A global cellular automaton (GCA) is a cellular automaton in which the evolution of each site is a function not only of the values of the site’s neighbors, but also of certain global features of the automaton. This notebook explores an extension of GCAs (abbreviated as a GCAN) in which each GCA is connected to other GCAs and each site of each GCA evolves not only according to the value of its local neighbors but also according to its own global features and those of its neighboring GCAs. GCANs may be useful in a variety of ways. First, they provide a vehicle for studying composition of cellular automata. Second, the concept of a GCAN, which is essentially one of complex agents interacting over a network, may be useful in emulating the computing systems or the interactions of complex agents.