NASA-CR-190677 19930000924

LIBRARY COpy

SEP I 6 1992

lANGLEY RESEARCH CENTER UBRARYNASA HAMPTON VIRGINIA

LARGE METEORITE IMPACTS AND PLANETARY EVOLUTION

Sudbury, Ontario, Canada August 31-September 2, 1992

Dl" l.DI.'.; DISPLAY 03/2/1 93N10112*# ISSUE 1 PAGE 105 CATEGORY 46 RPT#: NASA-CR-190677 NAS 1.26:190677 LPI-CONTRIB-790 CNT#: NASW-4574 92/00/00 93 PAGES UNCLASSIFIED DOCUMENT UTTL: International Conference on Large Meteorite Impacts and Planetary Evolution CORP: Lunar and Planetary Inst., Houston, TX. CSS: (Publications Services Dept. ) SAP: Avail: CASI HC A05/MF AOl . CIO: UNITED STATES Conference held in SUdb~ry, Ontario, 31 Aug. - 2 Sep. 1992; sponsored by Barringer Crater Co., Falconbridge Ltd., Geological Survey of Canada, Inco Ltd., International Union of Geological Sciences, Laurentian Univ., LPI, and Ontario Ministry of Northern Development and Mines MAJS: /*CANADA/*CONFERENCES/*HYPERVELOCITY IMPACT/*METEORITE COLLISIONS/* METEORITE CRATERS/*PLANETARY EVOLUTION MINS: / BRECCIA/ CANADIAN SHIELD/ GEOMORPHOLOGY/ ROCK INTRUSIONS/ STRUCTURAL PROPERTIES (GEOLOGY) ANN: The papers that were accepted for the International Conference on Large Meteorite Impacts and Planetary Evolution, 31 Aug. - 2 Sep. 1992, are presented. One of the major paper topics was the Sudbury project. For individual titles, see N93-l0ll3 through N93-l0l96.

•

PAPERS PRESENTED TO THE INTERNATIONAL CONFERENCE ON LARGE METEORITE IMPACTS AND PLANETARY EVOLUTION

August 31-September 2, 1992 Sudbury, Ontario, Canada

Sponsored by Barringer Crater Company Falconbridge Limited Geological Survey of Canada Inco Limited International Union ofGeological Sciences Laurentian University, Sudbury Lunar and Planetary Institute, Houston Ontario Ministry ofNorthern Development and Mines

LPI Contribution No. 790 ..

Compiled in 1992 by

Lunar and Planetary Institute 3600 Bay Area Boulevard Houston TX 77058-1113

Material in this volume may be copied without restraint for library, abstract service, education, or personal research purposes; however, republication ofany paper or portion thereofrequires the written permission ofthe authors as well as the appropriate acknowledgment of this publication.

TheLunarandPlanetary Institute is operated by the Universities Space Research Association under ContractNo. NASW-4574 with the National Aeronautics and Space Administration.

,. PREFACE

This volume contains papers that have been accepted for the International Conference on Large Meteorite Impacts and Planetary Evolution, August 31-5eptember 2, 1992, in Sudbury, Ontario, Canada. The Scientific Organizing Committee consisted ofB. Dressler and V. L. Sharpton, co-chairmen, and B. M. French, R. A. F. Grieve, and E. M. Shoemaker. Members ofthe Local Organizing Committee were H. Gibson (Chairman), S. Ball, C. J. A. Coats, B. Dressler, G. W. Johns (Treasurer), W. Meyer, W. V. Peredery, and D. H. Rousell.

Administrative support was provided by the Program Services Department staff at the Lunar and Planetary Institute. This abstract volume was prepared by the Publications Services Department staff at the Lunar and Planetary Institute.

,

Ne,3-JOI/~*

TABLE OF CONTENTS

Carbon and Oxygen Isotope Study 0/Carbonates from Highly Shocked Clasts o/the Polymict Breccia o/the Haughton Crater (Canada) P. Agrinier. I. Martinez. M. Javoy. and U. Scharer...... •...... •... 1

Research Core Drilling in the Manson Impact Structure, Iowa R. R. Anderson. J. B. Hartung. D. J. Roddy. and E. M. Shoemaker 2

A Quasi-Hertzian Stress Fieldfrom an Internal Source: A Possible Working Model for the Vrede/ort Structure L. A. G. Antoine. W. U. Reimold. and W. P. Colliston 3

Sudbury Projec{(University o/Munster-Ontario Geological Survey): (6) Origin o/the Polymict, Allochthonous Breccias o/the Onaping Formation M. E. Avennann 4

Sudbury Project (University o/Munster-Ontario Geological Survey): (1) Summary 0/Results-An Updated Impact Model M. Avermann. L. Bischoff. P. Brockmeyer. D. Buhl, A. Deutsch. B. O. Dressler. R. Lakomy. V. Muller-Mohr. and D. StOffler 5

Structural Aspects o/the Araguainha Impact Structure (Brazil) L. Bischoff. P. Brockmeyer, U. Jenchen, and R.-M. Swietlik 6

Sudbury Project (University 0/Munster-Ontario Geological Survey): (2) Field Studies 1984-1989-Summary o/Results L. Bischoff, B. O. Dressler, M. E. Avermann, P. Brockmeyer, R. Lakomy. and V. Muller-Mohr 7

Large Meteorite Impacts: The KIT Model B. F. Bohor 8

Shocked Zircons in the Onaping Formation: Further Proof0/Impact Origin B. F. Bohor and W. J. Betterton 9

Suevite Superposition on the Bunte Breccia in Nordlinger RieslGermany: New Findings on the Transpon Mechanism o/Impactites D. Bringemeier 10

Sudbury Project (University o/Munster-Ontario Geological Survey): (7) Sr-Nd in Heterolithic Breccias and Gabbroic Dikes D. Buhl. A. Deutsch, R. Lakomy, P. Brockmeyer, and B. Dressler 11

Hong Kong is an Impact Crater: Proo/from the Geomorphological and Geological Evidence C.-I. Chan, S. Wu. and X. Luo 12 .. v Melt Production in Large-Scale Impact Events: Calculations ofImpact-Melt Volumes and Crater Scaling . M. J. Cintala and R. A. F. Grieve ~ 13

Melt Production in Large-Scale Impact Events: Planetary Observations and Implications M. J. Cintala and R. A~ F. Grieve 14

Structural Review ofthe Vredefort Dome W. P. Colliston and W. U. Reimold 16

Intrusive Origin ofthe Sudbury Igneous Complex: Structural and Sedimentological Evidence E. J. Cowan and W. M. Schwerdtner 17

Enhanced Magnetic Field Production During Oblique Hypervelocity Impacts D. A. Oawford and P. H. Schultz 18

Impactite and Pseudotachylite from Roter Kamm Crater, Namibia J. I. Degenhardt Ir., P. C. Buchanan, and A. M. Reid 20

Sudbury Project (University ofMUnster-Ontario Geological Survey): (4) Isotope Systematics Support the Impact Origin A. Deutsch, D. Buhl, P. Brockmeyer, R. Lakomy, and M. Flucks 21

Noril'sklSiberian Plateau Basalts and Bahama Hot Spot: Impact Triggered? R. S. Dietz and J. F. McHone 22

Mobilization oftile Platinum Group Elements by Low-Temperature Fluids: Implications for Mineralization and the Iridium Controversy K. Dowling, R. R. Keays, M. W. Wallace, and V. A. Gostin 23

Does the Bushveld-Vredefort System (South Africa) Record the Largest Known Terrestrial Impact Catastrophe? W. E. Elston 23

Ruthenium/Iridium Ratios in the Cretaceous-Tertiary Boundary Clay: Implications for Global Dispersal and Fractionation Within the Ejecta Cloud N. J. Evans, W. D. Goodfellow, D. C. Gregoire, and J. Veizer 25

Diaplectic Transformation ofMinerals: Vorotilov Drill Core, Puchezh-Katunki Impact Crater, Russia V. I. Feldman 25

Breccia Dikes from the Beaverhead Impact Structure, Southwest Montana P. S. Fiske, S. B. Hougen, and R. B. Hargraves 26

vi New Perspectives on the Popigai Impact Structure J. B. Garvin and A. L. Deino...... •...... •...... ••...... •...... •...... 27

The Zhamanshin Impact Feature: A New Class ofComplex Crater? J. B. Garvin and C. C. Schnetzler ...... •...... •. 28

Asteroids and Archaean Crustal Evolution: Tests ofPossible Genetic Links Between Major Mantle/Crust Melting Events and Clustered Extraterrestrial Bombardments A. Y. Glikson 29

The Acraman Impact and Its Widespread Ejecta, South Australia v. A. Gostin, R. R. Keays, and M. W. Wallace 30

Optical and TEM Study ofShock Metamorphismjrom the Sedan Test Site A. J. Gratz 31

Simulated Meteorite Impacts and Volcanic Explosions: Ejecta Analyses and Planetary Implications A. J. Gratz and W. J. Nellis 31

Melt Production in Large-Scale Impact Events: Implications and Observations at Terrestrial Craters R. A. F. Grieve and M. 1. Cintala 32

Venusian Impact Basins and Cratered Terrains w. B. Hamilton 33

Where's the Beaverhead Beef! R. B. HargJ'aves 35

Report on the International Cambodian Crater Expedition-1992 J. Hartung, C. Koeberl, P. Lee, K. Pagnacith, and T. Sambath 36

The Distribution and Modes ofOccurrence ofImpact Melt atLunar Craters B. R. Hawke and J. W. Head 37

Large Impacts in the Baltic Shield with Special Attention to the Uppland Structure H. Henkel and R. Lilljequist 38

The Panther Mountain Circular Structure, A Possible Buried Meteorite Crater Y. W. Isachsen, S. F. Wright, F. A. Revetta, and R. J. Dineen 40

Geomechanical Models ofImpact Cratering: Puchezh-Katunki Structure B. A. Ivanov ...... •40

vii Thermobarometric Studies on the Levack Gneisses-Footwall Rocks • to the Sudbury Igneous Complex R. S. James, W. Peredery, andJ. M. Sweeny ~ 41

The Cretaceous-Tertiary (KIT) Impact: One or More SoW'ce Craters? C. K<>e1>erl ••••• "' •.••.••••••...•••••••.•••..••••••.•••..••.••••••.•..•.••••••••..••..•••••••••.••.••••.•...•....•.•••••...••••••...... •...• 41

Tektite Origin by Hypervelocity Asteroidal or Cometary Impact: The Questfor the SoW'ce Craters c. Koeberl "' : 42

Early Archean Spherule Beds ofPossible Impact Originfrom Barberton, South Africa: A Detailed Mineralogical and Geochemical Study

C. Kreher}, W. U. Reimold, and R. H. Boor 11 ••• 44-

U-Pb Isotopic Results for Single Shocked and Polycrystalline Zircons Record 55D-65.5-MaAgesfor a K-TTarget Site and 2700-1850-MaAges for the Sudbury Impact Event T. E. Krogh, S. L. Kamo, and B. F. Bohor ...... •..... 44-

Influence ofthe Preshock TemperatW'e on Shock Effects in Quartz F. :Langenhorst and A. Deutsch 4S

Search for a Meteoritic Component at the Beaverhead Impact StruetW'e, Montana P. I..ee and R. W. Kay 47

Ordovician Impacts at Sea in Baltoscandia M. LindstrOm, V. Pl!ura, T. Floden, and A. Bruun 47

Does the Sedimentology ofthe Chelmsford Formation Provide Evidence for a Meteorite Impact Origin ofthe Sudbury Structure? D. G. F. Long 48

Impact Origin ofthe Sudbury StructW'e: Evolution ofa Theory P. D. Lowman Jr 48

Imaging Radar Investigations ofthe Sudbury StructW'e P. D. Lowman, V. H. Singhroy, and V. R. Slaney 49

Phase Transformations in 40-60-GPa Shocked Gneisses from the Haughton Crater (CanmJa): An Analytical Transmission Electron Microscope (ATEM) Study I. Martinez, F. Guyot, and U. Sch1irer 49

.Impactites from Popigai Crater V. L. Masaitis 51

viii Sudbury Breccia and Suevite as Glacial Indicators Transponed 800 km to Kentland Astrobleme, Indiana J. F. McHone, R. S. Dietz. and W. V. Pered.ery 51

What Can We Learn About Impact Mechanics from Large Craters on Venus? w. B. McKinnon 8Ild J. S. Alexopoulos 52

Sudbury Project (University ofMUnster-Ontario Geological Survey): (5) New Investigations on Sudbury Breccia V. Milller-Mohr 53

A History ofthe Lonar Crater, India-An Overview V.K.Nayak 53

Sudbury Igneous Complex: Impact Melt or Igneous Rock? Implications for Lunar Magmatism M. D. Nann.an S4

Melting andIts Relationship to Impact Crater Morphology J. D. O'Keefe and T. J. Ahrens 55

Impact Cratering Record ofFennoscandia L. J. Pesonen and H. Henkel 57

Bohemian Circular Structure, Czechoslovakia: Search for the Impact Evidence P. Rajlich 57

The Vredefon Dome-Review ofGeology and Deformation Phenomena and Status Repon on Current Knowledge and Remaining Problematics (Five Years After the Cryptoexplosion Workshop) W. U. Reimold 59

The Pseudotachylites from the Vredefort Structure and the Witwatersrand Basin W. U. Reimold and W. P. Colliston ...... •...... •...... ••....••...... •....•..... 6()

Coincidence in Time ofthe Imbrium Basin Impact and Apollo 15 KREEP Volcanic Series: Impact-Induced Melting? G. Ryder ~ 61

Apollo 15 Impact Melts, The Age ofImbrium, and the Earth-Moon Impact Cataclysm G. Ryder and G. B. Dal:rymple ..•...... 62

Searchfor the 700,OOO-Year-Old Source Crater ofthe Australasian Tektite Strewn Field C. C. Schnetzler and J. B. Garvin 63 l Recognizing Impactor Signatures in the Planetary Record P. H. Schultz and D. E. Gault 64

Paradigm Lost: Venus Crater Depths and the Role ofGravity in Crater Modification V. L. Sharpton 65

KIT Boundary Stratigraphy: Evidence for Multiple Impacts and a Possible Comet Stream E. M. Shoemaker and G. A. Izett 66

Geological Evidence for a 2.6-Ga Strewn Field ofImpact Spherules in the Hamersley Basin ofWestern Australia B. M. Simonson : 68

SAR in Support ofGeological Investigations ofthe Sudbury Structure v. Singhroy, R. Mussakowski, B. O. Dressler, N. F. Trowell, and R. Grieve 69

Viscosity Determinations ofSome Fractionally Generated Silicate Melts: Implications for Slip Zone Rheology During Impact-Induced Faulting J. G. Spray 69

The Large Impact Process Inferredfrom the Geology ofLunar Multiring Basins P. D. Spudis 69

Sudbury Project (University ofMilnster-Ontario Geological Survey): (3) Petrology, Chemistry, and Origin ofBreccia Formations D. StOffler, A. Deutsch, M. Avermann, P. Brockmeyer, R. Lakomy, and V. MUller-Mohr ; 71

I'Bronzite" Granophyre: New Insight on Vredefort A. M. Therriault and A. M. Reid 72

A Comparison ofthe Chemistry ofPseudotachylyte Breccias in the Archean Levack Gneisses ofthe Sudbury Structure, Ontario L. M. Thompson and J. G. Spray 73

40Ar-39Ar Ages ofthe Large Impact Structures Kara and Manicouagan and Their Relevance to the Cretaceous-Tertiary and the Triassic-Jurassic Boundary M. Trieloff and E. K. Jessberger ...... •...... 74

40Ar-39Ar Dating ofPseudotachylites from the Witwatersrand Basin, South Africa, with Implications for the Formation ofthe Vredefort Dome M. Trieloff, J. Kunz, E. K. Jessberger, W. U. Reimold, R. H. Boer, and M. C. Jackson 75

x Al Umchaimin Depression, Western Iraq: An Impact Structure? J. R. Underwood Jr 77

A Late Devonian Impact Event and Its Association with a Possible Extinction Event on Eastern Gondwana K. Wang and H. H. J. Geldsetzer ~ 77

Electron Petrography ofSilica Polymorphs Associated with Pseudotachylite, Vredefort Structure, South Africa J. C. White 78

Floor-fractured Crater Models ofthe Sudbury Structure, Canada R. W. Wichman and P. H. Schultz 79

Variation in Multiring Basin Structures as a Function ofImpact Angle R. W. Wichman andP. H. Schultz 80

Self-Organized Rock Textures and Multiring Structure in the Duolun Crater S. Wu andJ. Zhang 81

Geochemical Aspect ofImpact Cratering: Studies in Vernadsky Institute O. I. Yakovlev and A. T. Basilevsky 82

,

xi

!PICOfIII'iblllion No. 790 1 " Papers Presented to the International Conference on Large Meteorite Impacts and Planetary Evolution

CARBON AND OXYGEN ISOTOPE STUDY OF CAR­ • polymict brc

It is knownthatthereleaseofvolatilesonimpactis an important· D controlling factorincrateringprocessesincarbonateterranes andin the mobility ofchemical elements [1,2,3]. • Inorderto assess thenature and the role ofcarbon- and oxygen­ bearing volatiles during impact-induced metamorphism of sedi­ • mentarY rocks, the 13C/l'JC and 11()/16() ratios~d carbonatecontents were determined for 30 shocked clasts from the Haughton Crater -10 +:..._--.--_-....-_._...... -_-....--_--, polymictbrecciaaswellas for someunshockedcarbonates from the o 20 40 60 80 100

sedimentarY cover adjacent to the crater. Shock-induced C01 loss Carbonate content weight '" during decarbonation of calcite is known to be a function ofpeak pressure and ambient partial pressure of the volatile species [2,3]. FIg.t. 313C vs.carbonatecontentdiagram. Thenumbers along the In our clast samples, shocked from 20 to 60 GPa, we expect about curve correspond to the degree ofRaleigh degassing. 20% to 100% C01 loss and preferential depletion in 13C and II() in theresidualcarbonate[4]. Rayleighmodel(progressivelossofCOJ and batch model (single-step loss ofCOJ curves for this depletion are shown in Figs. 1 and 2. The magnitudes of the BC and 18() In the shocked clasts, carbonates represent between 0.5 to depletions increase with the increase of the CO2 loss. In addition, 10 wt% of the rock and display isotopic composition of carbon these isotopicdepletions shouldbecorrelatedwith an enrichmentin between -4 to +9%". When compared to the unshocked samples, CaO and MgO in the residual solid. they are mostly enriched in 13C (313C up to 99'01» or Wlchanged

TABLE 1. Carbon IIld oxygen isotopic compositions of carbonliles from clasts of the Haughton Cnter polymict breccia and from the adjacent sedimentary cover. Rocks were auaclted in phosphoric aeid [51 in order to analyze ClIJbonates only. Clasts Carbonate auc &'0 Reference Carbonate &'C &1'0 Sample Content CLPDB tLSMOW Sample Content CLPDB CLSMOW

HAHlI 0.5 8.8 15.88 DI52a 36.1 -5.4 14.17 HAH8 4 -3.5 15.15 DI52b 60.2 -4.7 14.09 HAH2S 21.8 -1.1 21.71 7275 21.7 -4.3 12.67 HAH68 5.5 2.7 17.61 DI9 88.9 -2.2 20.55 HAH69 6.4 5.2 17.78 71314 13.1 -4.3 9.70 HAH5 0.5 4.6 19.45 DI21 99.6 -1.9 21.37 HAH15 4.3 4.7 16.82 1984 41.2 -1.8 20.8 HAH17 1.8 3.8 18.70 DIG2 92.9 -3.7 28.48 HAH19 3.1 0.8 19.44 DIGI 76.2 -3.8 27.32 HAH26 1 4.9 18.45 DIG3 74.6 -3.7 28.99 HAH28 8.9 -21 19.01 DIG4 50.4 -2.1 24.10 HAH74 4 0.9 19.75 DIG5 61.2 -1.5 24.74 HAH73 3.3 1.9 18.53 DI95 10.7 -0.7 22.6 HAH66 5.3 2.5 18.50 DI92 100 0.6 21.06 HAH19 3.1 0.8 19.44 013 63.1 -1.8 24.29 HAH39 5.1 -0.6 16.20 DI95 12.3 -1 24.83 HAH34 14.8 -2.2 15.26 DI93 14.5 -1.4 23.06 HAH21 8.1 1.7 16.23 DI96 7.6 -0.5 23.43 HAH71 7.4 2.8 17.40 0197 9.1 -0.4 22.7 HAH29 7.8 7.5 17.24 DI19 17.1 -6.2 25.44 HAH30 4.6 2.6 15.02 DI45 6.2 -2.7 24.24 HAH67-1 13.8 3.8 15.50 DIG12 16.5 -1.8 24.61 HAH67·2 1.7 0.6 19.66 814 80.9 -2.6 21.20 HAHum24 0.2 2.7 15.86 72614 46.1 -3.9 18.14 2 1992 Sudbury Conference .. •.polymict breccia clllSt RESEARCH CORE DRILLING IN THE MANSON IMPACT l 1 o unshockcd dolomite STRUCTURE, IOWA. R. R. Anderson , J. B. Hartung , D. J. A unshockcd slll1dstonc Roddy1, and E. M. Shoemaker, Ilowa Department of Natural Resources Geological Survey Bureau, 109 Trowbridge Hall, Iowa ..... CaCO:J -+ Cao + CO:z 10 a CaJbon. l.ClO22 City IA 52242-1319, USA, 2U.S. Geological Survey, Branch of 8 • a OKygcn. 1.CIO/iO AstrogeologicStudies,2255 NorthGeminiDr., FlagstaffAZ 86001, 6 • USA. 4 • The Manson impact structure (MIS), located in north-eentral 83C 2 •• Iowa, has a diameter of35 km and is the largestconfmned impact 0 • structure in the United States. The MIS has yielded a ~ArptAr age of 65.7 Ma [1] on microcline from its central peak, an age that is -2 "indistinguishable"from the ageoftheCretaceous-Tertiary bound­ -4 • DIXI ary. -6 to Inthe summer of 1991 the Iowa Geological Survey Bureauand U.S. Geological Survey initiated a research coredrilling projecton -8 theMIS.Thefll'St core(M-1) was locatedontheedgeoftheCentral 8 10 12 14 16 18 .~20 22 24. 26 28 30 Peak (Fig. 1).Beneath55 m ofglacial drift, thecore penetrated a 6­ Btl'O m layered sequence of shale and siltstone and 42 m of Cretaceous shale-dominated sedimentary clast breccia (Fig. 2). Below this breccia, the core encountered two crystalline rock clast breccia Fig. 2. &13(: vs. &180 diagram. Thenumbers along the curve cor­ units. The upperunit is 53 m thick, with a glassy matrix displaying respond to the degree ofRaleigh degassing. various degrees of devitrification. The upper half of this unit is dominated by the glassy matrix, with shock-deformed mineral grains (especially quartz) the most common clast. Clast content instead of being depleted as they should be if they were residues increases toward the baseoftheunit. Theglassy-matrixWlit grades resulting from the 13(:-enriched CO2 losses (Fig. 1). On the other downward into the basal unit in the core, a crystalline rockbreccia hand, these carbonates are systematically depleted in 180. The witha sandymatrix,thematrixdominatedbyigneous andmetamor­ magnitude of the 180 depletion is variable (Fig. 2). Thus no phic rock fragments ordisaggregated grains from those rocks. The systematic correlation between 180 and 13(: was observed as ex­ unit is about 45 m thick, and grains display abundant shock pected from the Rayleigh model or the batch model for CO2 loss deformation features. Preliminary interpretations suggest that the (Fig. 2),norwas anyclearrelationshipobservedbetweenthe carbon crystallinerockbreccias are the transientcraterfloor,liftedupwith and oxygen isotopic shifts and CaD + MgO contents of the shock the central peak. Thesedimentaryclastbrecciaprobably represents clasts. The spreadofthe carbonand oxygenisotopic composition is a postimpact debris flow from the crater rim, and the uppermost probablyduetoa varietyofprocessesthatmayaffectC andoduring layered unit probably represents a large block associated with the the shock. Several explanations can be suggested for the volatile flow.

release from sedimentary rocks: (1) degassing of CO2 with a The second core (M-2) was drilled near the center of the crater peculiar carbon isotope fractionation coefficient (a < I, CO2pref­ moatin an area where anearly cratermodel suggested the presence erentially concentrates 12C), (2) other reactions with production of of postimpact lake sediments. The core encountered 39 m of CO orC, or (3) oxygen isotope exchange with coexisting silicates. sedimentary clast breccia, very similar to that in the M-1 core. Moreover, the absence ofCaD +MgO enrichments in the shocked Beneaththebreccia, 120mofpoorlyconsolidated,mildlydeformed clasts, which is a general feature, may indicate that substantial and sheared siltstone, shale, and sandstone was encountered. The amounts of CaD + MgO are mobilized during the shock processes since original sedimentary samples contain up to 66 wt%. The modalities of this inferred mobilization of CaD + MgO are still unknown. However, for somesamples,latesecondaryprocesses mayhave partly altered the primary characteristics (nature) of the residues M-1 M-2 resulting from the volatile release because carbonate crystals are w E observed along cavity walls (bubbles, cracks). Itsuggests thatsome

C-rich fluids (C02?) were pervasive during the formation and the cooling ofthe polymictbreccia. In these samples, the observed 13(: enrichments can therefore be partly explained by the trapping of some heavy CO2 released during the shock process itself. References: [1] Kieffer S. W. and Simonds C. H.(1980) Rev. Geophys. Space Phys., 18, 143-181. [2] Lange M. A. and Ahrens T. J. (l986)EPSL,77,409-418. [3] TyburczyJ.A.andAhrensT. J. (1986)JGR. 91,4730-4744. [4] Bottinga(I968)J.Phys.Chem.• 72, Fig. 1. East-west ClOSS section of the Manson Impact Structure 800-808. [5] McCrea J. M. (1950) J. Chem. Phys., 18,849-857. showing the location ofthe Manson M-I and M-2 cores. .. [PI ConlributUm No. 790 3

MANSONM·1 MANSONM·2 feet feet o o . o· 00 0 \ 0 / ...... 0 0 "o/0/ °/ . '\. . t= o/ 0...... • 0 100 - ° 100 - 0 --- 0 • 0 ,,0 o. / o / 0 0 . .~'(? .-"i_i .n <:::.-:>Of\. 00 ?~o?~ as- 400 - .~ '§i lJ m:=E ~~X ~i~ -C!J 500 - (( ~~~ o 0_

as- ~:~~ CI) .-.~ ~ 600 - • 0 .5 8 if o· 0 1lDu' #J. 0 ) lJ d5:=E • • 0 ~ ... "'O '0 <% . o le~ ~. 0 0- 700 - o 0 0 0 0

!Z2l shale I:::}:::::::::::::! sandstone 800 - D siltstone ~ carbonate

Fig. 2. Generalized lithologic logs of the Manson M-l and M-2 cores.

basal unit in the core was another sequence of sedimentary clast A QUASI-HERTZIAN STRESS FIELD FROM AN INTER­ breccia, 51 m thick, and similartotheupperinterval inthe core. The NAL SOURCE: A POSSmLE WORKING MODEL FOR two sedimentary clastunits, likethe lithologically similarunitinthe THE VREDEFORT STRUCTURE. L. A. G. Antoinel , W. U. 2 M-1 core, probably formed as debris flows from the craterrim. The Reimold , and W. P. Colliston', IDepartment of Geophysics, Uni­ middle, nonbrecciated interval is probably a large, intact block of versityoftheWitwatersrand,PrivateBag3,Wits 2050,Johannesburg, Upper Cretaceous strata transported from the crater rim with the South Africa, 2Economic Research Unit at the Department of debris flow. Alternatively, the sequence may represent the elusive Geology, University of the Witwatersrand, Private Bag 3, Wits postimpact lake sequence. 2050, Johannesburg, South Africa, 'Department ofGeology, Uni­ Additional drilling is planned for the late spring and summerof versity ofthe Orange Free State, P.O. Box 339, Bloemfontein9300, 1992. Targets include structurally preserved Upper Cretaceous South Africa. strata on the Terrace Terrane, a zone of complete melting, and postimpact lake sediments in the Crater Moat. The Vredefort structure is a large dornal feature approximately Reference: [1] Kunk M. J. et al. (1989) Science, 244, 110 km southeast of Johannesburg, South Africa, situated within 1565~1568. and almost central to the large intracratonic Witwatersrand Basin. 4 1992 SUdbury Conference

&

former model, for the Vredefort case. argue that itcould provide a 'j"'-"-----v----.-::-..::---­ mechanism for deformation phenomena widely regarded as evi­ c--~_---::.....,.. dence ofshockmetamorphism (pseudotachylite, quartz with planar w-----.f. microdeformations, and shatter cones). Conversely. these same deformationphenomenaare currently being debated [4,5] and ithas OGG Contact surlace: eftecI:lve Mace unknown, beenhypothesized thatthey could beformed byhigh-strain tectonic - Cone fracture p"ob*y within the OGG processes. ( IX Angle of pr_tion(u""'-"l In Fig. I, a Hertzian stress field is sketched (after [6]), both in / Appfoxlmate lim" ot compressive regime plan- and cross-section. The stress component GJ is compressive, (Al while G., the principal component, is tensile and subparallel to the overlying strata. The lrajectories of these stress fields em well account for overturning of collar stratigraphy and the subvertical The crust - on • edge model attitude of the gneissic fabric in the core. In the collar rocks

WEST EAST pseudotachylite veins generally occur along bedding plane faults, •••••••.:: +? + ? + .:' :....,•.1.1: ..., while in the core they are parallel to the principal shear directions [78). In plan, G\ is tensile ndially, whereas thc intermediate G ~.;.~'~.=.~.=..~+. + /~f:\ + '+~'='~"~"~"~"~"~'~' 2 pIulon component is a "hoop stress." It is noteworthy that the lineaments ofgeophysical images comply with the G. orientation and the triad of alkali granitic complexes (Roodekraal, Rietfontein, and Lindequesdrif), intruded into the collar, describe an arc similar to (B) the G2 hoop stress. The location ofthese intrusives could be at the • Alkali intrusives intersection of the G\, G2 tensile stress components. Conefraeture This postulated quasi-Hertzian deformation model is dynamic, so the contact stress and resultant sttain would be expected to be complex and to modify in time (see, e.g., (9». Within the compres­ (e) basement sive regime (which may have a radius as much as the contact diameter ofthe indentor, in this case the diapir) [8,9J the G\ stress component is compressive and may aa:ount for the radial inwud riding of sedimentary strata. The observed polygonal geometry FIg.t. Schematicillustratingadiapiricquasi-Hertzianstressfield results from the outwud rupture and relative brittle strain of as a possible working model for the Vredefort structure: (a) pre­ overlying strata outside the compressive regime (Fig. la). deformation cross section showing the quasi-Hertzian stress field, In conclusion, thc geometric and structural attributes of the (b) postdeformation cross section (the crust-on-edge model), and Vredefortstructureare consonantwith aquasi-Hertzianstress field. (c) postdeformation plan view with superimposed stress field. In particular, it corroborates the many observations of ubiquitous subhorizontal structures that have led investigators to deduce that the Vredefort structure was produced by subhorizontal fon:es (see, This structure consists ofan Archeancoreofca. 45 Ian indiameter. e.g., [8,10». consisting largely of granitic gneiss, surrounded by a collar of References: [IJ Hart R. J. et al.'(1991) Tectonophysics, 182, metasedimentary and metavolcanic supracrustal rocks of the 313-331. [2J Antoine L. A. G. et al. (1990) Tectonophysics,171, Dominian Group. Witwatersrand and Ventersdorp Supergroups. 63-74. [3] NicolaysenL. O. and FurgesonJ. (1990)Tectonophysics, and Transvaal Sequence (forgeological descriptions see, e.g., [I». 171,303-335. [4J Antoine L. A. G. and Reimold W. U. (1988) The interpretation of images of the gravity and magnetic fields Global Catastrophes in Earth History, 2-3, LPI Contrib. No. 673. over Vredefort has permitted the delineation of several important [5] Reimold W. U. and Wallmach T. (1991) S. A. J, Sci., 87. features ofthe structure and ofits immediate environment [2]. The 412-417. [6] LawnB. R. andWillshaw E. (1975)J.Mater.Sci., 10, polygonal, concentric outline of the collar strata is a prominent 1049-1081. [7] Reimold W. U. and Colliston W. P., this volume. feature ofboth the gravity and the magnetic fields. The Vredefort [8] Colliston W. P. and Reimold W. U., this volume. [9] Bahat D. structuresharesthis distinctivegeometrywithotherstructures (e.g., (1980) J. Geol., 88, 271-284. [IOJ Colliston W. P. (1990) Manicouagan, Decaturville, Sierra Madera) of debated impact Tectonophysics,171,115-118. origin. Inall these, successivelyolderstratawithsteepoutwarddips are encountered while traversing inwud to the center of the struc­ ture. A further attribute of these structures is the shortening of the outcrop of a particular stratigraphic unit compared to the original SUDBURY PROJECT (UNIVERSITY OF MirNSTER­ perimeter of that unit. ONTARIOGEOLOGICALSURVEY): (6)ORIGIN OFTHE Toaccountfor the geometricattributes ofthe Vredefortstructure POLYMlCT, ALLOCHTHONOUS BRECCIAS OF THE a mechanical scheme is required where there is radial movement of ONAPING FORMATION. M. E. Avermann, Institute of Plan­ horizontal strata towud, with uplift in. the center ofthe Vredefort etology, Univcrsity ofMiinster, Wilhelm-Klemm-Str. 10, W-4400 structure. Two models can be proposed: (I) one in which there is a MUnster, Germany. rapid rise and violent disruption of cover rocks in response to expansionofa fluid accumulation [3] and (2) one in which there is, The Sudbury structure has been interpreted as a deeply eroded in contrast. a nonexplosive, quasi-Hertzian stress field resulting remnant of a peak-ring basin [I]. The polymict, allochthonous from a diapiric process. Both models can accommodate the geom­ breccias ofthe Onaping Formation (OF) occur in the central partof etry and structural components ofVredefort. The proponents ofthe the Sudbury structure. which is surrounded by the l.85-Ga-old [2J LPI Contribution No. 790 5 .. "Sudbury Igneous Complex" (SIC). From bottom to top the OF can SUDBURY PROJECT (UNIVERSITY OF MONSTER­ be divided into Basal, Gray, Green, and lower and upper Black ONTARIO GEOLOGICAL SURVEY): (I) SUMMARY OF Members (3,4]. The breccias were mapped indetail intheeastrange RESULTS-AN UPDATED IMPACT .MODEL. M. Aver­ l l s l of the structure. The SIC and the lower part of the OF (Basal mann .2, L. Bischoff,P. Brockmeyer .2, D. BubI ,A. Deutsch , B. O. Member) are interpreted as the impact meltsystem [1,3,5; compare Dressler',R. Lakornyl,2, V. MUller-Mohrl .2, andD. StaffierI, Ilnstitut also 6]. fOr Planetologie and 2Qeologisch-Paliontologisches Institut, The Basal Member occurs as a fragment-laden polymict melt­ Universitit Milnster, Wilhelm-Klemm-Str. 10 and Corrensstr. 24, breccia on top ofthe grmophyre ofthe SIC and as isolated bodies W-4400 Milnster, Germany, btitut fOr Geologie. Universitit (formerly called melt bodies [7]) in the Gray and Black Members. Bochum, W-4360 Bochum,Germany, 4()ntario Geological Survey, TheBasalMembercontainsabundantrock fragments thatconsistof 77 Grenville Street, Toronto, Ontario M7A IW4, Canada. metasediments oCtile Huronian Supergroup and minor amounts of Archean crystalline basement. In the upper part of the granophyre In 1984 the Ontario Geological Survey initiated a reseuch a similar fragment population is observed. Therefore, these rocks project on the Sudbury structure (SS) in cooperation with the musthave been present in the target area at the time ofthe impact. University of Milnster. The project included field mapping Geochemical investigations of the main elements and rare earth (1984-1989) and petrographic, chemical, and isotope analyses of elements underline the close genetic relationship ofthe SIC and the the majorstratigraphicunits ofthe SS. Fourdiplomatheses andfour Basal Member and their formation ofcrustal rocks. doctoral theses (Avermann, Brockmeyer, Lakorny, MUller-Mohr) The overlying Gray Member is a breccia unit with a clastic were performed during the project(1984-1992). Specific results of matrix and has a sharp contai;t to the Basal Member. The polymict, the various investigations are reported in five accompanying ab­ allochthonous breccias of the Gray Member are characterized as a stracts [1-5]. As shown in Fig. I of[1], selected areas ofthe SS were suevitic breccia by a high amount ofmelt particles and fragments mapped and sampled: Footwall rocks, FootwallBrecciaandparts of with shockmetamorphic features. Signs ofmultiple brecciationand the sublayer and lower section of the Sudbury Igneous Complex intemal contacts found during the mapping provide evidence for (SIC), Onaping Formation and the upper section of the SIC, and turbulent movements during the emplacementofthe Gray Member Sudbury breccia and adjacent Footwall rocks along extended pro­ [4,8]. Basedonits petrographiccharacter, the lowerpartoftheGray mes up to 55 km from the SIC. All these stratigraphic units of the Memberis interpreted as a ground-surge deposit, which grades into SS had been studied in substantial detail byprevious worlc:ers [6,7J. fall-back breccias. The most important characteristic ofthe previous research is that it The Green Member is considered as a continuous uniform was basedeitheronavolcanic model orOIlamixed volcanic-impact breccia layer on top ofthe Gray Member and comprises the former model for the origin ofthe SS. The presentproject has been clearly "chlorite shard horizon" [7]. This breccia layer is characterized by directedtoward atestofthe impactoriginofthe SS withoutinvoking amicrocrystalline matrix, chloritized"glassy"particles, and ahigh an endogenic componenL In general, our results confmn the most content ofsmall mineral clasts. The Green Member is regarded as widely accepted stratigraphic division [6] ofthe SS. However, our gradational and fme-grained fall-back material, which was affected interpretation ofsome of the major stratigraphic units is different byhigh temperaturesduring thedeposition. The Green Memberand from mostviews expressedin [6]. Thestratigraphyofthe SS and its the chloritized particles originated by early excavation to high new interpretation is given in [3] as a basis for the following atmospheric regimes, condensation out of a vapor phase, and discussion. .deposition with the fmal fall-back material. The main conclusion to be drawn from our results is that (I) the The uppermostunitoftheOF(BlackMember) canbe subdivided 55is the erosionalremnantofapeak-ormultiringimpactbasinwith into a lower and an upper Black Member unit. The lower part an original diameter in the 200- to 240-km range and (2) the SIC is (100-150 m thick) still shows petrographic features of suevitic no endogenic intrusion but rather the main part of an impact melt breccias, small fragments of basement rocks, melt particles, sheet that occupies the central depression ofthe basin and has been chloritizedparticles, andbrecciafragments in adark, clastic matrix. produced by shock-induced total meltinS ofcrustal rocks (8-12]. These signs indicate that the lower unit has been transported from Independently, R. A. F. Grieve ofthe Geological Survey ofCanada its original position outside the crater into the central depression of has come to quite identical conclusions [9]. The individual strati­ the crater. The upward increase of sedimentary features, signs of graphic units or impact-related rocks can be characterized and multiple brecciation in the upperpartofthe Black Member, and the interpreted as follows. gradationalcontaettotheoverlyingOnwatinslatesofthe Whitewater FootwaU Rocks and Related B~: The Archean and Group indicate the subsequent change in the depositional environ­ Proterozoic crystallinebasementoftheSICdisplaysimpact-induced ment. Investigations ofthe carbonaceous material in the matrix of features upto aradial distance ofat least55 km, possibly 80 km [6], the Black Members and the Onwatin slates suggest an origin by from the SIC. At the contact to the SIC itforms a (mega)breccia that biogenetic material deposited at a slow rate lasting into a local, is thermally metamorphosed and partially molten by the SIC euxinic basin, which was created by the Sudbury evenL (Footwall Breccia). This brecciagrades into aweakly shocked (and, References: [I] Swffler D. et al. (1989) Meteoritics. 24, 328. further out, into unshocked) brecciated basement. All basement [2] Krogh T. E. et al. (1984) In Drat. Geol. Sur. Spec. Vol.J (E. G. units contain at least three generations of breccia dikes (Sudbury Pye et al., cds.). [3] Brockmeyer P. (1990) Ph.D. thesis, MUnster, breccias) formed by frictional melting and shearing during the 228 pp. [4] Avermann M. E. (1992) Ph.D. thesis, MUnster, 175 pp. compression, excavation, and modification stages of the crater [5] DeutsehA. etal. (1990) LPSCXX,282-283. [6] FaggartB.E.et formation-a feature typical ofall complex terrestrial craters (e.g., al. (1985)Science.230,436-439. [7] MuirT. L. and PerederyW. V. [13]). (1984) In Drat. Geol. Sur. Spec. Vol. J (E. G. Pye et al., cds.). SIC and Basal Member of the Onaplng Formation: This [8] Avermann M. E. and Brockmeyer P. (1992) Tectonophysics, in unit represents a layered complex ofrocks thatcrystallized from an press. impact melt. It comprises from bottom to top (I) the sublayer, including the offsetdikes, anoritic to quartz-dioritic,clast-richmelt

... 6 1992 Sudbury Conference .. breccia that covers the floor of what is considered the transient cooled and fully crystallized [20J and is therefore not foliated, cavity of the impact basin; (2) the norite, quartz gabbro, and whereas the rocks above and below were deformed (6). It is highly granophyre units that are interpreted as clast-free, differentiated probable that the section of the SIC exposed in the south range is impact melt; and (3) the Basal Memberofthe Onaping Formation, from a deeper position of the impact basin than the section of the a clast-rich melt breccia. The complete melt sequence has the northrange. (3) As theSSis notonly the largestimpactstructureand chemical and isotope signatures of a mixture of upper and lower the only peak- ormultiring basin on Earth but also the largest Cu­ crustal rocks [12,14,15). Ni deposit, ithas gainedconsiderableimportance as a modelfor the Gray, Green, and Black Members or the Onaplng Forma­ study of large impact basins on the tem:strial planets and as a tion: The rocks of these units form a layered sequence of poly­ pathway for exploring the geological processes and the genesis of mict,melt-bearing,clasticmatrixbrecciasshowings1rOng similarities crustaloredeposits intheearlyhistoryofthe Earth. The recognition to suevitic breccias. The lithic fragments of these breccias are ofa nearly 3-km-thick differentiated impact meltsheet at Sudbury derivedmainlyfrom theHuronianSupergroup andmorerarely from has far-reaching consequences for the interpretationofthe >4-b.y.­ the Archean basement The melt particles appear to represent old "plutonic" pristine rocks of the lunar highland crust. whole-rock melts ofa mixture ofsuchbasementrocks [14,15). The Referenc:es: [IJ BischoffL. etal.,this volume. (2) DeutschA. lower unit (Gray Member), which is interpreted as a ground-surge­ etal., this volume. [3J Staffleretal.,this volume. (4) AvermannM., type suevite, is affected by thermal metamorphism induced by the this volume. [5J MWler-Mohr V., this volume. (6) Pye E. G. et al., underlying melt complex. It is topped by a thin layer of melt-rich cds. (1984) TMGeologyandOreDepositsoltheSudburyStructure, clastic fall-back material (Green Member, formerly called the Ministry ofNatural Resources, Toronto, 603 pp. (7) DresslerB. o. chloritized shard horizon) that might have temporarily formed the (1987) In Research in Terrestrial Impact Structures. Earth Evolu­ impact basin floor before the suevitic brecciamaterial ofthe Black tion Series (1. Pohl, ed.), 39, F. Vieweg, Braunschweig. (8) StlSffler Member was deposited by slumping during the gravity-induced D. eta!. (1989)Meteoritics. 24,328. [9J GrieveR.A. F. etal. (1991) modification stage ofthe basin formation [16J. We believe that the JGR, 96, 22753. [10J Lakomy R. (1990) Meteoritics. 25, 195. lower part of the Black Member formed in this way whereas the (l1J Deutsch A. et al. (1989) EPSL. 93, 359. (12) Faggart B. E. upper part was deposited aquatically under euxinic conditions. (1985) Science. 230, 436. [13J Stamer D. et al. (1988) In Deep These conditions are responsible for the carbonaceous matter (de­ Drilling in Crystalline Bedrock. Vol. 1 (A. Boden and K. G. rived from organic material according to the C isotopes) in the Eriksson, cds.), 277,Springer-Verlag, New York. (14) Brockmeyer matrix of these breccias. P. and Deutsch A. (1989) IPSC XX, 113. (15) Deutsch A. et al. The distribution and stratigraphic relation as well as the petro­ (l990)IPSCXXI, 282. [16J Melosh H. J. (1989) Impact CrQleru.,: graphic, chemical, isotope, and shock metamorphic characteristics A Geologic Process, Oxford, New York,245 pp. (17) Milkereit B. of the rock units of the SS are clearly compatible with its interpre­ et al. (1992) Geology, in press. [18J Shanks W. S. andSchwerdtner tation as an impact basin whose transient cavity had a diameter Ole W. M. (1991) Can. J. Earth Sci.• 28, 411,1677. (19) Krogh T. E. in the 100- to 140-km range. These figures are derived from the et al. (1984) In TM Geology and Ore Deposits of tM Sudbury radial extension ofshock effects in quartz, ofshattercones, and of Structure (E. G. Pye et al., cds.), Ministry of Natural Resources, the Sudbury breccias, and from the position of the down-faulted Toronto. (20) Grieve R. A. F. (1992) personal commumication. "megablocks" of Huronian rocks within the Archean basement north of the structure [9J. They translate into a diameter of the apparent basin D. of at least 150 km [9J, more probably of 200 to STRUCTURALASPECTSOFTHEARAGU~A~PACT 240 km [8,IOJ, depending on what empirical relation between Dtc and D. is used. A major uncertainty in determing these dimensions STRUCTURE (BRAZIL). L. Bischoff, P. Brockmeyer, U. is caused by the postimpact deformation of the basin during the Jenchen, and R.-M. Swietlik, Institute of Geology, University of Penokean Orogeny. Recent geophysical and structural data [17,18J MUnster, Germany. have removed previous objections againSt a primary circularity of the SS. The present interpretation of the subbasin structure [l7J A report is given on the results of two months' fi~ld studies indicates a minimum diameter of the outer margin of the SIC of carried out in 1988. During that time the northeast segment was 60 km. Taking the deformation ofthe east range by the Wanapitei mapped and the structural setting ofthe central area was studied in impact structure into account [6J, we believe that this diameter was detail. In addition, the structure ofouter zones of the Araguainha at least65 km although thecurvatureofthe northrange would allow Crater was investigated along three radial sections. a diameter ofup to 80 km. In the latter case, D.could exceed even The Araguainha impact occurred 243 ± 19 Ma ago [lJ under 250km. shallow marine conditions on a mixed target consisting of Devo­ From the present interpretation ofthe SS additional conclusions nian, Carboniferous, and Permian mainly clastic sediments overly­ can be drawn: (1) The depth ofthe transient cavity was in the order ing a Precambrian phyllitic and granitic basement. The intrusion of of30to 40km and the depth ofexcavation near 12 to 20km [8,IOJ. the granite probably took place 449 ± 9 Ma ago [l)..A peripheral According to current models for melt zones in impact craters [16J, ring-fault system, 40 km in diameter, forms the outer boundary of the melt zone at Sudbury probably reached a depth ofabout 30 km, thecomplexstructure (2). Thecentralpartofthestructure, 6-6.5 kID which is the base of the transient cavity [8,10J. This may explain in diameter, consists ofa ring ofsteep hills rising 150 m above the why the Sudbury basin never had a morphologically expressed surrounding plain. The ring is made up oflarge uplifted blocks of central uplift [9J but instead a had a central depression bordered by Devonian sandstone, which surround a central 2.5-3-km-wide a peak ring with a diameterof80 to 90 km. (2) The Sudbury impact depression. The interior of the central depression consists of up­ occurred during an active orogeny (Penokean), 1.85 b.y. ago [19J, lifted alkaligranitic basement. The outer limit of the granite is during which the southeastern part of the basin and peak ring was mostly covered by polymict suevitic breccias forming another ring deformed by thrustfaulting tothenorthwest This deformation took of hills. place while the central part of the melt complex had not yet been LPIContribution No. 790 7

The granite, which is shocked to stage I, often shows in its upper quartz and shatter cones, can be observed up to a radial distance of part in sit"melting, which gradually increases·toward the top. This 9 km from the lowerSIC contactorupto 17 km (shattercones). The partly molten granite is not partofthe allochthonous coherent melt Footwall rocks were subjected to a thennal overprint by the SIC up sheet [3] thatcovered the granite and induced itsmelting. Except for to a distance of 1 km [8,12]. some relics this layer ofmolten target rock was eroded like most of The FB [3,6,10,12,13) is the remnant ofa clastic matrix crater the ground-surge and fall-back breccias. floor breccia. It occurs u irregularly or lenticularly shaped bodies The original bedding ofthe Paleozoicformations is deformedby atthecontactbetween the Footwall and theSIC. FurtheronFB dikes impact-inducedfracturing andfolding. Theconcentric urangement intrude into the brecciated crystalline basement rocks up to 250 m with increasing youngerfonnations from thecentertothemarginof away from theFB layer[10,12,13].The contactsbetweenFB and the the structure give the SllUcture the apparent domelike appearence. underlying Footwall rocks are always sharp, evenon a microscale. This is due to the combinationofa central uplift and the peripheral The fragment contentreflects alocalorigin. Fragments ofFB occur graben sllUcture. But, in fact, most contacts between the different enclosed in the sublayer and norite. The crystalline matrix was formations are of tectonic origin and each ring consists of a formed after localized melting of the originally clutic matrix due multitude of fault blocks. Bedding in these blocks has a steep to the thermal overprint by the SIC. Locally megaclasts (>25 m) dipping in the central ring, buthas moderate to gentle dipping in the adjacent to the country rock are embedded in huge interoalations of outer parts of the structure. Folds are best developed in the sand­ FB matrix. stones ofthe Devonian Furnas Formation of the interior hills. The The SB [5,7,9,11,12,14] occur as various types of clastic or axis ofthe steeply outward plunging open folds scatterremarkably crystalline matrix breccia forming irregular dikes whose orienta­ in different directions. They are sometimes arranged in a complex tions appear not to be controlled by the geometry of the Sudbury fanlike pattern. From the analysis of the fault pattern and the sllUcture. In the north range dikes have been observed up to a geometry ofthe folds, the principal stress components canroughly distance ofatleast S5 km from the lowercontactofthe SIC, butare be estimated with 0, radial to the center. abundant within a zone of 10 km adjacent to the SIC and a second Pebbles of some target rocks are heavily sheared. Shear plane zone located approximately between20 and 33 km north ofthe SIC. analysis may be used to n:conSllUct stress distribution during the The thickness of the dikes ranges from centimeters to several cratering process if the original position of the pebble-containing hundred meters. Most ofthe dikes exhibitsharpcontacts to the wall beds is known. rock. Several phases ofbrecciation canbedistinguished by breccia References: [1] Deutsch A. et al. (1992) Tectonophysics, in fragments in SB, whereas crosscutting ofdifferent breccia types is press. [2] Theilen-Willige B. (1982) Geol. Rdsch., 71, 318-327. extremelyrare. Mostofthe fragments ofthe SB are directly derived [3] Engelhardt W. v. et al. (1992) Meteoritics, in press. from the country rock, but in large dikes relative movements of clasts over several hundred meters can be found. The SB were formed by shearing and friction ofFootwall rocks. The breccias ofthe OF [4,15-17] can bedivided from bottom to SUDBURY PROJECT (UNIVERSITY OF MUNSTER­ top into Basal Member, Gray Member, Green Member, and lower ONTARIOGEOLOGICALSURVEY): (1) FIELDSTUDIES and upper Black Member according to their lithological composi­ 1984-1989--SUMMARY OF RESULTS. L. Bischoff', B. O. tion, matrix,clastpopulation, contacts, andtexture. Thetenn"Basal Dressler. M. E. Avennann',3. P. Brockmeyer',3, R. Lakomyl.3, and Member" comprises a clast-rich melt breccia on top ofthe SIC and V. Mdller-Mohr1,3,'Geologisch-PaliontologischesInstitut,Univer­ the "melt bodies" (18], as they were formerly called, whereas the sityofMOnster,Corrensstr. 24,W-4400MOnster,Germany,2()ntario term "Green Member" implies the former "chlorite shard horizon" Geological Survey, 77 Grenville Street, Toronto, Ontario, M7A [18]. The Black Member was subdivided due to obvious changes lW4, Canada, 'Institut fiir Planetologie, University of MOnster, upward infragment content, size, matrix texture, grade offoliation, WilhelmKlemm-Str. 10, W-4400 MOnster, Germany. and occurrence of both carbonaceous material and sedimentary features. The lower part of the polymict allochthonous breccias of In cooperation between the Ontario Geological Survey (8. O. the OF are interpreted as impact-melt breccias, the overlying Dressler) and the Institute ofGeology and Institute ofPlanetology, breccias as ground-surge breccias, and the uppennost as original or both of Westfilische Wilhelms-Universitit, Miinster, geological, resedimented fall-back breccias [see 1,2,4]. petrological, and geochemical studies were carried out on impact­ References: [1] Staffler D. et al., this volume. [2] Avermann related phenomena ofthe Sudbury sllUcture during the last decade. M. et al., this volume. [3] Deutsch A. et al., this volume. [4] Aver­ The main results of the field studies are briefly reviewed (see also mannM., this volume. [5] Milller-MohrV., thisvolume. [6] Lakomy [1-S)). As shown in Fig. I, Footwall rocks, sublayer, and lower R. (1986) Unpublisheddiplomathesis,MOnster,13S pp. [7] Brock­ sections of the Sudbury Igneous Complex (SIC) were mainly meyer P. (1986) Unpublished diploma thesis, MUnster, 101 pp. mapped and sampled in the northern (Levack Township) and [8] Avennann M. E. (1988) Unpublished diploma thesis, MOnster, western (Trillabelle and SultanaProperties)parts ofthe northrange. 113 pp. [9] Millier-Mohr V. (1988) Unpublished diploma thesis, Within these mapping areas Sudbury Breccias (SB) and Footwall MOnster, 106 pp. (10] Lakomy R. (1989) Ph.D. thesis, MOnster. 165 Breccias (FB) have been studied; SB have also been investigated pp. [11] Millier-Mohr V. (1992) Ph.D. thesis, MOnster. 139 pp. along extended profiles beyond the north and south ranges up to [12] DresslerB.O.(1984)InOnl.Geol.S"rv.Spec.Vol. 1(E.G.Pye SS km from the SIC [S]. TheOnaping Formation (OF) and the upper et al., eds.). [13] Lakomy R. (1990) Meteoritics, 25, 195-207. sectionofthe SIC were studied both in the north range (Morgan and [14] Milller-Mohr(1992) Tet:tonophysics, in press. [1S] Brockmeyer Dowling Townships) and in the southern east range (Capreol and P. (1990) Ph.D. thesis, MUnster, 228 pp. [16] Avermann M. E. Mclennan Townships) (see Fig. I). (1992) Ph.D. thesis, MUnster, 17S pp. [17] Avermann M. E. and The Footwall rocks [6-12] exhibit shock metamorphic phenom­ BrockmeyerP. (l992)Tectonophysics, in press. [18] MuirT. L. and enaofshockstage I, which is characteristicofan impactcraterfloor. Peredery W. V. (1984) In Onl. Geol. S"rv. Spec. Vol. 1 (E. G. Pye Shock metamorphism, including especially planar elements in et al., cds.). 8 1992 Sudbury Conference

Sudbury Structure Superior Province

• Sudbury

Legend 2 Chelmsford Fonnation Areas of investigation Onwatin Formation 1 footwall rocks/SIC 2 Sudbury Breccia Onapiog Formation 3 Footwall Breccia Sudbury Igneous Complex 4 Qnapiog Formation incl. Basal Member ~ Footwall Breccia & Sublayer

Fig. 1. Sketch map of the Sudbury structure showing the areas of investigation.

LARGE METEORITE IMPACTS: mEKIT MODEL. B. F. the putative crater. both in the western interior [4] and on a global Bohor. u.s. Geological Survey. Box 25046. MS 972. Denver CO basis [5]. Geochemically. the fll'Cballlayerhas a basaltic signature 80225. USA. andis mainlycomposedoflaminatedsmectiticclay. Itimmediately overlies the lower layer with a sharp textural contact. This lower The Cretaceousrrertiary (KIT) boWldary eventrepresents prob­ layer. which I have named the "melt ejecta" layer. has a very ably the largest meteorite impact known on Earth. It is the only subdued Ir anomaly, contains only a small amoWlt of shocked impactevent conclusively linked to a worldwide mass extinction. a minerals and no magnesioferrite. has a silicic geochemical signa­ reflection of its gigantic scale and global influence. Until recently. ture. and displays a turbated texture ofWlSOrted. altered. imbricate the impact crater had not been defmitively located and only the shards [6]. vitricclasts [7]. andmicrotektites [8] inamicrosphmilitic distal ejecta of this impact was available for study. However. kaolinite matrix [6). The melt ejecta layer thins radially away from detailed investigations of this ejecta's mineralogy. geochemistry. the putative crater location in the Caribbean region and cannot be microstratigraphy. and textures have allowed its modes ofejection identifiedbeyond -4000 kmofthe crater.Boththefll'Cball andmelt and dispersal to be modeled without benefit of a source crater of ejecta layers contain a similar suite of trace minerals [9). which known size and location [1.2]. argues for a mutual origin from a single impact. . Initially. only KIT boundary sites in marine rocks with a single. The partitioning of these impact components and signatures thin (-3-5 mm) layer ofIr-rich ejecta were known. Subsequently. between the two layers supports a dual-phase model ofejection and nonmarine sites were discovered in the western interior of North dispersal from a single impact [2). Moreover. the distinctive clay America with thicker (-3 em) boWldary claystone Wlits composed minerals fonned from the vitric components in each of these two oftwo distinct layers [3]. Theuppermostofthese two layers. which layers in the western interior lends further support to the dual-phase I have called the "fll'Cball" layer. contains a substantial Iranomaly. model [10). In this model. the fll'Cballlayer represents sedimenta­ skeletal Ni-rich magnesioferrite crystals. and a large population of tion from a radially expanding cloud of vaporized bolide and shocked minerals in the nonclay fraction. The maximwn grain size entrained target material dispersed above the atmosphere. The melt ofthese shocked minerals decreases regularly away from the site of ejectalayerrepresents melted targetrock ejected from thecraterand /.,pI Contribu#on No. 790 9

LARGE METEOROID IMPACT SMALL METEOROID IMPACT ______------Atmosphere scale height

AIRBLAST

" AIRBLAST' ;r'...... -- -Atmosphere scale height

(MOd(liedjro~M~losh"J989J (Mod(lVdftom MrlosPl, 1989J Target Materi~' /''IJ\elt ejecta"

Fig. 1. Vapor plume cloud (fireball) exceeds the scale height of Fig. 2. Vapor plume cloud (frreball) contained and dispersed atmosphere and expands globally above it. within scale height of the atmosphere. transported bothballistically and in a detached ejectacurtain within 54,456-466. [7] Bohor B. F. (1988) Meteoritics, 23, 258-259. the atmosphere [11], forming an areal1y limited, continuous ejecta [8] Bohor B. F. and Betterton W. J. (1991) Meteoritics, 26, 320. blanket. This dual-phase impact model applies only onplanets with [9] Bohor B. F. et al. (1989) Meteoritics, 24, 253. [10] Pollastro atmospheres whenthe size ofthe stabilizedvaporplumeexceeds the R. M. and BohorB. F. (1991) Clay Min. Soc. Prog. andAbst., 129, scale height of the atmosphere (Fig. 1) [12]. In this case, the cloud LPI Contrib. No. 773 [11] Barnouin O. and Schultz P. H. (1992) of vaporized bolide penetrates above the atmosphere and is dis­ LPSC XXIII, 65-66. [12] Melosh H. 1. (1989) Impact Cratering, persed globally, resulting in two-layer deposition within the limits Oxford, New York. [13] Lowe D. R. et al. (1989) Science, 245, of the continuous ejecta blanket. Only a single layer (fIreball) is 959-962. present elsewhere beyond the bounds of the basal melt ejecta blanket. This pattern of ejecta dispersal explains many ofthe early misconceptions about the nature ofthe KIT target rocks and impact site that were based on geochemical analyses of far-fIeld ejecta SHOCKED ZIRCO~§IN THEONAPING FORMATION: comprised solely of the fIreball layer. FURTHER PROOF OF IMPACT· ORIGIN. B. F. Bohor and Distal ejecta from other large terrestrial· impacts should re­ W. J. Betterton, U.S. Geological Survey, Box 25046, MS 972, semble this dual-phase KIT model. Thus, if distal ejecta frpm the Denver CO 80225, USA. Sudbury structure can be located, it would be expected to be . composed of two layers within the extent of the continuous ejecta The Onaping Formation fIlls the structural basin at Sudbury, blanket and single-layered beyond this limit. However, analyses of Ontario, Canada. This formation is composed of three members: a ejectadeposits from Archean impacts in South Africa and Australia basal, coarse, mainly quartzitic breccia (Basal Member), a light­ [13] indicate that most ofthese thick spherulitic beds are comprised colored, heavily included, polymict middle unit (Gray Member), ofbothfIreball (Ni-spinels andhighIr)andmeltejecta(microtektites) and a similar but dark-colored upper unit (Black Member). Two components. This comingling ofcomponents suggests eitherlackof different origins have been proposed for the Onaping: (1) volcanic a signifIcant atmosphere during the late bombardment period, or ash-flow sheet and (2) impact fall-back 'ejecta. These origins are small-scale impacts where the fireball is contained within the critically discussed in a review paper coauthored by proponents of atmosphere (Fig. 2) [12]. Other planets with atmospheres, such as each view [1]. Venus and Mars, may also show a dual-phase distribution ofejecta French [2] identified multiple sets of shock lamellae in quartz for large impacts. and feldspar grains from the Onaping Formation at Sudbury. We Acknowledgments: This work was partially supported by have also identifIed sets of shock lamellae (called planar deforma­ NASA Grant T5115P. W. J. Betterton provided valuable technical tion features, or PDF) in a single quartz grain from a thin section of support. the Black Member. These PDF usually consist of "decorated" References: [1] Bohor B. F. (1990) In GSA Spec. Pap. 247, lamellae that are much less distinct than those in younger impacted 335-342. [2] Bohor B. F and Betterton W. 1. (1992) LPSC XXIll, rocks and ejecta, such as the KIT, because of annealing by subse­ 135-136. [3] Izett G. A. and Bohor B. F. (1986) GSA Abst. with quent metamorphic events. Prog., 19, 644. [4] Bohor B. F. and Betterton W. J. (1991) Meteor­ Because it is more refractory than quartz and feldspar, zircon itics, 26, 321. [5] Bohor B. F. and Izett G. A. (1986) LPSC XVIll, should resistannealing by thermal metamorphism. We havealready 6869. [6] Pollastro R. M. and Pillmore C. L. (1987)J. Sed. Petrol., shown that some zircons from KIT distal ejecta display PDF when 10 J992 Sudbury Conference ..

Fig. 1. PDF (4 sets). Fig. 2. Granular texture. subjected to an alkaline etch [3]. We dissolved samples ofboth the the Onaping show no phase or compositional changes by X-ray Gray and Black Members of the Onaping with acids to free the analyses, in either diffraction or energy-dispersive modes. This is contained zircons, which were then etched in alkaline solutions to another indication that their texture is due to solid-state transforma­ reveal any PDF. These samples were collected from outcrops ofthe tion induced by shock, and not thermal melting. Onaping along Highway 144 south of Levack and west of the High It is instructive that these granular-textured zircons have been Falls on the Onaping River. In the process ofseparating the zircons, found only in ejected material, and not in shocked, in situ target we also recovered otherresistant trace minerals indicative oftarget rocks around craters. However, zircons bearing impact-generated rock lithologies. These include tourmaline, garnet, kyanite, rutile, PDF have been identified from three types of sites and materials: staurolite, chromite, pyrite, and pyrrhotite. K{f distal ejecta, Sudbury fall-back ejecta (Onaping), and Many of the etched zircons from both the Gray and Black Manicouagan target rocks. More importantly, PDF and granular Members display PDF when viewed in an SEM. Figure 1 shows an textures have never been seen in volcanic zircons. Therefore, area of a shocked zircon from the Black Member that displays at zircons can provide corroborating evidence of impact-generated least four sets of PDF. These shock lamellae in zircon are much shock; this discovery could prove very useful in evaluating the narrower than those in quartz and do not etch as deeply, probably metamorphic histories of quartz-poor lunar rocks and meteorites. because they contain less glass within them. Precession X-ray Acknowledgments: We thank B. Dressler for guiding one of photos ofzircons from K{fejecta with PDF show extreme broaden­ us (BFB) to the "type" locality of the Onaping. T. Krogh provided ing and streaking (asterism) ofdiffraction maxima, confirming that the Manicouagan zircons, and E. Foordperformed theX-ray analyses. they have been highly shocked [3]. References: [1] MuirT. L. andPeredery W. V. (1984) In Onto Krogh et a1. [4] reported zircons from the Onaping Formation at Geol. Sur. Spec. Vol.J (E. G. Pye et al., eds.),139-21O. [2] French Sudbury that show linear, crystallographically oriented fracturing. B. M. (1967)Science.J56, 1094-1098. [3] BohorB. F. etal. (1990) They ascribed these features to shock caused by impact, and this Meteoritics, 25, 350. [4] Krogh T. E. et al. (1984) In Onto Geol.Sur. conclusion is supported by U-Ph data. The discordance of these Spec. Vol.J (E.G. Pyeetal.,eds.),431-446. [5] ElGoresy A. (1968) zircons is crudely proportional to the amount of fracturing they In Shock Metamorphism ofNatural Materials (B. M. French and display, caused by the Sudbury event dated from a lower intercept N. M. Short, eds.), 531-553, Mono, Baltimore. age of -1836 Ma. We have also observed both linear and irregular fractures in our Onaping zircons, but the finer-scale PDF probably indicate a significantly higher level of shock than do these coarse, SUEVITE SUPERPOSITION ON THE BUNTE BRECCIA IN open features. NORDLINGER RIES/GERMANY: NEW FINDINGS ON In addition to fractures and PDF, Onaping zircons also show THE TRANSPORT MECHANISM OF IMPACTITES, D. another type of textural feature that indicates exposure to a high Bringemeier, Marie-Hedwig-Str. 15, W-3392 Clausthal-Z., Ger­ level of shock. We have called this texture, first noted in zircons many. from K{f ejecta [3], "granular" or polycrystalline (Fig. 2). Often, zircons displaying this texture are idiomorphic with some of the Research undertaken in the last decades in Nordlinger Ries, original crystal surfaces still visible. This indicates that the zircons Germany, has repeatedly emphasized the sharp contact between have not melted, but instead have been recrystallized due to shock. Bunte breccia and suevite. However, extensive investigations into Thus, these granularzircon grains canbeconsidered to bediaplectic­ this layer boundary have not yet been possible due to insufficient that is, shock-convertedby solid-state transformationinto polycrys­ outcrop ratios. talline zircon below their fusion temperature. On the other hand, New outcrops enabled an in-depth investigation into the zircons also can be melted by impacts, as shown by fused grains superposition of suevite on the Bunte breccia, which is assigned a partially or completely converted to baddeleyite in high-tempera­ key role in interpreting the transport mechanisms ofejecta oflarge ture impact glasses [5]. The granular zircons in K{f ejecta and from impact. LPI Contriblllitm No. 790 11

In two quarries lying several kilometers east and south-south­ annealing by the overlying hot impact melt sheet (SIC) at tempera­ west of the crater (Otting, Aufhausen/Seelbronn), the contact be­ tures >1000oC resulted in melting of the fme crushed material, tween the suevite and Bunte breccia was recorded in detailed followed by an episode of metasomatic K-feldspar growth and, sections on outcrops ofover 50 m in length. fmally, formation of low-grade minerals such as actinolite and Itwas possible to confum studies made in the 19605 by Wagner chlorite. Figures 1 and 2 show that the Rb-Sr method on thin slabs (1) that suggested a divisionofthe suevite into main suevite, rich in and mineral fractions clearly canseparatethese events,whereas the pancake bombs (also called "flidle"), and a relatively well-sorted, Sm-Nd system apparently did not respond to either the thermal or thin-base suevite consisting of fme gravel. A semiquantitative the "late"metamorphic episode [7,8]. This is due to the fact that the analysis of the just slightly consolidated base suevite revealed the highly refractory accessory phases that are the main REB carriers main constituent to be "fresh," bubble-abundant. albeit sometimes survived the melting in part [10). bubble-deficient, angular glasses. Secondary crystalline and sedi­ Isotope relationships in the Onaping breccias (Gray and Green mentary rock clasts and very rarely "flidle"were detected. Signifi­ Member) are much more complex ([10-14); compare also Fig. 1in cant to the transport mechanism ofthe base suevite is its content of (31). All attempts to date the breccia formation failed: Zircons are Bunte breccia fragments and the discovery of shell fragments. entirelyderived from countryrocks andlack the pronounced Pb loss Between the base suevite and the Bunte breccia is a crystalline caused by the heat of the slowly cooling impact melt sheet (SIC), breccia ofca. 0.1 m in thickness that is separated from the Bunte which is documented inthe crater basement [15J. Rb-Sr techniques brecciaby asharpboundary.Insome areas atransitionbedis visible usingeitherlithic fragments ofdifferentshock stages [12),carefully between the crystalline breccia and the base suevite. This transition separated,now recrystallizedmeltparticles and meltmatrices orthe bedindicates an erosivereworking ofcrystallinebrecciaby thebase thin slab method [13,14) just set time limits for the apparently suevite. In one of the sections (Aufhausen/Seelbronn) the base pervasive alkali mobility in these suevitic breccias. The data array suevite was not observed, as the main suevite lay either on the crystalline breccia or directly on the Bunte breccia. The crystalline breccia is highly altered and in the transition to main suevite contains disintegrated glasses. The SudbUry Event Inbothsections slrUCtures were established thatcanbeexplained SIldbury "-II Breccia only by an erosive reworking of the subsoil caused by a shifting Sfl.fB.658 whole lOCk IlIin slabs viscous suevite flow. Particulary on the flanks ofthe hummocks of all thin slabs Bunte breccia, lying several meters higher, the layers below the T- 1.68 :t .08 Ga main suevite have been plained, compressed, and mixed by the 108 .70399 [17] 4 .. - suevite flow. In the Aufhausen/Seelbronn section hook-shaped, Is. - matrix only decimeter-sized, fingerlike compressions of Bunte breccia and T- 1.825 :t .021 Ga crystalline breccia project into the main suevite. A clear erosive Is. - .70356 [5J discordance between the main suevite and the base suevite is visible in the Otting section. 704 •• Reference: (1) Wagner G. H. (1965) Jh. Oeol. LandeSamI Baden-Wilrttemberg, Vol. 7,199-222.

Fig. 1. Rb-Sr isochron diagram for thin slabs of the Footwall SUDBURY PROJECT (UNIVERSITY OF MUNSTER­ Breccia (northofStathconamine, Levack Township, North Range). ONTARIO GEOLOGICAL SURVEY): (7) Sr-Nd IN BET­ Thin slabs ofthe quartzdioritic (= melt) matrix yield the age ofthe EROUTHlCBRECClASANDGABBROICDIKES. O.Buhll , impact. A. Deutseh2, R. Lakomy2, P. Brockmeyer2, and B. Dressler', Ilnst. f. Geologie, RU Bochum, Postfach 102148,0-4630 Bochum I, Germany, 2Inst. f. Planetologie, Univ. MOnster, Wilhelm-Klemm­ I I i Plnoleean Orogeny I Str. 10, 0-4400MOnster, Gennany, 'OntarioGeological Survey, 77 Sudbufy FclcIhNI IIIwcl:ia Grenville Street. Toronto, Ontario M7A 1W4, Canada. SR-f8.4li8 minerWI 130 ClI*1ZdiOritic • QIWllIlIllyric One major objective ofour Sudbury project [1-6] was to defme oWR 15 *WR 18 origin and age of the huge breccia units below and above the T- 1.87 :t .01 GIl M'" Sudbury Igneous Complex (SIC), i.e., the Footwall Breccia (FB) Is, - .70390 [35] .. ,< ~ ~ lite Even" and the Onaping Formation, which caps the SIC. For the terminol­ .' ~ I ..- , WR15·1M1 ogy used here we refer to the companion abstracts in this volume . , ./:.. T- 1.430 :t .015 GIl [1,3); pelrographic descriptions of the FB and of the Onaping 110 ... breccias are found in [3,5,7-10J. .'" The heterolithic FB, which is up to 150 m thick. represents apart of the uplifted crater floor [7,8J. It contains subrmUlded fragments up to several meters in size and lithic fragments with shock features 8 (>10 GPa) embedded into a fUle- to medium-grained matrix [8,9]. £Nd-£Sr relationships point to almost exclusively parautochthonous Fig. 2. Same samples as in Fig. 1. K-spar-rich granophyric matrix precursorlithologies (see Fig. 1 in (31). The different textures of the date a metasomatic episode; mineral fractions defme a late, low­ matrix reflectthe metamorphic history ofthe breccialayer: Thennal grade event. 12 1992 Sudbury Conference

drastically.Therefore, the Onaping brec:cias giveonlyagelimits for alteration and low-grade metamorphism. The Sm-Nd system was notresetduring the Sudbury event; clasts as well as thematrix in the FB and in the Onaping breccias show preimpact "Archean" Nd isotope signatures. References: [1] AvermannM. etal., this volume. [2] Bischoff L. etal., this volume. [3] StafflerD. etal., this volume. [4] Deutsch A. et al., this volume. [5] Avermann M., this volume. [6] Mil11er­ Mohr V., this volume. [7] Deutsch A. et al. (1989) EPSL, 93, 359. [8] LaromyR. (1990)Meteoritics, 25, 195.[9] DresslerB.O.(1984) In The Geology and Ore Deposits o/theSudbury StrllCtlU'e (E. G. I. Pye etal.,eds.),97-136,Toronto. [10] BrockmeyerP. (1990) Ph.D. thesis, Univ. MUnster, 228 pp. [11] DeutschA. (1990) Habil-Schrift FB Geowiss Univ Mamiter. [12] Fullagar P. D.etal. (1971) CQII.J. Earth Sci., 8,435. [13] BrockmeyerP. and Deutsch A. (1989)LPSC XX, 113. [14] Deutsch A. et al. (1990) LPSC XXI, 282. [15] Krogh T. E. et al. (1984)InThe Geology andOre Deposits oftlte SlIdOury '.1 u· StrllCtlU'e (E. G. Pyeetal.,eds.),431-446,Toronto. [16] ShawH. F. and Wasserburg G. J. (1982) EPSL, 60, 155. [17] Hurst R. W. and Fig. 3. TUr vs. l/fRb diagram after [16] for Onaping breccias and Fahart 1. (1977) GCA, 41, 1803. [18] Gibbins W. A. and McNutt whole-rock samples from the granophyre (SIC). Data somces R. H. (1975)Can.J. Earth Sci., 12,1970.[19] DresslerB. 0.(1984) [10,12-14,17,18]. In The Geology and Ore DeposUs oftlte Sudbury Stf'llCtlU'e (E. G. Pyeetal.,eds.),57-82,Toronto. [20] DePaoloD. J. (1981)JGR,86, 10470. and the intercept in Turvs. 1/fRb plots [16] point to a major Rb-Sr fractionation aro\Uld 1.54 Ga ago (Fig. 3). This model age is in the same range as 1.63 ± 0.07 Ga obtained for the metasomatic matrix ofthe FB (Fig. 2). Ithas tobepointedoutthatRb-Srwhole-rockdata HONG KONG IS AN IMPACT CRATER: PROOF FROM for the granophyre ofthe SIC [17,18] also show this event (Fig. 3). THE GEOMORPHOLOGICAL AND GEOLOGICAL whereas the norite behaved as a closed system during this phase. EVIDENCE. Chu-Iok Chanl , Wu Siben1, and Luo Xiuquan1, Preliminary Sm-Ndresults for gabbroic dikes oftholeitic com­ IHong Kong AmateurAstronomical Society,Hong Kong, 2Institute position that intrude Chelmsford wackes [19] are listed in Table 1. of Mineral Deposits, Chinese Academy of Geological Sciences, They have an internal isochron age of 1648 ± 103 Ma (20) with China. £Nl- 16S0aof -5.4.The£-value,theirmodel ageT DM relative to the depleted mantle [20] of -2.7 Ga, and the intrusion age point to a HongKongis a famous cityinsouthemChina,22°19'N, 114°10'E, remobilization of old crust during the Penokean orogeny, but a within which the urban districts of Hong Kong, Kowloon, and mantle magma heavily contaminated with crustal material may be Victoria Harbour are situated. Hong Kong is surro\Ulded by also consistent with the data. These dikes apparently have no mountains with a diameter of 11 km. Three million people live connection with the SIC and postdate the impact event. inside the basin. Rb-Sr dating of a shock event in impact-related breccias seems The round structure of the mountains in Hong Kong has been to be possible only iftheir matrix had suffered total melting by the described as a granite dome thatis deeply eroded (batholith). In this hot melt sheet (FB) orifthey contain a high fraction ofimpact melt paper, thecircularityofthemountains,theexistenceofa centralhill, (suevitic Onaping breccias), whereas thedegreeofshockmetamor­ the inner slope ofthe mountains being greater than the outer slope, phism in rock or lithic fragments plays a minor role [11]. In the the presence ofdeep layer rock inside the basin, and the depth-to­ Sudbury case, however, the impact melt in the suevitic breccias is diameter ratio have been studied. All this evidence shows that the devitrified and recrystallized, which changed Rb/Sr ratios quite Hong Kong structure satisfiesthe geomorphologicalrequirementof an impact crater. SomeshockmetamorphicphenomenaoftherocksinHong Kong suchas planarfeatures, microspheriliticsilicaglassOechaterlierite), TABLE 1. Sm-Nd analytical results for gabbroic dykcs. fused marginsofrockfragments, concussionfrllCtures, impactglass Sudbury stNeturC. in which some schlierens are consistent with pyroxene spiculites. etc., were ftrSt discoveredin October 1990.In Hong KongIsland, an Sm Nd I4.'lNdJl44Nd· (ppm) (ppm) I47Smj144Nd ±20 impact melt sheet has been observed from the Victoria Peak to the southern shore. Quenching fractures of quartz in Kowloon fme­ WR 1.692 6.755 0.15139 0.511872 ± 16 grained granite has also been discovered. coarse-grained In ourwork, the K-Arage (83.34 +1.26m.y.)ofthe impactmelt plag 180-250mm 1.009 5.302 0.11501 0.511464 ± 10 rock, which is younger incomparison to the K-Arage (117 m.y.) in 9.490 0.15197 0.511860±7 mafics 180-2S0mm 2.386 Hong Kong and Kowloon granite, has been measured, and the WR 2.619 11.86 0.13311 0.511687 ± 23 phenomena indicate that after the granite body formed, there was .Normalized to I46NdJl44Nd = 0.7219: 20 erron refer to thc last significant anothergeologicevent. Maybe itis the Hong Kong impactcratering digits. event. LPJ ConIribution No. 790 13

MELT PRODUCTION IN LARGE-SCALE IMPACT TABLE 1. Impact-melt volumes at terrestrial enters. EVENTS: CALCULATIONS OF IMPACT-MELT VOL­ ~ Melt Volume UMES AND CRATER SCALING. Mark J. Cintalal and Rich­ Crater (km) (kmJ) ard A. F. Grieve2• ICode SN4. NASA Johnson Space Center, HoustonTX77058.USA,7QeophysicsDivision,OeologicalSUlVey Brent 3.8 2x1(}2 ofCanada. Ottawa. Ontario KIA OY3, Canada. Zapadnaya 4.5 1x1(}1 lI'nets 8 7x1(}1 Along withan apparentconvergenceinestimatesofimpact-melt Kaluga 1S 8 7xl(}24' volumes [1-3] produced during planetary impact events. intensive Logoislt 17 Lappajarvi 17 8 efforts at deriving scaling relationships for crater dimensions have Ries 24 2xl(}l* alsobeenyieldingresults [4]. Itisnowpossibletoexamine a variety Boltysh 2S 11 ofphenomenaassociatedwithimpact-meltproductionduring large Mistutin 28 20 cratering events and apply them toplanetary problems. This contri­ W. Clearwater 32 80 butiondescribesamethodofcombiningcalculationsofimpact-melt Kara 6S 480 production with crater scaling to investigate the relationships be­ Manicouagan 100 1200 tween the two. Popigai 100 1750 CalculationsofImpad-MeltVolumes: Thisstudyusesmelt­ Sudbury 200 80001 volume calculations that treatvertical impacts into a flat target; the *Values are low probably due to mixed nablre of target and presence of projectile and target are described by a modified, material-specific volatiles in sedimentary strata [9J. Murnaghan equation of state. the details of which have been 'Assumes Sudbury Igneous Complex and part ofoverlying Onaping Fonna­ described elsewhere [3,5]. It does not use the "constant-energy lion are a coherent melt sheet [10J. shell" assumption of Charters and Summers [6] and Gault and Heitowit [7], and yields a closer approximation to the results of more complex, fmite-difference models [1,2]. In an attempt to MeltVolumesand CavityScaling: Even though itis far from approximate the off-axis stress decay evident in more detailed complete. the terrestrial dataset is the best one available for the models of the impact process [1,2.8], the particle velocity in the determination of impact-melt volumes associated with large cra­ target behind the shock front is assumed to vary as cosll 9, where ~ ters. Values for these volumes forimpacts intocrystallinerockhave is·the ratio of target to projectile compression, and 9 is the angle beenestimatedfrominfonnationculledfromtheliterature(Table 1), from the axis of penetration (equivalent to the direction of the and arepresentedin Fig. 2 against transient-cavity diameters, which velocity vector) to the point of interest, measured at the center of were derived from the observed crater diameters with Croft's [11] flow in the target. Energy is added to the target until the projectile modification scaling relationship. In order to extrapolate from the decompresses. whereupon the detached shock front is treated as a terrestrial data to planetary events, provision must be made for the thin region whose intensity decays due to entropy production and effects of variations in gravity. impact velocity. and target and geometric effects. The results ofthis procedure incalculating melt projectileproperties. Calculations were perfonnedforthe impactof volumes for the impact of chondrite (simulated by an artificially chondritic projectiles into Hardhat granite, covering a velocity dense basalt) into Hardhat granite and Tahawus anorthosite are range of 10 to 50km S-I; chondrites are considered to be represen­ compared inFig. I to those for anorthosite into anorthosite, using a tative ofsilicate impactors, and the velocity range falls within that fmite-difference model [1]. allowable for the Earth. The scaling relationship as given by

'10 8 Granite Target CD 1000 50kmII ..... ".- E 10 6 Chondrite Projectife ::J CO)- ~ 100 E : •. SudbcHy ..¥: 10· ...... popigtli CD '-' 55 Q) 2 CJ 10 LapPllillrvl '. ." ..··Kafa CD 10 E ". , Mlslaslln e- ::J Kaluga ...... •····.. BOIfysh o. ~ z.pedn:'~~ .~~:::::.:.~~isk Ql ..·'·::·."· 0 E ..... 10.2 ·····-Brenl ::J G) ~ a..r..o.n OIGnolln MoM ::2: 0.1 o i ~ -{)- -tr--o- 10-4 Schmidt Scaling (Hard rock) CD ~ ···0·······6···· ····0·· Terrestrial Gravity ~ TrhI --(}-. .-/:)-. --0-. J: Ii 10.a a. 0.01 0.1 1 10 100 1000 10 100 1 Transient-Cavity Diameter (km) Impact Velocity (km 8. )

Fig. 2. Comparison between model calculations and actual Fig. 1. Volumes ofpuremelt andpure vaporproduced for impacts terrestrial datafor craters incrystalline rock. Note thedisplacement of chondrites into granite and anorthosite and for anorthosite into between the three cUlVes and the observed values. Because they anorthosite [1]. Thechondritecalculations were perfonned with the were fonned in mixed targets, the Ries and Logoisk structures were model described above. not included in the analyses. 14 1992 Sudbury Conference

Schmidt and Housen [4] was then used to estimate transient-cavity TABLE 2. P1l11et-specifie constants for use in equation (3). diameters fora rangeofprojectilesizes. Sincethe meltvolumes are originally expressed in lenns of impactor volume, it is a simple Planet matter to relate the volume ofimpactmelt to the dimensions ofthe Mercury 10.0 2.61xlO-2 transient cavity as given by the scaling relationship [4]. Three such Venus 4.0 (aslUDled) 2.S4xl()-2 curves arealsoincluded inFig. 2.Itis immediately apparent that the Earth 4.0 2.24xl()-2 theoretical curves overestimate the quantity of melt by, in some Moon 10.9 3.38xl()-2 cases,morethananorderofmagnitude. Webelieve thatthisreflects Man 3.1 3.39xIQ-2 an underestimate ofthe transient-cavity dimensions rather than an overestimate ofmeltproduction for the following reasons: (I) The scaling relationship used for cavity dimensions was formulated Naval Ordnance Laboratory, 200. [8] Thomsen J. M. et al. (1980) partlyonthebasisofthefinaldimensionsofcraters formed insand, Proc.LPSC10th, 2741. [9] KiefferS. W. andSimonds C. H. (1980) whichalmostcertainlyrepresentadjustedtransientcavities. (2) The Rev. Geophys. Sp(l(;e Phys., 18, 143. [10] Grieve R. A. F. et al, melt-volume estimates are accurate to within a factor of 2, with (199I)JGR, 96,753. [11] CroftS. K. (1985)Proc.LPSC 15th,828. underestimates being equally likely as overestimates, and thus [12] Shoemaker E. M. (1977) In lmp(l(;t and Explosion Craterillg cannot account for the differences. (3) Meltejection could account (D. J. Roddy et al., cds.), 617, Pergamon, New York. forremoval ofupto50%ofthetotalproducedatthesmallestcraters, but will have a vanishingly small effect in the cases of the largest craters in the figure. (4) The melt volumes calculated here, as MELT PRODUCTION IN LARGE-SCALE IMPACT evidenced by Fig. I, are in good agreement with those detennined EVENTS: PLANETARYOBSERVAnONSANDIMPUCA­ from the more complex models. Lacking a detailed physical basis nONS. Mark: J. Cintalal and Richard A. F. Grievez, ICode SN4, for changing the scaling relationship-which, it must be empha­ NASAJohnsonSpace Center,HoustonTX 77058,USA, Z

size increase in theorderMoon, Mars, Mercury, Earth, and Venus; "ChondrUe" Projectile venu. for thepurposeofillustration,weconcentratehereontheendmember ot " Mercury cases, Venus and the Moon. Forexample, an impact eventcreating >­ a l00-km crater on the Moon also results in about 1200 km' of 5"" impact melt, compared to about 3500 km' in a comparably sized ~ crateron Venus. Because the bulk of the impact melt inside large, .~ incrystallinetargetsoccursascoherentmeltsheets, c: .' complexcraters EarIh "'-, " this disparity means that features such as visible central structures . ,," Mars' •.../ andfloorroughness,whichreflectparautoehthonoustargetmaterial i'5l Moon' ofthetruecraterfloorandwalls,willbeless buriedbymeltandmost :::!: prominent in hmar craters compared to craters on the other terres­ '0 trial planets. .c Ithas become something of a tenet that the depth ofcraters and Io 0.1 L.-__-'-__...... _ ...... 'OW'...... _~ the diameters at which they undergo morphological transitions are 0.1 1 10 100 1000 inverse functions of planetary gravity [7,8]. Although there is a Final Crater Diameter (km) general progression toward increasing depth with decreasing grav­ ity,thedatashowconsiderablevarianceandthe inverserelationship Fig.2. Depthofmeltingrelative to the transientcavity'sdepth as isnotstrict[9]. Theeffectsofgravityindeterminingcraterdepth are a function of fmal crater diameter. Similar values appear on the generally assumed to be in limiting and maintaining topography vertical axis for the diameters -at which maxima occur in the during and after cavity modification [8]. This work indicates frequency ofpeak-ring basins. another potential role for gravity through its effect on relative impact-melt volumes. Ifrectangular cross sections were assumed, for instance, our standard lOO-km lunar crater would have an interiormeltsheetabout 125 m thick, compared tooneabout450m cavity modification and uplift, this material will be unable to fonn thick for its venusian counterpart. This disparity is a minimum, as a coherent central peak. The zone of melting will approach and the lower lunar gravity will favor relatively more ejection ofmelt. ultimately intersectthebaseofthe transientcavity (Fig. 2), at which Current estimates of the ratio of the apparent depth/diameter point the baseofthecavity will havenostrength. Theonlypossible relationships [7,10] forcomplex craters onthe Moon and Venus are central topographic fonn insuch a caseis an interiorring. When the between4 and 6, compared to a Ilg ratio for the two planets of5.5. diameter at which maximum development ofpeak rings in craters Taking the melt volumes into account and calculating true depths on the terrestrial planets is considered, and the corresponding (i.e., depth to the true crater floor) gives a ratio of true depth/ relative depths of melting are determined, they defme a very diameterof-2.5-3. This suggests that the called-for 1/grelation for restricted set ofmelting depths (Fig. 2). While we will not defend apparentdepths maybe fortuitous and thatgravity andotherplanet­ these actual values, we do suggest that the small range over which specificproperties (such as impact velocity andphysical properties they occur(Fig. 2) couldbeconsideredasa validationofthe general of the target, including the presence ofvolatiles on planets such as concept that impact melt could have a role in peak-ring fonnation. Mars), might playa complex and sometimes competing role. Whileonecould argue thatthis couldalsobe fortuitous, wenote the Crater and Basin Morphology: Previously, we have sug­ additional prediction ofthismodel that, ascraterdiameterincreases gested thattheincreaseddepthofmelting withincreasingcratersize and progressively more of the transient-cavity floor is melted, the will result in weakening ofthe baseofthe transientcavity [2,3]. On diameter of the inner (peak) ring will increase. Although not emphasized in the literature, this has been noted for a relatively small setof12lunarbasins, where thepeak-ring diameter/rim-crest Q) diameter ranges from 0.45 to 0.56, and is generally inversely E ::::I "Chondrits" Projectile venus correlated with size [11]. This changein relative ring diameterwith Metr;ury size, from 0.28 to0.67 [12],is moreobvious from a largermercurian ~ •.... ~ 0.1 ....., .....• datasetof45craters.Ithasmostrecently been showntooccuronthe '5 .' .' (1J basis of 16 venusian craters [13] changing from 0.29 to 0.67, with

~ " ... ~,' ~.' even smaller ratios occurring at the transition from central peak to .i! 0.01 peak-ring craters. c Weofferthe impact-meltingmechanismofpeak-ring fonnation (1J ...... ". ." " ,," as an alternative, atleast supplemental, explanation to the previous t:: Earth ~ 0.001 hypothesis that invoked the collapse of overbeightened central ::::I Mars ,." peaks [14]. Even ifit accounts for the fonnation ofrings, this latter ~... Moon hypothesis does not specifically allow for the disappearance of ~ 0.0001 L.-~...... L..~...... ~~...... '----'...... central peaks with increasing crater size. The model here also 0.1 1 10 100 1000 maintains functionality withvariations inplanetary gravity through Final Crater Diameter (km) the scaling ofcraterdimensions. While atfust glanceitmight seem to be at variance with othermechanisms suggested for ring fonna­ tion [15], itis an alternativemethod ofreducingrockstrength in and Fig. 1. Volume of impact melt relative to that of the transient around the transient cavity, as required in someofthose models. In cavity as a function of fmal crater diameter. Generated for the addition, wholesale, deep-seated melting in truly large basin-fonn­ conditions listed in Table I, a separated curve is presented for each ing events will result in inward flow during modification, possibly planet. leading to circumferential faulting and outer-ring fonnation. Thus, 16 1992 Sudbury Confereru:e • itmightreducetheneedforsuchweakening mechanisms as acoustic The effectofthe "Vredefortevent" is demonstrably large and is fluidization [16], and might provide a substitute for the "asthe­ evident within a northerly arc of about 100 km radius around the nospheric" flow required in these models ofbasin-forming events. granitic coreofthe structure. Northerly asymmetric overturning of A few centralupliftedstrueturesincomplexcratersareaamtrally the strata is observed within the first 17 km (strata is horizontal in located, which has been ascribed to preimpact structural control the south), followed by a 4O-km-wide rim synclinorium. Fold and [17]. Oblique impact. however, can produce asymmetric melt fault structures (normal, reverse. and strike-slip) arc locally as well zones, withincreased melting inthedirection ofimpact [18]. Thus, as regionally concentrically arranged with respect to the northern asymmetricimpactmelting followedbyupliftmaybeanalternative and western sides ofthe structure. mechanism offormation ofacentral peaks. Theunusualeategmyofbrittledeformation,theso-called"shock Impact Lithologies: There will be second-order differences deformation,"observed inthe collarstratahas attracted worldwide in the impact lithologies at comparable-sized craters on the terres­ attentionoverthe past twodecades. These deformationphenomena trialplanetsbecause oftheeffectofgravityonscalingrelations. For include the presence of coesite and stishovite, mylonites and example, the levels ofrecorded shock in uplifted central structures pseudotachylites, cataclasis ata microscopic scale. and the ubiqui­ and the ratio between melted and clastic material will be lower in tous developmentofmultiply striatedjointsurfaces (whichinclude lunar craters than in those onotherterrestrial planets, otherparam­ "shattercones," orthogonal. curviplanar, and conjugate fractures). eters being equal. These potential differences must be considered The macroscopic to microscopic deformation features have led when interpreting remote-sensing data [19]. Similarly, the various tothe formulationofvarious hypotheses toaccountfortheoriginof proportions ofimpactlithologies and their second-ordercharacter­ the Vredefort structure: (1) tectonic hypotheses: deep crustal shear istics will vary with the size of the event. At larger impact events, model [1], doming and N-directed thrust fault model [2], fold for instance, there will be less clastic debris available within the interference model [3], and diapir model [4]; (2) the exogenous transientcavity forincorporationintothemeltSuchimplications of bolide impact hypothesis [e.g., 5,6]; and (3) the endogenous differentialmelting anderateringhavebeenused toexplainsomeof cryptoexplosion model [7]. the observations at large tell'Cstrial impact melt sheets such as at Ongoing structural studies on the dome [8] have aided in Sudbury [20]. Similar arguments apply to lunar samples. The lack narrowing the field of possible hypotheses. The subvertical faults ofclasts is therefore an insufficient single condition to rule out an and shears associated withdiapirs oranendogenic cryptoexplosion impact-melt origin for relatively coarse-grained, igneous-textured couldnotbeidentifiedineitherthebasementorthecoUarrocks.The rocks in the samples from the lunar highlands. subverticalconjugatenorthwest- andnortheast-trendingshearzones References: [1] CintalaM. J. (l992)JGR, 97, 947. [2] Grieve that occur in the migmatitic basement predate the extrusion of the R. A. F. and Cintala M. J., this volume. [3] Grieve R. A. F. and ca. 3.07-Oa-old Dominion Group volcanics. Toward the southern Cintala M. J. (1992) Meteoritics. submitted. [4] Strom R. G. and extremity ofthe structure,subhorizontal gneissic fabrics. which arc Neukum G. (1988) In Mercury (p. Vilas et al., eds.), 336, Univ. deformed by the subvertical shears, become more prominent The Ariz., Tucson. [5] Schmidt R. M. and Hausen K. L. (1987) IN. J. majority of the macrostructural deformation (faulting, folding) in Impact Eng., 5, 543. [6] Croft S. K. (1985) Proc. IPSC 15th, 828. the collar is related to the Vredefort event, and the remainder to [7] Pike R. J. (1980) Icarus. 43, 1. [8] Melosh H. J. (1989) Impact reactivation of pre-Vredefort structures. Pseudotaehylite occur­ Cratering: A GeologicProcess, Oxford, 245 pp. [9] CintalaM. J. et rence is not exclusive to the Vredefort structure and is found al. (1977)Proc.IPSC8th,3409. [10] GarvinJ.B. andSchaberG. G. throughout the northern and northwestern Witwatersrand Basin. (1992) IPSC XXlll, 399. [11] Head J. W. (1977) In Impact and Several pseudotaehylite generations were produced over a wide Explosion Cratering (D. J. Roddy et al., cds.), 563, Pergamon, New interval from 2.2 to 1.1 Ga (pre- to poJt-Vredefort event) [9]. This York. [12] McKinnon W. B. (1981) In Multi-Ring Basins, 259, suggests the regional occurrence of episodic brittle deformation Pergamon, New York. [13] Alexopoulos 1. S. and McKinnonW. B. events with associated high-strain intensities. (1981)JGR. inpress. [14] Grieve R. A. F. etal. (1981)InMulti-Ring It has been identified that the multiply striated joint surfaces Basins, 37,Pergamon,New York. [15] MeloshH.J.(1982)JGR.87, postdatetheoverturning andrelatedfaulting inthestructure,aswell 371. [16] Melosh H. J. (1979) JGR. 87, 7513. [17] Schultz P. H. as a phase ofpostoverturning pseudotachylite development. These (1972) Moon Morphology, Univ. Texas, Austin, 626 pp. (18] observations do not conform to the generalizations prOposed by O'KeefeJ.D. andAhrensT.J.(1986)Scieru:e.234, 346. [19] Pieters otherworkers who assume a horizontal stratigraphy prior to shatter C.M.(1982)Scieru:e, 215,59. [20] GrieveR. A. F.etal.(1991)JGR. conedevelopmentbyanimpact-generatedshockwave[e.g., 10,11]. 96,753. It also places doubt on the validity of using shatter cones as a diagnosticcriterion for impactstructures. Although the presenceof coesite and stishovite cannot yet be fully explained, it is suggested STRUCTURAL REVIEW OF THE VREDEFORT DOME. that these high-pressure polymorphs and multiply striated joint W. P. Collistonl and W. U. Reimold~, IDepartment of Geology, surfaces may also be produced in a tectonic regime by Mohr­ University of the Orange Free State, P.O. Box 339, Bloomfontein Coulomb fracture within varying local stress fields. . 9300, South Africa, ~Economic Geology Research Unit at the According to regional gravity and aeromagnetic data the domal DepartmentofGeology•UniversityoftheWitwatersrand.P.O.Wits structure is interpreted to be located at the intersection of a north­ 2050, Johannesburg, South Africa. west-trending anticlinalarch(whichuplifts lowererust)and anortb/ northwest-axisofcrustal downwarp (corresponding to the long axis The structure of the 0Ider-than-3.2-Ga Archean basement and of the Witwatersrand Basin) [12]. Reflection seismic data along a Archean-to-Precambrian sedimentary/volcanic rocks (3.07 to ca. line roughly parallel to the northwest-anticlinal arch confmns 2.2 Ga) inthe centeroftheWitwatersrandBasintothe southwest of regional structural data and interpretations of the structure [13,1]: Johannesburg (South Africa) is dominated by the ca. 2.0-Oa mega­ The deep structure in the basement reveals only subhorizontal scopic Vredefort "Dome" structure. reflectors. whichundergoachangeindip(overtumed withthecollar LPI Contribution No. 790 17 ..

rocks) in the northwest. Structural information suggests that the Sudbury Igneous Complex (SIC) [3]. Such origin implies that the structure is opento the southeast. From this itmay be inferred that SIC was emplaced before deposition of the Whitewater Group, in contractional forces acted from south to north. contrasttoorigins inwhich the SICpostdates the lithificationofthe In conclusion, the structural studies coupled with the geophysi­ Onaping Formation. Structural and sedimentological evidence is cal results suggest that the Vredefort structure was produced by swnmlrized herein that supports an intrusion of the SIC after subhorizontal forces. Nomacro-ormegascopicSbUctural deforma­ lithification of all Whitewater Group strata, and confficts with the tion thatcould berelated to a 2-Gacentral catastrophic event could hypothesis advanced by Grieve et al, [3]. be identified. The SIC has the map pattern of an oval ring, and dips inward at References: [1] Colliston W. P. (1990) Tectonophysics, 171, the present erosion level. The bilobate eastern part of the SIC 115-118. [2] Du Toit A. L. (1954) TM Geology ofSouth Africa, resembles fold interferencepatternsfiguredbyStaufferand Usleet Oliver and Boyd, Edinburgh, 611 pp. [3] BrockB. B. and Pretorius al, [405], yetthe granophyre, gabbro, and oorite have undergone no D. A. (1964)Geol. Soc. S. Afr. Spec. Publ. 1,549-599.[4] Ramberg solid-state deformation at most localities. This rules out the folia­ H. (1967) Gravity, Deformation and tM Earth's Crust, Academic, tion pattern in Fig. 18, which is consistent with the impact-melt London, 214 pp. [5] Dietz R. S. (1961) J. Geol., 69, 499-516. hypothesis [3]. If the SIC acquired its foldlike shape during or [6) Hargraves R. B. (1961) Trans. Geol. Soc. S. Afr.• 64,147-154. immediately after emplacement, metamorphic-foliation trajecto­ [7] Nicolaysen L. O. (l972)GSA Mem., 132,605-620. [8] Colliston ries in the Onaping Formation would continue as igneous-foliation W. P. and Reimold W. U. (1990) Econ. Geol. Res. Unit Inf. Cire. trajectories into the granophyre, gabbro, and norite (Fig. Ib) [6,7]. 229, 31 pp. [9] Reimold W. U. et al, (1990) Tectonophysics. 171, This is hUe in the northeast lobe of the SIC, and rules out the 139-152. [10] Manton W. I. (1965) N.Y. Acad. Sci. Ann., 123, possibility ofpost-fold sheet injection (Fig. lc) [8]. 1017-1049. [11] AlbatH.M. (1988) S. Afr. J. Geol.• 91,106-113. The Chelmsford Formation, a turbidite deposit with nearly [l2]ComerB.etal.(1990) Tectonophysics. 171.9961.[13) Colliston invariant bed thickness (I.2 m average), detrital composition, and W. P. and Reimold W. U. (1989) Abstr. First Tech. Meet. S. Afr. high sand/mud ratio, was deposited by uniformly southwest­ Geophys. Assoc., 13-14, BPI Geophysics. directedcurrents [9,10],andwaspartofaverylargenonchannelized foreland basinturbidite system[11]. Thelackoffacies changein the northwest-southeastdirection implies that thepreservedChelmsford INTRUSIVE ORIGIN OF THE SUDBURY IGNEOUS stratawere far from the original foreland basinmarginorfrom a site COMPLEX: STRUCTURAL AND SEDIMENTOLOGICAL of syndepositional tectonic disturbance. This suggests that the EVIDENCE. E.I.CowanandW. M. Schwerdtner, Departmentof South Range Shear Zone [12], which probably had a geomorphic Geology, University of Toronto, Toronto, Ontario, Canada M5S expression at surface, postdates the Chelmsford Formation and its 3B1. lithification. The combined sedimentary and structural evidence constrains Inrecentyears,many geoscientists havecometobelieve that the the time ofemplacementofthe SIC and its consolidation. Turbidite Sudburyeventwas exogenicratherthan endogenic [1-3}. Criticalto complexes have sedimentation rates of 1~1000 m/m.y., with a recent exogenic hypothesis is the impact melt origin of the foreland basin-fill systems typically ranging 400-900 m/m.y. b c

Onwatin Fm.

solid-state schistosity trajectories Pre-fold emplacement of Syn-fold emplacement of Post-fold emplacement of the Sudbury Igneous the Sudbury Igneous the Sudbury Igneous Complex: Complex: Complex: Solid-state strain whithin SIC with igneous-foliation SIC with igneous-foliation the SIC with schistosity trajectories concordant to trajectories concordant to trajectories concordant to schistosity trajectories of intrusive contacts that of the sedimentary core the sedimentary core

Fig. 1. Eastern Sudbury structure: SIC (shaded), Whitewater Group (white). 18 1992 Sudbury Conference

[13-15]. The minimwn stratigraphic thickness of the Chelmsford tion of the magnetic field and electron energy within imp8Ct­ Formationis 600m,sothat thedepositionaltimerequired is >1m.y. generatedplasmademonstratethat thepeak magneticfield strength The fastest deposition rate ofhemipelagic deposits is 300 m/rn.y., may only weakly depend on projectile size. Because the ratio of so that 600-m thickness of the Onwatin Formation corresponds to projectile size to crater size increases at larger scales for gravity­ >2 m.y. [15], and possibly as much as 6 m.y. (D. G. F. Long, limited growth, the peak strength of transient imp8Ct-generated personal communication, 1992). This amounts to a time interval of magneticfields probably increases with increasing cratersize at the atleast 3 m.y. betweenthe Sudburyeventand the tectonic deforma­ same diameter-scaled distance. A conservative estimate (from tion ofthe Whitewater Group and the SIC. extrapolated experimental data) for IQ-l00-km craters formed by Ifthe SIC was formed as an impactmelt, the sequence ofevents vertically incident meteorite impacts at 25 km/spredicts magnetic proposedhereinrequires theSICdeform inanunconsolidatedstate, field strengthsofatleast0.03-0.1 G forseveralminutesormore [6]. 3-10 m.y. after its intrusion. Plutons emplaced at mid to upper This is within the range of paleointensity values detennined for crustal levels are thought to take 1-10 m.y. to cool to the ambient certain relatively young (3 Ma to 1.5 Ga) lunar samples [7-9] and wall-rocktemperatures; however, consolidation ofa pluton takes a more generally may help account for the lunar magnetic record fractionofthattime[7]. Impactmeltsheets ofthesizeofthe Igneous during the last -3.5 b.y. Recently acquired experimental evidence Complex wouldcrystallizewellwithin 1m.y. oftheirformation [3]. suggests that impact at oblique incidence may further enhance This duration is shorter than the time interval required for the magnetic field production by as much as Ul order ofmagnitude. deposition of the Onwatin and Chelmsford Formations. Together Experimental investigations of magnetic field generation and with the evidence for magmatic folding of the SIC. this time evolutionduringhypervelocity impacts have beenconductedatthe constraintrenders the impactmelthypothesis ofthe SICuntenable. NASAAmesVerticalGun Range,MoffenField,California[10-12]. References: [1] Peredery W. V. and Morrison G. G. (1984) The vertical gun is a two-stage hydrogen light gas gun capable of Onto Geol. SIUV. Spec. Publ. 1,491-512. [2] Faggart B. E. et al. launching macroscopic projectiles at up to 7 km/s with the angle of (1985) NatlU'e, 230, 436-439. [3] GrieveR. A.F.etal. (1991)JGR, impact varying from nearly horizontal to vertical in increments of 96,22753-22764. [4] Stauffer M. R. (1988) Tectonophysics. 149, 15°. Thelarge impactchamber, whichcanbeevacuated to less than 339-343. [5] LisleR. J. etal. (1990)Tectonophysics.172,197-200. -1 Torr, is large enough to accommodate, surrounding the impact [6] Paterson S. R. et al. (1989) J. Struct. Geol.• 11, 349-363. point, a mu-metal shield thatreduces the 35-..,Tterrestrialmagnetic [7] Paterson S. R. et al. (1991) Min. Soc. Am. Rev. Min., 26, field to 450 ± 80 nT-comparable to lunar surface field strength. 673-722. [8] Schwerdtner W. M. et al. (1983) J. Struct. Geol., 5, Impacts of alwninwn projectiles into powdered dolomite 419-430. [9] Cantin R. and Walker R. G. (1972) Geol. Ass. Can. (M8euCBo.,CO,) targets readily produce a self-luminescent,slightly Spec. Pap. 10, 93-101. [10] Rousell D. H. (1984) Onto Geol. SIUV. ionized vapor cloud that we infer to be the source of impact­ Spec. Pub. 1,211-218.[11] LongD.G.F.,this volwne. [12] Shanks generated magnetic fields [3,6]. Oblique impacts demonstrate en­ W. S. and Schwerdtner W. M. (1991) Can. J. Earth Sci., 28, hancedvaporyieldproducing a vaporcloudthatretains a portionof 411-430. [13] RicciLucci F. andValmori E. (1980)Sedimentology, the impactormomentum witha leadingedge thattravels downrange 27,241-270. [14] HiscottR. N. etal. (1986)/ntl.Assoc. Sediment. at a significant fraction of the impact velocity [13]. Geol.Spec.Publ. 8,309-325.[15] Pickering K. T.etal.(1989)Deep The configuration and duration of impact-generated magnetic Marine Environments, Unwin Hyman, 416 pp. fieldsobservedduringlaboratoryhypervelocity impactsarestrongly dependent on impact angle (Figs. 1-3). Magnetic search coil data from many experiments under identical impact conditions were ENHANCED MAGNETIC FIELD PRODUCTION DURING combinedtoproduce theplotsshown..Theobservedmagnetic fields OBLIQUE HYPERVELOCITY IMPACTS. D. A. Crawford exhibit a regular transition from a cylindrically symmetric field and P. H. Schultz, Department of Geological Sciences, Brown configuration at vertical incidence to a strong bilaterally antisym­ University, Providence RI 02912, USA. metric field configuration at high obliquity (Figs. I and 2). The stronger magnetic fields observed during oblique impacts (see The natural remanent magnetization of the lunar surface as Fig. 3) could result simply from ,the close proximity ~f impact­ displayed in returned lunar samples and the data returned by the generated plasma to the target surface, from a fundamental change Apollo subsatellitemagnetometerhas anunexpectedlyhighmagni­ in the field production mechanism within the plasma or from tude and exhibits spatial variation at all scales. The origin of the increased vaporization [13] yielding a greater volwne of magne­ lunar remanent fields may be due to crustal remanence of a core tized plasma; however, this could not be resolved with the data dynamo field occurring early in lunar history prior to extensive obtained.Inaddition toimpact angle. experiments demonstrate that modificationbyimpact[1] orremanenceoftransientfields,particu­ the configuration andduration ofimpact-generated magnetic fields larly associated with impacts, occurring on a local scale throughout are dependent on impact velocity and projectile/target composition lunar history [2-5]. The presence of an early core dynamo field [11]. wouldhave strong consequences for the fonnation and early evolu­ A remnant of the impact-generated magnetic fwld could be tion of the Moon, yet to deconvolve the role that an internally induced within the target material during passage of the impact­ generated core dynamo field may have had, it is necessaI)' to induced shock wave [14,15] or by cooling through the Curie point understand how the magnetic state of the lunar surface has devel­ of small portions of impact melt or hot target material. During oped through time. Impact-inducedmagnetismmaybeanimportant oblique impacts, spalled fragments of the projectile may impact component ofthe present magnetic state of the lunar surface. furtherdownrangeathypervelocities [16], therebyinducing a shock New theoretical considerations suggest that transient magnetic and/or thermal remanence significantly offset from the crater rim. fields within plasma produced by hypervelocity meteorite impacts Because of these dependencies, remnant impact-generated mag­ may have greater significance at larger scales than previously neticfields couldbea usefulgeophysical toolforthestudy ofimpact thought [6]. Self-similar, one-dimensional solutions for the evolu- craters on the Earth and planetary surfaces by helping to detennine the impact angle, direction, and composition of impactors. The LPI Contribution No. 790 ·19

Magnetic Field (vertical, 5.03 km/s) Magnetic Field (45 degrees, 5.06 kIn/s) 2.0 4. 25 2 15 2. 1.0 1 5 y o. YO o.Ot~ -5 -5 -1 -10 -2. -1.0 -15 -15 -2 -20 -25L..,-~~ -25 2 0 -25-20-15-10 -5 o 5 10 15 20 25 cm-4. nT -25 -20-15 -10 -5 o 5 10 15 20 25 cm- . nT X X - 1nT - 2 nT

Fig. 1. Average magnetic field observed during the time interval 0.3--D.5 ms after hypervelocity (-5 km/s) impacts ofO.64-cm aluminum projectiles into powdered dolomite targets. Each plotrepresents the magnetic field as seen in a horizontal plane 9 cm below a vertical impact (left) and an oblique impact, 45° from horizontal (right). Shading represents the vertical component of the field whereas vectors represent the horizontal component with white vectors appearing where the field is directed into the page (down). The crater(shown as a white circle) defines the origin (shown as a white cross) of the coordinate system. A white triangle indicates the projectile trajectory prior to impact.

Magnetic Field (15 degrees, 5.17 km/s) Magnetic Field (t5 degrees, 5.17 kIn/o) 6. 6.0

3. 3.0

o. -10 - ~ -15 -3.0 -3. -20 - -25 -30 o 5 10 15 20 25 cm-6. nT -25 -20-15 -10 -5 o 5 10 15 20 25 cm'"6.0nT X X -10nT -20nT

Fig. 2. Magnetic field during an oblique 15° impact at the same time as in Fig. 1 (left) and 0.4 ms later (right). many possible time-dependent impact-related magnetic field pat­ Average rield Strength tcrns and the various possible rcmanence acquisition mechanisms 20 (some or all of which may come into play during a given impact event) yield a broad speetrum of possible crater-related remnant field patterns. An understanding of this complex contribution of impact-induced magnetism to the magnetic state of solid body 15 surfaces, in general, and the lunar surface, in particular, is necessary to assess the role of internally derived fields such as from a core dynamo and may help to define future magnetic survey missions to E- 10 solid surface planets, satellites, and asteroids. '" References: [1] Runcorn S. K. (1983) Nature, 304, 589-596. [2] Gold t. and Soter S. (1976) Planet. