Communications networks are interconnected by a bridge which allows the stations on each network to communicate with those on the other netowrk. Each station has an identifying address unique in the system. Data is transmitted in the form of packets each containing a source address and a destination address. The bridge includes a table holding addresses of stations in the first network. Whenever the bridge receives a packet from the first network, it compares its destination address with the contents of the table, and if no match is found, transmits the packet on to the second netowrk. The bridge also compares the source address of the packet with the contents of the table and if no match is found, enters that source address into the table. Thus, the contents of the table can be built up by the bridge by a learning process. The bridge handles transmission of packets in the opposite direction by a similar method, using a second table, holding the addresses of stations in the second network.