By studying some relatively complex chemical reactions and a relatively simple neural network, the principal investigators hope to derive new insights into how the different parts of a network interact to produce rhythmic behavior. Ultimately, the principal investigators seek to explain how groups of neurons work together to generate and modulate simple rhythmic behaviors such as walking, chewing or breathing. The principal investigators propose to construct and study the properties of networks of coupled components, each of which is independently capable of oscillatory, bistable or other complex dynamical behavior. Two classes of systems will be investigated: a) inorganic chemical reactions in continuous flow reactors coupled through mass transfer or through computer- controlled concentration-dependent modulation of input flows; and b) neurons of the crustacean stomatogastric ganglion grown in tissue culture and coupled through electronically-controlled artificial synapses and electronic connections. By employing a variety of different components and varying the nature of the connections between them, the principal investigators shall be able to explore questions that have not bee answered by studies of neural networks consisting of large numbers of simple, identical units. Such issues include a) the relationship between emergent network properties, which depend primarily on the coupling between units, and network properties that result from the properties of individual units; b) the range of behavior modes that can be displayed by coupled oscillator networks; c) how the details of the coupling influence the behavior of the network; and d) the role of delayed coupling in oscillator networks. The proposed experimental studies will be supplemented by ongoing theoretical work on both the chemical and neural systems. While the chemical systems are better understood mechanistically, dynamic similarities already apparent between the two classes of systems should make for a fruitful interplay between the two sets of experiments and ultimately lead to a clearer picture of the operation of real neural networks such as those found in rhythmic motor systems in general, or the stomatogastric nervous system in particular.