of which are set out in recent overviews (Malfait et al., 1993; Larson et al., 1997). Other achievements, just as significant, are the cumulative result of many legs of drilling. I review here a sampling—probably reflecting my own interests—of some of the most important results, findings that truly changed our way of thinking. I also mention a few problems that remain as very important but unresolved by drilling.

It has been said that the special philosophical contribution of the geological sciences is the establishment of the immensity of geologic time. The elaboration and refinement of time scales have developed apace with new methods to measure the passage of relative and absolute time: superposition of strata, cross-cutting relations among rock bodies, biostratigraphy, radiometric decay, magnetic reversals, variations in isotopic compositions of strata, and rhythmically deposited sediments. Because cores of pelagic sediments from ocean drilling are commonly exceptionally rich in the remains of the most important planktonic microfossils—radiolarians, foraminifers, and coccolithophorids—the cores provide the basis for very detailed biostratigraphic zonations, based on first and last appearances and joint occurrences of taxa. The web of drill sites in the different oceans enables the establishment of a virtually global biostratigraphic scheme for the past 150 million years. The continuous cores also provide material for determination of magnetic-reversal sequences, which can in turn be linked to the sequence of seafloor magnetic anomalies.

The direct radiometric dating of volcanic ash beds in the sediments and of drilled oceanic crust, using laser technology that can yield 0.1 million-years resolution, plus radiometric dating of biostratigraphically constrained ash beds and igneous rocks on the land, has improved resolution by an order magnitude since the Ocean Drilling Program began. In the last decade, these scales, with resolution of about 0.5-2 million years, have been further refined by an order of magnitude by the realization that rhythmic sedimentation in step with the rhythmic changes in the Earth's orbital parameters is a common feature of pelagic sediments. We are now close to the definition of a time scale for the last 30 million years with 20,000- to 100,000-year resolving power. The road is open to extend this precision back into the Jurassic via our drill cores. Having a time scale with such fine resolution makes it possible to address a host of rate problems: rates of sediment accumulation, rates of evolution, rates of change of environment. It makes possible the detailed ordering of related events on a global scale and the unraveling of cause-and-effect, chicken-and-egg problems.

The planktonic microfossils in pelagic sediments fall to the seafloor from overlying near-surface waters and thus reflect prevailing environmental conditions in these waters, while benthic fossils reflect conditions at the seafloor. This simple picture is distorted by the effects of dissolution: calcareous fossils tend selectively to dissolve in cold deep waters, owing to the greater dissolved carbon dioxide content there. Thus, to make paleoceanography quantitative, we need an independent method of estimating paleodepth. Almost concurrent with the start of drilling, an empirical relation was established, using an early version of the magnetic anomaly time scale, for the age of oceanic crust and its depth below the sea surface. The empirical curve, with correction for isostatic loading by sediments, fits closely to a simple curve D = D₀ + K(Age^1/2), where D is the depth of oceanic crust, D₀ is the depth at the spreading center and K is a constant, generally about 350. The curve is applicable out to crust about 80 million years old, where it begins to flatten. The immediate payoff was the charting of the regional and temporal fluctuations in the depth where carbonate supply and dissolution rates balance, the calcite compensation depth (CCD). A first-order finding was that there was an abrupt deepening of the CCD by about 1,000 m near the beginning of the Oligocene, about 35 million years ago, at about the time of the earliest continental-scale Antarctic glaciation. Global paleodepth maps of the CCD now exist for many levels in the post-Jurassic.

A paleoceanographical surprise emerged with the coring of organic carbon-rich layers at several levels in the mid-Cretaceous in both the Atlantic and the Pacific oceans. Paleodepth estimates for these sediments yielded a broad range of depths, excluding the abyssal waters of the Pacific, suggesting that the anoxic conditions were associated with a broadening and intensification of the oxygen minimum, possibly owing to relatively strong density stratification of the oceans during these extreme "greenhouse" times of raised global sea level and warm ocean temperatures.

Determination of ¹⁶O/¹⁸O in precisely dated mid-Cretaceous-Recent planktonic and benthic foranminifers has allowed construction of a detailed history of oceanic surface-and bottom-water temperatures and an estimate of the changing volumes of continental ice. What the isotope record shows, besides the contrast between the generally warm "greenhouse" ocean climates of the Cretaceous and the colder (in high latitudes) "icehouse" climates of the Neogene, is a stepwise history of long periods of relatively stable conditions and abrupt transitions to new, but different, stable conditions. The record also shows that tropical sea-surface temperatures have been relatively stable; it has been the high-latitude oceans (and the deep waters derived from these latitudes), that have changed the most. What we do not understand are the "why's" of the stepwise history. One promising avenue was explored in the South Atlantic by coring the summit and flanks of Walvis Ridge in a highly successful attempt to document the history of bottom-water temperatures along an oceanic depth profile (Shackleton et al., 1984). The depth-profile approach has not since been much exploited,