Limb-Somite Relationship: Origin of the Limb Musculature

/. Embryo/, exp. Morph. Vol. 41, pp. 245-258, 1977 245 Printed in Great Britain © Company of Biologists Limited 1977

Limb-somite relationship: origin of the limb musculature

By ALAIN CHEVALLIER, MADELEINE KIENY AND ANNICK MAUGER1 From the Laboratoire de Zoologie et Biologie animale, Universite scientifique et medicate de Grenoble (Equipe de recherche associee au CNRS n° 621, 'Morphogenese experimentale')

SUMMARY Quail-to-chick grafting experiments performed during the third day of incubation demonstrate that somites can contribute to limb development. In orthotopic recombinations, migrating cells originating from the grafted unsegmented or segmented somitic mesoderm adjacent to the wing or leg field end up in the musculature respectively of the wing or the leg, where they express exclusively myogenic properties. Thus, in these heterospecific recombinations, the anatomical muscle has a double origin: muscle bulk of somitic origin; tendons and connective tissues of somatopleural origin. Similar features are observed in heterotopic recombinations with (segmented or unsegmented) somitic mesoderm located cranially or caudally to the limb levels. In the reverse chick-to-quail grafting experiments, the somitic participation to the limb mesoderm can also be observed. But it is less regular than that obtained in the quail-to-chick recombinations, and the muscle bulk is made up in various proportions of graft-originated somitic cells and of host somatopleural cells. The possible existence of juxtaposed and interdigitated myogenic and tendinogenic compartments is discussed in view of the dissimilarity between the results of the two kinds of heterospecific recombinations.

INTRODUCTION The limb-bud arises as a local thickening in the somatopleural mesoderm of the body wall at somite level 15-20 for the wing-bud and 26-32 for the leg- bud. These thickenings subsequently elongate along their proximo-distal axis, i.e. perpendicularly to the cranio-caudal axis of the embryo. During this early period of development the mesenchyme of the limb-bud looks homogeneous and appears to be composed of a single cell type. Later, from stage 24 of Ham- burger & Hamilton (1951) on, these mesenchymal cells differentiate into muscle, cartilage, bone and connective tissue cells. Pursuing our investigations on the relationships between the somites and the limb-bud (Kieny, 1969, 1970, 1971, 1972), the question arose whether bird 1 Author's address: Laboratoire de Zoologie et Biologie animale, Universite scientifique et medicale de Grenoble, 38041 Grenoble Cedex, France. 246 A. CHEVALLIER, M. KIENY AND A. MAUGER somite cells participate in limb myogenesis in the same manner as that des- cribed in lower vertebrates [Selachians: Dohrn (1884); Braus (1899, 1906); Borisov (1970), Teleosts: Corning (1894), Amphibians: Borisov (1970)]. While it is well established that the bird limb skeleton and the dermis (Dhou- ailly & Kieny, 1970, 1972; Kieny, 1960, 1971; Pinot, 1970) arise from the somatopleural mesoderm, it is not clear yet whether the musculature is of somatopleural or somitic origin or whether limb muscles form from a combination of somitic and somatopleural cells. Up to a recent date, the assumption of the somitic participation in limb musculature, hence the previous penetration of somite cells into the limb-buds, could not be confirmed experimentally. Grafting (Hamburger, 1938) as well as carbon-marking (Saunders, 1948) and tritium-labelling (Seichert, 1971) experiments failed to demonstrate a somitic contribution to the development of the limb. All these experiments were carried out at late stages, when the limb- bud was, at the earliest, about to bulge out. Despite the in vitro recombination experiments of Gumpel-Pinot (1974) which indicated that somitic material takes part in the formation of the wing-bud, the participation of somite cells in limb myogenesis remained unsolved. Since the presumptive myoblasts are histologically indistinguishable from other mesenchymal limb-bud cells at early stages of development, only longlasting labelling experiments could reveal the origin of the limb musculature. Using the biological cell labelling technique advocated by Le Douarin & Barq (1969), Christ, Jacob & Jacob (1974#) on the one hand, and our group [Chevallier, Kieny & Mauger (1976)] on the other hand, performed similar experiments that consisted in replacing the somitic mesoderm of the wing level of the chick host by somitic mesoderm from a quail donor. These experiments showed that indeed somitic cells participate in limb myogenesis. The experiments reported in this paper were undertaken to answer the following questions. (1) What is the extent of the somitic participation in the morphogenesis of the limb musculature ? (2) Are the somite-limb musculature relationships restricted to the somitic mesoderm of the corresponding limb level ? (3) Are these relationships the same in both kinds of xenoplastic grafting experiments between chick and quail ?

