. (1998) suggested that the flexorpollicis longus and extensor pollicis brevis may have co-evolved toenable the members of the subtribe Hominina to maintain themetacarpophalangeal joint in extension while flexing the distalphalanx of the thumb as occurs in stone tool making and particu-larly usage. However, the presence of the flexor tendon attachmentsite on distal phalanges of all known fossil taxa of the subtribeHominina suggests that some of this musculature was present earlyin human evolution and indeed it may be the primitive conditionfor the Pan-Homo LCA.

The functional significance of the adductor pollicis accessorius ismore obscure, because this structure usually connects the ulnar sideof the base of metacarpal I to the wing tendon of the extensorapparatus of digit 1. It is not very substantial and it is difficult to see(e.g., Morrison and Hill, 2011) how it could confer significantbiomechanical advantage across the joints of the thumb. Detailedcomparative morphological and functional analyses, ideallyincluding electromyographic studies, should be carried out inmodern humans and also in gorillas and chimpanzees in order to

R. Diogo et al. / Journal of Human Evolution 63 (2012) 64e7876

better understand the functional and evolutionary significance of thisenigmatic structure. In addition, further studies ofmore specimens ofother primate taxa are needed to check if this structure may bepresent in at least some specimens of non-hominin taxa as well.

The adoption and consistent use of the names proposed in thispaper is not amere nomenclatural detail: it is a return to an old, andDarwinian, tradition in which it is important to understand thephylogeny and homology of each and every anatomical structure. Itis also a way to stress that modern humans usually have 11 muscles(abductor pollicis longus, extensor pollicis brevis, extensor pollicislongus, flexor pollicis longus, abductor pollicis brevis, flexor pollicisbrevis, opponens pollicis, adductor pollicis, dorsal interosseus 1,adductor pollicis accessorius, and flexor brevis profundis) attachedto metacarpal I and the pollical phalanges, and not nine as it oftenstated in atlases and textbooks. We hope that this paper will attractthe attention of the authors of human anatomy atlases and text-books and will thus result in the inclusion, in these atlases andtextbooks, of information about the musculus adductor pollicisaccessorius and the musculus flexor brevis profundus 2.

Acknowledgements

We thank R. Walsh and F. Slaby (Department of Anatomy,George Washington University), R. Bernstein and S. McFarlin(Department of Anthropology, George Washington University),N. Rybczynski (Canadian Museum of Nature), H. Mays (CincinnatiMuseum of Natural History), A. Aziz (Department of Anatomy,Howard University), F. Pastor (Department of Anatomy, Universityof Valladolid), A. Gorow, H. Fitch-Snyder and B. Rideout (San DiegoZoo) and J. Fritz and J. Murphy (Primate Foundation of Arizona) forkindly providing the non-primate and primate mammalian speci-mens dissected during this project. RD was supported by a GeorgeWashington University (GW) Presidential Merit Fellowship and bya Howard University start-up package, BW by the GW UniversityProfessorship in Human Origins, the GW Provost and the GWSelective Excellence Program, and BGR by the National ScienceFoundation (BCS-0725122).

References

Abdala, V., Diogo, R., 2010. Comparative anatomy, homologies and evolution of thepectoral and forelimb musculature of tetrapods with special attention to extantlimbed amphibians and reptiles. J. Anat. 217, 536e573.

Abramowitz, I., 1955. On the existence of a palmar interosseous muscle in thethumb with particular reference to the Bantu-speaking Negro. S. Afr. J. Sci. 51,270e276.

Allen, H., 1897. Observations on Tarsius fuscus. Proc. Acad. Nat. Sci. Phila 49, 35e55.Almécija, S., Moya-Sola, S., Alba, D.M., 2010. Early origin for human-like precision

grasping: a comparative study of pollical distal phalanges in fossil hominins.Plos One 5, e11727.

Almécija, S., Alba, D. M., Moyá-Solá, S. The thumb of Miocene apes: New insightsfrom Castell de Barberá (Catalonia, Spain). Am. J. Phys. Anthropol., in press.

Aziz, M.A., Dunlap, S.S., 1986. The human extensor digitorum profundus musclewith comments on the evolution of the primate hand. Primates 27, 293e319.

Barnard, W.S., 1875. Observations on the membral musculation of Simia satyrus(Orang) and the comparative myology of man and the apes. Proc. Am. Assoc.Adv. Sci. 24, 112e144.

