From this point, the rocks cooled slowly to below their solidus. U-Pb ages on zircons from granitic veins average 11.93 ± 0.14 Ma (John et al., submitted [N1]). This is ~0.2 m.y. older than the age of magnetization of the rocks as estimated from the placement of Hole 735B in crust that formed slightly earlier than Anomaly C5r.2n, which spanned 11.48-11.53 Ma (Dick et al., 1991b). This was the time it took after the last igneous minerals crystallized for the rocks to drop below their Curie temperatures. Other minerals, namely amphibole, biotite, and titanite obtained from shear zones in oxide gabbros and felsic veins, give younger ages that indicate conductive cooling at a rate of 65°C/m.y., following an initial rapid cooling of 800°C in the first 0.5 m.y. at depths between 2 and 3.4 km beneath the rift valley floor. About 30% of the original heat was present after the uplift of this portion of Atlantis Bank from beneath the rift valley floor (John et al., submitted). The gabbros of Hole 735B therefore took a long time to cool. Why?
Thickening of the lower crust by asymmetrical intrusion as described here is one way of sustaining geothermal gradients for some distance off axis. We believe that it extended to the rift valley walls because master faults that reached downward from those walls penetrated partially molten rock that included a great thickness of underplated gabbro. The distance from the present ridge axis to the rift valley wall to the south is only 4.3 km. Spreading toward this wall occurs at 1.0 cm/yr (Dick et al., 1991b); thus, the rocks of Hole 735B evidently spent ~0.4 m.y. beneath the rift. In this time, all crystallization, all high-temperature deformation and metamorphism, and all locking in of magnetic structure took place. The rocks have somewhat scattered but nevertheless consistent magnetic orientations throughout and are sufficiently magnetized to account for the intensity of the magnetic anomaly measured over the site (Dick et al., 2000). Possible long-lived, long-offset faults marked by brittle-ductile shear zones at 490, 560, and 690-700 mbsf noted by John et al. (submitted) do not disturb the magnetic orientation of the rocks and, in fact, do not mark important differences in core composition or lithology. They were evidently active mainly before magnetic properties were locked in, and thus did not disturb them. The entire section in Hole 735B consequently cooled through the Curie temperature virtually as a unit. This was because crystallization and crystallization differentiation earlier had ceased in all the rocks at about the same time and at nearly the same temperature. This was a consequence of the thermal regime imposed by thickening of the lower crust, the condition that allowed very late stage melts to crystallize as oxide gabbros with associated granitic veinlets throughout the core.
The gabbro drilled in Hole 735B continued to behave as a coherent structural block from the time the magnetic structure was set up until the present, leaving it with a single, consistent magnetic inclination throughout (Kikawa and Pariso, 1991; Pariso et al., 1991; Kikawa and Ozawa, 1992; Pariso and Johnson, 1993; Shipboard Scientific Party, 1999c; Dick et al., 2000). This spans the time from initial exposure of the boundary of the block at the rift valley wall, uplift to sea level, erosion and partial sedimentation of the summit, loss of the last 30% of heat the block contained to its present levels (John et al., submitted), and subsidence to the present depth of 719 m. The vertical displacement of the top of the hole from an original location in a rift valley ~3 km below the seafloor to its maximum elevation was ~9 km. The total displacement along whatever master fault this required must have been significantly greater than this. This did not significantly perturb the thermal structure. Exposure of the block did not accelerate the cooling rate (John et al., submitted) or disrupt it structurally other than tilting it by ~19°.
The coring success in Hole 735B has been ascribed to serendipity, in reference to the way in which the site was seemingly targeted by chance during Leg 118. The site selection was more thought out than this (Natland and Dick, History Chap., this volume). However, an epitaph for Hole 735B might be that if there was serendipity involved in the site selection, it consisted in dropping the hard-rock base on such a coherent block, one that cored magnificently, and that this good fortune finally allowed us, 17 yr after the drilling was first proposed, to reconstruct much of the early history of accretion of the lower ocean crust at this slowly spreading ridge.