Click for next page ( 160


The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 159
v CONTINENTAL EVOLUTION

OCR for page 159

OCR for page 159
Cenozoic Volcanism in the Western United States: Implications for Continental Tectonics INTROD UCTI ON PETER W. LIPMAN U.S. Geological Survey Volcanic rocks can provide insights into major features of continental tectonics, for which in some cases little other record may exist. Thus, study of Cenozoic volcanism in He western United States may reveal structural and com- positional constraints on the lithosphere of the American plate and on the geometry of past interactions with other plates. This chapter focuses on four features of conti- nental tectonics: (1) the composition and geometry of underlying and, as yet unexposed, cogenetic magma chambers and batholiths; (2) lateral compositional varia- tions in the lithosphere of the American plate; (3) struc- tural discontinuities in this lithosphere; and (4) the geometry of past interactions between the American plate and various Pacific oceanic plates. SUBVOLCANIC INTRUSIONS The study of Cenozoic volcanic activity in the western United States provides information on the distribution 161 and composition of young intrusions at shallow depth. Such intrusions constitute significant upper-crustal conti- nental-tectonic features of major potential economic sig- nificance for mineral and geothermal resources. The in- trusions are also products of evolving magmatic systems that may have caused long-lived perturbations of the lower crust and upper mantle during processes of partial melting, magma migration, and fractionation. The close relationship between continental volcanism and emplacement of shallow granitic intrusions was dis- cussed by Hamilton and Myers (1967~; they concluded that major silicic volcanism is typically the surface mani- festation of batholithic intrusive activity at high crustal levels and that many large batholithic complexes are shal- low, grossly lenticular bodies, roofed largely by cogenetic volcanic rocks erupted during immediately preceding stages in the rise and emplacement ofthe batholith. Publi- cations by these authors and others in the past 10 years have developed generally similar interpretations for volcano-plutonic associations in the western United States, especially for the Boulder batholith of Montana (Hamilton and Myers, 1974; Klepper et al., 1971) and

OCR for page 159
162 large parts of the Siesta Nevada batholith in California (Schweikert, 1976; Fiske et al., 1977~. By analogous rea- soning, the distribution and compositions of volcanic rocks, together with pertinent geophysical data, can be used to infer the geometry of subvolcanic intrusions or even magma chambers beneath young volcanic fields into which erosion has not exhumed widespread intrusive rocks. Oligocene volcanic activity in the San Juan field, south- westem Colorado, is interpreted as recording the rise, differentiation, and consolidation of a composite batho- li~, largely of intermediate composition and covering an area of roughly SO km by 100 km (Steven and Lipman, 1976; Lipman et al., 1978~. Early eruptions that fanned stratovolcanoes of intermediate composition in the San Juan field were accompanied by intrusion of small stocks into the cores of these volcanoes (Figure 14.1A). No evi- dence exists for any shallow magma bodies of batholithic dimensions at this stage, although some of the early volca- noes are clustered, apparently marking sites of concen- trated accumulations of magma. Within several of these PETER W. LIPMAN clusters, sufficient magma subsequently accumulated at shallow depth to permit formation of silicic differentiates, eruption of ash-flow tufts, and resulting caldera collapses (Figure 14.1B). At least 15 ash-flow calderas, mostly formed within a 2-million-year (m.y.) interval (29-27 m.y. ago), are clustered within terrane characterized by a large, stee~sided, flat-floored, Bouguer gravity low (Figure 14.2) that is interpreted as defining the final shape of the batholith, win the calderas marking isolated higher cu- polas. The dimensions of the Bouguer gravity anomaly in the San Juan Mountains (Figure 14.2) and, by inference, those of the concealed batholith are comparable with those of the Upper Cretaceous Boulder batholith of Montana. In Montana the gross stratigraphy of the Elkhom Mountains VoIcanics, erupted penecontemporaneously with em- placement of Me Boulder batholith, is also similar to that of volcanics in the San Juan field; a thick sequence of intermediate~omposition lavas and volcaniclastic rocks is capped by culminating more silicic ash-flow eruptions. Ash-flow-relate<1 caldera collapses must also have accom- v s_ ~ ~ ,~ _~y~~S s~~ EX PI~ANA T~ ON ~~~ 1 a. 35-30 m.y. i,,,,,/, j~'~ v ~ -~ B. 30 - 26.5 m.y. 1 sit 1 . , ~ . note intermediate -composition rocks j~ vl, volconi class ic sedimentory rocks ~3 \2 sl, strotovoiconoes f.,2,ei`''.2 Lote intrusions (Numbered in order of emplacement) Silicic ash-flow sheets Eorly intrusions ~ ~ s 1 Eorly intermediote-compositian rocks v, volconiclastic sedimentary rocks s, strotovolcanoes FIGURE 14.1 Schematic model for evolution of Oligocene subvolcanic batholith in San Juan Mountains (Lipman et al., 1978, courtesy the Geological Society of America). A, Time of early intermediate volcanoes. Clusters of interrnediate~omposition stratovolcanoes are surrounded by interfingering aprons of volcaniclastic debns. Small intrusions fonn at shallow levels in volcanic pile, but it is uncertain whether a large high-level intrusive complex has developed by this time. B. Time of ash-flow eruptions and caldera Connation. Eruption of complex sequence of ash flows and associated caldera collapses is triggered by accumulation at shallow depth of batholithic-size magma bodies of intermediate to silicic composition. Many of these shallow accumulations are localized within clusters of earlier stratovolcanoes. Some calderas are composite, with younger activity nested within older collapse structures, and many caldera collapses are followed by resurgent doming, indicating renewed upwelling of silicic magma.

OCR for page 159
Cenozoic Volcanism in the Western United States ~- 3Y - 163 10 , ( in, faultBar and ball on downthrown side (3 Caldera __ ,` ,, Buried or inferred caidera ~ J (~ ~-;^ / _ _ I 0 10 20 30 40 50 K'L - ETRES 1 1 1 1 1 ~ -300 - Elouguer gravity co ntoursHachures point toward closed gravity lows. Contour interval 10 milligals. Gravity data from Plouff and Packer {1972, fig. 3) , . . DE~VE Ro ~ CoLORADO l it S.A Jeer ~ . _>'-~-'Ce^" "-'6 ~ LOC^710N Of 5~N ~~N VOLCANIC flELD CALDE RAS (in order of increasing age) Lake City Creede Coct~etopa Park San Luis Bachelor La Garita Mount Hope S. Iverton San Juan U ncompaNgre Lost Lake Ute Creek Summ.tv~lle Platoro Bonanza FIGURE 14.2 Calderas in the San Juan volcanic field in relation to Bouguer gravity field (Steven and Lipman, 1976). panted emplacement of the Boulder batholith, perhaps within the region where only intrusive rocks are currently exposed. At another major but much younger caldera complex, centered in Yellowstone National Park and related to large-scale ash-flow eruptions about 0.6 m.y. ago, exten- sive geological and geophysical evidence indicates the presence of a large shallow body of silicic magma (Eaton et al., 1975~. This body is thought to underlie an area more than 85 km long and 55 km wide, as indicated by surface geology and by gravity studies. Seismic-attenuation and P-wave-residual studies suggest that the top of the magma body is only a few kilometers below the surface and is still molten. The molten magma is interpreted to be underlain by an even larger volume of mechanically and thermally disturbed crustal and mantle rocks that contains pods of basaltic and silicic magma and extends at least 50 km into the mantleprobably through the lithosphere of the American plate (Eaton et al., 1975~. These features have been interpreted as the geophysical expression of a "grav- itational anchor" (Straw and Jackson, 1973), which re- sulted from sinking of dense residue from partial melting episodes during formation of the Yellowstone magmas (Christiansen and McKee, 1978~. Here, volcanism is a sur- face manifestation of geophysical discontinuities that ex- tend to great depth. Questions as yet unaddressed are whether such discontinuities are preserved under older volcanic fields and whether such discontinuities can af- fect later tectonic and igneous events. Putting together fragments of the magmatic record in this way, we can determine the main loci of relatively recent platonic activity, for which erosion has not yet provided direct exposure. Much more should be determinable about such economically important en-

