but holds great promise in mapping the paleotemperature structure of the oceans.
Beginning with DSDP Leg 27, attempts were made to drill in very high latitude waters, mainly for paleoceanographic objectives. In spite of daunting conditions, drilling around Antarctica and in the seas off northeast Greenland and in the Labrador Sea has elucidated the Paleogene beginnings of continental glaciation and clarified the plate tectonic events that opened a circum-Antarctic Ocean and led to the formation of Antarctic Bottom Water. In the north, drilling has enabled reconstruction of the history of formation of North Atlantic Deep Water.
Drilling on the two sides of the Isthmus of Panama has established the timing of the late Neogene closure of the isthmus, isolating Atlantic from Pacific marine biotas and forming a land bridge for terrestrial animals.
Two spectacular events have captured public imaginations, the impact of the cosmic bolide that struck Earth at the end of the Cretaceous and the drying-up of the Mediterranean near the end of the Miocene. The K/T bolide story has depended as much on data obtained from land outcrops as from the ocean drill cores, which have served mainly to provide an especially detailed record of the sequence of events in regions relatively close to the impact site, on the Yucatan Peninsula. The discovery of the Mediterranean events, on the other hand, was almost purely the result of drilling on DSDP Legs 13 and 42A, which showed that the salt deposits that accumulated in shallow salt marshes and brine basins at the bottom of several Mediterranean depressions are both underlain and overlain directly by deep-sea biogenic sediments. Only small tectonic movements were required to isolate the Mediterranean from the Atlantic, and near-total evaporation, which may have been repeated many times, was likely very quick. These two catastrophes are now so well documented that, taken together with the evidence about very rapid shifts in ocean temperatures and the long-standing evidence of catastrophic floods on land (e.g., the rapid emptying of Lake Missoula to create the scablands of Washington), they are softening the rock-hard beliefs of the Earth science community in traditional gradualism. We must now admit the possibility of rare and powerful events, the amplifying effects of critically located small events, and the wide range of possible rates of change. James Hutton, the father of classical uniformitarianism and his disciple Charles Lyell may be uneasy in their graves.
Solid hydrates of methane are stable in the pore spaces of sediments where the temperatures are cold or the confining pressures sufficient. Vast regions of the arctic tundra are underlain by sediments containing gas hydrates, and drilling has confirmed that continental margin sediments containing concentrations of biogenic methane also contain crystalline gas hydrates where temperatures and confining pressures are right. These concentrations are commonly visible on reflection seismic records as "bottom-simulating reflectors." Drilling has permitted preliminary estimates of the locations of these buried hydrates and an appreciation of the quantities of methane that might be released into the atmosphere if bottom-water temperatures were to rise significantly.
Although evidence of bacteria has been recovered in cores from oil exploration and from pores in volcanic glass under 400 m of mid-Atlantic sediments, ODP drilling in plant-rich layers in turbidites of the Amazon deep-sea fan in the Atlantic Ocean has recovered bacteria that are actively reproducing at subbottom depths of hundreds of meters. Taken together with the evidence of living bacteria from the high-temperature vents along spreading ridges, we can agree with Reiche (1945) that "The infernos envisioned by medieval theologians can [hold] only limited terrors for such creatures." We are still exploring for the outer limits of the biosphere.
Early attempts to drill into very young oceanic lithosphere, close to the active spreading centers where post-emplacement alteration of rocks should be minimal, were defeated. The brittle and fractured basaltic rocks broke up in front of the drill and stopped progress. Except in areas of strong hydrothermal alteration, we have still not been able to sample more than a few meters into "zero-age" oceanic crust. The most successful drilling has been at a site off Costa Rica on crust about 6 million years old, covered by about 275 m of sediment. Here coring was successful to a depth of 1,836 m into pillow basalts and sheeted dikes. Surprisingly, seismic velocities commonly associated with Layer 3, generally believed to be gabbro, are measured at this hole in part of the zone of sheeted dikes and basalt flows.
The deeper parts of the oceanic lithosphere can be reached only where spreading was very slow and magma supply so skimpy that basalts are thin or absent and spreading has allowed gabbro and mantle rocks to emerge at the seafloor. Gabbros were cored almost continuously at a site on the slow-spreading Southwest Indian Ridge, southeast of Africa, yielding virtually the full suite of oceanic plutonic rocks and partly validating models erected on the basis of scattered dredge samples and from studies of supposed oceanic lithosphere tectonically emplaced onto continents—the ophiolites. The excellent drilling conditions at this site suggest that, in principle, one might reach the dreamed-of Moho here.
The mantle itself has been cored at a few places, (e.g., in the Atlantic off Iberia in tectonically disturbed locales, where ultrabasic rocks have been serpentinized and uplifted in dia