Beattie, J., 1927. The anatomy of the common marmoset (Hapale jacchus Kuhl). Proc.Zool. Soc. Lond 1927, 593e718.

Begun, D.R., Teaford, M.F., Walker, A., 1994. Comparative functional anatomy ofProconsul phalanges from Kaswanga primate site, Rusinga Island, Kenya. J. Hum.Evol. 26, 89e166.

Bello-Hellegouarch, G., Aziz, M. A., Ferrero, E. M., Kern, M., Francis, N., Diogo, R.'Pollical palmar interosseous muscle' (musculus adductor pollicis accessorius):attachments, innervation, variations, phylogeny, review of the literature, andimplications for human evolution and medicine. Anat. Rec., Submitted.

Bischoff, T.L.W., 1870. Beitrage zur Anatomie des Hylobates leuciscus and zueinervergleichenden Anatomie der Muskeln der Affen und des Menschen. Abh. BayerAkad. Wiss Miinchen Math. Phys. Kl 10, 197e297.

Bischoff, T.L.W., 1880. Beitrage zur Anatomie des Gorilla. Abh. Bayer Akad. WissMiinchen Math. Phys. Kl 13, 1e48.

Brooks, H.S.J., 1886. On the morphology of the intrinsic muscles of the little finger,with some observations on the ulnar head of the short flexor of the thumb.J. Anat. Physiol. 20, 644e661.

Bufill, E., Agusti, J., Blesa, R., 2011. Human neoteny revisited: the case of synapticplasticity. Am. J. Human Biol. 23, 729e739.

Burmeister, H., 1846. Beiträge zur näheren Kenntniss der Gattung Tarsius. GeorgReimer, Berlin.

Chapman, H.C., 1900. Observations upon the anatomy of Hylobates leuciscus andChiromys Madagascariensis. Proc. Acad. Nat. Sci. Phila 52, 414e423.

Cihák, R., 1972. Ontogenesis of the skeleton and intrinsic muscles of the humanhand and foot. Ergeb. Anat. Entwicklungsgesch 46, 5e194.

Day, M.H., Napier, J., 1961. The two heads of the flexor pollicis brevis. J. Anat. 95,123e130.

Day, M.H., Napier, J., 1963. The functional significance of the deep head of flexorpollicis brevis in primates. Folia Primatol. 1, 122e134.

Deniker, J., 1885. Recherches anatomiques et embryologiques sur les singesanthropoides, foetus de gorille et de gibbon. Arch. Zool. Exp. Gén 3, 1e265.

Diogo, R., 2007. On the Origin and Evolution of Higher-Clades: Osteology, Myology,Phylogeny and Macroevolution of Bony Fishes and the Rise of Tetrapods.Science Publishers, Enfield.

Diogo, R., 2009. The head musculature of the Philippine colugo (Dermoptera:Cynocephalus volans), with a comparison to tree-shrews, primates and othermammals. J. Morphol. 270, 14e51.

Diogo, R., Abdala, V., 2007. Comparative anatomy, homologies and evolution of thepectoral muscles of bony fish and tetrapods: a new insight. J. Morphol. 268,504e517.

Diogo, R., Abdala, V., 2010. Muscles of Vertebrates e Comparative Anatomy,Evolution, Homologies and Development. Science Publishers, Enfield.

Diogo, R., Wood, B., 2011. Soft-tissue anatomy of the primates: phylogenetic anal-yses based on the muscles of the head, neck, pectoral region and upper limb,with notes on the evolution of these muscles. J. Anat. 219, 273e359.

Diogo, R., Wood, B., 2012. Comparative Anatomy and Phylogeny of Primate Musclesand Human Evolution. Taylor and Francis, Oxford.

Diogo, R., Wood, B. Violation of Dollo’s law: evidence of muscle reversions in primatephylogeny and their implications for the understanding of the ontogeny,evolution and anatomical variations of modern humans. Evolution, in press.

Diogo, R., Abdala, V., Aziz, M.A., Lonergan, N., Wood, B., 2009. From fish to modernhumans - comparative anatomy, homologies and evolution of the pectoral andforelimb musculature. J. Anat. 214, 694e716.

Duckworth, W.L.H., 1904. Studies from the Anthropological Laboratory, theAnatomy School, Cambridge. C. J. Clay & Sons, London.

Duckworth, W.L.H., 1915. Morphology and Anthropology, second ed. CambridgeUniversity Press, Cambridge.