OCR for page 159
164 vironments by combined geological, geochemical, and geophysical studies of transitional volcano~lutonic com- plexes, especially where interpretation of ~ird- dimensional relations can be augmented by drill-hole data. COMPOS ITIONAL VARIATIONS IN THE CONTINENTAL LITHOSPHERE Recent volcanic studies using a plate-tectonic framework have emphasized the importance of magma sources below the lithosphere in extensional environments, from ups welling asthenosphere along spreading oceanic ridges; in convergent environments, from descending slabs along subduction systems below magnetic arcs; and in intra- plate settings, Tom upwelling deep "mantle plumes" (Wilson, 1965; Morgan, 1972; Martin and Piwinskii, 1972~. Geochemical studies, especially Sr isotopic data, have tended to emphasize the importance of mantle sources for silicic as well as for magic volcanic rocks and to rule out major contributions from upper~rustal ma- tenals (Hurley et al., 1962; Peterman et al., 1970; Kistler and Peterman, 1973~. Such conclusions, initially drawn largely Tom studies of volcanism in ocean basins. have also been widely held to be generally applicable to continental igneous activity. However, long-recognized contrasts in the nature of volcanic activity between conti- nental and oceanic regions (Gilluly, 1955), as well as addi- tional arguments outlined below, indicate that conti- nental volcanism is strongly influenced by structural and compositional features of He immediately underlying lithosphere, even though the activity may ultimately have been initiated at greater depths. In particular, continental volcanism seems sensitive to the age and composition of the immediately underlying lithosphere, as well as to the geometry of major structural flaws. Implications of the compositions of continental volcanics for compositions of the lower crust and attached lithospheric upper mantle are discussed elsewhere (see Chapter 12~. Accordingly, aspects of compositional relations are outlined here only to the degree necessary to support other interpretations of relations between continental volcanism and tectonics. Recent Pb and Sr isotopic studies of continental vol- canic rocks in the western United States indicate that the isotopic composition of volcanic rocks reflects the age and composition ofthe underlying lithosphere, demonstrating major compositional control by the lower crust or the li~- osphenc upper mantle. For example, isotopic composi- dons of both Tertiary volcanic rocks of the Absaroka vol- canic field and Quatemary volcanics of the Yellowstone Plateau in the same region (Petennan et al., 1970) lie along a well-defined secondary isochron that defines an apparent age of the source region of 2.8 billion years (b.y. - he same as that of the underlying Precambrian basement (Figure 14.3~. In the San Juan Mountains and adjacent parts of the Rio Grande rifle in sounded Colorado and northem New Mexico, Pb isotopic data for rocks rang- PETER w. LIPMAN . 7 1 5 8 Q CL o Cal CL o 150 Cal 15.4 o ~_. ~~ / 46- 35 ,4.5 155 165 206pb/2o~pb 17 5 ,8 s FIGURE 14.3 Lead isotope relations for the Tertiary Absaroka volcanic field (solid circles) and Quatemary basalts and rhyolites (open circles) Mom Yellowstone National Park (Peterman et al., 1970). ing in age from Oligocene to Quaterna~y yield a secondary isochron that indicates an apparent source age of about 1.7-1.8 m.y.again the dominant age of the underlying Precambrian basement (Lipman et al., 1978~. In these two regions the rocks analyzed are from both compressional and extensional tectonic environments and range from basalt to rhyolite. These relations strongly suggest that, whatever He ultimate origin of the volcanism and its thermal requirements, the compositions of the rocks are dominated, at least for Pb, by contributions from the lith- osphere of the American plate. In He case of the middle Tertiary rocks of intermediate composition in both the Absaroka and San Juan volcanic fields, for which complex subduction models have been proposed (Lipman et al., 1971; 1972), any initial compositional signature from melting of the subducted slab or the immediately overly- ing asthenospheric mantle has been masked by interac- tions of the rising magmas with the American plate. For the San Juan field, at least, detailed consideration of the isotopic data suggests a major contribution from lower- crustal sources; only minor interactions with upper- cn~stal rocks are compatible with the Sr isotopic data. Isotopic compositions of upper Tertiary and Quatemary basaltic rocks in the Yellowstone Plateau and Rio Grande no indicate dominant sources of Pb from mantle regions that have been a part of the American plate since forma- tion of the associated parts of the continental craton, re- spectively, about 2.8 b.y. and 1.7-1.8 b.y. ago. If the vol- canism of the Yellowstone Plateau~nake River Plain trend represents the trace of a deep mantle plume, as Morgan (1972) and Suppe et al. (1975) conjectured, then isotopic compositional identity of the deep source has been lost. Variations in Pb isotopic compositions of Cenozoic ig- neous rocks across the western United States delimit an abrupt discontinuity between a West Coast region and the eastern Cordillera; this discontinuity is interpreted as masking the limit of Precambrian crust (Doe, 1967; Zart- man, 1974~. Analogous Sr isotopic studies also show strong regional gradients and boundaries that have been interpreted as reflecting major structural boundaries in the underlying continental crust (Kistler and Peterrnan, 1973; Armstrong et al., 1977~. In some regions, however, interpretation of the isotopic data is age-dependent; in

OCR for page 159
Cenozoic Volcanism in the Western United States southern Arizona and New Mexico, mafic Oligocene vol- canic rocks are relatively radiogenic (87Sr/86Sr - 0.706- 0.708), whereas Miocene and Pliocene basaltic volcanics are significantly less so (87Sr/86Sr = 0.703~.7041. This compositional difference may reflect a change in the com- position of the source region, perhaps related to upwell- ing of deeper mantle material in late Tertiary time, during major lithosphere extension and development of Basin- Range structure (Lipman and Mehnert, 1975~. Interpretation of this sort eventually should increase our understanding of the complex structure and composi- tion of the continental crust and upper mantle of the west- em United States, where, during Cenozoic time, a plate- convergent regime was superceded by an extensional one as a result of changing plate boundaries (Atwater, 1970~. In this region the compositional and structural zones, with which rising and evolving Cenozoic magmas could have interacted, include (1) sialic upper crust that ranges re- gionally in age from Phanerozoic to as much as 3 b.y. old and that is relatively radiogenic both in Pb and Sr; (2) mor mafic lower crust, probably of less radiogenic iso- topic character; (3) lithospheric upper mantle ofthe Amer- ican plate, having a depth of about 50-100 km between the Moho and the top of the low-velocity zone, that may have been little modified chemically since formation of the overlying craton; (4) asthenospheric mantle, below about 100 km, of relatively worldwide compositional homogeneity; and (5) at various times and places (as dis- cussed later), a subclucted plate descending slowly to the east through the asthenosphere to a depth of several hundred kilometers. This inferred subducted plate con- sisted of oceanic lithosphere, generated at the East Pacific Rise, and was subjected to little-understood processes and interactions with asthenosphere beneath the Ameri- can plate for 10-20 m.y. before thennally equilibrating and losing its geophysical identity. STRUCTURAL DISCONTINUITIES IN THE CONTINENTAL LITHOSPHERE Structural discontinuities in the lithosphere also seem im- portant in controlling the distribution and type of Ceno- zoic volcanism in the western United States. Especially conspicuous as controls for Cenozoic volcanism seem to be several no~east-trending zones of different ages and origins: the Snake River-Yellowstone zone, the Springer- ville-Raton zone, and the Colorado mineral belt (Figure 14.41. Other possible crustal flaws, not discussed here, may also be important in interpreting aspects of conti- nentaI tectonics. Proposed interpretation of the Snake River Plain- Yellowstone zone as the trace of a mantle plume or melt- ing anomaly, mentioned earlier, is based largely on the age progression of volcanism during the last 15 m.y., from the Idaho-Nevada border to the Yellowstone caldera (Armstrong et al., 1975) and on agreement between this vector and the inferred absolute motion of the American 165 ~ ,sO ,~_,-~ sit ~ .0 , . . . ,' .~1/ oU I ~ \ :' ,, ~ \ I'm! / ~ as. O / By; vale' _ of\/ \ ~ I / IF \ , ~ So of `q 'a Cot FIGURE 14.4 Major northeast-trending Cenozoic volcano- tectonic zones in the western United States (shown by solid lines where best constrained, dashed lines where less certain). "Rail- way track" indicates middle Cenozoic subduction boundary be- tween American and Farallon plates (base map adapted from Atwater, 1970). plate (Minster et al., 1974). However, geological and geo- physical studies in the Yellowstone~nake River region cast doubt on the mantle-plume model. The Pb isotopic evidence, already mentioned, apparently requires a man- tle source for the basaltic magmas that has been partofthe American plate for the past 2.8 b.y. Recent geological and geophysical studies in the region have documented other complications. In addition to the nor~east-trending age progression during the past 15 m.y., there is an opposing northwest-trending progressiontoward Newberry cal- dera in Oregon (MacLeod et al., 1975~. The Yellowstone- Snake River zone lies along a regional northeast-trending structural and aeromagnetic lineament that extends from northeastern Nevada, beyond the Yellowstone caldera, into Canada. According to Eaton et al. (1975), "if the Yel- lowstone magma body marks a contemporary deep mantle plume, this plume, in its motion relative to the American plate, would appear to be 'navigating' along a funda- mental structure in the relatively shallow and brittle lith- osphere overhead. The concept that a northeastward- propagating major crustal fracture controls the migration path of the major foci of volcanism is at least equally favored by existing data." Another major northeast-trending igneous zone is the long-recognized Colorado mineral belt, which consists of aligned Upper Cretaceous~ower Tertiary (Laramide) in- trusives and locally preserved extTusives. No age progres- sion is evident along the Colorado mineral belt, but, as first pointed out by Tweto and Sims (1963), this belt fol- lows structural trends of Precambrian ancestry. The Colo- rado mineral belt has been proposed to constitute only part of a very long northeast-trending structural zone that extends from Arizona through Colorado, across the North- er;1 Plains, to join a major boundary between Precambrian age provinces in the Great Lakes region (Warner, 1978~. Farther to the south, another northeast-trending late )

