Knowledge of the hydrography of the English Channel and Celtic Sea is needed as a background for the life histories of the mackerel, pilchard and herring and, indeed, of every organism living within range of the Plymouth Laboratory. Such knowledge can never be fully attained until we know more of the exchanges with the deep Atlantic Ocean which take place 200–300 miles to the south-west and west over the continental slope. Moreover, at the slope we shall need to know not only what waters move in and out but what move up and down. Long ago Storrow (1925), and no doubt others, saw clearly some of the problems here to be discussed, but they were unable to bring factual evidence to bear. A critical reconstruction of the considerable but fragmentary observations in the neighbourhood of the continental slope of the Celtic Sea will be presented here and supplemented by observations made in 1950 with the generous co-operation of the vessels of the National Institute of Oceanography (Discovery Committee) and in 1948 by H.M. Surveying Ship Dalrymple. Though many of the conclusions remain tentative, it should now be easier to design observational work at sea to test specific hypotheses at the right time and place by the methods of experimental oceanography. The numerical results in 1950 will be published by the Conseil International pour l'Exploration de la Mer.