From this beginning the English system developed into what has been termed a “radial system;” that is to say, one principal and several minor central pumping stations have been established, and from these radiate tubes to numerous sub-stations (see Fig. 2). Some of the stations are connected with double lines for sending in opposite directions. The out-going tube from the pumping station is worked by compressed air, and the incoming tube by exhaustion. Other stations are connected by single tubes, and they are operated alternately by compression and exhaustion. Intermediate stations are located on some of the lines. For the central station the Varley valves were found too expensive and troublesome to keep in order, so they were replaced by the Wilmot double sluice-valves, which are operated manually. In recent years the sluice-valves have been in turn replaced by what are termed D-boxes, a simpler form of apparatus. At the sub-stations the tube terminates in a box into which the carriers drop. As the system has been gradually extended, tubes two and three-sixteenths inches inside diameter have been used for short lines, and three-inch tubes for long lines. The tubes are of lead laid inside a cast-iron pipe which serves as a shield, protecting them from injury. They are laid in twenty-nine foot sections, the joints being made by soldering over a steel mandrel, which is afterwards drawn out by a chain. The joints in the cast-iron protecting pipe are made by caulking with yarn and lead. “Electric signals are used between the central and sub-stations, consisting of a single stroke bell and indicator, giving notice of the arrival and departure of carriers, and to answer the necessary questions required in working. Where there are intermediate stations the tubes are worked on the block system, as if it were a railway. Experience shows that, where great exactness of manipulation cannot be obtained, it is necessary to allow only one train in each section of a tube, whether worked by vacuum or pressure. But where there is no intermediate station, and where the tube can be carefully worked, carriers may be allowed to follow one another at short intervals in a tube worked by vacuum, although it is not perfectly safe to do so in one worked by pressure. In working by pressure it has been found that, notwithstanding a fair interval may be allowed, carriers are apt to overtake one another, for no two carriers travel in the same times, because of differences in fit, unless they are placed end to end. If signalling be neglected and a carrier happens to stick fast, being followed by several others, a block will ensue which it will be difficult to clear, while the single carrier could readily have been dislodged.” (_Proceedings Institute of Civil Engineers, London_, Vol. XLIII. p. 61.) No changes have been made in the carriers from those used in the early experiments which have already been described.