PLATE TECTONICS -its influence on man-

an overview

By

WARREN HAMILTON, Geologist

U. S. Geological Survey

This article was abridged from the author's paper entitled "Plate tectonics and man." which appeared in the U. S. Geological Survey Annual Report for 1976, p. 39-53.

INTRODUCTION

Coastal and inland California ground past one another for less than a minute, at 5:13 am on April 18, 1906, and produced the San Francisco earthquake. The coastal side lurched as much as 6 meters (20 feet) northwestward along a portion of the 430-Kilometer-long San Andreas fault. In those catastrophic moments, the two sides released strains which had been building for about a century as the part of the Earth's outer shell beneath the Pacific Ocean moved slowly but inexorably past that part beneath North America. The strain in the outer shell is continuing, and future great earthquakes are a certainty.

Like most earthquakes, the 1906 shock was a manifestation of the fragmentation of the Earth's outer shell into large and small plates, all moving relative to all others with steady velocities that reach 13 centimeters (5 inches) per year --- pulling apart here, slipping past one another there, sliding beneath another somewhere else, and in other places colliding slowly to build some of our most spectacular mountain ranges. Earthquakes are the most dramatic way in which these plate motions affect man, and our new understanding of the motions permits a comprehension of earthquakes which could not be approached until a decade ago.

Most volcanic eruptions are also produced by the plate motions. The distribution of mineral deposits and fossil fuels upon which our civilization depends has to a large extent been controlled by plate motions and interactions.

SEE FIGURE 1

Figure 1. Lithosphere plates of the world, showing boundaries that are presently active. Double line; zone of spreading, from which plates are moving apart. Line with barbs; zone of underthrusting (subduction), where one plate is sliding beneath another: barbs on overriding plate. Single line: strike-slip fault, along which plates are sliding past one another. Stippled area: part of a continent, exclusive of that along a plate boundary, which is undergoing active extensional, compressional, or strike-slip faulting. Compiled and adapted from many sources; much simplified in complex areas.

The basic understanding of plate motions was developed primarily during the 1960s in a conceptual revolution as profound for the earth sciences as was the earlier development of the concept of evolution in biology and of atomic and molecular structure in physics and chemistry. The new field is known as plate tectonics: the "plate" is the basic unit of the system, and "tectonics" (from the Greek work tekton, meaning builder) refers to the processes and products of motions within the Earth. Many of the concepts are accepted by most earth scientists, but some of the concepts are controversial.

DEVELOPMENT OF CONCEPTS

The concepts of plate tectonics were developed slowly at first. The reality of continental drift was reasonably well established during the 1930s, as Alexander Du Toit of South Africa and others followed up on the pioneering work by Alfred Wegener of Germany and Emile Angand of Switzerland a decade and more earlier. The distribution of indicators of past climates --- for example, the presence in now-tropical India of continental glacial deposits of the same 300-million-year age as fossilized tropical reefs and equatorial forests in lands now in the Arctic --- required that the continents had moved slowly about the globe. Reassembly of continents in patterns suggested by their shapes provided remarkable continuity for similar, truncated geologic terrains. The "drifters" mistakenly visualized continents as rafts sliding over the ocean floors. Most geophysicists correctly regarded this as impossible, and only a vocal minority of earth scientists supported the concepts. David Griggs, then at Harvard, had a preview of ideas to come when he suggested that the mountain systems around the Pacific were caused by the ocean floors plunging beneath the continents.

Paleomagnetism

Geophysicists changed sides in increasing numbers in the 1950s as the science of paleomagnetism expanded rapidly and provided powerful new evidence for drift. Many rocks, such as volcanic rocks crystallized from lava, contain tiny grains of magnetite and other magnetic minerals which retain the magnetic orientation of the Earth's magnetic field at the time when the rocks formed. Early studies of this phenomenon had been made mostly in France and Japan during the 1920s but scientists in other nations became involved subsequently. The magnetic orientations of rocks indicated fossil latitudes, which proved to be about the same for the various continents as those indicated by paleoclimatology and other geologic and paleontologic criteria.