OCR for page 159
166 FIGURE 14.5 Space time relations of Cenozoic volcanism along the Spnngervill - Raton zone. A, Distribu- tion of Pliocene and Pleistocene vol- canic field along the Spungervill~ Raton zone (stippled). B. Plot of age of volcanic activity versus position along the Spnngervill - Raton zone. Sloping line indicates a migration rate of 2.5 ctn/year, the observed migration rate for inception of silicic activity along the Snake River Plain-Yellowstone zone; no such age - Stance trend is evident for Me Spnngerville-Raton zone. Cenozoic volcanic feature (Figure 14.SA), referred to venously as the Spnngerville~aton zone, the Jemez zone, or the Raton volcanic chain, has also been proposed as marking the trace of a mantle plume (Suppe et al., 1975~. This major locus of late Cenozoic activity includes the Pinacate basalt field in northwest Mexico, the Spnngerville-White Mountain volcanic zone in eastern Arizona, the Mount Taylor and lemez volcanic fields in PETER W. LIPMAN 37o A 35r 1 1 4 1 09 or / ll F _r _ . 2 / ~ / 'V/ COLORADO . PLATEAU l 31 ! ~ ~ ~ ~. . .-. . . . . BASIN AND I RANGE C Capulin area CM Cimmaron Mountains G Gila Bend ~ Jemez Mountains O Ortiz Mountains P Pinacate volcanic field Q Ouesta area R Rabbit Ears area SF San Francisco volcanic field S S~ringerville area 0 O _i l MD Mogollon-Datit volcanic field MT Mount Taylor T Taos Plaueau W White Mountains Z Zuni Mountains l J ~ HIGH PLAINS j _1 ~ 11 \~ MEXICO ~ too 200 300 KILOMETERS 1 1 1 1 EXPLANATION 1 ~ 1 ! 1 ~ 1 c,0 30 c U" 0-25 z o ~ 20 z z '0 1 5 t _ ,,1 ol :~4 1 000 800 600 400 200 0 NE \ 800 600 400 SW DISTANCE ALONG VOLCANIC TREND. IN KILOMETERS B I- New Mexico, volcanics in the Rio Grande rip on the Taos Plateau and at the Cerros del Rio, and basalts of the High Plains of nor~eastem New Mexico. In contrast to the Yellowstone zone, no age progression is evident, either for Me upper Cenozoic basaltic rocks or for more silicic earlier Cenozoic activity along this zone (Figure 14.5B); and a mantle-plume origin seems less probable than con- trol by a crustal flaw (Mayo, 1958; Laughlin, 1976~. The

OCR for page 159
Cenozoic Volcanism in the Western United States Springerville-Raton zone follows general Precambrian structural trends of the region, such as the northeast- trending basement trends Great length on the Colorado Plateau (Shoemaker et al., 1974), and coincides in a gen- eral way with a major boundary between Precambrian age provinces. Analysis of patterns of Cenozoic volcanic activity sug- gests that these northeast-trending zones are key features in interpreting Cenozoic tectonics of the western United States- features identifiable only by the presence of vol- canic activity along them. Another aspect of tectonic control of volcanism that has implications for structure of the lithosphere is the clear pattern of recurrence of igneous (and tectonic) activity in the same places in the western United States through time. Such recurrence of activity was documented by Snyder et al. (1976), who delimited five major "magmatic loci" in the western United States. The distinction among several of these loci is not entirely convincing, for exam- ple, Weir Nevada locus is separated from the Idaho- Montana locus to the north only by younger rift-related volcanic cover along the Snake River Plain (Snyder et al., 1976~. In contrast, the Colorado Plateau was a stable "miniplate" during late Cretaceous-early Tertiary Lara- mide deformation and igneous activity, during the middle Cenozoic predominantly andesitic volcanism, and during the late Cenozoic fundamentally basaltic volcanism and associated extensional block faulting. The recurrence and general confinement of Cenozoic volcanism to regions where previous episodes of magnmatism and tectonism may have healed, annealed, and weakened the litho- sphere seems an important consideration in interpreting the distribution of Cenozoic volcanic activity in the west- em United States. PLATE-TECTONIC INTERACTIONS Continental volcanism also provides evidence for the na- ture of plate interactions in the western United States during the past 100 m.y. For relatively young interactions the seafloor-spreading record is sufficiently complete to permit fairly reliable reconstructions, but for the western Cordillera of North America the plate-tectonic history is complex, the seafloor record is largely missing, and many aspects can be reconstructed only from igneous and tec- tonic activity recorded on the continental plate. Interpre- tation of this record is still very incomplete and uncertain. A contrast exists in the western United Sates between earlier Cenozoic volcanic assemblages, which are pre- dominantly andesitic in composition and are inferred to be related to convergence and subduction along the west- em edge of the American plate, and a younger assemblage of fundamentally basaltic volcanism, in places including bimodal basalt-rhyol~te associations, that appears related to extensional tectonics within the American plate (Lip- man et al., 1971, 1972; Christiansen and Lipman, 1972; Snyder et al., 1976~. The change Tom predominantly an- desitic to fundamentally basaltic volcanism has been cor- 167 related in a general way with changing boundaries be- tween the American and various Pacific plates (Atwater, 1970~. Changes in volcanism accompanying the initial in- tersection of the Pacific and American plates are thought to be especially significant; these changes terminated the middle Tertiary and earlier subduction system and ini- tiated the evolving transfo~ boundary ofthe San Andreas Fault system, bounded at both ends by migrating triple junctions. Since first application of this plate model to changing patterns of volcanism in the western United States (Lip- man et al., 1971, 1972; Christiansen and Lipman, 1972), new data have accumulated, and furler evaluation is ap- propriate. Refinements of Cenozoic volcano tectonic pat- ten~s in terms of generally similar plate models for the western United States include those of Snyder et al. (1976), Coney and Reynolds (1977), Cross and Pilger (1978), Dickinson and Snyder (1979~. The re-examination presented here focuses on the predominantly andesitic volcanic suite, inferred to reflect subduction-related re- gimes. Aspects of the succeeding fundamentally basaltic volcanism and associated extensional tectonics are dis- cussed elsewhere (see Chapter 9~. Important variables that could affect distribution and compositional variation in subduction-related andesitic suites include dip of the Benioff zone, rate of plate con- vergence, angle between plate boundaries and conver- gence direction, temperature and thickness of descending slab, warps or breaks in the descending slab, depth of magma generation, composition of andesitic magma, dis- tance from the trench to volcanic front, and width of the volcanic belt. Compositions of convergent volcanic suites correlate with depth to the Benioff zone, becoming more potassic and alkalic with increasing depth, and vary also from intracontinental to intraoceanic settings (Dickinson, 1975~. Volcanoes occur at present above active Benioff zones over a depth range of about 75 300 km. Relatively rapid convergence favors more gently dipping Benioff zones (Luyendyk, 1970) and delayed heating of the de- scending slab; accordingly, such environments favor in- creased distances between trench and volcanic front, broader volcanic belts, and more alkalic volcanism. Oblique convergence, transitional toward transform mo- tion, would reduce the effective convergence rate. In ac- tive convergent systems, offsets of the volcanic front cor- relate with segmentation of the Benioff zone, requiring transverse breaks or flexures in the descending slab (Carr et al., 1973; Stoiber and Carr, 1974~. The behavior of the descending slab may also reflect the age of the oceanic lithosphere, which becomes cooler, thicker, and more dense with age and distance from the spreading ridge; older lithosphere should therefore sink more rapidly and preserve its physical identity to greater depths in a sub- duction environment than thin, hot, buoyant, young lith- osphere (Molnar and Atwater, 1978~. Complex, and as yet imperfectly understood, interplay of these variables seems capable of accounting for much of the diversity of arc volcanism. Some possible effects of

OCR for page 159
168 change in rate of plate convergence are illustrated dia- grammatically in Figure 14.6. In a subcontinental en- vironment, the structural and compositional complexities of Me continental lithosphere, discussed earlier, also seem capable of fiercer complicating the volcanic pattern, as illustrated by relations from the western United States. The following discussion is based on a series of figures Mat show in generalized fashion Me distribution of ig- neous activity Trough Cenozoic time in 10-m.y. incre- ments (Figure 14.7~. Through most of Mesozoic time and continuing to about 80 m.y. ago, igneous activity was confined to near the western margin of Me American plate (Figure 14.7A). Gra- nitic and cogenetic volcanic rocks as young as about 8() m.y. occur in Me Sier a Nevada batholith, but Mesozoic igneous rocks are sparse east of western Nevada. The bend in the igneous trend in the Pacific Northwest is probably due largely to Cenozoic deformation (Hamilton and Myers, 1966; Simpson and Cox, 1977~. Local east- A TYPICAL CONVERGENCE B INCREASED CONVERGENCE 1 ;~ c DECREASED CONVERGENCE - W~ i so FIGURE 14.6 Possible effects of change in convergence rate on condnen~l-margin subduction and related volcanism. A, Typical convergence rate: subduction zone dips about 30 degrees, and related volcanoes are in a relatively narrow arc, are low in potas- sium content, and are located relatively close to trench. B. In- creased convergence rate: continental plate ovemdes descend- ing slab at a lower angle, resulting~in a broader, more diffuse volcanic arc located farther from the Bench. Depending on depth to subduction zone, volcanic rocks can vary widely in composi- tion from low to high potassium. If the dip of the subduction system becomes very shallow, volcanism may cease entirely, as has probably happened in modern southern Peru (Barazangi and Isacks, 1976). C, Decreased rate of convergence: downgoing slab sinks gravitationally or breaks and reestablishes itself in a steeper orientation, causing arc volcanism to migrate toward the trench. Increased sinking of the descending slab requires coun- tedlow of asthenosphenc mantle into the region above the slab, possibly heating the base of the continental lithosphere and driv- ~ng extensional rifting and basaltic volcanism within the region of earlier arc volcanism. PETER W. LIPMAN west variations in distribution and composition of granitic rocks in the central Sierra Nevada are compatible with shifting depths of magma generation along a subduction system of relatively constant geometry (Dickinson, 19701. More limited, similar data along over transects across the batholithic belt suggest possibly more complex shifts in geometry of the subduction system, especially in the Klamath Mountains, where low-potassium quartz diorites and trondhjemites tend to be concentrated on the eastern side of the plutonic belt (Hotz, 1971), and also in the Peninsular Ranges batholith to the south (Silver et al., 1975~. Nevertheless, through late Mesozoic time, a long- lived, relatively steeply dipping, subduction system ap- parently was maintained, along the western margin of the American plate, concurrently with consumption of an eastern Pacific plate. Starting at about 80 m.y. ago, concurrently with inferred accelerated convergence between the American and east- em Pacific plates (Coney, 1972), igneous activity mi- grated eastward in the western United States (Figure 14.7B), as recognized long ago by Lindgren (1915~. The most dramatic eastward shift between 80 m.y. and 70 m.y. ago- (Figure 14.7B) occurred in the northwest, where plu- tonic rocks of the Boulder batholith and associated Elk- hom Mountains volcanics were emplaced as far east as western Montana; however, igneous activity of this age also moved eastward in Nevada and Arizona. Pattems of igneous activity continued to evolve rapidly and had changed notably by 70 m.y. to 60 m.y. ago (Figure 14.7C). The locus of intense activity associated with the Boulder batholith in western Montana extinguished abruptly about 70 m.y. ago. Many K-Ar ages in the Idaho batholith could indicate continuation of igneous activity; more likely, they reflect reheating by pervasive Eocene activity (Armstrong et al., 1977~. To the south, major activ- ity of Laramide age continued in southem Arizona and began to extend into southwestern New Mexico, consti- tuting the peak connation of porphyry copper deposits in this region. A southeasterly age progression of these eco- nomically important intrusives, suggested as a mantle- plume trace (Livingston, 1973), seems more likely part of the regional eastward migration of igneous activity when considered with distributions of other intrusions of roughly similar age in the region. Between northern Idaho and southem Arizona, He only other significant locus of igneous activity at this time was the northeast- trending zone of intrusions and associated volcanics along the Colorado mineral belt. These represent the most east- em sweep of early Tertiary igneous activity in the central and southern Rocky Mountain region. Thus, early Tertiary igneous activity in the we stem United States covered a broad region but was defuse, discontinuous, and nonsynchronous in time, reaching the eastern margins of the Cordillera earlier in the northern Rockies than farmer south. Even with abundant new data, these patterns reasonably seem related to effects of flat- tening of He subduction system during the early Tertiary (Lipman et al., 1971; Coney and Reynolds, 19771. Flatten-

