Beschreibung: The first attempts to sample formation waters from basaltic basement in the oceanic crust were made at Sites 501, 504, and 505 of the Deep Sea Drilling Project. Two methods were used. In the first, the water that occupied the hole was sampled some time after pumping had stopped. In the second, water from both the hole and the surrounding rocks was sampled by sealing off a 3-meter section at the bottom of the hole and opening a large-volume sampler, thereby creating negative pressure. Neither method produced a sample that contained an appreciable and unambiguous component of true formation water, although the samples generally showed large compositional differences from seawater. Samples from Holes 501 and 505B were mainly mixtures of the surface seawater used as drilling fluid and pore water from sediment that fell down the hole. Samples from Hole 504B contained a large fraction of seawater that displayed large chemical changes due to reaction with basement basalts. Tritium analyses revealed the samples to be surface seawater that had been pumped into the formation a few days earlier and had reacted rapidly with basalt at the in situ temperature of 80°C. The solutions had gained Ca and lost Mg on a mole-for-mole basis, lost K and possibly SO4, and gained Si. The Si/Ca ratio increased with depth. These results are consistent with those of laboratory experiments in which seawater reacted with basalt at 70 °C and indicate that at 80 °C in situ reaction rates in the crust are sufficiently rapid to produce large changes in solution chemistry almost instantaneously. The absence of an 18O shift in the solutions indicates that the amount of rock that reacted with the solution during its brief residence in the crust was negligible.