Marine Geophysical Data

The 1950s also saw the gathering of an enormous amount of geophysical data from the oceans, particularly by research vessels from American oceanographic institutions. A globe-girdling system of interconnecting submarine ridges was recognized that seemed, on geologic grounds, to be moving apart (figure 1). Bruce Heezen of Columbia University suggested that new oceanic crust was being formed at the ridges and added to great plates moving apart. Great submarine trenches were also delineated, particularly along the convex oceanic sides of the volcanic arcs which make up the "ring of fire" around the Pacific. From these trenches, inclined zones of earthquakes dip as deep as 700 kilometers (450 miles) into the mantle beneath and behind the volcanic arcs.

Harry Hess of Princeton University was probably the first to see the broad outline of what was to emerge as plate tectonics. He reasoned that oceanic crust was formed at spreading ridges behind drifting plates and was destroyed at the same rate elsewhere as oceanic plates tipped down at the trenches and slid deep into the mantle along the seismic zones. Robert Coats of the U. S. Geological Survey and Robert Dietz of the National Oceanographic and Atmospheric Administration were among the first to suggest applications of these processes to the geology of the continents. Others saw the motions of the continents themselves clearly enough, but not the way these motions involved the floors of the oceans.

Magnetic Reversals

Paleomagnetic research demonstrated that the Earth's magnetic field has been "normal" during about half of geologic time and "reversed" (a magnetic compass would point south instead of north) the other half. It was also found that the intervals of contrasted polarity were extremely irregular. Allan Cox, Richard Doell, and Brent Dalrymple, all then of the U. S. Geological Survey, during the early 1960s devised a precise magnetic-reversal time scale for these polarity periods during the past 4 million years.

Frederick Vine and D. H. Matthews, English scientists, suggested that the alternating belts of highly magnetic and weakly magnetic oceanic crust, known by then to trend along the midocean ridges, represented magnetic recordings as new crust formed in the gap behind the separating plates. Thus, the belts would record symmetrically, on opposite sides of the young ridges, the periods of time represented by the magnetic-chronology time scale. The scientists tested the suggestion against several small ridge sectors and showed that it fit. James Heirtzler, Ellen Herron, Xavier Le Pichon, and Walter Pitman, III, of the Lamont Geological Observatory of Columbia University, showed in the late 1960s that ridge after ridge fit the pattern, and they were able to carry the magnetic time scale back 75 million years.

SEE FIGURE 2.

Figure 2. Tension rift on the spreading crest of the Mid-Atlantic Ridge, as exposed in southwest Iceland. The faults and fissures break a plain of basalt lava flows a few thousand years old. The foreground fissure has a maximum width of about 60 meters (200 feet), and a maximum depth of about 45 meters (150 feet) below the rim on the near side and twice that below the rim on the far side. View is northeastward along the Almannagja (Great Fissure). Photo by Bruce Heezen, Lamont-Doherty Geological Observatory of Columbia University.

Additional Data

Also in the late 1960s Le Pichon (French), Dan McKenzie (English), and Jason Morgan (American, at Princeton) deduced from magnetic, bathymetric, and other data the geometric principles by which the great pieces moved. Bryan Isacks, Jack Oliver, and Lynn Sykes, all then of Lamont, showed that seismic data, including the distribution and slip direction of earthquakes, fit the new concepts in detail. John Bird (then at the State University of New York at Albany), Gregory Davis (University of Southern California), John Dewey (English and now at Albany), William Dickinson (Stanford University), Tanya Atwater and Daniel Karig (both then at the Scripps Institution of Oceanography, University of California), and George Plafker and Warren Hamilton (U. S. Geological Survey), among others, showed how the geology of continents and volcanic island arcs could be explained in terms of the new concepts.

By 1970, the concepts had been developed, proved, and broadly applied. The 1970s have produced an enormous amount of additional evidence. The research drilling of the ocean floors by the Deep Sea Drilling Project, for example, has confirmed the ages of the magnetic "stripes," the recognition of which provided the breakthrough in the initial concepts. There have already been many refinements and extensions of our knowledge and understanding. Soviet and Eastern European earth scientists have recognized the importance of the new field, only slowly, and have yet to contribute significantly to its concepts.

THE CONCEPTS

Lithospheric Plates

The Earth's crust is broken into moving plates of "lithosphere" (figure 1). There are seven very large plates, each consisting of both oceanic and continental portions, and a dozen or more small plates (not all of which are shown on figure 1). Each plate is about 80 kilometers (50 miles) thick and can be pictured as having a shallow part that deforms by elastic bending or by brittle breaking and a deeper part that yields plastically, beneath which is a viscous layer on which the entire plate slides. The plates tend to be internally rigid, and they interact mostly at their edges.

