The fact that lithospheric plates were beginning to respond (2.5 b.y. B.P.) to deformation, intrusion, and deposition in a mode comparable to that of today is indicated by the development of linear/arcuate orogenic belts bordering continental plates, leaving stable interiors with little-deformed cratonic sequences and linear dike swarms, the development of aulacogens and continental-rise sequences, island arcs, Andean arcs and back-arc basins, back-arc thrust belts adjacent to high-potash minimum-melting granites and slip-line indentation fracture systems bordering linear/arcuate orogenic belts, and geochemical patterns of igneous rocks comparable to modern tectonic equivalents; all these features indicate that modern-style plate tectonics began in the early Proterozoic. Archean-type greenstone belts and granulite-gneiss belts continued to form throughout the Proterozoic, probably in marginal basins and the deeper levels of Andean belts, respectively. In the early Proterozoic a large number of orogenic belts formed, which are increasingly being interpreted in terms of Wilson Cycle processes. In the mid-Proterozoic (1.7 to 1.2 b.y. B.P.), major abortive rifting gave rise to anorogenic sodic anorthosites and rapakivi granites. In the period 1.0 ± 0.2 b.y. B.P., the Grenville and Dalslandian belts in the North Atlantic region formed with prominent Andean-type and Himalayan-type stages of development; today we see deeply eroded levels of these belts. The 0.8 to 0.57 b.y. B.P. Pan-African/Braziliano/Cadomian belts are widely acclaimed as having formed by Wilson Cycle tectonics. The Proterozoic was a period of substantial lateral plate motion, accretion, and subduction and of corresponding crustal growth, although less intense than in the Archean. It was also a period of differentiation of the continental crust and the formation of potash granites in upper levels.

Studies of recent years, notably from metamorphic terrains, have resolved the magnitude and direction of apparent polar wandering (APW) motions over much of Proterozoic times. This paper re-examines the case for a Proterozoic supercontinent, and shows that both the polarities and the positions of the Proterozoic paleopoles from the major shields conform to a single APW path using a unique reconstruction. These data imply that the continental crust was amassed together as a single lens-shaped body as heat loss by extensive small-scale mantle convection correlating with low rates of APW was replaced by a large-scale mantle convection system responsible for the Proterozoic mobile belt regime and correlating with high rates of APW. The later greenstone belts (ca. 2900-2200 Ma old) formed in permobile environments and early Proterozoic straight belts (> 2200 Ma old) formed between larger stabilized divisions of continental crust are oriented parallel to the long axis of the continent. This alignment is preserved by younger mobile belts and most tectonic trends. There was a progressive contraction of rapakivi-massive anorthosite magmatism towards one margin of the supercontinent as temperature gradients declined; most features linked to ocean lithosphere subduction are associated with this margin. Large-scale fracturing with the formation of aulacogens took place along the opposite colder margin. Peripheral parts of the supercontinent broke up apparently without major separation 1100 Ma ago, and the central parts broke up at the beginning of Cambrian times with the formation of a large ocean basin between the Gondwanaland and the Laurentian, Fennoscandian, and Siberian Shields; this event marking the end of Proterozoic times is defined by a number of chemical changes linked to the faunal diversification. Both of these episodes are defined by widespread alkaline magmatism and rifting, and the latter is linked to subdivision of a simple roll mantle convection system which appears to have pertained in some form during most of Proterozoic times. The characteristic signature of the Proterozoic APW path is a closed loop with a hairpin near the apex. These features can no longer be correlated with tectonic/magmatic episodes and appear to reflect overturn in the driving mantle system. The ensialic mobile episodes follow periods when APW movements were relatively small and may be linked to a thermal blanketing effect of the continental crust. Since 2700-to 2200-Ma-old paleopoles from Africa, Laurentia, Australia, and India are not significantly different from one another on the supercontinent reconstruction, the continental crust has evidently been a highly coherent unit since late Archean times. Movements across later mobile belts do not appear to have been on a scale large enough to be detected by paleomagnetism.

