regarded as the major contributor of dissolved material. Wind transport of dust was recognized as adding to the particle flux borne by streams, and the input of altered volcanic material to deep-sea sediments was clearly identified by the 1960s. Within the past 15 years, deep-sea hydrothermal activity, predominantly at spreading centers, was found to be important as both a source and a sink of elements. In addition, a significant flux of dissolved material may be expelled from sedimentary wedges and underlying oceanic crust as they descend into the mantle through subduction zones. Cosmic dust and cosmic ray-produced nuclides are not a major input to the ocean in terms of volume, but are important as tracers and for understanding episodic influxes of significance to the history of our planet.
Elements are redistributed within the ocean in dissolved form by horizontal and vertical advection, diffusive mixing, and incorporation into particles. Chemical species are removed from the ocean as particles settle to the seafloor; they also react directly with the seafloor as the result of diffusion and circulation through sediments and the oceanic crust. Seafloor sediments and altered rocks undergo secondary reactions over time, and their composition may change substantially. From a chemical point of view then, the oceans are a large chemical reactor with multiple feeds and outputs. Chemical oceanographers want to understand the present reactor and then, from examination of changes in the outputs over time, determine variations in the reactor's behavior and the compositions and fluxes of inputs in the past. With this information, the limits of future oceanic changes under given climatic and tectonic scenarios can be estimated.
Geochemical study of the oceanic water column and the output of elements to the sediments is now relatively mature. The descriptive phase is largely complete, and studies of mechanisms are growing more numerous. For inputs of elements to the ocean, however, quantitative research is difficult and, in the case of inputs from the continents, may not correctly reflect disturbances in terrestrial inputs caused by humankind. These disturbances have occurred on time scales of a few years to several centuries, and the ocean has not yet equilibrated to the altered inputs.
Quantitative measurement of river inputs is difficult because measurements of fluid discharge from rivers are uneven in quality, frequency, and distribution. Because the best data are avail-