J・Anthrop・Soc・Nippon人 類 誌、 100(4):391-404(1992)

REVIEW ARTICLE

Evolution of the Girdle with Special Reference to the Problems of the

Norihisa INUZUKA

Department of , The University of Tokyo, Faculty of Medicine

Abstract Persistent problems involving the clavicle include the reason for its reduction in many , its function in hominids, the last ossification, and so on. In this paper these problems are considered from the point of views of comparative and functional morphology. The clavicle is an element of the dermal that is in the process of reduction through evolution. The reason why the clavicle disappeared or was reduced in cursorial mammals is probably that it disturbed the reciprocal movement of the . The human clavicle works also as a supporter of the forelimb. The characteristic double-curved form of the clavicle may be related to its function. The sternal epiphysis of the clavicle fuses last among the long bones of the human , in spite of the first ossification of the clavicular body. The delay of determination of the clavicular length should be convenient to slight adjustment for retaining dynamic stability in human bipedal walking.

Key Words: Clavicle, Shoulder girdle, Evolution, Comparative osteology, Functional morphology

truly homologous to the pubis? Introduction In the present paper I will try to take into There is no bone as peculiar as the clavicle account many factors that may help us to get among the bones that comprise the human answers to the problems about the clavicle. In skeleton. It is the only dermal bone in the order to solve the problems of the clavicle, the postcranial skeleton. Its ossification begins first shoulder girdle must be considered as a whole. and the sternal epiphysis closes last among the The free limbs must be also considered altogether girdle and limb bones. In mammalian forms the to understand the girdles. It is necessary to com- vary in their presence or absence and pare humans with other primates in order to degree of development. The human clavicle has understand some features of the human shoulder the characteristic S-shaped double curve of girdle. The , and primates. Why is the clavicle so peculiar? Some other must be followed to know the authors have noted that pectoral and pelvic derivation of their shoulder girdles. Therefore, girdles are serially homologous. Is the clavicle under the title "Evolution of the shoulder girdle"

Article No. 9203 Received July 28, 1992 392 N. INUZUKA problems of the clavicle are considered, including the pelvic girdle and free limbs.

Dements of the Shoulder Girdle and Their Evolu tionary Trends As the shoulder girdle consists of dermal and endoskeletal elements, each element is separately described following ROMER and PARSONS (1977), KENT (1978), WAKE (1979) and STARCK (1979). The dermal shoulder girdle is derived from the pectoral armor of placoderms. As the cros- Fig. 1. The dermal and endochondral sopterygians took to land, the operculum was shoulder girdle; cl: clavicle, cth: , lost; the subsequent loss of the posttemporal, cor: coracoid, gl: glenoid, ic: interclavicle, which was situated most dorsally among the der- pc: procoracoid, pt: posttemporal, sc: scapula, sct: supracleithrum. mal shoulder girdle elements, resulted in the dis- connection of the dermal shoulder girdle from the cranium. The dermal shoulder girdle includes have a tendency to become reduced. It is con- the posttemporal, supracleithrum, postcleithrum, cluded that the dermal elements as a whole have cleithrum, clavicle and interclavicle (Fig. 1). a tendency to become reduced or disappear. Among these elements, all except the interclavicle The endoskeletal shoulder girdle lies within were originally present and only the interclavicle and behind the dermal girdle and bears the appeared after the transition to land (ROMER (ROMER and PARSONS, 1977). The and PARSONS, 1977). The dorsal elements (post- endoskeletal girdle includes the scapula, pro- temporal, supracleithrum, and postcleithrum) are coracoid and coracoid (Fig. 1). The scapula is well developed in aquatic forms but tend to be present above the glenoid fossa, and the others reduced in terrestrial forms. Table 1 shows the are below it. The procoracoid is also called the presence or absence of the ventral elements epocoracoid and the coracoid is also called the (cleithrum, clavicle and interclavicle). According metacoracoid. Table 2 shows the presence or to the table, even the ventral dermal elements absence of the endoskeletal shoulder girdle.

Table 1. Distribution of features of ventral elements of the dermal shoulder girdle; 0: presence, x : absence Evolution of the Shoulder Girdle 393

Table 2. Distribution of features of the endochondral shoulder girdle; 0: presence,* : fuse, x : absence