Dunlap, S.S., Thorington, R.W., Aziz, M.A., 1985. Forelimb anatomy of New Worldmonkeys: myology and the interpretation of primitive anthropoid models. Am.J. Phys. Anthropol. 68, 499e517.

Dylevsky, I., 1967. Contribution to the ontogenesis of the flexor digitorum super-ficialis and the flexor digitorum profundus in man. Folia Morphol. 15, 330e335.

Forster, A., 1917. Die mm. contrahentes und interossei manus in der Säugetierreiheund 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.

Gibbs, S., Collard, M., Wood, B.A., 2000. Soft-tissue characters in higher primatephylogenetics. Proc. Natl. Acad. Sci. 97, 11130e11132.

Gibbs, S., Collard, M., Wood, B.A., 2002. Soft-tissue anatomy of the extant homi-noids: a review and phylogenetic analysis. J. Anat. 200, 3e49.

Gommery, D., Senut, B., 2006. The terminal thumb phalanx of Orrorin tugenensis(Upper Miocene of Kenya). Geobios 39, 372e384.

Groves, C.P., 1986. Systematics of the great apes. In: Swindler, D.R., Erwin, J. (Eds.),1986. Comparative Primate Biology: Systematics, Evolution and Anatomy, vol. 1.A.R. Liss, New York, pp. 187e217.

Haines, R.W., 1939. A revision of the extensor muscles of the forearm in tetrapods.J. Anat. 73, 211e233.

Haines, R.W., 1946. A revision of the movements of the forearm in tetrapods. J. Anat.80, 1e11.

Haines, R.W., 1950. The flexor muscles of the forearm and hand in lizards andmammals. J. Anat. 84, 13e29.

Hartmann, R., 1886. Anthropoid Apes. Keegan, London.Haughton, S., 1864. Notes on animal mechanics, 2, on the muscles of some of the

smaller monkeys of the genera Cercopithecus and Macaca. Proc. R. Ir. Acad. 8,467e471.

Haughton, S., 1865. Notes on animal mechanics, 7, on the muscular anatomy of theMacacus nemestrinus. Proc. R. Irish Acad. 9, 277e287.

Henkel-Kopleck, A., Schmidt, H.M., 2000. Das Ligamentum metacarpale pollicis.Handchir. Mikrochir. Plast. Chir. 32, 1e8.

Hepburn, D., 1892. The comparative anatomy of the muscles and nerves of thesuperior and inferior extremities of the anthropoid apes: I - myology of thesuperior extremity. J. Anat. Physiol. 26, 149e186.

Hill, W.C.O., 1953. Primates e Comparative Anatomy and Taxonomy, I, Strepsirhini.University Press, Edinburgh.

Hill, W.C.O., 1955. Primates e Comparative Anatomy and Taxonomy, II, Haplorhini:Tarsioidea. University Press, Edinburgh.

Hill, W.C.O., 1957. Primates e Comparative Anatomy and Taxonomy, III, Pithecoidea,Platyrrhini (Families Hapalidae and Callimiconidae). University Press,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 Americanprimate. Trans. Am. Phil. Soc. New Ser. 49, 1e116.

Hill, W.C.O., 1960. Primates e Comparative Anatomy and Taxonomy, IV, Cebidae,Part A. University Press, Edinburgh.

Hill, W.C.O., 1962. Primates e Comparative Anatomy and Taxonomy, V, Cebidae, PartB. University Press, Edinburgh.

Hill, W.C.O., 1966. Primates e Comparative Anatomy and Taxonomy, VI, Catarrhini:Cercopithecoidea, Cercopithecinae. Interscience Publishers Inc, New York.

Hill, W.C.O., 1970. Primates e Comparative Anatomy and Taxonomy, VIII, Cyn-opithecinae: Papio, Mandrillus, Theropithecus. University Press, Edinburgh.

Hill, W.C.O., 1974. Primates e Comparative Anatomy and Taxonomy, VII, Cyn-opithecinae: Cercocebus, Macaca, Cynopithecus. University Press, Edinburgh.

Howell, A.B., 1936a. Phylogeny of the distal musculature of the pectoral appendage.J. Morphol. 60, 287e315.

Howell, A.B., 1936b. The phylogenetic arrangement of the muscular system. Anat.Rec. 66, 295e316.

