The Galileo System, a variant of metric multidimensional scaling, is used in this paper to analyze over-time changes in social networks. The paper first discusses the theoretical necessity for the use of this procedure and the methodological problems associated with its use. It then examines the air traffic network among 31 major cities in the United States over the 14-year period of 1968-1981, demonstrating how the proposed method provides insights into activities within the network and how exogenous factors such as the physical distances among the nodes, changes within the airline industry, and economic conditions affect the changing network structure. (Author)