According to the table, the scapula never dis- Finally, no matter what the actual ossification appears, but instead develops. To the contrary, pattern is, the clavicle of eutherians, including the ventral elements of the endoskeletal girdle humans, is truly homologous to the original have a tendency to become reduced. The dermal dermal shoulder girdle. The reason why only the girdle is more developed than the endoskeletal clavicle among the postcranial skeleton derives girdle in aquatic and is less developed in from dermal bone is that only one bone survived terrestrial . The reason why such a among the many original dermal elements, and condition occurred is that the caudal margin of that the clavicle was originally a part of the the gill opening must be hardened in fishes and dermal cranium in spite of being a part of the that tetrapods need many more girdle muscles "shoulder girdle" in human anatomy . As MIKI because of the transformation of the main (1981) expressed it, "the dermal shoulder girdle locomotory organ from the trunk to the limbs. was a remnant of the posterior margin of the There have been many opinions about whether helmet in armored fishes." the ossification pattern of the human clavicle is really dermal or endochondral. According to The Evolution of the Shoulder Girdle in a LESSERTISSEURand SABAN (1967), HUXLEY, Functional View SABATIER, BROOM, FAWCETT (1913), 1. Early tetrapod stage: lateral type limb ANTHONY, WATSON, HANSON and ROMER The early tetrapods and amniotes have preferred a dermal origin, while GEGENBAUR, laterally oriented limbs. REWCASTLE(1981) calls GOETHE, HOFFMAN, TODD and d'ERRICO it "sprawling limb posture" . The "lateral" type (1928), and KOCH (1960) preferred an endo- limb is also called the "horizontal" or "trans- chondral one. In a third view the clavicle is con- verse" type. LESSERTISSEURand SABAN (1967) sidered a combination of both, but the relation separate the advanced "type horizontal" from between dermal and endochondral bones is the more primitive "type transversal" , but I unite variously understood. For example, according to them as the lateral type; they are common in LESSERTISSEUR and SABAN (1967), FITz- having the stylopodium protruding horizontally WILLIAMS referred to the clavicular body as from the trunk. Bodies with the lateral type limb dermal bone and only to the sternal end as endo- develop adductor muscles which run from the chondral. APOSTOLAKIS(1934) referred to the ventral part of the joint of girdles to the body as dermal bone and to both ends as endo- and femur in order to bear the trunk off the chondral. GARDNER (1968) said that the clavicle ground. The amphibians and early reptiles walk ossified without a cartilaginous stage, but that by moving both stylopodia fore-and-aft in a cartilage occurred late on both ends. horizontal plane for their lateral type limbs . This 394 N. INUZUKA movement enlarges the area of muscles in the type. Actually the transformation was established pelvic girdle which protract and retract the limbs. by rotation of the elbow backward and of the Thus, in both girdles the area ventral to the knee forward. joint with the stylopodia becomes broader than There have been some theories about the the dorsal area, and the area anterior to the joint reason why the stylopodia in fore- and hindlimbs extends forward and the posterior area backward. rotated in opposite directions: for example, in an "Both pectoral and pelvic girdles can be reduced animal that extends its limbs outward from the to a common pattern, of one dorsal and two trunk, it is dangerous when a predator attacks, ventral elements" (YAPP, 1965). Therefore, the or it conserves body heat to hold the limbs under clavicle is not homologous to the pubis but the trunk, or it is better for walking on narrow analogous to it. branches (MATTHEW, 1904; GREGORY, 1910; While the shoulder and the pelvic girdles BOKER, 1935). But amphibians and reptiles converged on a common pattern, they were to retain the lateral type limbs even now, notwith- take a contrasting shape in their connection with standing the presence of many predators. It does the axial skeleton because of the presence of the not seem to have been so cold as to require the thorax. That is, the shoulder girdle unites with animals to conserve body heat, when and where the thorax via the clavicle or the interclavicle (lost the "-like reptiles" evolved the inferior later and changed to the sternoclavicular joint), type limbs. And there is no evidence that they and the pelvic girdle unites with the sacrum via were arboreal animals. the ilium by a bony connection. The shoulder The most important reason for the trans- girdle forms a striking contrast to the pelvic girdle formation from the lateral to the inferior type in the relative position of its bony connection between the axial skeleton and the articulation with the limbs: it is more ventral in the shoulder girdle and more dorsal in the pelvic. In other words, the body weight hangs from the shoulder girdle in the anterior half of the body and rides on the pelvic girdle in the posterior half. This point is one of the factors that later differentiates the shoulder and pelvic girdles in shape.