OCR for page 159
Cenozoic Volcanism in the Western United States ,~ ~ ..2'.,'.~2 '. ,'~'..2 . ,2 ,~ ' ':: .' .' " ." ~: '''' . 2. i:: .:...2 .., W: ~~: /'' }/ / \// ~ FIGURE 14.7 Generalized distribution in the western United States of predominantly andesitic volcanic suites, inferred to be related to subduction. Distributions are based on compilations (Lipman et al.' 1972; Snyder et al., 1976; Stewart and Carlson, 1976; Armstrong et al., 1977; Cross and Pilger, 1978) and on descriptions of local areas too numerous to cite individually. The base maps and diagrammatic plate geometry are from Atwater (1970) and Atwater and Molnar (1973). No attempt has been made to remove effects of late Cenozoic extensional and rotational deformation, even though such effects are probably large (Hamilton and Myers, 1966). Northeast-t~ending lines mark approximate traces of the Snake River-Yellowstone zone, the Colorado mineral belt, and the Springenrill~Raton zone. 169

OCR for page 159
An Outline of the Tectonic Characteristics of China paraplatform, such as the Alashan region, were consoli- dated by the end of the Proterozoic through the Yangtze cycle. Geologists now agree that the main portion of the paraplatfortn developed through three stages, as revealed by three major unconformities (Ch'eng et al., 1973~. These are (1) the unconformity between the Archean Fuping Group and the lower Proterozoic Wutai Group, represent- irlg the Fuping orogeny about 235(~2550 m.y. ago*; (2) the unconformity between the Wutai Group and the upper part of the lower Proterozoic, or the Huto Group, representing the Wuta~ orogeny about 2000 m.y. ago; and (3) the unconformity between the Huto Group or its equivalents and the upper Proterozoic Sinian Suberathem,! representing the Chungtiao orogeny. The blanket of the platfonn consists of Sinian and Cambro- Ordivician neritic sediments (largely carbonates), Perrno- Carboniferous terrestrial sediments (with marine inter- calations), as well as purely terrestrial Meso-Cenozoic sediments. The Sinian System is present in a few regions, while Silurian, Devonian, and lower Carboniferous are absent. An abundance of Mesozoic continental volcanics and granitoid intrusions occurs in the eastern part of the paraplatfo~l~, especially in the Yenliao depression, Shangtung, and other regions. Cenozoic basalts are widespread. Tectonic disturbances of the Indosinian cY- cle were limited in Inner Mongolia, Liaotung, and the Yenliao depression. YANGTZE PARAPLATFORM The Yangtze paraplatform (Figure 16.~) comprises the greater part of the Yangtze Basin from eastern Yunnan to Kiangsu, also including the southern part of the Yellow Sea. Previously, the consolidation ofthe Yangtze paraplat- forrn was considered contemporaneous with that of the Sino-Korean paraplatfonn. Recent investigations using stromatolites, micropaleoflora, and isotopic geochronol- ogy of the basement rocks indicate that it was consoli- dated about 700 m.y. ago (Yangtze orogeny). The blanket of the paraplatforrn consists chiefly of carbonate and clas- tic deposits ranging in age Tom Sinian System to Triassic, with Devonian and Carboniferous normally absent in Szechwan and North Kweichow, whereas terrestrial Jurassic, Cretaceous, and younger deposits occur in Szechwan, Central Yunnan, Hupeh, Kiangsu, and over regions. Tectonism was strong chiefly in the Yenshanian cycle with the formation of well-developed blanket folds. The Kam-Yunnan Axis and the Lower Yangtze, however, suffered polycyclic deformation and magmatism; the *The oldest basement rocks are dated at 320(~3400 m.y. ago. IThe author uses the classification of the Sinian as given in the new Geological Map of China (1:4,()00,000). The tenn "Sinian System," with its standard section in the Yangtze Gorges, com- prises sediments ranging in age from 600 800 m.y. ago. The tend "Sinian Suberathem," with its standard section in ChihEsien near Peking, comprises sediments ranging in age from 600-1700 m.y. ago. The term sinian in this paper corresponds to Sinian Suberathem. ~7 former belongs essentially to the Variscan cycle, while the latter belongs to the Yenshanian and Indosinian cycle. Formerly, the Huaiyang .\Iassif was considered a portion of the Sino-Korean paraplatform; recent data show that it belongs to the Yang~e paraplatforrn. TARIM PLATFORM The Tarim platform (Figure 16.1) is defined by the Tienshan fold system in the north and by the Kunlun fold system in the south and is largely covered by Cenozoic deposits. Its basement, together with its Paleozoic blanket, crops out along the northern border as seen at Kelpin and Kuruk Tagh. In the laker region, late Protero- zoic (S~nian) tillites and stromatolites are found, showing that the Tarim platform took shape near the end of the Proterozoic just as the Yangtze paraplatforrn did. TIENSHANEHINGAN GEOSYNCLINAL FOLD SYSTEM AND ARGON FOLD SYSTEM These two systems are components of the great central Asiatic-Mongolian arcuate fold region, which extends be- tween the Sibenan platform, on the north, and the Sino- Korean paraplatfonn and the Tarim platform, on the south (Figure 16.1~. This fold region is separated into two halves by the Derbugan depth Eacture (eastern extension of the Mid-Mongolian depth fracture); the southem half is the Tienshan-Khingan geosynclinal fold system, while the northern haIfincludes the Argun fold system. The laker is Hsingkaiian, and the fondler consists, within the Chinese territory, of the Altai (Caledonides and Variscides), the Dzungarian (Vanscides), the Tienshan (Variscides), and the Kirin-Heilungkiang (Variscides) fold systems. All of them, except the southern Tienshan, are eugeosynclinal in nature. During the Caledonian cycle, volcanic activity, chiefly submarine, was widespread in the Ordovician and Silurian. Radiolarian cherts were found closely associated with spilites and ultrabasic intrusives forming ophiolitic suites, such as in the western Dzungaria. Dunng the Variscan cycle, the Devonian and Carboniferous were characterized by calc-alkaline volcanics (mainly an- desites). From Carboniferous to the end of the Permian,* the geosynclinal coronations accompanied by granitoid in- trusives were brought into intense and complicated folds, thus converting the geosyncline into a craton. During the Yenshanian cycle, a general tendency of increasing tec- tonic intensity from east to west can be recognized, es- pecially in the Kirin-Heilungkiang and Great Khingan regions, where strong remobilization and magmatism took place. During the Himalayan cycle, the general tendency of increasing intensity of tectonism was inverted. Fault- ing and uplift along the general strike of the Tienshan were strong and widespread, foxing lofty mountain ranges with sunken northern and southern foredeeps as *This is the Variscan orogeny. Caledonian erogenic: movements are also present but they are very limited in distribution.