MATERIAL AND METHODS The experiments were carried out on chick embryos either of a White Leghorn strain or of a crossing of Wyandotte x Rhode Island Red and on Japanese quail (Coturnix coturnix japonica) embryos, during the 3rd day of incubation. The exact stage attained at theJime of operation was specified by the number of pairs of somites. The experimental design comprised replacing the somitic mesoderm of the wing or the leg level of one species by somitic mesoderm from the other species Origin of limb musculature 247

Table 1. Distribution of experimental cases according to origin and stage {in pairs of somites) of grafted tissues and host embryos

Grafts Hosts operated on the Wing level Leg level Cephalo- caudal Limit stages of donors Limit stages Number Limit stages Number level of of of origin Chick Quail Chick Quail cases Chick Quail cases Cervical — 14-18 14-23 — 8 — — — Wing 12-20 12-17 9 — 17-24 3 10-23 12-22 20 Flank — 20-24 13-15 — 6 — — — Leg 16-22 15-20 4 — 18-22 4 18-26 15-16 3 21-25 — 5

according to a procedure that has become routine and basic to many experiments in our group (Chevallier, 1975; Kieny, 1972; Mauger, 1912a, b). Figs. 1 and 5). When the translocations were performed at the wing level, the chick or quail hosts ranged in developmental stages from 12 to 17 pairs of somites (with the exception of 5 out of 50 cases ranging in stages from 20 to 23 pairs of somites) so that unsegmented or only partially segmented somitic mesoderm was excised; whereas, when performed at the leg level, only unsegmented somitic mesoderm was extirpated, the hosts ranging in developmental stages between 17 and 25 pairs of somites. The grafts composed of a row of five to seven somites or presumptive somites were obtained from the cervical region (somites 5 through 10 inclusive), the wing region (presumptive somites or somites 15 through 20 inclusive), the flank region (presumptive somites or somites 21-25 inclusive) or the leg region (presumptive somites 26 through 32 inclusive). The donors ranged in developmental stages between 10 and 26 pairs of somites (Table 1). The orientation of the graft was recorded for each operation and only aadd homopleural operations were made on the left or right side of the embryos so that in no case was the medio-lateral axis of the graft reversed. The operated embryos were allowed to develop until day 6 to 12 of incubation. They were then sacrificed, fixed in Helly's solution and immediately embedded in paraffin. Longitudinal or transverse sections of the limbs, 7/*m thick, were stained according to the nuclear reaction of Feulgen and Rossenbeck. 248 A. CHEVALLIER, M. KIENY AND A. MAUGER

Donor

Somite no. 14

Graft: somite Somite no. 14 mesoderm from level

Ectoderm

Discarded

Fig. 1. Experimental scheme of orthotopic replacement of the somitic mesoderm from the wing level by heterospecific somitic mesoderm. Combinations between chick host and quail donor, and vice versa.

RESULTS In this paper, we focus our histological investigations on the intrinsic limb musculature. The observations that were made simultaneously on the limb girdles will be published separately (Chevallier, 1977).

