Through the deconvolution of ocean chemical data it has been shown that the ratio of O2 utilization to phosphate production in the sea is 175¿6 (rather than 138, as proposed by Redfield et al. (1963)). We find no evidence that this ratio changes significantly with location or depth in the sea. We have used this new ratio to calculate the initial phosphate concentrations for the waters sampled during the GEOSECS and TTO programs. The important application of these new results is in constraining the origin and flow patterns of deep waters in the ocean. We believe that a strong case can be made that Antarctic salinity maximum water (i.e., the common water of Montgomery (1958)) is produced by the mixing of waters entering the Antarctic from mid-depths in the Indian, Pacific, and Atlantic with Weddell Sea bottom water (southern component). Antarctic common water consists of about 45% Weddell Sea bottom water, 30% intermediate waters from the Pacific and Indian oceans, and 25% deep water originating from the northern Atlantic.