OCR for page 159
188 well as intermontane depressions such as the famous Tur- fan Basin. KUNLUNNANSHANTSINLING GEOSYNCLINAL FOLD SYSTEM AND TIBETYUNNAN GEOSYNCLINAL SYSTEM These two tectonic units, separated from one another by the Chinshak-~ang-Red River depth fracture, occupy the vast territory south of the Tarim platform and Sino-Korean paraplatfonn, west of the Yangtze paraplatform, and north of the Tsangpo depth fracture (Figure 16.1~. The Kunlun-Nanshan-Tsinling geosynclinal fold system is a typical polycyclic geosynclinal fold system embracing three erogenic events. During the Caledonian cycle, a eugeosyncline came into existence mainly in the Nanshan, with well-developed ophiolite suites and glaucophane schist belts. During the Variscan cycle, a large portion of the system was miogeosynclinal, while eugeosyncIines were maintained only in Burkhanbuddha, Amnemachin, and along the Chinshakiang. During the Indosinian cycle, miogeosynclinal development con- tinued in ~e Sungpan-Kantze, Tsinling, and South Ko- konor Range, while the area along the Chinshakiang re- mained eugeosynclinal. Different parts of the system differ in age of folding. The Nanshan is Caledonian, the Kunlun is Variscan, while the Sungpan~antze anc] Tsin- ling are basically Indosinian. The Yenshanian and Hima- layan cycles are characterized by remobilization ac- companied by a variable amount of magrnatism. That Cambrian volcanics were over~rust upon late Tertiary red bed~s in the Lachishan of Chinghai and Jurassic sandstones were ove~rust upon Quatemary gravels at the Hungliu Gorge near Yumen of Kansu demonstrate that intensive horizontal compression has been a factor even in more recent geoIogical time. The Tibet-Yunnan geosynclinal fold system, situated to the south and west of the Chinshakiang-Red Biver depth fscture, is a Mesozoic arc uate fold system, consisting of ~e Sankiang* or Three-River (Indosinides), the Tangla (Yenshanides), and the Lhasa (Yenshanides) fold systems. The existence of the proposed Precambnan Tibetan Massif (Terman, 1973) is negated by recent observations. The Sankiang fold system includes western Yunnan and the Changtu district of Tibet and is characterized bY sreo- synclinal folds strictly controlled by depth Eactures among which the Chinshakiang depth fracture, the Lant- sangkiang depth fracture, and the Nukiang depth fracture are of prime importance. Along these depth fractures, eugeosynclinal activity took place, with polycyclic tec- tonism accompanied by intensive polycyclic magmatism. Not only the Paleozoic and Mesozoic, but also, in some localities, the Tertiary formations underwent tectonism and metamorphism in various grades and formed several tectono-magmato-metamorphic belts (the Ailaoshan, the *So named because it embraces the Chinshakiang, the Lantsang- lciang, and the Nukiang drainage systems. T. E. HUANG Lantsangkiang, and the Kaol~kungshan metamorphic belts). HIMALAYAN GEOSYNCLINAL FOLD SYSTEM This tectonic unit (Figure 16.1), situated to the south of the Tsanggo and Indus Rivers, consists of Sulaiman (in Pakistan), the Himalaya, and the Arakan Yoma (in Burma). It must be pointed out that true geosynclinal deposits are absent in the Himalaya proper but appear immediately to the south of the Tsangpo, where a more or less continuous belt of ophiolites is well exposed. The lofty ranges of the main Himalaya form in fact the northern border of the Indian craton and associated Hsingkaiides, the metamor- phic and sedimentary rocks of which were deeply in- volved in the H~malayan folding when the Indian craton collided with Tibet (Chang and Cheng, 1973; Gansser, 1964, 1966; Molnar and Tapponnier, 1975~. SOUTH CHINA GEOSYNCLINAL FOLD SYSTEM This tectonic unit is located to the south of the Yangtze paraplatfonn (Figure 16.1~. The author believes that this unit is not entirely Caledonian as had been prev~ously regar~led but consists of two parts of different character. The major part, still called the South China fold belt, is truly Caledonian, but the subordinate part, including coastal regions of Chekiang, F'ukien, and Kwangtung and the neighboring continental shelves as well as a greater part of the Hainan Island, belongs to the Vanscan, to be narned the Southeastern Maritime fold system. Because the latter was strongly transfonned by Mesozoic tec- tonism and widely covered with Jurassic and Cretaceous continental volcanics, its tectonic character has long been misjudged. NATANHATA EUGEOSYNCLINAL FOLD BELT AND UPPER HEILUNGliCIANG MIOGEOSYNCLINAL FOLD BELT These are components of the Mesozoic geosynclinal fold systems located to the east of the Siberian platform (Figure 16.1). The Natanhata belongs to the Sikhote-Alin fold system, while the Upper Heilungkiang belongs to the Mongolo~Okhotsk fold system. Both are Yenshanides. DEEP-SEATED STRUCTURES AND DEPTH FRACTURES IN CHINA DEEP- SEATED STRUCTURE From a comprehensive examination of available geo- physical and seismic data, it is preliminarily concluded that the lithosphere within China can be subdivided into several heterogeneous layers, as elsewhere in the world. This layering character is demonstrated by recent investi- gations (Hsi et al., 1974a; 1974b; 1975) of seismic sound-

OCR for page 159
An Outline of the Tectonic Characteristics of China FIGURE 16.2 Cross section of the Moho from Himalaya to Nanshan. 1, Tsangpo depth fracture; 2, Lantsang- kiang depth fracture; 3, Chinshakiang depth fracture; 4, East Kunlun depth fracture; 5, depth fracture along north- e~n margin of Tsaidam; 6, Altyn depth fracture; (by, Himalaya fold system; @), Lhasa fold system; 0), Sankiang fold system; (9, Tangla fold system; (I), Sungpan-Kantze fold system; 6), Kunlun fold system; by, Nanshan fold system. FIGURE 16.3 Cross section of He Moho from Lanchow to Talien. 7, depth fracture zone of North Nanshan; 8, depth fracture of western margin of Ordos; 9, depth fracture zone of Taihangshan; 10, Tancheng-Lukiang depth fracture zone; 6), Nanshan fold system; 0), Sino-Korean paraplatfo~. ,,,, ~~ - .so. ~ st. . ~ imalaya Tangla Kuniun Km' '~Lhasa 2 3 4 Gormu !_ _ ~ _ _ 30 - 1 50 - ~1 God / 8oi Lanchow Taiwan Km, .x ,; t: ~ 'A _ it) ~ I ing along a profile from Yuanshih to Tsinan in the Norm China Plain and another profile across the eastern section of the Tsaidarn Basin. In these regions the earth's crust possesses layers of velocity gradients in addition to low- velocity zones. A blocklike pattern is developed for the Chinese main- land, especially its eastern part, when the areal variations of the depth to the Moho (Mohorovicic discontinuity) is considered. Two.cross sections show the nature of the depth to the Moho (Figures 16.2 and 16.3~. Each block is confined generally by depth fractures of persistent devel- opment. For example, the Yinchuan-Liupanshan-Karh- Yunnan gravity gradient belt (Kunming-Yinchuan depth fracture system) separated eastern China from western China in the geological past. The Great Khingan-Tai- hangshan-Wulingshan* gravity gradient belt, a dividing line between the eastern and western belts of the Marginal-Pacif~c tectonic domain, came into existence long since the late Mesozoic. The northern section of the above-mentioned gravity gradient belt, i.e., the section from the Great Khingan to Taihangshan, coincides with a depth Eacture zone. The North Tsinling-North Huaiyang depth fracture zone is a geological dividing line between North and South China. The gravity gradient belt of West *No NO directed depth fractures are present along the Wulingshan. 189 N anshan Aksai s 6i 44. ~ 1~ ~ : ~ i ~ , ~ 3 ~> , (I) i ~ 1 ,. O It)o Con .,of coo s`3oA'o1 .9 Cohn _~ Talien ~ lij Kunlun-Altyn-North Nanshan essentially coincides with the depth fracture zones between the Tarim platforms and the Alashan Massif on the north and the geosynclinal fold belts on the south. Moreover, the present geomorphic features of China form the mirror image of the Moho, as shown in Figures 16.2, 16.3, and 16.4; mountains and plateaus correspond to the "downwards" of the Moho, while plains and basins agree with its "uplifts." This appears to be a result of a combined effect of the Pacific plate and the Indian plate acting against the continental block of China since the Indosinian, particularly since the Himalayan cycle. In these movements, not only the crust but also the upper mantle were involved. DEPTH FRACTURES OF CHINA The author classifies depth fractures into three classes according to their depth (Table 16.2~. Translithospheric fractures either came into being during a definite geo- logical period or continued in motion till today marked by a series of deep-focus earthquakes (Benioff zones). Lithospheric fractures are generally characterized by basic and ultrabasic complexes but without the develop ment of extensive ophiolites. Crustal fractures are of much

OCR for page 159
190 ~ ~ En -~ 'my ~ ~~- W-o ~ Ale-'- - - N :-? I: brow ~ ~. - all it  ~ --a At e, If ~ ] q_f en. '--] To ~~ yak li tare P388~ ee ~ - - o ~ - c: ~ - o c) c) en Hi - et At; in 3 o _ V, et ._ ._ U. CO ~4 C) _

OCR for page 159
An Outline of the Tectonic Characteristics of China TABLE 16.2 Classification of Depth Fractures According to Depth 191 Examples Class Depth World China Sialic crusted fracture Simatic depth fracture Lithosphenc fracture Translithospheric fracture Dissecting the sial but not clearly extending into the sima Dissecting the whole crust but not extending into the mantle Dissecting the lithosphere but not extending into the asthenosphere Dissecting the lithosphere and extending deep into the asthenosphere Many Many RiR Valleys of East Africa Benioff zones along island arcs of He Pacific Heyuan depth fracture East Tsangchow depth fracture Tancheng-Lukiang fracture Tsangpo depth fracture smaller magnitude and are usually accompanied by acid and/or intermediate magmas (sialic fractures) or by basalts (simat~c fractures). Depth fractures can also be divided into three classes* according to their mechanical behavior: tension~epth fracture, compression-depth fracture, and shear-depth fracture. Fractures along midoceanic ridges and along the great rift valleys are tension-depth fractures. Fractures along the island arcs of the western Pacific (marked by deep-sea trenches) are examples of compression- or com- pression-shear-depth fractures. The San Andreas Fault of North America, the Altyn Depth Fracture, and the Tancheng-Lukiang Depth Fracture of China are ex- amples of shear-depth fractures. Depth fractures are not unchangeable; on the contrary, their formation, develop- ment, transformation, and extinction came one after another in succession during the long history of the evolu- tion of the earn. Numerous facts note that different sec- tions of the same depth-firacture zone might have ongi- nated in different geological ages, belong to different types, and possess different charactenstics. Based on the available geological and geophysical data, the main depth fares are listed in Table 16.3, and a brief description of them follows. Those belonging to the translithospherrc depth frac- tures are as follows: 1. The Tsangpo depth-~acture zone (Huang, 1960) ex- tends westward following the Indus River valley and was termed the Indus suture (Fitch, 1972~. It is now consid- ered as the suture line between the Indian plate and the Eurasian plate and is marked by well-developed ophio- lites extending about 2000 km within the confines of China (Chang and Cheng, 1973; Gansser, 1964~. 2. The longitudinal depth-fracture zone of Taiwan forms only a small section of the long depth fractures (BenioE zones) along the island arcs of the western Pacific. Its presence is marked by ophiolites, glaucophane schists and melanges, and frequent deep-focus earth- *ln fact, there are transitional types, such as compression-shear and tension-shear depth fractures, which are attributable respec- tively to compression, tension, and shear depth fractures accord- ing to their dominant mechanical behavior. quakes. This depth fracture is sinistral shear in mechani- cal behavior in contrast to the predominantly compres- sional Benioff zones. Available data indicate that the Norm Nanshan depth fracture, the Chinshakiang-Red River depth fracture, and the Derbugan (Mid-Mongolian) depth fracture are trans- lithospheric, or ancient plate sutures. In China a great majority of depth fractures in geo- synclinal regions and a pa* of those in platforms belong to lithosphenc depth fractures. Among them the domi- nant ones are the following: (a) The Tancheng-Lukiang depth fracture system is perhaps the major depth-fracture system of eastern Asia and is composed of the Tancheng-Lukiang, the Mishan- Tunhua, and the Yilan-Yitung depth fractures with a total length of about 2400 km. From south to north, it cuts through the Yangtze paraplatforrn, the Sino-Korean para- platforrn, and the Kirin-Heilungkiang fold system. It also appears to be an important volcanic, metallogenic, and seismic zone, especially during Meso-Cenozoic times. Sinistral shear occurred in the geological past, but dextral shear is apparent from analyses of modem earthquake mechanisms. Some geologists consider that this fracture system existed in Pre-Sinian times, while others suggest that it was formed in the Indosinian cycle. The fracture system continues northward into the Far East of the U.S.S.R. (b) The Kunlun-Tsinling depth fracture system includes the North Nanshan-Nor~ Tsinling-North Huai- yang fracture zone, the northern margin of Tsaidam- South Kokonor Range-North Tsinling-North Huaiyang fracture zone, and the east Kunlun-Tsinling fracture zone. These complicated fracture zones controlled the origi- nation and development of the Kunlun-Tsinling geo- synclinal fold system and constitute a geological dividing line between northern and southern China. This divid- ing line shifted from the line of North Nanshan-North Tsinling-North Huaiyang during the Caledonian cycle to the line of the northern margin of Tsaid~n~outh Koko- nor Range-North Tsinling-North Huaiyang during the early Variscan cycle, and finally to the line of East Kunlun-North Tsinling-North Huaiyang since the late Variscan cycle. It is important to note that these fracture zones deepen from east to west. Widespread submarine

OCR for page 159
192 TABLE 16.3 The Pnncipal Depth Fractures in Chinaa T. lc. HUANG Number Oepth Fracture Zone Depth Character Age of Activity Magmatism and Metamorphism 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 23 24 25 26 27 28 29 30 31 32 Darbut Irtish Karameili North Margin of Central Tienshan Derbugan Silamulun Cherchen No~ern Margin of Inner Mongolian Axis Altyn North Nanshem-Northem Margin of North Tsingling~orth Huaiyang Northen~ Margin of Tsaidam~outh- em Margin of No~ Tsingling- North Huaiyang East Kunlun Kantze-Litang Chinshakiang-Red River Lantsangkiang Nukiang Anningho Hsiaokiang Lungmenshan Lingshan Wuchuan-Szehwei Heynan Lishui-Haifeng Changle-Amoy Taihangshan East Tsangchow Liaocheng-Lankao Tancheng-Lukiang Yilan-Yitung Fushun-Mishan Tsang+Indus Longitudinal Valley of Taiwan aNumbers correspond to those circled on Figure 16.4 T. Translithosphenc fracture zone L, Lithosphenc fracture zone C, Crustal fracture zone c-s, Compression and compression-shear t-s, Tension and tension-shear s, Shear s-s, Sinistra1 shear ~ . L L L L L L L L L L L L L L L C L C C C C C C C C L C L T T s c-s c-s c-s c-s c-s s-s c-s s-s c-s c-s c-s c-s c-s c-s c-s c-s c-s c-s c-s a a c-s c-s c-s t-s t-s s-s t-s s-s c-s s-s Pz Pz Pz Pt?,Pz Pz Pz Pz Pt?,Pz Pz Pt,Pz Pz Pz Pz Pz Pz Pz Pz Pz Pz Pz Pz ~z Pz ~z Mz Mz Mz Pt?,Mz? ~z Pz Mz Mz OS I,gs OS,gs 5,m 5,m I,m y y y y OS,m OS,m,gs Pt, Active since Proterozoic Pz, Active since Paleozoic Mz3 Active since Mesozoic OS' Oph~olite suites I, Basic and ultrabasic complexes y, Gran~toid plutons ,B, Basalts a, Compression~hear and tension~hear, aItemating m, ~Ielanges as, Glaucophane schists volcanics occur along the westem section, where eugeo- synclinal conditions prevailed, and especially in No~ Nanshan, where remarkable zones of ophiolites 700 km long and zones of gIaucophane schists more than 100 km long were recently discovered, indicating the existence of an ancient plate suture. On the contrary, the eastern sections of these hactures, being characterized by flysch formations and scanty ultrabasic rocks, appear miogeosyn- clinal. This is perhaps the reason why the Tsinling geo- syncline became appendixlike toward the east between the Sino-Korean and the Yangtze paraplatfortas. Crustal fractures are numerous in China as repre- sented by large-scale acid-intermediate plutons with ex- tensive volcanics of the "Pacif~c" type, lack of ultrabasic rocks, and zones of strong compression followed by strong tension along the same zones, thus giving rise to a series of Cretaceous-Tertiary basins Blled w~th red beds along

OCR for page 159
An Outline of the Tectonic Characteristics of China growth faults* of depth-fracture nature. ARerwards, in late Tertiary and Quaternary, they were subjected to compres- sion again. In this respect, the Heynan depth fracture is one of the fairly good examples. Another type of crusty fracture in eastem China is represented by numerous tension fractures along which originated and developed Mesozoic and/or Tertiary fault basins, filled with terrestrial elastic deposits. Among them the East Tsangchow depth fracture, the Liaocheng- Lankao depth fracture, and the Weiho-Fenho graben (young Tertiary) are the most important. Unlike the Heyuan depth fracture, they originated as tension frac- tures in Cretaceous or old Tertiary time and are buried by young Tertiary and Quatemary deposits. Thus' they are typical growth faults, forming with elastic deposits, the so-called "dustpan-like basins" accompanied by basalt flows. Geophysical data show that these depth fractures dissect the entire earth's crust but do not extend deep into the upper mantle and therefore should belong to simatic fractures. THE TECTONIC DEVELOPMENT OF CHINA Two major stages oftectonic development of China before Paleozoic times can be distinguished. The first is the stage from Archean to early Proterozoic or the Pre-Sinian stage, through which the Sino-Korean paraplatform came into being about 1700 m.y. ago. The second is the late Proterozoic or Sinian stage, through which the Yangtze paraplatform and the Tarim platform came into being about 700~00 m.y. ago. The great importance of the Yangtze orogenic cycle should be emphasized here. Available data show that platform` areas formed through the Yangcze orogeny were more extensive than the pres- ent Tangle paraplatfiorm and Tarim platform. Platform- type fo~l~ations correlatable with the Sinian System lying unconformably upon metamorphosed Sinian Subera- them, are met with In East Kunlun, Tsinling, Altyn, aIld along the northern border of the Tsaidam Basin. More- over, phosphatic sediments of early to middle Cambrian age have been discovered in the Tienshan and Peishan. These facts suggest that a gigantic craton, to which the name of Chinese proto-platform is temporarily given, might have been created by and maintained through the Yangtze orogeny with a time span of about 200 m.y. (Sinian System to lower Cambrian). The tectonic development of China since Paleozoic time could be divided into two stages, the Paleozoic stage and the Meso-Cenozoic stage. In spatial distribution, three major units (tectonic domains) were developed dur- ing this long time period. They are the Pal-Asiatic (~) tectonic domain, the Marginal Pacific (~-P) tectonic do- main, and the Tethys-Himalayan (T-H) tectonic domain. The PEA tectonic domain took shape and developed *Also termed synsedimentary faults. 193 through the Hsingkaiian, the Caledonian, and the Varis- can Progenies. The process was roughly as follows: as soon as the Sayan-North Mongolian-Argun Geosynclinal System was folded and uplifted at the end of the early Cambrian (Hsiangkaiian), the Chinese proto-platform was disintegrated and partially transformed to form the Tienshan, the Nanshan, the Tsinling, and other geosyn- clines, all of which were again folded, transfol~ed, and consolidated at the end of the Variscan cycle. As a result, these newly consolidated geosynclinal folds joined and "cemented together" the pre-existing four platforms, the Siberian platform, the Tarim platform, the Sino-Korean paraplatforrn, and the Yangtze paraplatform, to form a gigantic craton called the Pal-Asia. During the Meso- Cenozoic stage, the tectonic development of China was under the control of the Marginal Pacific tectonic domain and the Tethys-Himalayan tectonic domain. The former may be subdivided into an inner or Cenozoic tectonic belt and an outer or Mesozoic tectonic belt. The Taiwan fold system is the only representative of the inner belt within the Chinese territory. The outer belt of great importance was superimposed upon older tectonic units of various ages. In addition to the Southeast Coast Variscides and the Mesozoic Northeast Asiatic geosynclinal fold system, they are, from north to south, the Inner Mongolian~reat Khingan Variscides, the Kirin-Heilungkiang Variscides, the Sino-Korean paraplatforrn, the Yangtze paraplatform, and the South China fold system (Caledonides). The belt is characterized by large swells and depressions trending NE or NNE, with Mesozoic blanket folds and faults, ex- tensive "Pacif~c-type" volcanics, and large-scale granitoid intrusives. THE POLYCYCLIC TECTONIC EVOLUTION OF CHINA For many years, Stille's concepts of tectono-magnetic cy- cles have been prevalent in models of geosynclinaI de- velopment: 1. A geosynclinal period with initial magmatism, mainly basic and ultrabasic rocks; 2. An erogenic period with synorogenic magmatism, mainly granitoid intrusions; 3. A quasicratonic phase with subsequent magmatism, mainly porphyries and andesites; and 4. A full cratonic period with final magmatism, mainly basalts. Since then, many well-known tectonic and economic geologists have accepted this concept (monocyclic concept) especially in the Soviet Union and Western Europe, for example, Beloussov (1962), Rittrnann (1960), deSitter (1964), and Aubouin (1965~. In the authors explorations of many key regions of the geology of China, he arrived at the conclusion that geo- synclines developed polycyclically, both in orogeny and

