Since 1963 several winter cruises have been made off the Gulf of Lions in order to study the MEDOC area (41°-43°N; 3°30-6°E) in the western Mediterranean. From 1972 to 1975, these studies were intensified with the use of CTD recorders which allow a better resolution in the measurements than the conventional hydro-casts. In addition VCM (Vertical Current-Meters) were used. Here we present the hydrological results collected during the cruise MEDOC 72. This cruise was divided into two periods: from January 13, 1972 to February 6, 1972 a "period of reconnaissance" took place on board the R/V LE NOROIT. There was a general study at a rather large scale (10 to 30 miles) during which the convection areas were localized; - from February 12, 1972 to March 13, 1972 on board the JEAN CHARCOT a study at a smaller scale (3 to 5 miles) was made for the first time in this area to try to explain the motions of the VCM floats in this region (J.C. GASCARD, 1973). During the previous cruises, a large cyclonic eddy had been found; in 1972, near its center an area of surface density nearly equal to the density in depth was present, which indicated that deep convection phenomena were taking place. As the distance from the center increases, northward or southward, the surface density gradually decreases and the hydrological structure near the shelf of the Gulf of Lions and symmetrically southward, is the one normally found: relatively fresh and cold water in the surface; deeper, the layer of intermediate water; then the deep water. The central dense patch was, during the NOROIT cruise, elongated in an EW direction. Now, during the CHARCOT cruise, about one month later, the system was found to have evolved in a rather complex way. It appeared as if the limits of the former homogeneous water had undergone modifications in such a way that the former dense patch was split in several small cells, only a few miles in diameter, lasting a few days only, and in which vertical convection took place down to about 800 m. The processes which take place in the change of scale of the dense patches are not clear and need further detailed observations. Current measurements with floats and with moored current-meters were made, and showed interesting features.