A. Quail grafts to chick hosts 1. Orthotopic transplantations Wing level. There were 20 chick embryos in which the somitic mesoderm of the wing level was orthotopically (Fig. 1) replaced by quail somitic mesoderm. FIGURES 2-4 Quail grafts to chick host. Fusiform muscles. Fig. 2. Forearm muscle six days after the orthotopic replacement of the somitic mesoderm at the wing level. Quail donor and chick host: 17 pairs of somites, x 284. Figs. 3 and 4. Two cases of forearm muscles seven days after the heterotopic replacement of the somitic mesoderm of the wing level by somitic mesoderm from the leg level. Quail donors: respectively 16 and 20 pairs of somites. Chick hosts: respectively 15 and 16 pairs of somites, x 672. The interrupted lines represent the limit between chick (C) and quail (Q) cells. CT, surrounding muscular connective tissue comprising chick cells. 250 A. CHEVALLIER, M. KIENY AND A. MAUGER In all wings of the operated side, quail cells of somitic origin were found in the upper arm, lower arm and hand. While the skeleton and dermis were always and uniformly composed of chick somatopleural cells, the musculature, on the contrary, was made of both somite-originating quail cells and somatopleural chick cells. However, these cells were not distributed at random, but the bispeci- ficity was highly organized. Indeed, 4-7 days after the operation, the bulk of the fusiform muscles were mainly composed of quail cells whereas the attached tendons, at both ends, were constituted by chick cells (Fig. 2). Moreover, in the bulk, the presence of quail nuclei was selectively restricted to the differentiating muscle fibres proper. We found only one exception to this picture; namely, in one of these 17 specimens, the bulk and the tendons of one fusiform muscle from the zeugopod were completely made up by chick cells. In the case of penniform muscles, the muscle fibres of quail constitution laid at an angle to a central tendon, that was composed solely of chick cells (Figs. 6, 7). Finally, in the particular case of the elbow muscles (anconeus and brachialis muscles), which are directly inserted to the skeletal elements without intervening tendinous tissues, quail muscle fibres were found to abut directly against the perichondrium of the humerus and of one of the zeugopodial bones. In the older specimens sacrificed at a total incubation time of twelve days (three cases), the architectural arrangement of the musculature was more ad- vanced and permitted the observations of another pattern of intermingling between quail and chick cells. The complete muscle organ was wrapped up by external connective tissues (perimysium) composed of chick cells. And, in the same way, the bundles of quail muscle fibres became surrounded by the internal endomysial connective tissues of chick constitution. Leg level. In five cases, orthotopic transplantations were performed at the leg level. Here also, the quail cells originating from the grafted somitic mesoderm were exclusively found in the bulk of the muscle, whereas the tendons comprised chick cells.

2. Heterotopic transplantations The question arose whether the somitic mesoderm adjacent to the prospective limb region is qualitatively determined and penetrates exclusively the limb of the corresponding cephalocaudal level or whether any somitic mesoderm can invade a growing limb-bud, the somitic mesoderm being thus endowed with general invading and myogenic properties. The experimental design consisted in replacing the somitic mesoderm of the wing level by a portion of the somitic mesoderm taken from the neck level (eight cases, Fig. 8), flank level (six cases) and leg level (three cases, Figs. 3, 4) (Fig. 5). These 17 translocation experiments gave the same results regardless of the cephalocaudal origin of the somitic quail graft: cells originating from the quail segmental plate penetrated far down to the autopodial level and in all Origin of limb musculature 251 Donors

Somite no. 14 Grafts: somite mesoderm from

Host

Somite no. 21.

Somite no. 24

Fig. 5. Experimental scheme of heterotopic replacement ofthesomitic mesoderm from the wing level by heterospecific somitic mesoderm from the neck, flank and leg levels. Combinations between chick host and quail donor, and vice versa. E, discarded ectoderm. 252 A. CHEVALLIER, M. KIENY AND A. MAUGER

&mxh Origin of limb musculature 253 three wing segments formed exclusively the muscle fibres proper. Again J;he tendinous and connective enveloping tissues comprised chick host cells. These results are similar to those observed in the series of orthotopic transplantations and they attest that there is no somitic region alization for the limb musculature in general. B. Chick grafts to quail hosts The question then arose whether the beforegoing observations correspond to the processes of normal development. In particular, is the somitic participation due to intrinsic morphogenetic properties of the somitic cells or merely to the developmental asynchrony between the two species involved (hatching after 16 days for the quail and 21 days for the chick). To answer this question, the reverse experiments were undertaken in which chick somitic mesoderm was grafted in a quail host. Orthotopic (13 cases) and heterotopic (7 cases) transplantations were performed at the wing and at the leg level. As in the previous set of experiments skeleton and dermis were composed of host cells. As concerns the musculature, again the tendons and connective tissues were always composed of host cells (Fig. 9). But the muscle fibres proper were constituted in various proportions of both chick and quail cells (Figs. 10-12). It appeared that no muscle bulk was exclusively composed of chick grafted somitic cells, whereas some muscles were constituted exclusively of quail host cells. Most muscle bulks were of mixed constitution, with a majority of quail host cells. Although the graft-originated cells were scarcer in the wings of these chick-to-quail grafting experiments than in the reverse ones, their muscular tissue specificity could be ascertained by their exclusive presence inside muscle bulks.

CONCLUSIONS In birds, somites can contribute to the development of the appendages. By means of heterospecific recombination experiments involving quail and chick embryos we were able to show that the somitic contribution is selectively restricted to the musculature in the developing wing or leg.