OCR for page 159
194 in magmatism (Huang, 1945, 1960; Huang and Chiang 1962; Huang et al., 1965, 1974~. Van Bemmelen (1949) also advocated a polycyclic development for the Indo- sinian geosynclines. Recent geological mapping in the Tienshan and Nanshan again confirms the view of poly- cyclic development of geosynclinal fold belts, which is briefly summarized below. 1. The Tienshan fold system, especially the North Tienshan consists of eugeosynclinal Vanscides. The Nanshan fold system, especially the Norm Nanshan, con- sists of eugeosynclinal Caledonides. Both exhibit un- questionable polycyclic Progenies; seven Progenies, three being the most important, are found in the Tienshan, while four Progenies (Figure 16.5), of which the last is Me most important, are found in the Nanshan. Roughly coinciding with the Progenies, intrusion of grani- toid plutons took place, indicating that magrnatism gen- erally corresponds to important Progenies. 2. Submarine volcanic eruptions are also polycyclic. In the Tienshan, intermediate volcanics (andesites) prevail and are found in four cycles corresponding to four oro- genic cycles. The Nanshan volcanics, generally basic in nature (tholeiites and spilites), can also be divided into FIGURE 16.5 Diagram showing the polycyclic development of the Nashan eugeosyncline.,B: basic submarine exclusives, mainly ~oleiites and spi- lites; a: andesite; Ha: andesite having an acidic tendency; A: andesite having a basic tendency; p: rhyolite; F: fly- schoid deposits, often alternating win ,B and c'; M: molasse, divided into two stages, D'_2 and D3, in Nanshan; I: basic and ultrabasic rocks, generally serpentinized; OS: ophiolitic suites, formed of serpentinites, basic ex~u- sives (with pillow structures) and radi- olarites; G: basic intrusives; y: granitic plutons. The size of the letters repre- senting the different rocks corresponds to the degree of their importance. T. lC. HUANG four cycles corresponding to the four most important oro- genic cycles. 3. Basic and ultrabasic rock complexes are Iilcewise polycyclic. In the Nanshan they are best developed and divisible into five cycles, each corresponding to an oro- genic cycle (Figure 16.~), of which the second cycle (Each) is the most important. Basic and ultrabasic rocks, together with radiolarian chert, form three typical oph~- olitic suites in three different ages (Z. E2, Off. On the contrary, ultrabasic rocks are poorly developed in the Tienshan, where ophiolites are absent. 4. Marine flyschoid sediments are well developed in both regions, and together with submarine volcanics form polycyclic volcano-flyschoid formations. However, they are andesite~lyschoid in the Tienshan and basaltic- flyschoid in the Nanshan. Continental molasse is devel- oped in both regions at the close of the geosynclinal evo- lution. The Devonian molasse of the Nanshan is the most typical. It is important to emphasize that polycyclicity of geo- synclinal development must not be considered as simple repetitions of geological processes as some writers be- lieve them to be, but rather they are processes of vectorial Ophiolitic Suites _ _ . ,~ ~ J Bare-- F Oslo 1 ! - 1 ~ 1 1 I l l ~ 1 ,~,\ ~ I ! 1 Us I 1 Cycle Age Marine Sediments. S u b~narine Extrus ives . V D-C ~ ~ _ F U A' S ~e 2 F urn 03 ~ ~5 ~ ~ _ O. ~~ II2 =~ I C ~ ~ O 5 ~ ~ :'~I Z: ~ .~ 1 Basic-U Itrabasic G ran itic ~ Complexes Plu20n~ ~ Remarks -gags I j Basic- ultrabasic complex chiefly | in S. Nanshan ~ 1 r I 1 Possible break beta een (2 and (: _ I ram I r _ . ~ Major angular unconforll~it~ ~ .\ngular unconformity ~ Stratigraphic break

OCR for page 159
An Outline of the Tectonic Characteristics of China . , , Granitic Marine Sediments. i Basic and Ultrabasic Cycle Age Plutons Submarine Extrusives I [ntrusives Remarks 1 -- .~1 1 1 i ~~ !~-- ,,, 21 ! it' Y-1 1 P'1 1 ' i c l~1 , m ~ it ~ ~ (-', 1~ ! ~ 1~ 'J D3 I Y , > TT D2 ~ cr D ~ al ,c~ 4~ j c, I J slj _ol ?- FIGURE 16.6 Diagram showing the polycyclic development of the Tien- shan eugeosyncline. Symbols are as in Figure 16.5. Ash spiral-like evolution with a regular arrangement. In the Nanshan geosyncline, for instance, ophiolites began to develop in Cycle I, reaching their acme of development in Cycle II, especially in Subcycle II2.* In Cycle III, although basic and ultrabasic rocks are present, no ophi- ol~tes were formed. In Cycle IV, they dwindled away. On the contrary, flyschoid sedimentation, poorly developed in Cycle I and Subcycle II', strengthened in Subcycle II2 and Cycle III, and reached its acme of development in Cycle IV. Folding, faulting, and granitoid intrusion all played an important role in each cycle, but it is obvious that they possessed a tendency of increasing activity from Cycles II to III and reached their peak of development in Cycle IV. Cycle V indicates the close of geosynclinal sedi- mentation, when extensive molasse Finned. The polycyclic evolution toward a definite direction of the Tienshan geosyncline is even more prominent (Figure 16.61. There, granitoid magmatism was not im- portant in Cycles I and Il.? but was greatly strengthened in Cycle III, and became best developed in the later part of Cycle III, i.e., in late Carboniferous time. Granitoid mag- matism rapidly decreased in Cycle IV. Moreover, grani- toid plutons of Cycles I and lI show gneissic structures absent in those of Cycles III and IV. It must be pointed out that each of the five granitoid magmatisms is clearly The author divides an erogenic cycle into subcycles (see Fig- ure~, 16.5 and 16.6). 195 l JO a F F (Ida I Basic and intermediate j extrusion es continental I Cyclic acid-intermediate I exlrusives G: Basic intrusives mainly stocks Two movements of uplift in C: l- s G ~ G: Basic intrusives | with gneissic structure, maim, stocks p in southern Tienshan Andesites, rh$olites as intercalations connected with an orogeny, while the occurrence of large- scale granite batholiths is generally contemporaneous with principal Progenies. From the petrochemical point of view, granitoid intrusives of early Variscan abound in pla- gioclase granites and albite granites; those of middle Variscan are generally normal granites, while those of late Variscan are characterized by kali-granites and alaskites, and even syenites appear. In other words, the petro- chemical characteristics of granitoid intrusives change from acid-inte~ediate to acid, then to acid~lkaline, and finally to alkaline. From the above discussion, the author arrives at the preliminary conclusion that the Nanshan geosyncline is characterized by poIycyclic basic extrusives (tholeiites and spilites), which, together with ultrabasic rocks, form polycyclic ophiolitic suites. Thlls, the two types of geo- synclines are quite different from each other in character. Poly~yclic geosynclina1 development is also distinctly revealed in other geosynclines of China, among which the Tsinling geosyncline and the East Kunlun-Tangla geo- syncline are typical (Huang et al., 1974~. Many of the famous geosynclinal systems in the world, such as the Appalachian, the Cordilleran, the Uralian, and the Tasman geosynclines' are also characterized by poly- cyclic development. Consequently, polycyclic develop- r~ent of geosynclines is, without doubt, to be considered as the general rule. For the sake of simplicity, a preliminary model for the