Spherical geometry requires that any motion between two portions of a spherical surface be expressed as the rotation of one portion relative to the, other, defined by an angular displacement and by a pole of relative rotation, and that all trajectories of relative motions be along small circles to that pole. (Latitude lines are small circles relative to the Earth's geographic pole). These theoretical constraints are largely met by all of the criteria by which plate motions are demonstrated. The magnetic "stripes" along the midocean ridges define spherical angles: velocities of separation of adjacent plates are constant in angle, not in linear value. The direction of slip in earthquakes along plate boundaries and the orientation of strike-slip faults that form or offset those boundaries are in the corresponding small-circle directions.

SEE FIGURE 3

Figure 3. Cross section through a subduction system where an oceanic lithosphere plate ids converging with and sliding beneath a continental plate. Details and dimensions are those for western Java and the Java trench system, but other continental margin systems are similar.

Plate Movement

All plates are moving relative to all others. There are grounds for suggesting that the African plate may now be approximately fixed relative to the deep mantle, but if so it is the only such plate. Velocities of relative motion between adjacent plates range from less than 1 centimeter (0.4 inch) to about 13 centimeters (5 inches) per year. Although these velocities are slow by human standards, they are extremely rapid by geologic ones: a motion of 5 centimeters (2 inches) per year, for example, adds up to 50 kilometers (30 miles) in only 1 million years, and some plate motions have been continuous for 100 million years.

Plates are now pulling apart primarily along the system of great submarine ridges in the world's oceans (figure 2). Hot material from the deeper mantle wells up into the gap, and some of it melts and is erupted on the surface as lava or is injected near the surface to crystallize as other igneous rocks. The ridge stands high because its material is hot, and hence low in density. As the plates move apart, the ridge material gradually cools and contracts, and its surface sinks. Ridges generally form steplike alternations of spreading centers perpendicular to the direction of motion and of strike-slip (transform) faults parallel to that direction. The actual sense of movement along these transform faults, conceptualized first by Canadian Tuzo Wilson, is opposite to the apparent direction of offset of the ridge.

Where plates converge, one tips down and slides beneath the other. Generally, an oceanic plate slides ("subducts") beneath a continental plate (for example, along the west coast of South America) or another oceanic plate (for example, the east side of the Philippine Sea plate). A trench is formed where the undersliding plate tips down, and the ocean-floor sediment it carries is scraped off against the front of the overriding plate (figure 3). Much is now known about the mechanics of these junctions from geophysical studies and particularly from seismic-reflection profiles made across them with instruments developed for oil field exploration. Farther back under the overriding plate, zones of earthquakes, inclined down into the mantle to depths that reach 700 kilometers (450 miles), show the trajectory of the descending plate. Typically, a belt of volcanoes lies above that part of this inclined earthquake zone which is about 125 kilometers (80 miles) deep.

New oceanic-plate (lithosphere) material is generated by the upwelling processes at spreading ridges. Old lithosphere is consumed, and recycled deep into the mantle, at the same rate at the convergent trenches. The balance is global only: the formation of lithosphere at the Mid-Atlantic Ridge is compensated by subduction primarily in the western Pacific.

Plates slide past one another along strike-slip faults, which can be either on land or at sea. The best known of these faults is the San Andreas fault of California (figure 4).

SEE FIGURE 4

Figure 4. Aerial view northwestward along the San Andreas fault in central California. The Carrizo Plain is on the left. The fault marks the major boundary between the Pacific and North American plates, which are moving inexorably past one another at a rate of about 5 centimeters (2 inches) per year; about two-thirds of the total motion is taken up on the San Andreas fault. Photo by Robert Wallace. U. S. Geological Survey.

Large parts of some of the continents depart from ideal rigid-plate behavior, and undergo much internal deformation (figure 1). The western conterminous United States is now being broadly stretched and sheared; although most of the relatively northwestward motion of the Pacific plate past North America is taken up along the San Andreas fault, a part is distributed far inland, and the continental crust is stretched and shattered into the many block ranges and basins and other structures. Slight extension is affecting eastern Africa. Most complex is Eurasia, into which Arabia and India are now being rammed, with a broad belt being deformed in front of the advancing pieces. As India is pushed north into and beneath interior Asia, a huge region is being shoved obliquely northeastward out of the way.

Applications to Continental Geology

Although the integrated concepts of plate tectonics were proved primarily by geophysical studies and deep-sea drilling of the ocean basins, the concepts have revolutionized our understanding of continental geology. By seeing how the geologic features of modern environments relate to present plate boundaries, we can deduce how similar features in the ancient geologic record related to ancient boundaries. Plate motions have dominated tectonic and magmatic processes for the past 2,500 million years. During that time, however, the interior of the Earth has become cooler and less mobile, and details of the processes have changed gradually.

All present oceanic plates will sooner or later vanish beneath other plates. The oldest ocean floor remaining anywhere in the world is less than 200 million years old. Continents stand high because they are composed of thick, light material --- too light to be dragged deep into the mantle, along with ocean --- plate material, so continents are jammed together when an intervening ocean floor disappears beneath one of them. If present major plate motions continue for another 50 million years, Australia will be crowded against China, and the island complexes of Indonesia and the Philippines will be squashed into a mountain system between the colliding continents. Many such past collisions can be recognized in the geology of the continents.

Plate Motion Mechanisms

The propulsive mechanism for plate motions is still a matter of speculation, and those working with the problem hold conflicting opinions. The greater part of the motions represent very slight differences in spin velocity about the Earth's axis: spreading ridges and subduction zones tend to be aligned more north-south than east-west, and poles of relative rotation between moving plates are concentrated at high latitudes. These facts indicate that the Earth's rotation is an important part of the mechanism.

EARTHQUAKES

The recognition that earthquakes are confined primarily to narrow connecting zones which have distinctive bathymetric settings was one of the major developments that led to the concepts of plate tectonics. Narrow zones of earthquakes follow the spreading centers and strike-slip faults shown on figure 1. Inclined zones of earthquakes extend deep into the mantle from the indicated subduction-zone trenches. There are also many earthquakes within those continental terrains undergoing distributed deformation. Shallow earthquakes represent sudden slippages that release strain stored elastically in rock over long periods. Whether deep-mantle subduction-zone earthquakes represent similar elastic rebound or an abrupt contraction of part of a descending lithosphere plate into rock of higher-density minerals is uncertain; these deep earthquakes are mostly small enough, and distant enough from the surface, so that their potential for human disaster is small.

Oceanic spreading centers are so hot at shallow depth that the solid rock above them cannot store enough elastic strain to produce great earthquakes, and the infrequent large earthquakes that do occur in ridge systems are mostly on the longer strike-slip (transform) faults, along which spreading on one ridge segment is stepped to that on the next. Great earthquakes occur primarily along convergent (subducting) and strike-slip plate boundaries and within those parts of the continents undergoing intraplate deformation (figure 5). These are the settings where fault ruptures of adequate size are possible.

SEE FIGURE 5

Figure 5. World map showing the location of earthquakes of magnitude 8.0 and higher. 1897-1976. These great earthquakes occur primarily along convergent or strike-slip plate boundaries and in zones of compressive or strike-slip deformation within continents. Data from the "World Seismicity Map. " compiled by Arthur Tarr and printed by the U. S. Geological Survey in 1974, with newer locations added.

Subduction-zone Earthquake: Alaska, 1964

A great subduction-zone earthquake of magnitude 8.5 struck south-central Alaska on March 27, 1964. Here Pacific lithosphere that is moving northwestward tips down at the Aleutian trench and slides, with gentle inclination, beneath a melange wedge similar to that illustrated by figure 3. The base of the wedge is being dragged very slowly northwestward. Compressive strain stored slowly at the base of the wedge resulted in sudden faulting that broke violently through to the surface.

Strike-slip Earthquakes: San Andreas Fault System

The Pacific and North American plates are sliding past one another at a steady 5 or 6 centimeters (2 or 2½ inches) per year. About two-thirds of this movement is taken up by slippage along the San Andreas fault and related structures, whereas the rest is taken up by other structures, mostly farther inland in the Basin and Range province. The sides of each active fault, like the San Andreas, generally remain stuck together until enough strain has accumulated in the flanking, bending rocks to rupture the bond and permit the lagging parts to suddenly catch up with the rest. Earthquakes on the San Andreas fault rarely reach deeper than about 15 kilometers (10 miles), and presumably the two sides flow smoothly past each other, without sudden ruptures, in the hot region of easy recrystallization beneath this depth. In some areas, notably that near Hollister, south of San Francisco, the two sides of the surface fault slide past one another almost constantly with only small earthquakes, steadily increasing the offset of such features as fences and roads.