2. Mammal stage: inferior type limb A typical mammal has limbs of the "inferior" type. It is also called the "parasagittal" or "vertical" type , and is a body form in which the stylopodia extend beneath the trunk. REw- CASTLE(1981) calls it "erect limb posture". This Fig. 2. "Gravity center approach theory". posture had already been gained in some When the limb type transforms from the "mammal -like reptiles". How did the lateral type lateral (light feet) to the inferior (dark feet) limb become the inferior type? A transformation type, contact points of feet on the ground approach the center of gravity (G) of the from the lateral to the inferior type limb is not body. Thus, the forelimb rotates backward simple; the toe tips point outward in the lateral and the hind limb forward. Evolution of the Shoulder Girdle 395 limb is to bring the contact points of the feet to inferior, the limb girdles also changed. As the the ground toward the center of gravity of the elbow rotated backward and the knee forward, body (Fig. 2). This theory is preferable, because the humerus inclined forward and the femur it explains with a single reason not only that the backward, and so the scapula inclined backward distance between each foot becomes narrow but and the ilium forward. There have been different also that the distance between fore and hind feet opinions about the reason why the dorsal becomes narrow. Thus the fore and hind stylo- elements of the girdles which extended upward podia rotated in opposite directions. in the animals with the lateral type limb were SWINNERTON (1923) and RADINSKY (1987) oriented in such a way. BOKER (1935) assumed very simply noted this the "gravity center an arboreal animal as a primitive type for approach theory" in spite of the presence of mammals and thought that the fore- and hind some different theories mentioned above. I limbs and the axial skeleton would comprise an happen to have the same opinion as his, and not arch as a whole to support the body weight of only support the theory but will emphasize its the animal. So, he illustrated a scheme like Fig. significance more and more, because I believe 3. BABA (1984) attributed the angle of forward that the transformation to the inferior type limb inclination of the pelvic girdle to the direction of played the most important role in the evolution a resultant of anti-gravitational direction in of later mammals and in the differentiation of standing and thrust direction in moving. But the fore- and hind limbs in shape. For example, inclination of the scapula cannot be explained along with the inferior type limb can be ex- with this opinion, and BOKER's view includes plained the rise of the elbow, knee and heel only a support function in standing, not in peculiar to mammals. moving. To explain the inclination of the girdles, When the limb type transforms from lateral a theory is more desirable by which both fore- to inferior, the direction of the long axis of the and hind limbs in both standing and moving stylopodium approaches the direction of gravity, positions can be explained. and so the role of muscle decreases in its support Now, a translation from the lateral to the of body weight, while the body weight that can inferior type of limbs brought the muscles much be supported by muscles of the same quantity alteration. In an animal with the lateral type limb increases. Therefore, body size can enlarge and three kinds of muscles were necessary: adductor limb bones elongate to heighten the center of muscles to pull inward the stylopodium under the joint for support of the body weight, retractors gravity of the body. A high position of the center of gravity brings better moving ability instead of for thrust, and protractors for recovery. When stability. Unstable posture would develop a an animal with the inferior type limb advances, it is enough to move the limbs fore-and-aft in the reflex tract of the nervous system and balance function of the brain, which will raise the sagittal plane. As a result, the same muscles could efficiency of movement more and more. Thus mammals finally achieved a high level of func- tion of movement and a larger brain. independently evolved the inferior type limb, while amphibians and other reptiles retain the lateral type limbs for adaptation of stability. Fig. 3. Bow structure in mammalian skele- When the limb type changed from lateral to ton. After BOKER (1935). 396 N. INUZUKA work for support and progression. An extensor and the deltoid ridge of the humerus. To pull the that passes before the shoulder joint or behind femur backward (extend the hip joint) the the hip joint, should be able to play a role of development of the gluteus medics and biceps supporting body weight in the "inferior" type femoris muscles is necessary and this develops the animals. gluteus surface and the ischial tubercle of the hip Dorsal and ventral elements of the girdles are bone. As a result, the dorsal elements of the markedly different except that they are situated girdles, scapula (GREGORY, 1912; WOLFFSON, dorsally and ventrally to the joint. The ventral 1950) and ilium are enlarged in animals with element extends inward to the joint, rather than inferior type limbs, while the ventral elements are just downward, and tends to join with another reduced with loss of the support function of the element in the median plane. On the other hand, adductor muscle. the dorsal element extends upward and never The clavicle, a ventral element of the shoulder joins with another. The ventral element from girdle, varies in its degree of development among which the adductors originate is suited to support mammals and is completely lost in many forms . the body weight in an animal with the lateral type Table 3 shows the presence, absence or degree limb, and the dorsal element that extends of reduction of the clavicle in each order sagittally is suited for muscle attachment that according to LESSERTISSEURand SABAN (1967). extends and flexes the inferior type limbs. Looking at the ordinal distribution from the In an animal with the inferior type limb the point of view of life style, the clavicle is very body weight is distributed so that the limbs are important in forms in which the anterior limbs flexed. That is, the body weight is distributed so move in three dimensions (the flyer, the glider, that the humerus inclines more forward because and the climber), except primitive it is set to be inclined anteriorly and the femur and insectivores. On the other hand, fully aquatic more backward because it is set to be inclined and cursorial forms lose their clavicles. In posteriorly. Therefore, it is important for the cursorial carnivores, the clavicle is present as a muscles to pull the humerus forward in the remnant, but is buried in the tendon and does not shoulder girdle and the femur backward in the work as a brace of bone. Less specialized forms, pelvic girdle in order to support the body weight. such as , rodents and edentates, retain To pull the humerus forward (extend the the clavicle and some of them lose it. Accordingly shoulder joint) the development of the we can conclude that the clavicle generally is supraspinatus and deltoideus is necessary and this reduced or lost in cursorial or aquatic forms. develops the supraspinatus fossa of the scapula There have been several views about the cause

