The development of functional motor innervation in supernumerary chick hindlimbs transplated to host embryos at stages 16--18, was studied throughout the embryonic period from stage 27 when peripheral synapses were first formed. 2. The innervation pattern, defined as the number and sequential order of spinal nerves innervating individual muscles, were determined for supernumerary hindlimbs innervated by thoracic, lumbosacral-thoracic, or wholly lumbosacral spinal cord segments. Spinal nerves were electrically stimulated and muscle contraction was scored visually or by tension measurement and compound action-potential recordings were made from muscle nerves. 3. There was no tendency for spinal nerves to grow to or to synapse with the muscles which they normally innervate. Rather, they formed functional connections with inappropriate muscles, which were maintained throughout the developmental period studied. The localization of the motoneuron cell bodies was confirmed with retrograde transport of horseradish peroxidase (HRP) and shown to be adjacent to the spinal nerves through which their axons exited. Motoneurons innervating specific muscles occupied a similar medial-to-lateral position in the cord in both control and supernumerary limbs. 4. In all supernumerary limbs a definite peripheral innervation pattern was formed, which approximated the normal pattern. Specifically, the relative craniocaudal position of motor columns innervating individual muscles was conserved, even though the actual spinal nerves innervating the transplant were different from animal to animal. In limbs reversed along the anterior-posterior axis, the innervation pattern was also reversed. 5. The results suggest that motoneurons have not been rigidly specified to innervate certain muscles at the time when the limb buds were transplanted. Further, the limb itself can apparently influence the innervation pattern in an important manner, possibly by specifying motoneurons or by subsequently directing axon outgrowth and synapse formation.
