Experimental facilities include a 3.5?0.2?0.6m flume, a 5.0?2.5?0.6m basin, a sand feeding device and a water circulating system. The sands, with grain-sizeranges 0.25 to 0.42mm, were tuken from the Yangtze River. The independent slope of the flume and basin was predetermined, but the water discharge (Q), the sand discharge (Qs), and the water levels in the basin were varied between runs. The results of six experiments showed that the sandbodies developed as lingulate form at the beginning. When the sand was fed, the progradation and lateral deposition deve-loped almost at the same rate, and the rate depended on Q and Qs. When no sand was fed, there was no progradation but the lateral deposition rate was great, and this rate also depended on Q. However, the water discharge (Q) was not related to the rate of area. In all the six-run experiments, the maximum width of the sandbodies located at a distance of 60cm from the river mouth; the maximum thickness located at a distance of 150cm from the mouth. A balance point of erosion and deposition existed between 30-50 cm from the mouth. When sand was fed, the near mouth side received deposition and the other side eroded; when sand was not fed, things were opposite. The erosion and deposition cycleresulted in the creation of new channels. The main channel began to separate between 30-60cm and rejoined at 120 cm from the mouth. In other words, the distributary mouth bars were easy to form at a distance of 30-60cm from the mouth.