Table 3. Development of the clavicle and mode of locomotion in mammalian order Evolution of the Shoulder Girdle 397 of the clavicle's reduction. HILDEBRAND (1960) zeugopodia and metapodials lengthen, and in the noted the advantage of no clavicle, which how- proximal portion of the forelimb the scapula ever may not be a cause of it: "In the carnivores lengthens to become a functional part of the limb. this freedom is increased by the reduction of the In this case, because the first limb segment of the collarbone to a vestige; in the ungulates the forelimb inclines oppositely to the corresponding collarbone is eliminated." ROMER (1966) also segment of hind limb, for the reason mentioned noted, "The clavicle disappears, thus freeing the above, the scapula corresponds to the femur shoulder from any bony connection with the functionally (Fig. 4). This idea occurred to me body breaking the jar of landing on the front feet while I thought about the reason of clavicle after a bound." LESSERTISSEUR and SABAN reduction (INUZUKA, 1988), but EATON (1944) (1967) said; "Reduction of the clavicle may be had long since noted the correspondence between related to decreasing of pressure of the fore- the forelimb plus scapula and the hind limb, and limb." YOUNG (1975) said, "the forelimb retains BABA (1988) probably got the same idea in- greater freedom than the hind and has tended to dependently without referring to the work as well. acquire other functions. ..the region to which the In other words, the scapula tends to move forelimb is attached (the thorax) must participate anteroposteriorly in the sagittal plane, like the in respiration, necessitating independent move- femur. At this time, the acromial end of the ments. For all reasons, therefore, power to move clavicle, which connects the to the the limb in various directions is more important scapula, draws a locus of an arc with an axis than firm attachment...." ROMER and PAR- revolving about the sternoclavicular joint and SONS (1977) said, "complete freedom of the produces horizontal component, so the presence shoulder blade from the body skeleton is de- of the clavicle must be an impediment for the sirable to relieve the jolts transmitted from front scapula to try to reciprocate sagittally (Fig. 5). legs to body and to increase the functional length I think that an experiment using a aclaviculate of the limb...." rat (JENKINS, 1974) proved this explanation. As A bony connection between the pelvic girdle a result, the clavicle would begin to reduce first and the spine, however, is not lost in bipedal from both articular ends to abandon its function forms to absorb the shock and to progress using only the hind limbs. In most mammals the body weighs more on the forelimbs than on the hind limbs. The more function the forelimb has, the more developed is the clavicle. It is hard to attribute advantages without the clavicle to the original cause of reduction. Therefore, in con- sidering the cause of clavicle reduction, a theory that does not contradict the life function of living forms should be chosen, and disadvantages for Fig. 4. Fore- and hind limb segments with life with the clavicle rather than advantages opposite inclination in mammalian without the clavicle should be looked for in order skeleton; broken line: analogy in both limb not to fall into teleology. segments, dotted line: in both limb segments. The elbow corresponds to Limb bones lengthen in mammals that are the heel functionally. adapted for running. Specifically, the stylopodia, 398 N. INUZUKA