Howell, A.B., Straus, W.L., 1932. The brachial flexor muscles in primates. Proc. U.S.Natl. Mus. 80, 1e31.

Howell, A.B., Straus, W.L., 1933. The muscular system. In: Hartman, C.G., Straus, W.L.(Eds.), The Anatomy of the Rhesus Monkey. Williams andWilkins Co, Baltimore,pp. 89e175.

Huxley, T.H., 1864. The structure and classification of the Mammalia. Med. TimesGaz. 1864, 398e468.

Jacobi, U., 1966. Die Muskulatur des Unterarmes and der Hand bei Macaca mulatta.Z. Morphol. Anthropol. 58, 48e73.

Jouffroy, F.K., 1971. Musculature des membres. In: Grassé, P.P. (Ed.), Traité de Zoo-logie, XVI: 3 (Mammifères). Masson et Cie, Paris, pp. 1e475.

Jouffroy, F.K., Lessertisseur, J., 1960. Les spécialisations anatomiques de la main chezles singes à progression suspendue. Mammalia 24, 93e151.

Kaneff, A., 1959. Über die evolution des m. abductor pollicis longus und m. extensorpollicis brevis. Mateil. Morphol. Inst. Bulg. Akad. Wiss 3, 175e196.

Kaneff, A., 1968. Zur differenzierung des m. abductor pollicis biventer beim Men-schen. Gegenbaurs Morphol. Jahrb. 112, 289e303.

Kaneff, A., 1969. Umbildung der dorsalen Daumenmuskeln beim Menschen. Verh.Anat. Ges 63, 625e636.

Kaneff, A., 1979. Évolution morphologique des musculi extensores digitorum etabductor pollicis longus chez l’homme. I. Introduction, méthodologie, m.extensor digitorum. Gegenbaurs Morphol. Jahrb. 125, 818e873.

Kaneff, A., 1980a. Évolution morphologique des musculi extensores digitorum etabductor pollicis longus chez l’homme. II. Évolution morphologique des m.extensor digiti minimi, abductor pollicis longus, extensor pollicis brevis etextensor pollicis longus chez l’homme. Gegenbaurs Morphol. Jahrb. 126,594e630.

Kaneff, A., 1980b. Évolution morphologique des musculi extensores digitorum etabductor pollicis longus chez l’homme. III. Évolution morphologique du m.extensor indicis chez l’homme, conclusion générale sur l ’évolution morpho-logique des musculi extensores digitorum et abductor pollicis longus chezl’homme. Gegenbaurs Morphol. Jahrb. 126, 774e815.

Keith, A., 1894a. The myology of the Catarrhini: a study in evolution. Ph.D. Disser-tation, University of Aberdeen.

Keith, A., 1894b. Notes on a theory to account for the various arrangements of theflexor profundus digitorum in the hand and foot of primates. J. Anat. Physiol. 28,335e339.

Keith, A., 1899. On the chimpanzees and their relationship to the gorilla. Proc. Zool.Soc. Lond 1899, 296e312.

Kimura, K., Tazai, S., 1970. On the musculature of the forelimb of the crab-eatingmonkey. Primates 11, 145e170.

Kohlbrügge, J.H.F., 1890. Versuch einer Anatomie des Genus Hylobates. In:Weber, M. (Ed.), 1890. Zoologische Ergebnisse Einer Reise in NiederländischOst-Indien, vol. 1. Verlag von EJ Brill, Leiden, pp. 211e354, pp. 138e208 (Vol. 2).

Landsmeer, J.M., 1986. A comparison of fingers and hand in Varanus, opossum andprimates. Acta Morphol. Neerl. Scand. 24, 193e221.

Lewis, W.H., 1910. The development of the muscular system. In: Keibel, F., Mall, F.P.(Eds.), 1910. Manual of Human Embryology, vol. 1. Lippin-Cott, Philadelphia,pp. 454e522.

Lewis, O.J., 1989. Functional Morphology of the Evolving Hand and Foot. ClarendonPress, Oxford.

Lindburg, R.M., Comstock, B.E., 1979. Anomalous tendon slips from the flexor pol-licis longus to the digitorum profundus. J. Hand Surg. 4, 79e83.

Linscheid, R.L., An, K.-N., Gross, R.M., 1991. Quantitative analysis of the intrinsicmuscles of the hand. Clin. Anat. 4, 265e284.