FIGURES 6-8 Quail graft to chick host. Penniform muscles. Figs. 6 and 7. General view and detail of a forearm muscle seven days after the orthotopic replacement of the somitic mesoderm at the wing level. Quail donor and chick host: 16 pairs of somites, x 160 and x 672. Fig. 8. Forearm muscle eight days after the heterotopic replacement of the somitic mesoderm of the wing level by somites from the neck level. Quail donor: 16 pairs of somites. Chick host: 14 pairs of somites, x 672. Fig. 9. Chick graft to quail host. Penniform muscle. Lower leg muscle ten days after the orthotopic replacement of the somitic mesoderm at the leg level. Chick donor and quail host: 19 pairs of somites, x 672. The central tendon (T) is composed of chick host and the muscle fibres (Mw) are made up by quail cells. 17 EMB 41 A. CHEVALLIER, M. KIENY AND A. MAUGER

11 Figs. 10—12. Chick graft to quail host. Penniform muscles. Two neighbouring forearm muscles six days after the heterotopic replacement of the somitic mesoderm of the wing level by somitic mesoderm of the leg level. The one (left muscle of Fig. 10 and Fig. 11) contains chick and quail (encircled areas) cells approximately in the same proportions; the other (Fig. 12) contains mainly quail cells. Only few scattered chick cells (arrows) can be detected. Chick donor: 19 pairs of somites. Quail host: 15 pairs of somites, x 250 and x 672. Origin of limb musculature 255 Thus, the muscle in the anatomical sense of the term appears as a composite structure. The tendons and other muscular connective tissues are of host, i.e. somatopleural origin, whereas the muscle cells are of graft, i.e. somitic origin. These results raise the question of a possible organizing role of the tendon blastemata in the organogenesis of the anatomic muscles. The above-mentioned features are conspicuous when the graft derives from the faster developing quail embryo, but they are less prominent when the graft is obtained from the slower growing chick embryo. In the case of quail somitic grafts, the muscle bulks proper of the chick limbs on the operated side were always exclusively composed of grafted somitic cells. Contrariwise, in the case of chick somitic grafts, the muscle bulks of the quail wings displayed a range of various cellular compositions. There were apparently no muscles exclusively composed of grafted chick cells, but there were some muscles exclusively composed of host cells. Anyhow, the majority of the muscle bulks were of mixed chick-quail constitution, with an evident predominance of quail host cells. The invasiveness of the somitic cells certainly plays a role in this phenomenon. The cells of the faster developing quail embryo conceivably invade the outgrowing chick limb-bud at an abnormally high rate with respect to the slower developing chick environment, and prematurely occupy available spaces. Thus, these quail cells come to participate in the formation of the muscles in a higher proportion than the chick somite-originated cells do in a quail environment. Nevertheless it must be emphasized that the participation in the muscle bulk does not occur haphazardly, but is highly selective; the cells from the slower growing chick also end up exclusively in the muscle fibres. This system thus bears some resemblance to compartmentalization in insect development (Garcia-Bellido, 1975). Indeed the quail-to-chick grafting experiments are reminiscent of those using Minute mutants in Drosophila (Morata & Ripoll, 1975) in which the M + /M+ recombinant cells overgrow the slower proliferating heterozygous M/M+ cells, filling up most or all of one compart- ment, the boundaries and extension of which thus becoming revealed. It would appear then, that in vertebrates as well as in insects, compartments exist (cf. Morata and Lawrence, 1977). However, the compartments of vertebrates would be of a different nature to those of insects, not being separated from one another by simple straight or curved lines, but rather being intimately interdigitated into one another. Consequently, the anatomical muscle would be composed of at least two highly interdigitated compartments, one containing prospective myoblasts, the other filled up by prospective tendinoblasts and possibly other types of specialized fibroblasts. The fact that the reverse combination of chick graft into quail host does not lead to an exclusive graft origin for the muscle bulk may also reflect the postu- lated greater invasiveness and faster growth properties of quail cells. The latter may originate either from the host somitic or host somatopleural mesoderm. At present, on the basis of our experimental knowledge, both origins appear 17-2 256 A. CHEVALLIER, M. KIENY AND A. MAUGER equally likely. On the one hand, the somitic origin may be explained by a possibly incomplete extirpation of the somitic mesoderm from the host or by a compensatory longitudinal immigration of somitic cells from anterior or posterior levels. On the other hand, unpublished experimental results indicate that grafted somatopleural mesoderm may participate in limb myogenesis. However that may be, somitic cells can participate in limb development, and when they do so they selectively contribute to the muscle bulk. The myotomes (and perhaps other somitic cells) not only contribute to the limb musculature, they also give rise to the body wall musculature (Christ et al. 19746; Chevallier, 1977). For limb, as well as for trunk, myotomal derivatives, no cephalo-caudal regionalization exists; they develop according to their new implantation site. This does not hold for the other somitic derivatives. Indeed, for the dermatome (Mauger, 1912 b) and for the sclerotome (Kieny et al. 1972; Chevallier, 1977; Chevallier, Kieny & Mauger, 1976; Chevallier, Kieny, Mauger & Sengel, 1977), the morphogenetic capacities of the somitic mesoderm are strictly level-dependent. The translocated portions of the somitic mesoderm develop according to their original cephalo-caudal level by forming the portion of the spinal feather tract and that of the vertebral column with ribs and girdle elements specific to the tested levels (Sengel, 1972). Prior to these somite-somatopleure relationships which involve a cellular contribution of the somitic mesoderm to the limb-bud once it is bulging out (unpublished data), there are other relationships, that do not seem to depend on a cellular contribution of the segmental plate to the limb-forming somatopleure. These other relationships exist during a limited period of time and are strictly restricted to the still unsegmented somitic mesoderm adjacent to the limb level. Their effect comprises the acquisition of qualitative limb-inducing properties by localised portions of the somatopleural mesoderm (Kieny, 1969-72). It is thus clearly established that, in birds, somitic mesoderm intervenes in limb organogenesis in at least two different ways. Firstly, it exerts an inductive influence on the adjacent limb somatopleural mesoderm, on which it confers the capacity to initiate limb development. Secondly, it provides the outgrowing limb-bud with a certain number of cells the progeny of which exclusively end up as myoblasts in the musculature.

