Each Lamont‐Doherty sonobuoy located on well‐dated crust has been carefully analyzed to determine crustal structure down to oceanic layer 3. Results from the Atlantic and the Pacific are compiled separately in order to study crustal structure as a function of plate age in both oceans, since they have very different spreading rates. Layer 2A (refraction velocity about 3.6 km/s) in the North Atlantic is 1.5 km thick at the ridge crest and thins consistently to about 100 m as the crust ages to about 60 m.y. Layer 2A in the east Pacific is 0.7 km thick at the ridge and thins to about 100 m at about 30 m.y. This difference in thickness is probably attributable to the much faster spreading rate in the Pacific. A poorly refractive acoustic basement layer about 200 m thick with similarities to layer 2A but not necessarily composed of the same materials is measured sporadically in the Pacific M‐Series plates and even less consistently in the Atlantic. This layer is not recorded in the Cretaceous or the Jurassic quiet zones. Regressions of refraction velocities in layer 2A as a function of age show that its velocity increases from about 3.3 km/s at the ridge crests to that of layer 2B on crust about 40 m.y. old. There is no corresponding increase of velocity with age in any of the deeper layers. The high resolution of the air gun/sonobuoy records shows that the layer 2B refraction line breaks directly to layer 3 velocities (46 times in the present work) or to a line with a velocity of 6.1 km/s (114 times in the present work), which we call layer 2C. The variance of the velocities in 2C is one fourth that of 2A and 2B, which indicates relative lithological uniformity in 2C. It does not seem likely that layer 2A really thins; what appears to be a thinning of the layer may actually be the result of an increase of its refraction velocity with age. However, the 2A/2B interface is well‐defined by large amplitude refractions from 2B on crust that is younger than about 30 m.y., which seems to rule out a transitional zone at the base of layer 2A where it ‘converts’ to 2B. The seismic observations seem to require a diagenetic process or repeated basaltic intrusions; both processes raise serious objections.

Multichannel and wide‐angle seismic data collected off Virginia during the 1990 EDGE Mid‐Atlantic seismic experiment provide the most detailed image to date of the continent‐ocean transition on the U.S. Atlantic margin. Multichannel data were acquired using a 10,800 in3 (177 L) airgun array and 6‐km‐long streamer, and coincident wide‐angle data were recorded by ten ocean bottom seismic instruments. A velocity model constructed by inversion of wide‐angle and vertical‐incidence travel times shows strong lateral changes in deep‐crustal structure across the margin. Lower‐crustal velocities are 6.8 km/s in rifted continental crust, increase to 7.5 km/s beneath the outer continental shelf, and decrease to 7.0 km/s in oceanic crust. Prominent seaward‐dipping reflections within basement lie within layers of average velocity 6.3–6.5 km/s, consistent with their interpretation as basalts extruded during rifting. The high‐velocity lower crust and seaward‐dipping reflections comprise a 100‐km‐wide, 25‐km‐thick ocean‐continent transition zone that consists almost entirely of mafic igneous material accreted to the margin during continental breakup. The boundary between rifted continental crust and this thick igneous crust is abrupt, occupying only about 20 km of the margin. Appalachian intracrustal reflectivity largely disappears across this boundary as velocity increases from 5.9 km/s to >7.0 km/s, implying that the reflectivity is disrupted by massive intrusion and that very little continental crust persists seaward of the reflective crust. The thick igneous crust is spatially correlated with the East Coast magnetic anomaly, implying that the basalts and underlying intrusives cause the anomaly. The details of the seismic structure and lack of independent evidence for an appropriately located hotspot in the central Atlantic imply that nonplume processes are responsible for the igneous material.

The results of twenty‐eight seismic refraction profiles recorded in the various physiographic provinces of the Philippine Sea as part of the United States and Japan Science Cooperation Program are presented in four schematic structure sections. The basins of the Philippine Sea have fairly normal oceanic crust that includes, between the sea floor and layer 2, a layer of about 3.5‐km/sec velocity controlling the characteristic rough topography. Crustal thickening beneath the Nansei Shoto, Oki‐Daito, Kyushu‐Palau, and the Honshu‐Mariana ridges is associated mainly with an increase in thickness of the 3.5‐km/sec layer and a thick underlying section of material with a velocity between 5.5 and 6.0 km/sec. Beneath the Nansei Shoto trench and the Honshu‐Mariana trench, there is a tendency for layer 2 to increase and layer 3 to decrease in thickness as the trench is approached from the adjacent oceanic sector.

The data from 48 seismic refraction profiles in the western Caribbean Sea and in the Gulf of Mexico are presented in the form of structure sections crossing the Colombian basin, Nicaraguan rise, Cayman trough, Cayman ridge, Beata ridge, Yucatan basin, Campeche bank, and Sigsbee deep. The Cayman trough has a remarkably thin crust, which suggests that it is a tensional feature. Although parts of the basins have a relatively thin crust, similar to the oceanic type, the shallower areas are intermediate or almost continental in structure. In the Gulf of Mexico the main basin is similar to typical ocean basins in structure except that the high‐velocity crust is overlain by very thick sediments. The depth to the mantle is appreciably greater in the Gulf than in an ocean basin. This may be partly the result of loading by the sediments, but large scale tectonic activity is a more likely cause. The Sigsbee escarpment, the northern boundary of the main basin, appears to be the surface expression of a fault or sharp flexure in the layers beneath the unconsolidated sediments.

No abstract