Fig. 5. Cause of reduction of the clavicle. The presence of the clavicle which draws a locus of an arc (left), which interrupts the Fig. 6. "Freelimbization" of the clavicle movement of the scapula as it tries to in the brachiator. reciprocate sagittally (right). as a brace between the sternum and the scapula, forelimb. In the brachiators the clavicle reaches and to be buried in the tendon joining them. This its greatest length (MILLER, 1932). In order to condition is shown in some carnivores. It seems increase the range of reach of the forelimb, a that the clavicle has become progressively more more mobile clavicle acting as part of a free limb reduced to become lost completely in ungulates. is favourable, that is to say, the center of forelimb movement is transferred from the shoulder joint 3. Primate stage: brachiation to the sternoclavicular joint. Thus, the clavicle The shoulder girdle of the primates, parti- of primates, particularly apes, is developed as the cularly of the apes, has been investigated in detail most proximal segment of the forelimb (Fig. 6). for a long time (MIVART, 1867; AEBY, 1878; In cursorial mammals the scapula itself also MOLLISON, 1911; BOLTZE, 1926; FICK, 192.6; rotates, that is, "the shoulder blade pivots from SCHULTZ, 1926, 1930, 1937; WATERMAN,1929; about midway in its length, and the shoulder joint MILLER, 1943; INMAN et al., 1944; SMITH and at its lower end is free to move forward and SAVAGE, 1955; ASHTON and OXNARD, 1964a, backward with the swing of the leg" (HILDE- 1964b; OXNARD, 1967, 1969; ROBERTS, 1974; BRAND, 1960). The shoulder girdles of both JENKINS etjal., 1978). However, a few func- primates and cursorial mammals are common in tional studies of the clavicle have been done. having the center of forelimb movement off the Arboreal primates have various modes of shoulder joint: the sternoclavicular joint on the locomotion: vertical clinging and leaping in one hand and the scapula on the other. Thus, prosimians and brachiation and knuckle generalizing the points of view that the scapula walking in apes, as well as a quadrupedal of cursorial mammals and the clavicle in primates walking on the ground and in branches. Among are functionally incorporated into the limbs, a these, brachiation is considered here, because it concept "freelimbization" of the shoulder girdle is developed in apes. is obtained. The shoulder girdle, which originally The importance of the clavicle to connect the filled the role of connecting the axial with the sternum with the scapula is great because appendicular skeleton, changed its main function brachiators must bear the body weight by the from supporting to moving as each animal Evolution of the Shoulder Girdle 399 adapted itself to distinctive modes of locomotion, simply pull the whole shoulder inward." terrestrial running or brachiation among branches BREATHNACH (1965) also said, "...thus func- in forests. By the way, while the clavicle is lost tions that call for lateral motions in the limb in ungulates in which the scapula is "free- require a clavicle to support the scapula." limbized", why is the scapula not reduced in CAMPBELL(1966) said, "The clavicle maintains primates in which the clavicle is "freelimbized"? the distance of the scapula from the sternum and, There are three presumed reasons: the scapula acting as a strut, allows movements of the scapula cannot be lost because of the glenoid cavity which at a constant radius from the manubrium." must articulate it with the humerus; the scapula MOSELEY (1968) noted several functions of the develops as an attachment area of the shoulder clavicle, as follows: 1) It acts as a rigid base for girdle muscles; the develops as a part muscular attachments; 2) It forms a strut holding to articulate with the clavicle. the glenohumeral joint in the parasagittal plane; Such a long clavicle in primates is very con- 3) It increases the power of the arm-trunk venient for locomotion among branches, but may mechanism in certain areas of motion; 4) It pro- prevent the forelimb from reciprocating sagittally vides a protection for the major vessels; 5) It during quadrupedal locomotion for the reason serves a cosmetic function. SINCLAIR (1970) said mentioned above. The least possible load on the that its function was "to thrust the shoulder joint forelimb would be better for the shoulder joint away from the trunk so that the arm can move which shakes side by side with every step in its freely without being obstructed by the thorax." horizontal component. This condition may be BASS (1971) said, "Its function is to act as a strut related to the gait peculiar to primates or prop to the shoulder, thereby holding the (HILDEBRAND, 1967) that make them different scapula and upper limb laterally, backward and from other mammals. slightly upward." AKASHI (1973) said, "because missing clavicles doesn't obstruct one's life, the 4. Hominid stage: erect bipedal walking clavicle works to guard blood vessels and nerves Many authors have referred to the function and as an attachment of supplementary muscles of the hominid clavicle. HOLDEN (1855) said, of respiration." SHIPMAN et al. (1985) said, "Its chief use is to keep the apart "The major function of the clavicle is to hold the , that the arm may enjoy a freer and wider range of shoulder joint and the upper limb away from the motion. By moving the shoulder, you will find thorax, permitting free movement of the limb." that the clavicle acts as a prop, the fixed end of All these explanations are not wrong, but it the prop being at the sternal joint." TODD (1937) doesn't seem that they are sufficient to interpret said, "Its mechanical function is to give side the form of the hominid clavicle functionally. support to the shoulder joint for the wide and The function of the clavicle to keep the shoulder varied movements of the arm. It acts like a yard- joint at a certain distance from the sternum is arm, keeping the shoulder joint free from the common not only to hominids but to mammals chest, and has a definite though limited action with non-reduced clavicles. For example, LE in itself." YAPP (1965) said, "Where, as in man, GROS CLARK (1971) noted, "This bone, ... the arm can be moved inward, the clavicle or serves as a strut which, ... allows free movements collar-bone is large. If it were not present, con- of the limb away from and towards the body" traction of the breast muscles, instead of rotating or EWER (1973) said, "The main function of the the humerus round the glenoid cavity, would clavicle is to stabilize the lower end of the scapula 400 N. INUZUKA during abduction of the arm ...." This function clavicle. The clavicle of apes is primarily acted should be seen as inherited from primates rather on by body weight both in walking and in than one unique to hominids. The hominid brachiation: each clavicle is under the influence clavicle has most remarkable double curvature of the whole body weight in brachiation and both (Fig. 7). "A sigmoid vertical (in Man horizontal) clavicles are forced to carry the weight of the curvature is not generally well marked; it is most upper half of the body in knuckle walking. so in Man, .., the sternal curvature is much less OXNARD (1984) noted that the clavicle bore the than in Man, even in Troglodytes and Simia" tensile forces. On the other hand, hominid upper (MIVART, 1867). And "the lateral end of the limbs became free with erect posture, so that the clavicle lacks the pronounced cranial twist that shoulder girdle switched its role from support of characterizes it in the apes" (AIELLO and DEAN, the body weight to support of the free upper 1990). These features should be considered from limbs. the peculiar functions of the hominid clavicle that The free upper limbs are not supported only make it different from those of other primates, by bones of the shoulder girdle, scapula and especially other apes. clavicle. If only these bones supported them, the The mode of locomotion of apes is primarily clavicle would work as a cantilever and bear characterized by brachiation among trees and bending stress. I think that co-operation of the knuckle walking on the ground, whereas that of clavicle and trapezius muscle is rather important. hominids is erect bipedal walking. Let us consider The trapezius muscle originates broadly from the the effect that this change of mode of locomo- cranium and the spinous processes of the cervical tion has on the shoulder girdle, especially the and thoracic vertebrae and converges to insert on

Fig. 7. Clavicles of man and apes. From left to right, sternal end, cranial view, and acromial end. Drawn to scale except for Hylobates (x 2) Evolution of the Shoulder Girdle 401 the clavicle and the scapular spine. Among the curve, most distinctive in man. By the way, the muscle bundle the part that originates from the idea that bones respond to plastic deformation highest position, the cranium, mainly works as in the short term of phylogeny is obtained from a support of the free limb. This is evidenced by observation of the parietofrontal crest of the exceptional thickness of the free edge of the NAUMANN's elephant (INUZUKA, 1977) and of human scapular spine. Fig. 8 shows their the tibia of Desmostylus (INUZUKA, 1984). relative position schematically. The figure is According to MOSELEY (1968), "This posi- reminiscent of a crane; the clavicle and trapezius tion of the clavicle is characteristic to each in- muscle work co-operatively as an arm and a cable dividual and shows considerable variation. An of a crane to hang the free upper limb. In this upward slope of 10° to 12° from the horizontal situation the load of the free upper limb must be is a fair average." There should be an optimum borne by tension in the trapezius and compres- angle at a position a little above the horizontal sion in the clavicle. so that the stress is equally distributed through Thus the freeing of the forelimb with erect the cross section of the clavicle. If the idea that posture in hominids must have converted the the human clavicle is under compression is true, clavicle's main function to rather compression this idea could explain also the slope of the than tension or bending. I suspect that this con- clavicle. version compressed the clavicle from side to side When a man walks, the torsion of pelvis by to make a double-curve, sternal and acromial the fore-and-aft swing of the lower limbs is

