THK HOMOLOGIES OF THE ARM AND liEG 51. R. DRENKAN Tfie Anntom!j Dcpartmmt of the lJnicrr8ity of Capetoirn EIGHT FIGURF,S The general morphological resemblaiice between the arm and leg is obvious, but wheii attempts are made to draw a detailed comparison between them there is great divcrsitj- of opinion as t o what are corresponding structures. This is not t o be wondered at, seeing that anatomists are iiot agreed as t o the basis 011 which the comparison must be made. Ilorcover, the modifications in each limb, adaptiiig it to its special function, have considerably blurred the resemblance between them. Especially is this tlie case in the liumaii arm arid leg, so that a comparison between these is particularly difficult, yet it is largely from a study of their affinities that deductions have been drawn and applied to other vertebrates. I t is important, therefore, on this account that anatomists should t r y to reach some agreement a s to tlie nature of the relationship which exists between the arm and leg. Uiitil this is done, it seems idle to go into very great detail in the comparison, and there is a dangcr. of attemptiiig t o establish homologies where none exist. SEBIAL IIOMOLOGY The orthodox view on this matter, for which Flower( 1), Huxley( 2), aiid others, are largelj- responsible, is that the arm arid leg a r e serial1;v homologous, and it is comparisons made on this interpretation, which a r e incorporated in the various text-books, which allude to this subject. In demonstrating this method of comparing the limbs it is usual t o 113 Tllh A I A T O N I C A L I<ECOKD, \ O L . .js. NO. 2 114 M. R. DREXXAN extericl them at riglit angles to the axis of the body in a position as similar a s possible to that occupied by them in the primitive and embryonic condition. That is to say, the borders containing the thumh and big toes are directed toward the head, while the other borders are directed caudally. At the same time the extensor groups of muscles in each liml) look dorsally, while the flexor groups look vent rally. The cranial border of each limb is called tlie greaxial border, and the caudal border is called the postaxial borcler. these terms tieing used in reference to the axes of the 1iml)s themselves. The flexor muscles are referred to by some authors as preaxial and the extensors as postaxial in ref erence to the axis of the trunk, hut in this paper these terms will not be used in this latter, but always iii tlie former sense'. According to this serial view of the relationship between tlie limbs, structures situated on the preaxial border of one limh are considered to be homologons to structures on flit. preaxial border of the otlier limb, arid so for the postaxial borders. That is to say, the arm is regarded as the exact morpholog<cal counterpart or cast of the leg of the same side, a i d both limbs from the same side of the body coiild he fittwt into the same type of mold (fig. 1). SYMMETRICAL HOMOLOGY Another view, which is lieltl by a small minority of anato mists, notably Humphrey(3), Parsoiis(4), Geddcs(S), wild Keith( G ) , but u-hicall has found little support in the text-books, is that the arm is symmetrically liomologons to the leg. According to this view, there is a relationship between the arm and leg of the same nature a s exists between a riglit limh and a left limb. That is to say that, just as there is a hilateral symmetry iii tlie body manifcsting itself hy each limb being the mirror image of its fellow, so there is a craniocaudal or hiterminal symmetry which is espressetl i i i a mii*rorimage or symmctrical homology hctweeii tlic forelimh and the liincllimb on each side. 115 HOMOLOGIES OF T H E A R M A N D LEG Viewed in this light, the preaxial border of the arm must be homologized with the postaxial border of the leg. Accordingly, Geddes has designated these borders as being apomphalic, or distal to the umbilicus, while he refers to the other two comparable borders of the limbs as paromphalic (figs. 1 and 2). Fossa. Do rsurn Scapulae. PRE-AxIAI; (Apomphalic) f- p S ub 8 cnpul.ar T r o chant e r - ---medial ; j ! Ulna. (Paromphalic) + BORDER. BORDER. Radius. k I 11 i Pig. 1 Diagrammatic representation of the bones of the arni and leg to show homologous parts. Fig. 2 Dingrain showing the hiterminal syinmetrp between the arm and leg of a primate. 116 M. R. DKENNAN DISC~TSSIONOF ‘rij E TWO HOMOT~OGIES The author of the present paper is unable to accept the theory of serial homology as usually applied to tlie limbs, a i d he advocates thc substitution of the theory of symmetri(*ti1homology f o r this scrial interpretation iii any comparison which is instituted. Wit11 this object in view, ail adrersc witicism of the serial theory is submitted, and a detailed comparison is made between the arm and leg on a symmetrical basis. Previous writers have only adumbrated this view, or they liave applied it merely to oiic part of the limbs. 