The NRM (natural remanent magnetization) of a metamorphic rock is the vector sum of primary NRM and one or more secondary overprints. Magnetic overprinting is a thermal or chemical effect resulting from (1) burial and uplift (regional heating), (2) intrusive activity (local heating), or (3) hydrothermal supercritical solutions (local heating and/or alteration. We present a number of examples drawn from the Proterozoic of North America illustrating how paleomagnetists decipher magnetic overprints and use them to date and interpret tectonic events. Granites from the western Superior Province record the 2600 Ma Kenoran orogeny and an event at 1250 Ma, perhaps related to Mackenzie dike intrusion elsewhere in the Shield. Diabase dikes and country rocks in the Abitibi subprovince, particularly those near incoherent dike contacts and fault zones, bear thermal overprints whose probable ages are 1900 to 1700 Ma. The overprints agree in direction with primary NRMs of similar ages from the Churchill Province, implying that the Hudsonian orogeny involved little relative motion of Superior and Churchill cratons. Multiple overprints dating from slow uplift and cooling of the Grenville orogen between 1050 and 800 Ma postdate the collisional phase of the Grenvillian orogeny, but 1100-Ma-old surviving primary NRM of the Tudor gabbro appears to record precollisional divergence between “Grenvillia” and the rest of the Shield.

A brief review of secular trends of Proterozoic sedimentation, volcanism, K20/Na20 ratio of supracrustal rocks, massive sulfide mineral deposits, isotopic Sr ratio in seawater, geochronology, and paleomagnetism clearly shows the episodic nature of orogenic activity. Brief episodes of deformation and generation of continental crust, around -1800 Ma and around -1000 Ma were separated by long (800 Ma) periods of limited continental growth. Rapid thickening of continental lithosphere and growing instability of oceanic lithosphere led to the start of subduction around -800 Ma. Subduction-controlled plate tectonics has dominated orogenic evolution ever since. A theory of unsteady mantle convection limited to the upper mantle (Tozer, 1965) can explain Proterozoic orogenies as the consequence of convective over-turns separated by “warm-up” periods. Present plate motions would control the distribution and motions of mantle currents rather than vice versa. Proterozoic plate motions of limited lateral amplitude led to formation of ensialic orogens. Paleomagnetic polar wander paths suggest that local crustal extension preceded phases of compression responsible for orogenies. The change-over from the Archean to the Proterozoic would correspond to a change in convection regime of the upper mantle from steady to unsteady convection.

Proterozoic foldbelts older than about 1 Ga lack the distinctive signatures of the contemporary Wilson cycle although they frequently contain thick sedimentary assemblages that resemble those in Phanerozoic orogens. Mafic volcanic rocks are rare, however, and occur intestratified with sedimentary strata; they are not layered and contain no sheeted dikes and can therefore not be interpreted as ophiolites. These characteristics as well as paleomagnetic and isotopic constraints preclude an evolution of these belts during closure of extensive oceans and subduction of significant amounts of oceanic lithosphere. An alternative plate tectonic model is developed that invokes rifting, heating, and stretching of the crust as a result of lithospheric thinning over a mantle plume. This mechanism eventually leads to a “geosynclinal” basin entirely floored by continental crust. The rise of asthenosphere enhances gravitational instabilities in the old and dense subcrustal lithosphere that, on fracturing after crustal stretching, may delaminate spontaneously. Hot asthenospheric material rises to take the place of the detached and sinking lithospheric slab, thereby inducing A-subduction and inter-stacking of continental crust. The much thickenend crust is partially melted at depth, intruded by synorogenic and postorogenic granites, and finally uplifted and eroded to its present level of exposure. Episodic thermal anomalies during orogeny are caused by the rise of asthenospheric magmas to the base of the crust and by radioactive self-heating after crustal interstacking. The model is entirely compatible with the concept of horizontally moving plates but differs from the Wilson cycle in that no wet oceanic crust is generated during basin formation and none is consumed during orogeny. Instead, dry subcrustal lithosphere sinks down but does not cause calc-alkalic magmatism. Towards the end of the Proterozoic, Wilson cycle signatures become widespread in the global rock record and signify a worldwide change from predominantly intraplate orogeny to predominantly plate margin orogeny that characterizes the Phanerozoic geodynamic pattern. Plate tectonics, therefore, is a nonuniformitarian process.

Plate tectonic interpretations are commonly used for Phanerozoic tectonic features, but there are still differences of opinion regarding the best model for Proterozoic tectonic features. We suggest that it will be more fruitful to apply the plate model as used in Phanerozoic examples than to build a special model based only on Proterozoic data, or to decide ad hoc what modifications of the plate model may be necessary. As a stimulus for discussion and further work we present plate tectonic interpretations for three widely discussed problems in the Proterozoic terranes of eastern North America: the search for a Grenville suture, the relationship between the Grenville orogeny and Keweenawan rifting, and possible relationships between the Labrador Fold Belt and the Canadian Southern Province. We emphasize the separate stages of the Wilson Cycle of ocean opening and closing; examine some of the available data appropriate for plate tectonic interpretations, particularly isotopic dates; and point out new avenues of investigation suggested by the model.

Geochemical parameters and isotopic indices of Precambrian silicic igneous rocks record largely ensialic crustal environments during 2.5 to 1.0 b.y. ago. Simatic crustal regimes have evidently coexisted (i.e., southwest and central United States, Fennoscandia, Venezuela) though their extent remains unclear. The significance of initial Sr 87 / 86 ratios (Ri) and large-ion-lithophile (LIL) element data is considered: both tend to be higher in Proterozoic than in Archean silicic igneous suites, suggesting, respectively, predominance of ensialic and ensimatic anatectic processes. The ensialic nature of middle Proterozoic crustal reworking is supported by Sm/Nd study of composite North American samples yielding Archean model ages. Estimates of average crustal compositions are allowed by Precambrian sediments. K/Na ratios, rare earth elements (REE), and Ri values of Proterozoic sediments indicate derivation from differentiated sialic source terrains, including a high proportion of K-rich silicic igneous rocks. This contrasts with the mafic and Na-rich nature of Archean greenstone-tonalite terrains and derived sediments. The relative paucity of oceanic crustal signatures during 2.5 to 1.0 b.y. ago, calculations of minimum crustal growth rates, limits on the thickness and volume of the Proterozoic sial, and paleomagnetic apparent polar wander paths (APWP) render the evidence difficult to reconcile with plate tectonics processes and/or with present-day Earth surface dimensions. The observed coincidence of APWP for Laurentia, Greenland, Africa, and Australia during 2.3 to 1.6 b.y. ago implies a preexistence of equivalents of the Atlantic, Indian, and Pacific ocean basins in approximately their modern positions—a conclusion faced by major geological, geochemical, and isotopic objections. It may appear from the Proterozoic record that during 2.5 to 1.0 b.y. ago almost three fourths of the continental crust has already existed. If so, five alternative models are considered with regard to the nature of the other three-fourths of the crust: (1) global sial of continental thickness; (2) a thin global sial; (3) dispersed sial and sima plates; (4) a hemisphere-size simatic regime; (5) global sial on a smaller radius Earth. An apparent inability of models 1, 2, 3, and 4 to account for the unrecorded Proterozoic crust is indicated, which directs attention toward model 5. Objections to and constraints on radial expansion are considered. Rapid expansion rates in the order of 10 mm/year during the Mesozoic-Cenozoic are unlikely. However, slow expansion rates in the order of 0.5 to 1.0 mm/year during early and middle Precambrian times may be capable of accounting for the questions posed by the geochemical, isotopic, and paleomagnetic data. This enigma remains open for further investigation, notably by paleomagnetic and isotopic studies of the worldwide system of ca. 2.4-b.y.-old basic dykes, which could provide a definitive time/place reference grid.

The uniformity of rare-earth element (REE) patterns in clastic sedimentary rocks, due to the low solubility and short residence times of REE in the oceans, provides overall average upper crustal compositions for these elements. Effects of weathering, diagenesis, and metamorphism are minor. Local provenance is recorded in first-cycle sediments, but is rapidly erased with sediment maturity. The REE patterns in Archean sedimentary rocks indicate that the Archean crust was not highly evolved. It appears to have been dominated by basaltic and Na-rich granitic rocks (tonalites and trondhjemites). The REE patterns in Proterozoic and later sedimentary rocks indicate a major episodic break at the Archean/Proterozoic boundary. This is consistent with a change to a more differentiated upper crust, dominated by granodiorites, with negative Eu anomalies. These are inferred to result from intra-crustal melting, during which Eu is retained in a plagioclase-rich lower crust. Detailed descriptions of the change in REE patterns at the Archean/Proterozoic boundary are given for the Huronian (Canada) and Pine Creek Geosyncline (Australia) successions. Evidence from these successions, the Hamersley basin (Australia), and the Pongola (South Africa) sequence indicates that the change in upper crustal composition was not isochronous, but extended over a period from about 3.2 to 2.5 Ga ago. Abundance tables are given for the Proterozoic continental crust, upper and lower crusts, and the Archean crust. A model for continental crust evolution suggests that the evolution of the present crust began in the Archean and that most of the volume of the crust was formed from the mantle between 3.2 and 2.5 Ga. This was followed closely by intracrustal melting which produced the Proterozoic upper crust.

Anorogenic magnatic activity characterizes much of the late to mid-Proterozoic, from 1030 to 1770 m.y. ago, in a broad belt trending from the southwestern United States, northeastward through Labrador, across southern Greenland, and into the Baltic shield. The association of gabbroic to anorthositic rocks, a separate mangeritic series of primarily intermediate composition, and granite of definite rapakivi affinity comprise an anorogenic “trinity” of world-wide occurrence. With the exception of the 1.76 b.y. Montello batholith (Wisconsin), this episode in North America is restricted to the interval 1.0 to 1.5 b.y. and occurs in three distinct events. Over 70 percent of Proterozoic anorogenic magmatism occurs in a 1.41 to 1.49 b.y. old 600-1000 km wide belt trending from southern California to Labrador that volumetrically and age-wise is totally a North American phenomenon. Renewed anorogenic granite magmatism occurred form 1.34 to 1.41 and from 1.03 to 1.08 b.y. ago in lesser proportions. Although anorthositic and mangeritic rocks are abundant in some provinces (e.g., Labrador), rapakivi granite (in the broad usage of the term) represents by far the most abundant magma-type generated during this nonorogenic period. The modal and mineral composition of these granitic rocks is distinctive and reflects the potassic and iron-enriched character of the magmas and the unique conditions under which crystallization occurred. Principal rock types include biotite ± hornblende granite to adamellite although numerous peraluminous, two-mica (biotite + celadonitic muscovite ± garnet) granites also occur. Crystallization of these epizonal granitic magmas occurred over the range of 640 to 790° C at low total pressures (most less than 2 kb) and at relatively dry conditions. A dramatic difference in crystallization conditions lies with the level of oxygen fugacity which ranges three orders of magnitude from low (ca. QFM) to high (above Ni-NiO), resulting in systematic differences in Fe-Ti oxide mineralogy and mafic silicate composition. Compositionally, the granite magmas are subalkalic and marginally peraluminous (peralkaline varieties are rare to nonexistent). Although some hastingsite or riebeckite-bearing granites may have been derived from fractional crystallization of the mangeritic series, most are primary melts derived from fusion of lower crust material. The high potassium, Fe /Mg, Ba, and rare earth element (REE) composition of the granites is consistent with small degrees of fusion (10-30 percent) of calc-alkaline crust of quartz dioritic, tonalitic, and granodioritic material. Initial Sr isotopic ratios average 0.7051 ± .0025. The relatively low ratios are the result of short residence times (commonly 170 to 340 m.y.) with much of the crustal source being formed in a preceding orogenic event. An earlier melting episode need not have occurred for the source. The dry nature of the magmas is due to vapor under-saturated melting of a metaigneous source with a total water budget less than 1 percent and tied up in relatively stable residual hydrous phases. The derivation of a marginally peraluminous melt from a metaluminous source is a consequence of variable amounts of residual hornblende ± clinopyroxene. The generation of isolated crustal-derived magma under anorogenic conditions is considered to be the result of localized thermal doming in the mantle. The mantle-derived anorthositic and mangeritic magmas may have played an active role in generating the necessary heat of fusion at lower crustal levels. For North America, the 1.4 to 1.5 b.y. event does not have the consistent age progression of a track and is probably an incipient rift that failed to integrate into a world-wide plate system. At a more mature stage, the 1.0 to 1.1 b.y. Keweenawan episode and the midcontinent gravity high represents another unsuccessful rifting attempt of the Proterozoic North American craton. The isolated granite complexes of this age (Pikes Peak, Enchanted Rock, Red Bluff), as well as the earlier, localized 1.76 b.y. and 1.34 to 1.41 b.y. magmatic episodes, may be further representations of a thermal perturbation at mantle depths during this fragile period of crustal “stability.”