Fig. 8. Support function of the clavicle in Fig. 9. Dynamic stability and regulation of the hominid skeleton. moment by clavicular length. 402 N. INUZUKA negated by the reverse torsion of the shoulder by of Tsukuba, Editor-in-Chief of this Journal, for

the reverse swing of the upper limbs. This is called giving me a chance to submit. I am grateful to the dynamic stability of the bipedal walking (Fig. Dr. Tasuku KIMURA of the Primate Research 9). The quantity of moment is given by a pro- Institute of Kyoto University for giving me a duct of the mass and the moment arm length. The chance to give a talk in the Hominization reason why the moment of the upper limb is Research Association. I also thank Dr. Nobutoshi balanced by one of the lower limb, which is much YAMAZAKI of Keio University and participants heavier than the upper limb, is that the shoulder of the Association for discussion and many width is larger than the distance across the suggestion. acetabula. As one grows older, the length and I thank Prof. Kevin PADIAN of the University weight of the upper and lower limbs increase. The of California, Berkeley and Dr. Gen SUWA of growth of limb bones in length ends with the the University of Tokyo for checking my English fusion of the epiphyses. The clavicle ossifies manuscript. I also thank Dr. Howard HUTCHI- earliest among all limb bones and its sternal SON of the Museum of Paleontology, University epiphyseal line closes last. The free limb bones of California, Berkeley and the Agassiz Museum of the upper and lower limbs close their epi- of Harvard University for giving me access to physeal line to setting the approximate weight of specimens of mammalian clavicles for compari- adult limbs during the later term of puberty. Even son. if the free limb increases just a little in length, 抄 録 it increases by a cube in mass. Therefore it ought to be more effective to adjust the moment arm 肩帯 の 進 化 一鎖 骨 を 中 心 と して 一 length than the mass in order to balance the moment of upper limbs with that of the lower 犬 塚 則 久 limbs. This is presumably a reason why the 鎖 骨 は特 異 な骨 で あ る.例 え ば,真 獣 類 の か らだ の clavicle closes its sternal epiphysis last, which 骨 で は唯 一 の 皮 骨 性 骨 で あ る.哺 乳 類 の 中で 退 化 し た delays the decision of its length than any other も の と発 達 して い る も の とが あ る.ヒ トの鎖 骨 は 二 重 free limb bones. が 最 も強 い.骨 化 点 出現 は早 い の に胸 骨 端 の癒 着 弯は 肢 骨 の中 で最 も遅 い.こ れ らの理 由 を探 る た め,肩 帯 Postscript の進 化 を比 較 解 剖 学 的 お よ び機 能 形 態 学 的観 点 か ら検 I introduce and criticize previous theories and 討 した. propose my opinion about some problems of the 皮 骨 性 肩 帯 は 四足 動 物 で は全 般 に退 化 傾 向 に あ る. shoulder girdle, especially the clavicle, in the 皮 骨 性 肩 帯 の実 体 は皮 骨 性 頭 蓋 の後 縁 で あ り,鎖 骨 は present paper. My opinion shown here is not そ の最 後 の名 残 で あ る.肩 帯 と腰 帯 は相 同物 で はな く, definitive and is only a tentative idea from a point 側 方 型 体 肢 を もつ 四足 動 物 の祖 先 が 獲得 した相 似 形 象 で あ る.体 肢 が側 方 型 か ら下 方型 へ 転 換 した の は,足 of view of comparative and functional morpho- の接 地 点 を重 心 に近 づ け るた め で,こ の 結果,肘 は 後 logy. I know of possible objections to many ろ,膝 は前 に 回転 す る こ とに な った.こ れ が 哺 乳 類 に points but my goal in publishing it here is to お け る前 ・後 肢 の形 態 差 の 発 端 で,肢 帯 や 基 脚 の 逆 傾 stimulate further discussion. 斜,肘 ・膝 ・踵 の 出 現 を 説 明 す る.哺 乳 類 の 肩 帯 は I would like to thank Prof. Banri ENDO of 自由 肢化 』 し,走 行 性 哺 乳 類 で は肩 甲 骨 の 自 由 肢 化『 the University of Tokyo for his helpful advice and が鎖 骨 の 退 化 を 促 した.一 方,樹 上 性 哺 乳 類 で は鎖 骨 a critical reading of the manuscript. I sincerely が 自由肢 化 し,新 た な 機 能 を 獲 得 した ため に,皮 骨 性 thank Prof. Morihiko OKADA of the University 肩 帯 の 退 化 傾 向 か ら一 転 して 発 達 す る よ う に な った. Evolution of the Shoulder Girdle 403