OCR for page 159
196 development of geosynclinal fold belts as shown in Fig- ure 16.7 is proposed. It can be seen Tom Figure 16.7 that Me development of geosynclinal fold belts includes the following: 1. Early geosynclinal cycles, including Cycle I, Cycle II, and possibly Cycle III; 2. Principal geosynclinal cycles, including Cycles III and IV; and 3. Postgeosynclinal cycles, including Cycles I and II. It must be noted that each geosynclinal cycle may include initial magmatism (ophiolites), synorogenic magmatism (granites), and subsequent magmatism (porphyries and andesites) but usually without final magmatism (plateau basalts). Moreover, flysch and molasse formations are also polycyclic, while the postgeosynclinal cycles are charac- tenzed by block-faulting accompanied by folding, again with polycyclic granites and polycyclic molasse (such as in Eastern China). If we accept the theory of plate tectonics to interpret the origin and development of geosynclines, as quite a number of geoscientists are doing, we come to the conclu- sion that plate motions are likewise polycyclic in nature. This is particularly manifested in the case of the Tasman Geosyncline. Scheibner (1972) pictured the development of that geosyncline as a series of successively eastward- accreting continental blocks, subducted by a series of eastward-retreating oceanic crust. From his palinspastic maps, it appears clear that the Tasman fold belt was formed by polycyclic plate motions. FIGURE: 16.7 Diagram showing polycyclic development of geosyn- ,'0~O clinal foldbelts. ,6000 ~ '.ooo a 1 2000 1~0 ~ 1 to) 6000 2000 T. E. HUANG PRELIMINARY OBSERVATIONS ON PLATE TECTONICS IN CHINA TSANGPO DEPTH FRACTURE ZONE Geological and geophysical data indicating the applica- bility of plate tectonics along the Tibetan part of the Himalayan geosyncline were collected and analyzed by Chinese geologists (Huang et al., 1974; Chang and Cheng, 1973~. Recent observations in the Tsangpo Valley reveal the occurrence of abundant ophioIites, and to the east of Shigatze typical melanges with exotic bloclcs of limestone containing Triassic and Jurassic fossils are found. It is thus probable that the Tsangpo depth-fracture zone is a subduction zone with the southern oceanic plate underthrusting the northern Asiatic plate. Such subduct- ing activities might be polycyclic in nature. NANSHANTSINLING REGION The eugeosynclinal character of the northern Nanshan was described in previous articles (Wang and Liu, 1976; Huang et al., 1965, 1974), while its polycyclic develop- ment has already been stressed. The glaucophane schist zone chiefly consists of quartz Muscovite glaucophane schist and garnet-epidote glaucophane schist, and the majority of He tholeiites of the ophiolitic suites are char- acterized by very low K2O contents (usually <0.3 butt). It is interesting to note that all of these features are similar to those of other paleogeosynclines in the world. The 20000 M East tics ' Principal cycles Po*~-c~les .. cop ~ _~ <2 i ads, coca 1 I C>c~ ~ I ~N 1 ~ >an pats - Sg - Worm I ~ ~ ! too ) ^" ~ f55~ q Posit 1 & - blazon ~ Cycle ~ . Cycle ~ ,Cyc~ 1 Cay ~ LEGEND Ale 0' Ophwliiic Suiles (hoot li7zes S~Ul7Z9 depth f7.aci~r.es) F`'Fz'F3 Bush: ull7abas~c comic:: ice of, go alone character Flinch formafio7`s I',M`,M~ ~,~,~37~ ~~ ~05' (g`,gz ~ems" ~ H.~l folding `iles3 early 7, ~ i7~ler~ncli - , ~~ Folding of secondary ~mp~- la~r /, f alkaline character 1 Sac ~Cl, ~2~3 Po - ~e. AL minor m`~ Ads, eat BY, Of band, la~ Mola~se forays \\ Back faming

OCR for page 159
An Outline of the Tectonic Characteristics of China ophiolites are polycyclic, with successive motions being from the south to the north (Wang and Liu, 1976~. The Northem Tsinling geosyncline is the continuation of the Northern Nanshan. Observations show that the Caledonides of the Tsinling extend from south of P~ochi to Nanyang Basin, probably including ophiolitic patches. Li (1975) advocates that the Tsinling is characterized by plate-tectonic interactions. It is possible that the eastern Tsinling fold belt was formed by the mutual approach between the Sino-Korean and the Yangtze paraplatfo~ms. CHINSHABIANGRED RIVER DEPTH- FRACTURE ZONE Recent mapping in the Chinshakiang-Red River region discloses evidence for paleoplate motions. South of Batang, in the western side of the upper Chinshakiang fracture zone, occur typical ophiolitic melanges charac- terized by serpentinites, spilites with pillow structure, and various kinds of basic to ultrabasic rocks intercalated with radiolarian cherts, while exotic blocks of omphacite- eclogite; dillage~innamon stones; and particularly De- vonian, Carboniferous, and Permian limestones are found in the matrix. In the eastern side of the same zone, wild- flyschlilce deposits characterized by a matrix of argilla- ceous and arenaceous rocks with various exotic blocks yield many fossils ranging in age from Silurian to Per- mian, which are, without exception, older than those in the matrix. Judging from all the facts observed so far, it is probable that a western oceanic plate approached and collided, from Permian to Late Triassic time, with the eastern suboceanic plate (this belongs to the Indosinian orogeny). Along the Red River valley, the existence of ophiolitic melanges and glaucophane schists is also prob- able. Apparently, the Chinshakiang zone and the Red River zone belong to the same system of convergent plates. Since in western Yunnan and northern Burma there occur a series of lithospheric fractures, i.e., Chin- shakiang-Red River, Lantsangkiang, Nukiang, and Naga- Arakan Yoma, and since the principal ages of these Eacture zones seem, respectively, to be Indosinian, early Yenshanian, late Yenshanian, and Himalayan, the hy- pothesis is suggested that polycyclic plate subductions happened from east to west in a successively repeating manner. RE FE RE NC E S Aubouin, J. (1965). Geosynclines, in Development in Geotec- tonics, Elsevier, Amsterdam. Beloussov, V. V. (1962). Basic Problems in Geotectonics, McGraw-Hill, New York, 809 pp. Chang, C. F., and H. L. Cheng (1973). Some tectonic features of the Mt. Joimo Lungma area, southern Tibet, Sci. Sin. 16, 257-265. 197 Ch'eng, Y. C., F. D. Chung, and Y. C. Su (1973). The Chen-Tan series in northern and northwestern Chinese areas, Acta Geol. Sin. 1973, no. 1 (in Chinese). Chinese Academy of Geological Sciences (1976). The Geological Map ofthe Peoples Republic of China (1:4,000,000), Map Pub- lishers (in Chinese). deSitter, L. U. (1964). Structural Geology, 2nd ea., McGraw-Hill, New York, 530 pp. Fitch, T. J. (1972). Plate convergence, transcu~Tent faults and internal deformation adjacent to southeast Asia and the west- ern Pacific, 1 Geophys. Res. 77, 4432~460. Gansser, A. (1964). Geology of the Himalayas, Interscience, New York. Gansser, A. (1966). The Indian Ocean and the Himalayas: a geo- logical interpretation, Eclogae Geol. Hel?;. 59, 831 848. Hsi, C. W., S. F. Feng, C. S. Li, H. P. Chen, K. T. Wen, C. J. Chang, and C. C. Hsiung ( 1974a). The background of the deep structure in the middle of the Northern China Plain and the Hsing-Ta earthquake, Acta Geophys. Sin. 17, no. 4 (in Chinese). Hsi, C. by., S. F. Feng, C. S. Li, H. P. Chen, K. T. Wen, C. J. Chang, and C. C. Hsiung (1974b). The reflection wave of the deep earthquake in Chai-Ta-Mu Basin and crustal structure, Acta Geophys. Sin. 17, no. 2 (in Chinese). Hsi, C. W., S. F. Feng, C. S. Li, H. P. Chen, K. T. Wen, C. J. Chang, and C. C. Hsiung (1975). The background of the deep structure in the middle of the Northem China Plain and the Hsing-Ta earthquake, Acta Geophys. Sin. 18, no. 3 (in Chinese). [Iuang, T. K. (1945). On major tectonic fonns of China, Geol. Mem. Ser. A, no. 20, National Geological Survey of China. Huang, T. K. (1964)). The main characteristics of the structure of China: preliminary conclusions, Sci. Sin. 9, 494 544. Huang, T. K., and C. F. Chiang (1962). The primary study of the crust rules developed from the polycyclic point of view, Acta Ceol. Sin. 40, no. 1 (in Chinese). Huang, T. K., C. K. Chang, C. M. Chang, and K. M. Chen (1965). On eugeosynclines and miogeosynclines and their polycyclic development, Chin. Acad. Geol. Sci. Selec. Prof. Papers Reg. Tectonics, Sec. C, Peking (in Chinese). Huang, T. K., C. S. Jen, C. Chiang, C. M. Chang, and C. K. Chang (1974). The new knowledge of some characteristics for Chinese plate tectonics, Acta Geol. Sin. 1974 (in Chinese). Lee, J. S. (1939). The Geology of China, Thomas Murby & Co., London, 528 pp. Li, C. Y. (1975). The primary analysis for the development of some parts of Chinese regional structure using plate tectonic theory, Acta Geophys. Sin. 18, no. 1 (in Chinese). Molnar, P., and P. Tapponnier ( 1975). Cenozoic tectonics of Asia: effect of a continental collision, Science 189, 419~26. Rit~mann, A. (1960). Vulkane and ihre Tatigleit 2. umgearioeite- ten and erw., Add., Enke, Stuttgart, 336 pp. Scheibner, E. (1972). Actualistic models in tectonic mapping, 24th Sess. let. Geol. Congr. Can. 405~22. Terman, M.J. (principle compiler) (1973). Tectonic Map of China and Mongolia, (1:5,000,000), Geol. Soc. Am. van Bemmelen, R. W. (1949). The Geology of Indonesia, Govern- ment Printing Office, The Hague, 732 pp. Wang, C., and H. Y. Liu (1976). The ancient oceanic crust in western China's Ch'i-Lien mountains and its plate tectonic meanings, Geol. Sci. 1974, no. 1 (in Chinese).