The deformational style of the San Andreas fault changes in the Imperial Valley region of southeastern California, and with it earthquake patterns, in accord with the changing geometry of the lithosphere plate boundary. The Pacific plate is moving northwestward relative to the North American plate (which includes northern Mexico), thus moving Baja California away from the mainland and opening the Gulf of California (figure 1). This pulling apart is accomplished by a series of short spreading centers stepped to one another by transform faults, and the northernmost spreading centers have produced much of Imperial Valley itself.

Compressive Deformation Within The Continent:

San Fernando, 1971

Both the Pacific and North American plates are moving relative to the deep mantle; thus, so also is the San Andreas fault boundary itself, which is slowly changing its shape as the adjacent plates deform. In southern California, the sector of the fault from north of Los Angeles to east of San Bernardino has rotated slowly counterclockwise, from a northwest to a west-northwest trend. The result is that in this sector pure strike-slip (sideways) motion is becoming more difficult, and oblique compression of the continental crust is taking place. The San Fernando earthquake of February 9, 1971 (M 6.5), was a small manifestation of this changing pattern.

Tensional Deformation: The Basin and Range Province

The San Andreas fault is the most important single structure in the broad boundary zone between the Pacific and North American plates, but some north-westward motion is distributed far inland across the western United States. The resulting oblique extension of the continental plate has produced the Basin and Range province which includes eastern California, Nevada, and western Utah, and similar terrains in western Montana, Idaho, eastern Oregon, southern Arizona, and New Mexico (figure 1). The province is characterized by a thin crust, by high temperatures at shallow depth, and by volcanism of a type which typically accompanies extension; these features may be products of the thinning of the crust by extreme extension. Tensional faults have blocked out mountains and basins which characterize the province. Surface faults tend to follow the pre-extension structural grain of near-surface rocks: the mechanical properties of the crust determine the details of its response to regional stresses.

VOLCANOES

Most volcanoes are products of lithosphere-plate motions. The "ring of fire" around the Pacific represents one type of this volcanism. The chains of volcanoes in the island arcs (such as the Aleutian Islands) and continental margins (such as the Andes) around much of the ocean form above undersliding oceanic plates. The main volcanic axis is typically about 125 kilometers (80 miles) above the inclined zone of earthquakes that marks the descent of the lithosphere plate into the deep mantle (figure 3), so processes related to the descent and to that depth must control the melting of the magmas. The melts that arrive at the surface to erupt in volcanoes have been profoundly modified by reactions with the mantle and crustal rocks through which they have risen. Lavas formed in this setting have distinctive compositions and systematic variations that relate directly to their height above the subducting plate. These characteristics permit geologists to recognize rocks formed in similar settings in the geologic past and to estimate even the depths to the long-dead seismic zones above which they formed. In ancient terrains, the volcanic rocks have been eroded away and granites and other rocks which crystallized slowly within the crust from similar magmas are now exposed.

MINERAL DEPOSITS

Many mineral deposits have plate-tectonic explanations. The great bulk of the metallic minerals mined in the western United States --- copper, molybdenum, tungsten, gold, silver, lead, and so on --- formed from magmas above subduction zones like that of figure 3, during the period from 100 ??? million to 15 ??? million years ago. Pacific Ocean lithosphere was then sliding rapidly beneath North America, and belts of volcanoes formed, with granites crystallizing from the large magma chambers beneath the volcanoes. The metals were concentrated in the last-remaining liquids in the magma chambers, after crystallization of the voluminous silicate minerals. Where conditions were favorable, as generally they were not, these enriched liquids altered and replaced the igneous rocks or the nearby wall rocks to form ore deposits. Different combinations of metals concentrate at different levels in the chamber, so the depth of erosion into an igneous complex controls the type of ore deposit that may be found within it. There are also provincial variations in types of deposits, and it is likely that depth of melting of the initial magma (which would be very different in composition from the final magma reaching the surface) and the composition of the mantle and crustal rocks through which the magma rises and reacts are important in determining what, if any, metals are concentrated within it.

PETROLEUM AND NATURAL GAS

Oil and natural gas occur in sedimentary strata in wedges and basins that are by-products of the motions of lithosphere plates. Most oil, including that of the Persian Gulf (which represents more than half of the known global total of reserves), occurs in continental-shelf strata. When a continent is split and the pieces are separated as an ocean spreads between them, the edge of the continental crust slowly spreads oceanward, thins, and subsides. Sedimentation keeps the top of the subsiding shelf subhorizontal, whereas deeper layers tend to have progressively greater, but still gentle, oceanward dips. Oil generated in such strata migrates up-dip, toward the continent. Oil along the coast of the Gulf of Mexico, both onshore and offshore, formed in this setting. The exploration now underway on the offshore Atlantic Outer Continental Shelf of the eastern United States is based on the theory that there may be large quantities of oil there, but drilling of possible trapping structures identified by geophysical methods will of course be necessary to test the theory.

Another important setting for oil accumulation is the foreland basin, which in various forms develops on the landward side of subduction systems (figure 3). In the United States, small oil fields in western Wyoming exemplify this setting, and the large Prudhoe Bay field of northern Alaska is in part of the same type. The latter field may prove to be the largest ever found in North America.

Oil occurs in broad basins within continents. Some of these basins are products of the wrenching of continental plates during collisions with other plates, whereas others are not related directly to plate interactions and perhaps have subsided because of cooling at the base of the lithosphere plate.

COAL

Coal is formed from wood and plant debris buried beneath other sediments and transformed slowly, under the influence of heat and pressure, from a mixture of cellulose and carbohydrates to a high-carbon residue. Most coal represents swamp deposits, and practically all is younger than 350 million years old.

Coal swamps originated primarily through interactions of lithosphere plates. The Appalachian coal of the United States, for example, formed about 275-310 million years ago in a foreland basin, analogous in its early stages to that of figure 3, but complicated in later stages by the collision of Africa with North America. (The modern Atlantic Ocean began to open about 200 million years ago, rifting approximately along the previous collision suture). The basin was alternately flooded by the sea, then drained. Coal swamps formed during the emergent periods. The fluctuations of sea level were themselves caused by glaciation and deglaciation of a large south-polar continent, Gondwanaland, which owed its position to plate motions.

CONCLUSION

The geologic processes which affect society---both harmfully, through earthquakes and other catastrophic events, and beneficially through mineral and fuel resources---are largely controlled by the motions of lithosphere plates. Plate-tectonic concepts permit an understanding of the interrelationships between the processes operating within the Earth. We have begun to integrate into a genetic, unifying framework the descriptive findings of the earth sciences.

SUGGESTIONS FOR FURTHER READING

Bally, A. W., 1975, A geodynamic scenario for hydrocarbon occurrences: Proceedings Ninth World Petroleum Congress, v. 2, p. 33-44.

Burke, Kevin, 1977, Aulacogens and continental breakup: Annual Review of Earth and Planetary sciences, v. 5, p. 371-396.

Burke, Kevin, Dewey, J. F. and Kidd, W. S. F., 1977, World distribution of sutures---the sites of former oceans: Tectonophysics, v. 40, p. 69-99.

Dewey, J. F., 1977, Suture zone complexities---a review: Tectonophysics, v. 40, p. 53-67.

Hallam, Anthony, 1973, A revolution in the Earth sciences. Oxford, Clarendon Press, 127 p.

Hamilton, Warren, 1978, Mesozoic tectonics of the western United States: Pacific Section, Society of Economic Paleontologists and Mineralogists, Paleogeography Symposium 2, p. 33-70.

Irving, E., North, F. K., and Couillard, R., 1974, Oil, climate, and tectonics: Canadian Journal of Earth Sciences, v. 11, p. 1-17.

Klemme, H. D., 1975, Giant oil fields related to their geologic setting---a possible guide to exploration: Bulletin of Canadian Petroleum Geology. v. 23, p. 30-66.

Le Pichon, Xavier, Francheteau, Jean, and Bonnin, Jean, 1973, Plate tectonics. Amsterdam, Elsevier Scientific Publishing Company, 300 p.

Sawkins, F. J., 1972, Sulfide ore deposits in relation to plate tectonics: Journal of Geology, v. 80, p. 377-397.

Smith, A. G., 1976, Plate tectonics and orogeny---a review: Tectonophysics, v. 33, p. 215- 285.