樹 上 性 類 人 猿 で は体 を 支 え て いた 鎖 骨 は,直 立 し た Biologische Anatomie der Wirbeltiere. Gustav Fischer, 人 類 で は逆 に 上 肢 を 支 え る よ う に機 能 転 換 す る.ヒ ト Jena, p. 52. BREATHNACH, A.S., 1965: FRAZER's Anatomy of thf で は上 肢 の 支 持 は鎖 骨 と僧 帽 筋 との 協 同 に よ っ て な さ Human Skeleton. J. & A. Churchill, London, p. 63 れ るた あ,鎖 骨 に は均 等 に圧 力 が か か る よ う に な る. CAMPBELL, B.G., 1966: Human Evolution. AldinE ヒ トの 鎖 骨 の形 態 は,こ の よ うな ヒ ト独 自の 機 能 か ら Publishing Company, Chicago, p. 139. 説 明 され るべ きで あ る. EATON, T.H. Jr., 1944: Modifications of the shoulder girdle related to reach and stride in mammals. J 二 足歩 行 の動 的安 定 性 を維 持 す るの に,上 下 肢 の 質 Morph., 75: 167-171. 量 とモ ー メ ン トア ー ム の 長 さの調 節 が 欠 か せ な い.質 EWER, R.F., 1973: The Carnivores. Cornell Univ 量 よ りは長 さ に よ る調 節 の ほ うが 容 易 な の で,鎖 骨 の Press, N. Y., p. 27. 骨 端 閉鎖 期 の遅 れ は,上 肢 の モ ー メ ン トJ'`_ム 長 の 調 FAWCETT, M.D., 1913: The development and ossifica- 節 に貢 献 して い る だ ろ う. tion of the human clavicle. J. Anat. Physiol., 47 225-234. FICK, R., 1926: Maf3verhaltnisse an den oberen Glied- References maf3en des Menschen and den Gliedmaf3en dei Menschenaffen. Sitzungsber. d. Preuss. Akad. d. AEBY, C., 1878: Beitrage zur Osteologie des Gorilla. Wiss., Phys. Mathe. Klasse, 1926, pp. 417-451. Morph. Jahrb., 4: 288-313. GARDNER, E., 1968: The embryology of the clavicle, AIELLO, L, and C. DEAN, 1990: An introduction to Clinical Orthopedics and Related Research, 58: 9-16. Human Evolutionary Anatomy. Academic Press, GREGORY, W.K., 1910: The orders of mammals. Bull. London, p. 352. Amer. Mus. Nat. Hist., 27: 4-524. AKASHI, K., 1973: Kinesiology. Ishiyaku-shuppan, GREGORY, W.K., 1912: Notes on the principles of Tokyo, p. 53. (In Japanese) quadrupedal locomotion and on the mechanism of the [明 石 謙,1973:運 動 学.医 歯 薬 出 版,東 京,p. limbs in hoofed animals. Ann. N. Y. Acad. Sci., 22: 5R1 267-294. APOSTOLAKIS, G., 1934: La clavicule de 1'Homme. HILDEBRAND,M., 1960: How animals run. Sci. Amer., Arch. Anat. Histol. Embryol. Strasbourg, 18: 202(5): 148-157. 169-180. HILDEBRAND, M., 1967: Symmerical gaits of primates. ASHTON, E.H. and C.E. OXNARD,1964a: Locomotor Am. J. Phys. Anthropol., 26: 119-130. patterns in primates. Proc. Zoo!. Soc. London, 142: HOLDEN, L., 1855: Human Osteology. John Churchill, 1-28. London, p. 109. ASHTON, E.H. and C.E. OXNARD, 1964b: Functional INMAN, VT., J.B. SAUNDERS, M. DEC. and L.C. adaptations in the primates shoulder girdle. Proc. ABBOTT, 1944: Observations on the function of the Zool. Soc. London, 142: 49-66. shoulder joint. J. Bone it. Surg., 26:1-30. BABA, H., 1984: Evolution of the pelvic structure. INUZUKA, N., 1977: On the origin of Palaeoloxodon In: Anthropology, its Various Development. Nikkei naumanni - A comparative osteology of the cranium Science, Tokyo, p. 151. (In Japanese) - . J. Geol. Soc. Japan, 83: 639-655. (In Japanese with English summary) [馬場 悠 男,1984:骨 …盤構 造 の 進 化.日 本 人 類 学 会 INUZUKA, N., 1984: Skeletal restoration of the des- (編)人 類 学 一 そ の 多 様 な 発 展.日 経 サ イ エ ン ス, mostylians: herpetiform mammals. Mem. Fac. Sci. 東 京,p.151] Kyoto Univ., Ser. Biol., 9: 157-253. BABA, H.,1988:Adaptation to the erect bipedal INUZUKA, N., 1988: Restoration of the extinct walking. In:Lectures on Anthropology. Vol.9, mammals, desmostylians. Biomechanisms, 9: 7-19. Yuzankaku-shuppan, Tokyo, p.113.(ln Japanese) (In Japanese) [馬場 悠 男,1988:直 立 二 足 歩 行 へ の 適 応.香 原 志 勢 [犬 塚 則 久,1988:絶 滅 哺 乳 類 デ ス モ ス チ ル ス の 復 元, (編)人 類 学 講 座9,適 応.雄 山 閣 出 版,東 京,p. バ イ オ メ カ ニ ズ ム 学 会(編)バ イ オ メ カ ニ ズ ム9. 113] 東 大 出 版 会,東 京,p.14] BASS, W.M.,1971:Human Osteology. Missouri JENKINS, F.A. Jr.,1974:The movement of the shoulder ArchaeologicalSociety, Columbia, p.101. in claviculateand aclaviculatemammals. J. Morph., BOLTZE, P.,1926:Beitrage zur Anatomie des Knoch- 144:71-84. engeriistesvon Hylobates syndactilus.Morph. Jahrb., JENKINS, F.A. Jr., P.J. DOMBROWSKI and E.P. 56:317-401. BOKER, H.,1935:Einfiihrung in die Vergleichende GORDON,1978:Analysis of the shoulder in brachiatingspider monkeys. Am. J. Phys. Anthropol., 404 N. INUZUKA

