The role of peripheral sense organs and muscles in specifying the circuitry of the central nervous system during ontogeny was tested in larval lobsters. Presumptive locomotor appendages, the abdominal swimmerets, were extirpated before their differentiation. Electrophysiological recordings made 2-4 weeks later from the corresponding motor nerves showed that, despite the absence of the target muscles and sense organs, normal reflexes and normal patterns of rhythmic locomotor output appeared in the swimmeret motoneurons at the usual developmental stage. Therefore, target muscles and sense organs are unnecessary to the differentiation of normal motor output patterns in this simple invertebrate locomotor system.The role of the periphery in determining the circuitry of the central nervous system has been studied and debated for nearly a century (1-5). According to one view, peripheral sense organs and muscles provide the central nervous system with essential information for specifying and modifying synaptic connections. Sensory feedback from movements in a developing motor system, for example, could help organize the corresponding central connections. Similarly, a muscle could, in principle, "instruct" its motoneurons to form appropriate central connections by means of biochemical "messages" conveyed centrally from the muscle-the well-known hypothesis of myotypic specification. Such hypotheses are significant not only for ontogeny and regeneration, but also for neuronal plasticity in general; if the periphery can indeed influence central circuitry, the underlying mechanisms could have implications for theories of learning and memory.Evidence on the role of the periphery in establishing central circuitry consists largely of behavioral observations (3). Such evidence suggests that in some motor systems, at least, sensory input or feedback may not be necessary to the formation of normal movements during development (6-11) or after limb transplantation (12). Decisive tests using electrophysiological methods, however, are lacking. Myotypic specification seemed an attractive explanation for transplantation data (12-14), but more recent studies support alternative mechanisms (4, 15-17), and in any case a possible ontogenetic role for myotypic specification has not been studied previously.The present study was undertaken to test the role of peripheral structures in specifying central nervous connections during the ontogeny of a simple locomotor system, the abdominal swimmerets of the lobster. This motor system was chosen because its neuronal organization is well understood in adults (18-29) and because it develops largely after larvae hatch from the egg (5, 30, 31). Swimmeret sense organs and muscles are undifferentiated at hatching, but the appendages are fully developed and capable of rhythmic locomotor movements 3 weeks later, when the larvae molt to the fourth larval stage (5). Thus it was possible to interfere with the normal developmental sequence by extirpation of presumptive swimmeret tissue, and to examine t...