Synthetic unit hydrographs. For flood estimation purposes, no published procedure generally acceptable. New procedure presented investigating 1. base flow separation. 2. method of multiperiod storm analysis. 3. effect of choice of unit period. 4. effect of estimation errors, storm duration. 5. effect of size of storm from which unitgraph derived. From these studies, a procedure of flood estimation is derived and pm x t data without storm rainfall data has been devised. A brief description of establishing the South Creek experimental catchment.