Lorenz, R., 1974. On the thumb of the Hylobatidae. In: Rumbaugh, D.M. (Ed.), 1974.Gibbon and Siamang, vol. 3. Karger, Basel, pp. 157e175.

Loth, E., 1931. Anthropologie des Parties Molles (Muscles, Intestins, Vaisseaux, NerfsPeripheriques). Mianowski-Masson et Cie, Paris.

Lovejoy, C.O., Simpson, S.W., White, T.D., Asfaw, B., Suwa, G., 2009. Careful climbingin the Miocene: the forelimbs of Ardipithecus ramidus and humans are primi-tive. Science 326, 1e8.

Macalister, A., 1873. The muscular anatomy of the gorilla. Proc. R. Irish Acad. Ser. 2,501e506.

MacDowell, E.C., 1910. Notes on the myology of Anthropopithecus niger and Papio-thoth ibeanus. Am. J. Anat. 10, 431e460.

Marzke,M.W.,1992. Evolutionary development of the human thumb. Hand Clin. 8,1e8.Marzke, M.W., 1997. Precision grips, hand morphology, and tools. Am. J. Phys.

Anthropol. 102, 91e110.

Marzke, M.W., Shrewsbury, M.M., 2006. The Oreopithecus thumb: pitfalls inreconstructing muscle and ligament attachments from fossil bones. J. Hum.Evol. 51, 213e215.

Marzke, M.W., Toth, N., Schick, K., Reece, S., Steinberg, B., Hunt, K., Linscheid, R.L.,An, K.N., 1998. EMG study of hand muscle recruitment during hard hammerpercussion manufacture of Oldowan tools. Am. J. Phys. Anthropol. 105,315e332.

McMurrich, J.P., 1903a. The phylogeny of the forearm flexors. Am. J. Anat. 2,177e209.

McMurrich, J.P., 1903b. The phylogeny of the palmar musculature. Am. J. Anat. 2,463e500.

Michilsens, F., Vereecke, E.E., D’Août, K., Aerts, P., 2009. Functional anatomy ofthe gibbon forelimb: adaptations to a brachiating lifestyle. J. Anat. 215,335e354.

Miller, R.A., 1943. Functional and morphological adaptations in the forelimbs of theslow lemurs. Am. J. Anat. 73, 153e183.

Mivart, S.G., Murie, J., 1865. Observations on the anatomy of Nycticebus tardigradus.Proc. Zool. Soc. Lond 1865, 240e256.

Morrison, P.E., Hill, R.V., 2011. And then there were four: anatomical observations ofthe pollical palmar interosseous muscle in humans. Clin. Anat. 24, 978e983.

Moyà-Solà, S., Köhler, M., Rook, L., 1999. Evidence of hominid-like precision gripcapability in the hand of the Miocene ape Oreopithecus. Proc. Natl. Acad. Sci. 96,313e317.

Murie, J., Mivart, S.G., 1872. On the anatomy of the Lemuroidea. Trans. Zool. Soc.Lond 7, 1e113.

Napier, J.R., 1962. The evolution of the hand. Sci. Am. 204, 2e9.Parsons, F.G., 1898. The muscles of mammals, with special relation to human

myology: a course of lectures delivered at the Royal College of Surgeons ofEngland e lecture II, the muscles of the shoulder and forelimb. J. Anat. Physiol.32, 721e752.

Patterson, E.L., 1942. The myology of Rhinopithecus roxellanae and Cynopithecusniger. Proc. Zool. Soc. Lond 112, 31e104.

Payne, R.C., 2001. Musculoskeletal adaptations for climbing in hominoids and theirrole as exaptations for the acquisition of bipedalism. Ph.D. Dissertation,University of Liverpool.

Polak, C., 1908. Die Anatomie des genus Colobus. Verhandl. Akad. Wet. Amst 14,1e247.

Preuschoft, H., 1965. Muskeln und gelenk der vorderextremitat des gorillas. Morph.Jb 107, 99e183.

Raven, H.C., 1950. Regional anatomy of the Gorilla. In: Gregory, W.K. (Ed.), TheAnatomy of the Gorilla. Columbia University Press, New York, pp. 15e188.

Richmond, B.G., Jungers, W.L., 2008. Orrorin tugenensis femoral morphology and theevolution of hominin bipedalism. Science 319, 1662e1665.

Rolian, C., Lieberman, D.E., Zermeno, J.P., 2011. Hand biomechanics during simulatedstone tool use. J. Hum. Evol. 61, 26e41.

