Principal results: The primary objective of Leg 148 was to revisit Hole 504B in the eastern equatorial Pacific in order to deepen the hole into oceanic Layer 3 and to determine the nature of the transition from Layer 2 to Layer 3. Site 504 is located on 5.9-m.y.-old crust 201 km south of the Costa Rica Rift, the easternmost segment of the Galapagos Spreading Center (at 1°13.61 l'N, 83°43.818'W, at a water depth of 3460 m). Prior to Leg 148, Hole 504B was the deepest scientific drill hole in the oceans, extending 2000.4 meters below seafloor (mbsf; 1725.9 m sub-basement). It is the only hole to penetrate through the volcanic section into the underlying sheeted dike complex, and the site has become an important in-situ reference section for the physical and chemical structure of the upper oceanic crust. Results from recent drilling in Hole 504B during Leg 140 in late 1991, as well as seismic evidence, suggested that the bottom of the hole lay within the lower portion of the sheeted dike complex, close to the seismic Layer 2/Layer 3 boundary. Many scientists believe that the Layer 2/Layer 3 transition coincides with the change downward from sheeted dikes to underlying gabbros as is observed in ophiolites, but the seismic boundary may be a metamorphic transition within gabbros or within the lower sheeted dikes. Although the transition from sheeted dikes to gabbros has been observed by submersible (Auzende et al., 1989;Francheteau et al., 1992) in tectonic exposures of both Atlantic and Pacific ocean crust, this transition has never been observed in situ in undisturbed ocean crust, and its relation to the Layer 2/Layer 3 boundary remains unproven.Upon re-entering Hole 504B on 28 January, 1993, temperatures were logged from seafloor to just above the total depth of the hole (2000.4 mbsf) with somewhat ambiguous results caused by a calibration error or instru-1 Alt, J.C., Kinoshita, H., Stokking, L.B., et al., 1993. Proc. ODP, Init. Repts., 148: College Station, TX (Ocean Drilling Program).2 Shipboard Scientific Party is as given in the list of participants preceding the contents. mentation problem. The measured gradient through the cased section was close to the background value in the sediments, indicating that presently there is no downhole flow like that observed in the past. A slight dip in temperatures immediately below the bottom of the casing is similar to that recorded during Leg 140, indicating some hydrologic activity in the uppermost basement. Deep in the hole the gradient is generally linear and consistent with logs taken during Legs 137 and 140, indicating a maximum recorded temperature of about 180°C at the bottom of the hole. Following temperature measurements, borehole waters were sampled on 28-29 January. Water was collected in 6 of 8 runs, at 560, 796, 1031, 1266, 1501, and 1970 mbsf. One sample (1501 mbsf) clearly displays chemical characteristics of seawater that has reacted with basalt at elevated temperatures, whereas the compositions of other samples are close to seawater. Following drilling ...