Ce memoire represente une partie de la these qui sera soutenue par A. Chevallier devant l'Universite scientifique et medicale de Grenoble pour l'obtention du grade de docteur d'Etat es Sciences. It is a pleasure to acknowledge the critical reading and linguistic assistance of Professor Philippe Sengel. We wish to thank Nicole Cambonie, Joselyne Clement-Lacroix and Josette Ferring for expert technical assistance. The research was supported by grants from D.G.R.S.T. and C.N.R.S. Origin of limb musculature 257

REFERENCES BORISOV, I. N. (1970). Evolution of the sources of limb development in vertebrates. /. gen. Bioi 31, 327-336. BRAUS, H. (1899). Beitrage zur Entwicklung der Muskulatur und des peripheren Nerven- systems der Selachier. II. Die paarigen Gliedmassen. Morph. Jb. 27, 501-629. BRAUS, H. (1906). 1st die Bildung des Skeletes von den Muskelanlagen abhangig? Morph. Jb. 35, 240-321. CHEVALLIER, A. (1975). Role du mesoderme somitique dans le developpement de la cage thoracique de l'embryon d'oiseau. I. Origine du segment sternal et mecanismes de la differenciation des cotes. J. Embryo/, exp. Morph. 33, 291-311. CHEVALLIER, A. (1977). Origine des ceintures scapulaire et pelvienne chez l'embryon d'oiseau (en preparation). CHEVALLIER, A., KIENY, M. & MAUGER, A. (1976). Sur L'origine de la musculature de l'aile chez les oiseaux. C. r. Acad. Sci. hebd. Seanc, Paris D 282, 309-311. CHEVALLIER, A., KIENY, M., MAUGER, A. & SENGEL, P. (1977). Developmental fate of the somitic mesoderm in the chick embryo. Symposyum on Vertebrate Limb and Somite Morpho- genesis Glasgow, 1st to 4th September, 1976. Cambridge University Press. (In the Press.) CHRIST, B., JACOB, H. J. & JACOB, M. (1974a). Uber den Ursprung der Fliigelmuskulatur. Experimentelle Untersuchungen mit Wachtelund Huhnerembryonen. Experientia 30, 1446-1449. CHRIST, B., JACOB, H. J. & JACOB, M. (19746). Experimentelle Untersuchungen zur Ent- wicklung der Brustwand beim Hiihnerembryo. Experientia 30, 1449-1451. CORNING, H. K. (1894). Uber die ventralen Urwirbelknospen in der Brustflosse der Teleostier. Morph. Jb. 22, 79-98. DHOUAILLY, D. & KIENY, M. (1970). Participation du mesoderme somatopleural du fianc a l'edification d'un membre supplemental chez les oiseaux. C. r. hebd. Seanc Acad. Sci., ParisT> 271,519-522. DHOUAILLY, D. & KIENY, M. (1972). The capacity of the flank somatic mesoderm of early bird embryos to participate in limb development. Devi Biol. 28, 162-175. DOHRN, A. (1884). Studien zur Urgeschichte des Wirbeltierkorpers. VI. Die paarigen und unpaaren Flossen der Selachien. Mitt. zool. Stat. Neapel5, 161-195. GARCIA-BELLIDO, A. (1975). Genetic control of wing disc development in Drosophila. In Cell Patterning. Ciba Foundation Symposium 29, pp. 161-178. North-Holland: Elsevier. GUMPEL-PINOT, M. (1974). Contribution du mesoderme somitique a la genese du membre chez l'embryon d'oiseau. C. r. hebd. Seanc Acad. Sci., Paris D 279, 1305-1308. HAMBURGER, V. (1938). Morphogenetic and axial selfdifferentiation of transplanted limb primordia of 2-day chick embryos. /. exp. Zool. 11, 379-397. HAMBURGER, V. & HAMILTON, H. L. (1951). A series of normal stages in the development of the chick embryo. /. Morph. 88, 49-92. KIENY, M. (1960). Role inducteur du mesoderme dans la differenciation precoce du bourgeon de membre chez l'embryon de Poulet. /. Embryol. exp. Morph. 8, 457-467. KIENY, M. (1969). Sur les relations entre le mesoderme somitique et le mesoderme somatopleural avant et au cours de l'induction primaire des membres de l'embryon de poulet. C. r. hebd. Seanc Acad. Sci., Paris D 268, 3183-3186. KIENY, M. (1970). Sur le role du mesoderme somitique dans la differenciation de la patte de l'embryon de Poulet. C. r. hebd. Seanc Acad. Sci., Paris D 270, 2009-2011. KIENY, M. (1971). Les phases d'activite morphogene du mesoderme somatopleural pendant le developpement precoce du membre chez l'embryon de poulet. Annls Embryol. Morph. 4, 281-298. KIENY, M. (1972). Interactions between somitic and somatic mesoderm during early organogenesis of the limb in the chick embryo. Collogue International sur le Developpement du Membre. Grenoble, 28th August to 1st September. LE DOUARIN, N. & BARQ, G. (1969). Sur l'utilisation des cellules de la caille japonaise comme 'marqueurs biologiques' en embryologie experimentale. C. r. hebd. Seanc Acad. Sci., Paris D 269, 1543-1546. 258 A. CHEVALLIER, M. KIENY AND A. MAUGER MAUGER, A. (1972a). Role du mesoderme somitique dans Je developpment du plumage dorsal chez l'embryon de poulet. I. Origine, capacites de regulation et determinisme du mesoderme plumigene. /. Embryol. exp. Morph. 28, 313-341. MAUGER, A. (19726). Role du mesoderme somitique dans le developpement du plumage dorsal chez l'embryon de poulet. II. Regionalisation du mesoderme plumigene. /. Embryol. exp. Morph. 28, 343-366. MORATA, G. & LAWRENCE, P. A. (1977). Homoeotic genes, compartments and cell deter- mination in Drosophila. Nature, Lond. 265, 211-216. MORATA, G. & RIPOLL, P. (1975). Minutes: mutants of Drosophila autonomously affecting cell division rate. Devi Biol. 42, 211-221. PINOT, M. (1970). Le role du mesoderme somitique dans la morphogenes precoce des membres de l'embryon de poulet. /. Embryol. exp. Morph. 23, 109-151. SAUNDERS, J. W. (1948). Do the somites contribute to the formation of the chick wing? Anat. Rec. 100, 756. SEICHERT, V. (1971). Autoradiographic study of the relation of somites and wing primordia in chick embryos. Acta morph. neerl. scand. 9, 129. SENGEL, P. (1972). Regionalisation precoce du mesoderme somitique chez l'embryon de poulet: developpement compare du plumage et du squelette axial. Bull. Soc. zool. Fr. 97, 485-495. (Received 9 March 1977, revised 28 April 1977)