Sarmiento, E.E., 1994. Terrestrial traits in the hands and feet of gorillas. Am. Mus.Novit. 3091, 1e56.

Schultz, M., 1984. Osteology and myology of the upper extremity of Tarsius. In:Niemitz, C. (Ed.), Biology of Tarsiers. Gustav Fischer Verlag, Stuttgart,pp. 143e165.

Senft, M., 1907. Myologie der Vorderextremitäten von Hapale jacchus, Cebus mac-rocephalus, und Ateles ater. Ph.D. Dissertation, Bern University.

Senut, B., Pickford, M., Gommery, D., Mein, P., Cheboi, K., Coppens, Y., 2001. Firsthominid from the Miocene (Lukeino Formation, Kenya). C.R. Acad. Sci., Paris,Série IIa 332, 137e144.

Shoshani, J., Groves, C.P., Simons, E.L., Gunnell, G.F., 1996. Primate phylogeny:morphological vs molecular results. Mol. Phylogenet. Evol. 5, 102e154.

Shrewsbury, M.M., Marzke, M.M., Linscheid, R.L., Reece, S.P., 2003. Comparativemorphology of the pollical distal phalanx. Am. J. Phys. Anthropol. 121, 30e47.

Sonntag, C.F., 1924a. On the anatomy, physiology, and pathology of the orang-outan.Proc. Zool. Soc. Lond 24, 349e450.

Sonntag, C.F., 1924b. The Morphology and Evolution of the Apes and Man. John BaleSons and Danielsson Ltd, London.

Stout, K., 2000. Grip types and associated morphology of the hylobatid thumb andindex finger. M.A. thesis, Arizona State University.

Straus, W.L., 1941a. The phylogeny of the human forearm extensors. Hum. Biol. 13,23e50.

Straus, W.L., 1941b. The phylogeny of the human forearm extensors (concluded).Hum. Biol. 13, 203e238.

Straus, W.L., 1942a. The homologies of the forearm flexors: urodeles, lizards,mammals. Am. J. Anat. 70, 281e316.

Straus, W.L., 1942b. Rudimentary digits in primates. Q. Rev. Biol. 17, 228e243.Susman, R.L., 1994. Fossil evidence for early hominid tool use. Science 265,

1570e1573.Susman, R.L., 1998. Hand function and tool behavior in early hominids. J. Hum. Evol.

35, 23e46.Susman, R.L., Nyati, L., Jassal, M.S., 1999. Observations on the pollical palmar

interosseus muscle (of Henle). Anat. Rec. 254, 159e165.Swindler, D.R., Wood, C.D., 1973. An Atlas of Primate Gross Anatomy: Baboon,

Chimpanzee and Men. University of Washington Press, Seattle.Tocheri, M.W., Orr, C.M., Jacofsky, M.C., Marzke, M.W., 2008. The evolutionary

history of the hominin hand since the last common ancestor of Pan and Homo.J. Anat. 212, 544e562.

Tuttle, R.H., 1969. Quantitative and functional studies on the hands of theAnthropoidea, I, the Hominoidea. J. Morphol. 128, 309e363.

R. Diogo et al. / Journal of Human Evolution 63 (2012) 64e7878

Tuttle, R.H., 1970. Postural, propulsive and prehensile capabilities in the cheiridia ofchimpanzees and other great apes. In: Bourne, G.H. (Ed.), 1970. The Chim-panzee, vol. 2. Karger, Basel, pp. 167e263.

Van Horn, R.N., 1972. Structural adaptations to climbing in the gibbon hand. Am.Anthropol. New Ser. 74, 326e334.

Ward, C.V., Kimbel, W.H., Harmon, E.H., Johanson, D.C. New postcranial fossils ofAustralopithecus afarensis from Hadar, Ethiopia (1990e2007). J. Hum. Evol, inpress.

Williams, E.M., Gordon, A.D., Richmond, B.G., 2012. Manual pressure distributionduring Oldowan stone tool production. J. Hum. Evol. 62, 520e532.

Wood, B., Harrison, T., 2011. The evolutionary context of the first hominins. Nature470, 347e352.

Wood Jones, F., 1920. The Principles of Anatomy as Seen in the Hand. J. & A.Churchill, London.

Woollard, H.H., 1925. The anatomy of Tarsius spectrum. Proc. Zool. Soc. Lond 70,1071e1184.