There is no simple relationship between a PT point derived for a metamorphic complex from mineral assemblage and composition data and the equilibrium geotherm which may be calculated for the crust. Transient or complex geotherms of various slopes may be generated by magmatic intrusion, rapid burial, rapid uplift or tectonic thickening, or almost any combination of these processes. Proterozoic metamorphic complexes yield PT conditions ranging from about 6 kb, 600° in the Wakeham Bay area to 3-5 kb, 750° in the Bear Province and 10 kb, 980° in the Musgrave Ranges. The lowest apparent geotherms lie in linear metamorphic belts, the highest in domal regions in the NW Canadian shield, while intermediate values occur in most granulite terrains. These can be related to equilibrium surface heat flows in the range 80 to 120 mW/m 2 , with mantle heat flow of about 50 mW/m 2 for reasonable crustal compositions. The higher values can be simulated by intruding sloping sill-like bodies of basic magma into the crust at depths around 20 km; the complex heat transfer equations have not yet been solved for such situations, however, and calculations are inexact. Tectonic crustal thickening and erosion can produce similar effects; but in the absence of the former, the latter can most easily be caused by continental underplating by basic magma. Gradients are generally higher than in Phanerozoic orogenic belts, and no blueschist facies rocks have yet been substantiated in pre U. Proterozoic complexes.

The period of earth history which includes Proterozoic and late Archean time (roughly 1.0-2.7 billion years ago) was punctuated by metamorphism at high temperature (700°-900° C) and high pressures (8 ± 2 kbar for many terrains) in a thick continental crust. Most of the known charnockites (quartzofeldspathic orthopyroxene-bearing granulites) were formed in this period. Archean charnockites were most commonly produced by metamorphism of amphibole-bearing “gray” gneisses whose precursors probably were, in many cases, calc-alkaline plutonic suites. Charnockites showing relic igneous textures are more characteristic of the later Proterozoic than of earlier time. Many large Proterozoic charnockite bodies had plutonic precursors of an anorogenic, alkalic-calcic suite that included massif anorthosites and rapakivi granites. The definitive characteristics of most Archean and Proterozoic charnockites have resulted from granulite-facies metamorphism in the presence of CO 2 -rich, H 2 O-poor fluids. Chemical analyses, experimental petrology, fluid inclusion studies, and field relations in the Archean amphibolite-facies to granulite-facies transition region of southern India show that charnockitic metamorphism of gray gneiss may be either virtually isochemical or profoundly metasomatic. K-metasomatism and anatexis commonly attended charnockitic metamorphism, as H 2 O-rich fluids were driven upwards ahead of a wave of hot CO 2 . Dehydration and depletion of incompatible elements, especially Rb and U, followed migmatite formation. It is not yet clear whether the observed depletions can be entirely accounted for by leaching by a vapor phase or whether escape of anatectic melt is required to produce the most intense depletions. The Proterozoic Adirondack and Grenville charnockite bodies are the result of a high-pressure metamorphic overprint on plutonic rocks originally emplaced at relatively shallow levels. The quartzofeldspathic igneous rocks were not consanguinous with associated anorthosites but may have been derived by melting of the deep crust during passage of gabbroic anorthosite magmas. Some orthopyroxene in very Fe-rich charnockites may be a primary magmatic phase, but the amount of uniquely magmatic charnockite is probably small. The relationship of granulite metamorphism to periods of crustal accretion is not clear at the present time. A source of heat and CO 2 is necessary for the metamorphism, and this may require major magmatic additions to the crust. Alternatively, liberation of CO 2 , as well as high pressures, may have resulted from continental collision as a 15 to 30 km-thick continental segment overrode carbonate-bearing shelf sediments and evaporites of another. Opening and closing of a proto-ocean could explain the anorogenic magmatism of the Grenville province followed by granulite metamorphism. Similarities between Proterozoic and Archean granulite terrains suggest that such plate tectonic processes operated at least back to the late Archean.

