pirs that reached the seafloor, and in magma-starved segments of the Mid-Atlantic Ridge). In the Pacific, serpentinized mantle rocks were recovered by drilling in a tectonic rift zone. Because seismic data indicate that the Moho is present at depth even where altered mantle rocks are close to the seafloor, we are left with the original AMSOC question: What is the nature of this seismic discontinuity? Is it an original petrologic boundary, a tectonic boundary, or a level in the lithosphere marking the downward limit of alteration by circulating seawater? Or any one of these, depending on where you are? Repeated measurements over a period of several years at the Costa Rica drill site show that cool ocean water is being drawn down into the upper parts of the oceanic crust. Heat flow measurements and direct observation from submersibles show that hot waters, charged with ions from crustal alteration, emerge elsewhere at oceanic spreading centers and from outcrops of crustal rocks on abyssal hills. The drill has successfully recovered hydrothermal spring deposits close to an active spreading center, deposits that include tall chimneys of sparkling metal sulfides. Gradually, we are building up quantitative estimates of the rates and depths of circulation of seawater through the oceanic lithosphere and of the extent to which this flow moderates the composition of seawater. Drilling on crust ranging in age back to the Middle Jurassic shows that most hydrothermal alteration takes place while the lithosphere is very young.
Beginning with the clean test of seafloor spreading on Leg 3, the determination of the age of oceanic lithosphere has been made at many of the 1100 sites drilled, giving us a set of ties between the biostratigraphic scale and magnetic anomalies, back to the mid-Jurassic, and enabling the interpretation of magnetic anomaly patterns in terms of plate tectonic evolution.
The ruling theory for the formation of linear seamount chains is that they result from motion of a plate over a fixed melting anomaly, or hotspot, in the underlying mantle. Drilling along the Emperor Seamount Chain in the North Pacific and the Ninety East Ridge in the Indian Ocean showed that these fitted the model. Other drilled chains (e.g., the Line and Marshall chains in the Pacific) have messy records of progressive volcanism, and some undrilled chains, sampled by dredge and by hammer, (e.g., the Australs and the Puka Puka chains), show a scrambling of ages, inconsistent with fixed hotspots. Linear seamount chains remain a problem.
Several oceanic plateaus—great deep-rooted (tens of kilometers to the Moho), smooth-backed leviathans that rise to levels 1-3 km above the surrounding deep ocean floor—have been drilled, primarily for the continuous stratigraphy of the mainly calcareous pelagic sediments that blanket the basaltic basement. The origin of many of the plateaus is ascribed to mantle plumes, arising mainly during a short interval during the mid-Cretaceous from unknown depths. Beyond limiting the times of formation, drilling has got us almost nowhere on the plateau problem so far.
Dating of the age of emplacement of several of the major oceanic plateaus (Ontong Java, Manihiki, Kerguelen), of scores of seamounts spread over a large part of the western Pacific and the great volumes of basalt on the deep Pacific seafloor, far from any contemporary spreading ridge, points to a highly unusual time of massive volcanism during a relatively short time of only about 20 million years in the Early Cretaceous. The volumes are comparable to those produced along the entire global spreading system and suggest some very deep rooted cause, a veritable revolution in the Earth's mantle. The near coincidence of these events with the beginning of the long period of normal polarity of the Earth's magnetic field and of the widespread deposition of organic-rich black sediments and evidence for warm climates, has stimulated a search for causal connections among these effects.
The Atlantic Ocean is bordered by passive continental margins, segments of which have been subsiding and receiving continent-derived and carbonate sediments since the Middle Jurassic. The North American margin is covered by a prism of sediments too thick for full penetration with JOIDES Resolution, but the European-African margin has a much thinner cover, and the early history of the margin is thus within reach of the drill. Cores documenting the early history of the Morocco margin show a beginning with a Late Triassic proto-Atlantic saline basin below sea level and progressive Mesozoic evolution from fluviatile to deeper and deeper waters.
Farther north, in the Norwegian Sea, which opened much later than the Central Atlantic, drilling penetrated and sampled huge wedges of mainly subaerial early Tertiary basalts that were extruded from both sides of the widening rift between Norway and Greenland. Such marginal basalts are imaged on seismic records from many other segments of passive margin around the world and may be related to voluminous mantle plumes that may localize and even initiate seafloor spreading.
Concentrated drilling has been done on several active continental margins, where oceanic lithosphere is being sub-ducted beneath a volcanic arc. These places are the loci of major seismicity, and understanding processes in them should contribute to public safety. The clearest results have been obtained from a transect off Barbados, where drilling was carded through the surface separating the two opposing