111 t h e present paper the full implications of comparing tlic arm and leg on this interpretation are worked out. The results show clearly that a morphological symmetry exists arid tliat it must be regarded as a very important factor in any cwmparative study of the forelimb arid tlie hindlimb. Before dealing with tliis latter rdationship, it is iiecessarv to refer briefly to some of tlie ur1satisfactor:- features of the serial theory. The foundation of this interpretation of thc homology of ilie limbs is the assumption that, because of their geiitwd resemblance, the thumb tiiitl the big toe are liomologous. This eorrespoiitlence is usually taken for graiitcd, lmt it will hc shcrwii later that tlir* homolog-,.yof the pollcs and hallux is not only not beyond all doubt, as Huxley thought, but that it is e1-w highly problematical. Starting from this supposed 21omology of the thumb ant3 big toe, it is asserted that strnctui*cs0 1 1 the thumb ( p i w s i a l ) hortler of the arm are morpliologically rcIated to striic.tures 011 the big toe (preaxial) loorder of thcl leg. The following are some of the unsatisfactory ons sequences of the applicatioll of this dehatable criterion of comparison. ‘rhus it postnlatcs that the veil t ral su1)scapular fossa, with its one muscle lying iii front of tlic long liead of tllc triceps, is homologons to the dorsal surface of the ilium, with its several muscles lying behiiid tlic loiig head of the quadriceps (rectus femoris). The greater tuberclc of the humerus, \l?ith its threr miiscle attacliments, is homologized to tlre lesser t~ocliantcr of tlie femur with its siiigle t e i i d o ~attiichmerlt. ~ HOMOLOGIES OF THE ARM A N D LEG lli The i~ndoul~tedly homologous triceps and quadriceps muscales go to bones which are not homologous on the serial interpretation, namely, to the ulna and the tibia, respectively. The biceps muscles are inserted into the radius and the fibula, respectively, yet these are not regarded as homologous bones. The medial epicoiidyle of the humerus, from which the flexor group of muscles arises, is likened to the lateral epicondyle of the femur, from which the extensors arise in most animals. The flexor pollicis longus arises from the radius, whereas the flexor hallucis longus arises from the fibula, yet these are non-homologous bones, according to the serial interpretation. It is customary to adduce different functional requirements as the explanation of all tliese and other anomalies in the structural correspondence. At the same time numerous rotations and tortions, some ingenious and others clumsy, have been postulated to accoiint for the discrepancies in tlie comparison, but riot one of these seems to be an adequate explanation. Again the well-known plasticity of muscles arid their capacity for migration has been a convenient method of explaining the wholesale ‘migrations ’ indicated above. Further, there is no end to the chaos which may be introduced by this theory, because of the difficulty of knowing in each anomalous instance cited whether it is the arm or the leg which has remained true to type or has become modified. That is to say, it is difficult t o decide, for example, whether it is the tubercles of the humerus or the trochanters of the femur which have become reversed from the fundamental plan. Nothing in the above criticism of the unsatisfactory and unacceptable results from a serial comparison o r of the explanations of its modification, which are usually adduced, is to be taken as in any way indicating that the author does not accept the well-known and simple rotations by means of which tlie embryonic limbs gradually assume tlie foetal and adult positions. These changes in position are generally accepted, but they do not account for any of the above discrepancies. In seeking for a true standard of comparison between the limbs, one is struck by the fact that when a preaxial structure iii one limb does not resemble in form a corresponding preaxial structure in the other limb it does resemble the postaxial structure in that limb. This suggests the possibility of an exactly opposite relationship being the fundamental one or a t least tlie chief modifying factor in the morphology of the limbs. According to this view, which has been alreadv alluded to a s a biterminal symmetrical homology, preaxial ( apomphalic) structures in one limb correspond to postaxial (apomphalic) structures in the other limb. Bilateral symmetry is the modification which nature has adopted to enable the right and left limbs to take an equal share in halancing the body weight around the center of gravity. It would be very surprising if there was not something analogous in the relationship between the forelimb and the hindlimb of each side to enable them to support each end of the body and thus complete the balance rouiid the center of gravity (fig,2 ) . If this reason for a hiterminal symmetry exists in quadrupeds, and it should be noted here that the possession of just four limbs is a fundamental characteristic of the vertebrates, there is no reason to doubt its applicability t o a comparison of the human limbs. If this conception is applied to a study of the arm and leg, the first and most striking result is to get rid of practically d l the difficulties encountered in