A brief review of some major types of Proterozoic metal deposits indicates that each type is characterized by a distinctive lithologic and tectonic setting. The main types of metalliferous ores considered are: volcanic-hosted massive sulfide deposits, sediment-hosted massive sulfide deposits, stratiform copper deposits, carbonate-hosted lead-zinc deposits, magmatic copper-nickel deposits, and banded iron formations. Metal deposits in sedimentary and volcanic rocks dominate Proterozoic metal-logeny, and it can be demonstrated that most formed in tectonic environments characterized by rifting. This association is reinforced by the time-space relationship between specific tectonic environments and metal deposits during Proterozoic time. Although Proterozoic metal deposits are distinctly different from the dominant types of late Phanerozoic deposits, these differences are considered to be primarily a function of erosion levels, and initial deep burial of syngenetic-type deposits. Finally, there seems to be no compelling evidence that tectonic styles or ore-generating systems were markedly different in the Proterozoic versus those of the Phanerozoic.

Proterozoic massive sulphide deposits occur in a wide variety of geologic settings that range from those composed of volcanic flow rocks to those composed of clastic sedimentary rocks. Evidence of volcanism, contemporaneous with the mineralizing event, can be found in the host rocks to most of the deposits, including a number of those occurring in thick sedimentary sequences. Proterozoic massive sulphide deposits in volcanic terranes have metal ratios similar to those of Archean, Phanerozoic, and younger deposits when compared to deposits in similar geologic settings. Deposits found in sedimentary terranes are more lead-rich and copper-poor than those in volcanic terranes. Stratabound nickeliferous sulphide deposits at Out-okumpu, Finland, and Thompson, Manitoba, are examples of exhalative massive sulphide mineralization. A genetic model that appears to be compatible with empirical observations and recent isotopic studies on massive sulphide deposits involves the generation of magmatic ‘ore-fluids’ during the crystallization of a magma and the mixing of this fluid with connate water and/or seawater in the upper portions of the volcanic rock pile.

Numerous Pb-Zn deposits formed during the Proterozoic, in contrast with the situation in the Archean. The largest examples, which are reviewed here, are McArthur, Mount Isa-Hilton, Sullivan, Gamsberg, and Broken Hill. These huge stratiform deposits occur well above the bases of thick, predominantly sedimentary sequences which accumulated in major ensialic troughs. The sequences are dominated by carbonates or clastics, but the mineralization is localized within distinctive chemo-clastic facies which were deposited in basins of restricted extent. There is commonly evidence for synsedimentary faulting and for distal igneous activity. Biogenic sulfide was probably incorporated into some of the pyrite but is unlikely to have precipitated the bulk of the Pb and Zn. The base metals are of crustal derivation; they could have been supplied by ascending basinal brines and/or hydrothermal fluids migrating from distant centres of igneous activity. There is active debate, focussed on the relatively pristine McArthur deposit, as to whether the mineralization formed at or beneath the sediment-water interface. It is maintained here that all features are consistent with syngenetic accumulation of the great bulk of the sphalerite, galena, and some of the pyrite, complemented by growth of biogenic pyrite and other diagenetic minerals within the sulfidic muds.

