The following article , by Remi Despres, is the second on the history of X.25 systems to appear in this column. As noted by Dr. Despres, the previous article focused on the Canadian Datapac system. Earlier articles on packet switching in this column have included one on the history of the Arpanet/Internet and one on early British packet switching systems. What makes this article particularly distinctive, aside, of course, from the fact that it focuses on the major contributions of French engineers to the development of packet switching as well as to X.25 standardization, is that it carefully outlines the reasons for the choice of connection-oriented virtual circuits for the Transpac network, as contrasted with datagram-based packet switching adopted for Arpanet. Interestingly, Dr. Despres notes that the idea of using virtual-circuit connection-oriented packet switching in the Transpac development came from the British packet switching activity. It is to be noted that early commercial packet switching networks in the United States, such as Tymnet and Telenet, also adopted the virtual circuit paradigm.