applying the serial cornparison. Figure 3 shows tlie bony parts in the region of the left shoiildcr compared with the corresponding parts in the left pelvic girdle. No less striking is the likeness presented by the left shoulder muscles to the left pelvic muscles in figure 4. in figure 5 the similarity between the triceps and the quadriceps is obvious, and they are inserted into bones which on a symmetrical basis a r e homologous. The long head of the biccps femoris corresponds to tlic short head of the biceps brachii. The short head of the biceps femoris is represented I)y the occasional third head of the biccps brachii. The biceps muscles arc inserted into the radins and the fibula, rcspec- HOMOLOGlES OF THE ARM AND LEG 119 tively, in the arm and leg, and these arc symmetrically homologous bones. As we proceed toward the distal extremity of each limb, the symmetry becomes more blurred. This is largely because, Femoral Artery Fig. 3 Corrcspoiitling parts a, b, e, ctc., in the left shoulder and pelvic yegions. Fig. 4 Corresponding museles in the left shoulder and pelvic regions. according to this basis of comparison, the thumb has to be homologized with the little toe, and the little finger with the big toe. At first sight, this appears to be a difficult homology t o sustain, but in reality a comparison of the forearm and 120 M. H. DBENNAN harid iii one lirnl) with tlic leg ant1 foot in the other limh is more harmonious on this interpretation than on the serial one. Most attempts to homologize the limbs on eveii a serial hasis have broken down in the muscles of these regions. I t is riecessary to concede, liowcve1*, that there is a greater apparent rcsembliincc hetweeii tlic thuml) and the big toe tliaii lwtween the thumb and the little toe, but this can be accouiiteci for by regarding it as a secondary morlificatioii of a fullctiorlid riaturcl. Iri Gctldes’(5) opinioiithe evidence available from the nervous system goes to show that secondarily, that is subsequent to the estahlishintbiit of the doriiiiiaiiee of the segmented ceplialisetl part of the body, when the linibs are required to hinction as part of a n miirrial with a definite head and tail. the great toe approxiiiiates in type to the thumb, taking o v t ~ from the thnmb’s real hornologne, the little toe, a sufficiency of its n t w e supply and of its musculature fully t o equip it for its n w pohition (cf. aponiphalic, fibular. origin of flexor hallticis loiigiis ) . In other words, the correspontlence betmeeri the thmnb and gwat tor i.; a11etsaniple of heterogeneons homoplasy. rl?liwe is 110 evidence of this c+eplialization of the liaiitl ant1 foot having affected the preaxial borders of the limbs more proximally, as is assumed by tlie theory of serial homology. Even i n tlie l i a d a i d foot this cephalization is hy no meails ;i constant feature of tlie vertebrate orgaiiization. I n the vory large group of animals iii which cw-tain of the digits h a w been sacrificed f o r functional purposes, it is usually the pollex and the hallux which a r e the first to go, showing that arc! the least stable of all the digits, and therefore the lwst rc1ial)le of all the digits a s a founrlatioii for it morphology applicable to tlic rest of the limb. Kor is there ally evidence that in the origin and develop a i d ontogcriy) of the limbs there is aiiy mciit (pli~-logc11~7 tlcfinite serial modification. The ai-m a i d leg h d s have ti very similar origin in the embryo in close relationship to a definite but different number of segments in each case, but it is still a matter of dispute as to how far, if at all, tlic somitcs enter into tlie early coiistitutioii of the limbs, so that HOMOLOGIES OF T H E ARM AN11 LEG 121 arguments hasctl on the origin and early development of the limbs are of t~ouhtfnlvalue at the present juncture. ('ertainly, so f a r as our present knowledge goes, there is no reason to regard tlie preaxial part of the one limb bud as corresponding more to the preaxial than to the postaxial part of the other limb bud, except as regards situation and the later thumb and big-toe modification. I n this connection Geddes(5) has, in fact, advanced a theory as regards the origin of the limbs, which considers them to be symmetrical ab initio. A case can therefore be made out for attempting to unravel the symmetry between the forearm and leg and between the hand and foot on a morphological basis. In order to simplif5the comparison, the left arm has been auperimposed on the right leg in figure 8. The left arm, being the mirror image of the left leg, must therefore be a n exact morphological cop:- of the right leg, if the basis of comparison is correct. The correspondence between the shoulder and the pelvis aiid between the hracliium and the thigh is again ohvious. The olecranon epiphysis corresponds to the patella. The posterior border of the ulna corresponds to the anterior border of the tibia. The extensor carpi ulnslris corresponds to the tihialis anterior. They arise from symmetrically homologous bones, and they have this also in common in their insertions, that they both reach