The wide variety of genetic models for Proterozoic iron formations indicates a lack of agreement about constraints on models. Constraints are reviewed in this paper to help focus the evidence for modeling. Useful constraints are found in modern processes of iron concentration and in variation among iron formations through earth history, as well as in characteristics of Proterozoic iron formations.

The Proterozoic System contains many of the world’s major uranium deposits. The earliest are of the quartz pebble conglomerate-type and are typified by the deposits at Elliot Lake, which belong to the uranium-thorium association. They have uraninite with both uranium and thorium, and represent paleoplacers. Except for some very small occurrences, this type of deposit developed only before 2,000 Ma ago. The principal economic deposits that first appeared during the Middle Proterozoic contain uranium without thorium, and all three types (sandstone, vein, and unconformity) were formed in conjunction with fluvial sandstone. It appears that prior to 2,000 Ma ago the low level of oxygen in the atmosphere permitted thorium-bearing uraninite to be weathered from igneous rocks, transported, and then deposited as detrital minerals. With the increase of oxygen in the atmosphere, about 2,000 Ma ago, meteoric waters could oxidize and thus dissolve uranium during weathering, separating it from the thorium which remained as an insoluble residue. The uranium was carried in solution and deposited in reducing zones within fluvial environments. This also explains the geochemical association of uranium with Ni, Co, Mo, Cu, Zn, Mn, Fe, V, Ag, Si, Se, and As, which is a characteristic of these deposits.

The sialic plutonism that transformed prevailing petrographic, sedimentary, and paleogeographic styles from Archean to Proterozoic aspects lasted for 600 million years. It created the first extensive cratonal surfaces, with all their potential for interaction among a diversity of microbiotas and their physical surroundings. Earth, in fact, was essentially completed during the Proterozoic. Continental and oceanic crusts and waters achieved their approximate present dimensions and chemistries. O 2 -releasing photosynthesis became established. Life at the cellular level differentiated, culminating in eucaryotic heredity and leading to Metazoa at the Proterozoic-Phanerozoic transition. A kinetic lag between sources and sinks of O 2 eventually created an oxygenous atmosphere, its development being reflected by such sedimentary peculiarities as detrital uraninite, banded-iron-formation, distinctive soil profiles, red beds, and perhaps sedimentary phosphorites and metal sulfides. Combined with biological evidence, the temporal distribution of such materials suggests an O 2 growth curve connecting these provisional points: (1) near zero O 2 until ~2.3 × 10 9 years ago, (2) 1 percent present atmospheric level (PAL) O 2 2-2.3 × 10 9 years ago, (3) ~7 percent PAL 670 × 10 6 years ago, (4) ~ 10 percent PAL 550 × 10 6 years ago, and (5) present atmospheric levels beginning ~ 400 × 10 6 years ago.

Proterozoic chemical sediments are depleted in 18 O with respect to Phanerozoic analogs. The suggestion that this results entirely from weathering and low temperature exchange with groundwater is not tenable. Metamorphism also fails to explain observed low 18 O values. Two hypotheses that fit the isotopic data are: 1) precipitation of chert and carbonate from a hot Proterozoic ocean (60° C for Precambrian X), or (2) precipitation of chert and carbonate from a Proterozoic ocean depleted in 18 O. Both of these hypotheses have important implications. Close association of diamictite with the Kuruman Iron Formation (Precambrian X) suggests that hypothesis 2 is more likely. A detailed study of this diamictite and its stratigraphic relation to the iron formation, however, is required to confirm this observation. A Late Archean (Precambrian W) iron formation (Weld Range, Western Australia) is strongly depleted in 18 O. If this depletion results from deposition in a brackish water environment, it implies that the hydrologic regime during the Archean/Proterozoic transition was similar to the modern hydrologic regime.