Two clay-sized vermiculites, one montmorillonite, and one Al-hydroxy-vermiculite, all of which were extracted from four different soil samples, were Ca (super ++) -saturated and suspended in sea water for periods up to 10 months. Chemical analyses of the sea water immersed clays show that the adsorbed cation composition of the clays changes very little after only a few days of contact with sea water. A large proportion (up to 40%) of the interlayer cations become non-exchangeable after prolonged contact with sea water. After contact with sea water for 11 weeks, the ratio of adsorbed Mg (super ++) to adsorbed K (super +) varies from about 1:1 to 5:1 for these four samples, and the amount of adsorbed Na (super +) is less than 20% of the total adsorbed cation content for all these samples. X-ray diffraction patterns of sea water-immersed samples show that discrete illite does not form as had been expected. Instead, a "mixed-layer" phase develops. This phase is a random interstratification of illite-like and expansible layers. The 001 diffraction maximum produced by this "mixed-layer" phase is extremely broad. Because this material gives diffraction over such a large 2theta range, this phase when mixed with other clay minerals may have been undetected in many previous studies of Recent marine clays. The suggestion is offered that burial diagenesis would tend to transform this mixed-layer clay into discrete illite and/or chlorite. Computations show that expansible soil clays delivered to the oceans as part of the suspended load of rivers can adsorb up to 50% or more of all the K (super +) and Mg (super ++) delivered as dissolved load to the oceans.