symmetrically homologous bones, the former the fifth metacarpal an(! the latter the first metatarsal. The extensor yuinti digiti proprius, being in any case most probably a member of the deeper stratum of the primitive extensor tendons, is homologous to the deep extensor brevis hallucis. The tendon of the extensor digitorum communis, which goes to the little finger, corresponds to the extensor hallucis longus. The three remaining tendons of the extensor digitorum communis together with the extensor pollicis longus may be homologized with the extensor digitorum longus. The exteiisor indicis proprius corresponds to the most lateral tendon of the extensor digitorurn brevis on the foot. The tendons of this latter muscle which go to tlie 122 M. R. DRENNAN second and third toes a r e unrepresented in the hand, unless c~oi.rcsr)(~~i(lin~ short ex tensors to tli e medius and annu1ai.is digits happen to be present a s abnormalities, as is frequently t h e case. The extensor pollicis lnrevis ant1 the abductor pollicis loiigus, which are usually united into one muscle in forms lower than man, can he suitably homologized by the peroneus tertius alone or with the addition of the occasional peroneus quintns. The extensor carpi radialis brevis and the extensor carpi radialis longus are well represented by tlie peroneus brevis and tlie peroneus longus, respectively. The crossing of the sole by the peroneus longus tendon is merely another example of the way in which the big toe, being short of muscles of its own, has borrowed from the rich supply of muscles of the little toe. Until the phylogeny of each individual muscle is completely worked out, it is impossihle to maintain that the homologies stated are individually correct, hut they are sufficiently accurate to show that the little toe with its rich group of peroiisal muscles is more comparable to the thumb aiid the numerous muscles related to it than is the big toe (fig 6). On the flexor aspect of the limbs the flexor carpi ulnaris is homologous t o the tibialis posterior. The V-shaped arrangement of muscles formed on each side of the biceps hrnchii by the brachioradialis laterally a d the pronator teres and the flexor carpi radialis medially can be homologized to the lateral mid medial heads of the gastrocnemius. The palmaris longus is best represented by the plantaris. The flexor digitorum sublimis is homologous to the soleus togethcr with the flexor digitorurn brevis, although the tendon to the thumb has become lost in the arm and the tendon to the big toe in the foot. The flexor digitorurn profundus and the flexor pollicis longus together correspond to the flexor digitorum longus, the flexor hallucis longus, and tlie accessorius. This was originally oiie continuous sheet in each limb, but it has hwomc broken up differently in each case to serve a different purpose. hforphologically, how-ever, it is easy to compare the most fibular with the most radiaI tendon, however little tlie former ma? be differentiated a s compared with the latter. HOMOLOGIES O F THE ARM A N D LEG 123 UU 5 4 . " 3 2 7 Fig. 6 Corresponding parts in the left arm and thigh regions. Fig. 6 Corresponding parts in the lrft liand a n d foot regions. Fig. 7 (a) Arrangement of the bones in a primitive unmodified hand or foot. (1)) Arrangement o f niuscles in thr hand or foot of an ape. Arrows indicatc their position in the human hand and foot. (Modified after Keith.) 124 M. R. DRENNAK Tlic morphology of the hand and foot is a very large sub ject, hut for the present argument a very few facts mill suffice). In tlie most primitive and unmotlified types of hand ant1 foot, met with in certain reptiles arid amphibians, the various skeletl-ll ant1 muscular elements are alike in numlwr aiitl ( a ) represents such a arrangement in each case. Figure ‘i hand or foot, and it is equally notcworthy lhat in each of such hatids and feet there is very little specialization of the polles and hallux. Each manus ant3 pcs is practically symmc~trival round its central digit, so that in such forms the thnml) is readily comparable to the little toe. Jfoceover, Keith (6) has shown that the muscles of tlic haiitl and foot of the ape arc arimiged identically (fig. 7 (1)) ). Tho arrows show that the muscles have migrated in a similar mimner to form the human liand aid foot, and that it is only a slightJ diffcrence iii the form of the muscle and a cliaiige of 1I ii 111 wliich c onst it u t es tli e di tlt’cir cnc e 1 t ween the 11arid ai I ( I tlw foot. A g1,znce at this figure and siiriilar reasoning will sliow how it is possible for oiie to maintain that the muscles 011 tlic lateral side of the hand are homologous to those 011 tlic lateral side of the foot in spite of their apparent difference in form, for they are morphologically identical i l l arimigemeiit . Tlic reduction in iiumher of the phalaiiges of the pollex aiul liallux and the aiiomaloiis ossificatim of the first metwcarpal arid tlie first metatarsal bones do not present anp insuperable tli&culty to tlie theory of symmetrical homology. Abnormal liuman thumbs with three phalanges and little tocs with only two a r e sufficiently common abnormalities to show that the thumb is a n ordinary and an unstable digit and that reduction in the number of the phalanges can readily take place. McOregor(7), in a recent paper on the morphology of the thumb, has come to the conclusion that the anomalous ossification of the first metacarpal and of the first metatarsal is a functional adaptation and can be adequately explained on mechanical grounds. (1 HOXOLOGTES O F THE ARM A N D LEG 125 .Adductor Magnus ~ -0leeranon epiphysis = Patella -Posterior border of Ulna :Anterior border of Tibia ~ l e x o rcarpi LLInaris Tibialis p o g t e r i o s Quint i Di i t i Halluc is L i D finger Fig. 8 Diagmm showing how the muscles of the left arm correspond exactly t o those of the right leg. 126 M. R. DRENNAN As regards the boiies of tlie foot, therefore, we must homologize the navicular with the calcaneus, tlie lunate with the talus, the triyuetrum with the navicular, the greater multangular with the cuboid, the lesser multangular with the third cuneiform, the capitate with the second cuneiform, aiid tlie hamate with the first cuneiform, while the first, second, third, fourth, and fifth metacarpal bones and accompaiiying plialangcs must correspond t o the fifth, fourth, third, second, and first metatarsal bones and accompanying phalanges, respectively. Arteries are regarded as too variable morpliologically t o be of much value as standards of comparison. According to thra serial homologists, the 1)rachial artery is homologous t o the inferior gluteal artcrp, the uliiar is represented hy the posterior tibial artery, and the radial by the iiiterrial saphenous artery. On a symmetrical interpretation the hrachial artery becomes readily comparable with the femoral and its continuation the popliteal artery. The uliiar coi*responds t o the posterior tibial artery, the radial artery to the peroneal, and the posterior interosseous artery to tlic iuitcriur tibial artery. Altogether, tlie comparison o i l tlic symmetrical basis is much more harmonious than on ttic serial hypothesis. As regards using the spinal nerves as keys to homologies or as standards of comparison, it is only necessary to point out how they have failed as such in the serial comparison. Homologous fibers take a totally different course in the two limbs. The segmental distribution of the cutaneous iiervw is as favorable to a symmetrical a s to a serial view of the morphology of the limbs ; according to the former view, apomphalic areas in one limb correspond to apomphalic areas in the other limb, whereas the latter view makes preaxial areas in one limb correspond to preaxial areas in the other limb. In the case of the motor nerves, Geddes(5) adduces cxperimental evidence in support of the law “that of two limb muscles that which is the more superficial is usually supplied HOMOLOGIES OF THE ARM A N D LEG 12’1 by a nerve apomphalic to that which supplies the deeper muscle.” The usual statement in the text-books, that the fifth cervical nerve supplies the abductors and the lateral rotators of the arm, and the flexors and the supinators of the forearm, that the sixth cervical nerve supplies the opponents of these, that the seventh cervical nerve supplies the flexors and the extensors of the wrist, that the eighth cervical nerve supplies the flexors and extensors of the fingers, and that the first thoracic nerve supplies all the short muscles of the hand, would seem to indicate that the craniocaudal series of nerves find their expression on the limb morphology not in any preaxial and postaxial modification, but on the very characteristic proximodistal segmentation of each limb. SUMMARY 1. Tliere is no evidence of a general serial modification of the arm and leg, such as is usually described as being present. The ccphalization of the thumb and big toe is probably a secondary modification of a functional character, which is not universally present in all forms and which does not dominate the general morphology of the rest of the limbs, as it is usually supposed to do. 2. On the other hand, there are many facts which support the view that the limbs have been modified according to a craniocaudal or biterminal symmetrical plan. This may also he part of the fundamental plan of organization of the limbs, and in any case it must always be taken into consideration in making any comparison between them. 3. The segmental homology between the proximodistal segments of the arm and leg is the most definite existing homology, and it is suggested tentatively that this proximodistal segmentation, and not any preaxial and postaxial modification, is the expression of the genetic relationship of the limbs possibly to a series of somites arid certainly to a series of segmental nerves. 4. The arm and leg thus present a segmental homology, hut the elements in each segment are mirror images of each 128 M. R. DBBKNAN other, so that there is also a symmetrical homology. This symmetry is blnrrecl distally by tlic ccphalizatioii of thc 2iaiitl and foot to form the thumb and big toe-an example of heterogeneous homoplasy.