48: 65-76. primate scapula. In: Jenkins, F.A. Jr. (Ed.) Primate KENT, G.C., 1978: Comparative Anatomy of the Locomotion, Academic Press, N. Y., pp. 171-200. Vertebrates. C.V. Mosby, Saint Louis, p. 184. ROMER, A.S., 1966: Paleontology. Chicago KOCH, A.R., 1960: Die Fruhentwicklung der Clavicula Univ. Press, Chicago, p. 195. beim Menschen. Acta anat., 42: 177-212. ROMER, A.S. and PARSONS, T.S., 1977: The Vertebrate LE GROS CLARK, WE., 1971: The Antecedents of Body. W.B. Saunders, Philadelphia, p. 183-187. Man. Quadrangle Books, Chicago, p. 175. SCHULTZ, A.H., 1926: Variations in man and their LESSERTISSEUR, J. and R. SABAN, 1967: Squelette evolutionary significance. Amer. Naturalist, 60(669): Appendiculaire. Masson, Paris, p. 712, 754, 758-763. 297-323. MATTHEW, W.D., 1904: The arboreal ancestry of the SCHULTZ, A.H., 1930: The skeleton of the trunk and Mammalia. Amer. Naturalist, 38: 811-818. limbs of higher primates. Hum. Biol., 2: 303-438. MIKI, S., 1981: Phylogeny in vertebrates. In: Koseibut- SCHULTZ, A.H., 1937: Proportions, variability and sugaku-kakuron, vol. 4, Tsukiji Shokan, Tokyo, 21. asymmetries of the long bones of the limbs and the (In Japanese) clavicles in man and apes. Hum. Biol., 9: 281-328. 1981 :1 o) phylogenie = Age SHIPMAN, P., A. WALKER and D. BICHELL, 1985: The Human Skeleton. Harvard Univ. Press, Cam- bridge, p. 103. SINCLAIR, D., 1970: An Introduction to Functional MILLER, R.A., 1932: Evolution of the pectoral * girdle Anatomy, 4th ed. Blackwell Scientific Publications, and forelimb in the primates. Amer. J. Phys. Oxford, p. 381. Anthropol., 17: 1-56. SMITH, J.M. and R.J.G. SAVAGE, 1955: Some loco- MILLER, R.A., 1943: Functional and morphological motory adaptations in mammals. J. Linn. Soc. Zool., adaptations in the forelimbs of the slow lemurs. Am. 42: 603-622. J. Anat., 73: 153-183. STARCK, D., 1979: Vergleichende Anatomie der Wir- MIVART, G., 1867: On the appendicular skeleton of the beltiere auf Evolutionsbiologischer Grundlage, Vol. Primates. Philos. Trans., 157: 299-429. 2. Springer, Berlin, p. 470, 496. MOLLISON, T., 1911: Die Korperproportionen der SWINNERTON, H.H., 1923: Outlines of Paleontology. Primaten. Morph. Jahrb., 42: 79-304. Edward Arnold, London, pp. 331-332. MOSELEY, H.F., 1968: The clavicle: its anatomy and TODD, ME., 1937: The Thinking Body. Dance function. Clinical Orthopedics and Related Research, Horizons, New York, pp. 144-147. 58: 17-27. TODD, T.W. and J. ERRICO, Jr., 1928: The clavicular OXNARD, C., 1967: The functional morphology of the epiphyses. Am. J. Anat., 41: 25-50. primates shoulder as revealed by comparative WAKE, M.H., 1979: HYMAN's Comparative Vertebrate anatomical, osteometric and discriminant function Anatomy, 3rd ed. Univ. Chicago Press, Chicago, pp. techniques. Am. J. Phys. Anthropol., 26: 219-240. 252-255. OXNARD, C., 1969: Evolution of the human shoulder: WATERMAN, H.C., 1929: Studies on the evolution of some possible pathways. Am. J. Phys. Anthropol., the pelvis of man and other primates. Bull. Amer. 30: 319-332. Mus. Nat. Hist., 58: 585-642. OXNARD, C., 1984: The Order of Man. Yale Univ. WOLFFSON, D.M., 1950: Scapula shape and muscle Press, New Haven, p. 134. function, with special reference to the vertebral RADINSKY, LB., 1987: The Evolution of Vertebrate border. Am. J. Phys. Anthropol. New Ser., 8: Design. Univ. Chicago Press, Chicago, p. 141. 331-341. REWCASTLE, S.C., 1981: Stance and gait in tetrapods: YAPP, W.B., 1965: Vertebrates: Their Structure and an evolutionary scinario. Symp. Zoo!. Soc. London, Life. Oxford Univ. Press, New York, pp. 416, 420. 48: 239-267. YOUNG, J.Z., 1975: The Life of Mammals. Clarendon ROBERTS, D., 1974: Structure and function of the Press, Oxford, p. 86.

犬 塚 則 久 東京大学医学部解剖学教室 〒113・東京都文京区本郷7-3-1

Norihisa INUZUKA Departmentof Anatomy, Universityof Tokyo, Facultyof Medicine, Hongo, Tokyo 113,Japan