1. Leading particulars of the problem. The main characteristics considered for the selected tidal power project site were 1) the area and volume of the estuary 2) maximum and minimum flood and ebb tide discharge and 3) estuary be features. These showed that the damming of the Rance would be a particularly difficult major operation. 2. Progress of the work. The damming scheme basically involved the creation of two areas protected by cofferdams (one, of medium size, being designed to enable the gate sluices to be built and put into service by the earliest possible date), and the gradual damming of the estuary by closing successive gaps between prefabricated piers positioned on the estuary bed. The damming zone was originally 360 m long and required nineteen reinforced concrete caissons and eighteen metal sheet piling cells. Structural stability has to be checked to within an adequate safety margin at each construction stage. 2.1. CAISSONS : The caisson foundations were laid under water with the aid of a pneumatic caisson, the features of which "and associated working conditions are briefly described in the article. The nineteen caissons were placed on their foundations at neap tide by tipping them over into position with the aid of special equipment, sketches of which are included in the article. 2.2 CELLS : After closing the gaps between caissons with special "rut-off boards," the sheet piling rolls were built up around the special jigs shown on the photographs. Various difficulties arose during." the construction work, especially due to rapidly increasing erosion during the final phase. 3. Analysis of observed results. LEVEL DIFFERENCES. - The maximum level differences depended fairly closely on the size of the flow passage and the tide coefficient (see graph 31). They were greater at flood tide than at the ebb and occurred at varying time intervals. The flow capacity of the 8m wide openings due to differences in level exceeded 1,000 cu.m/sec. EROSION. - This became marked at velocities of 6 m/sec. It's development is discussed in the light of Isbah's formula. 4. Purpose and reliability of the scale model. A 1:50 undistorted scale model was used to predict the progress of the damming work and to ascertain the loads acting on the temporary structures. A comparison between conditions predicted experimentally and those observed at the site confirmed the model's complete reliability.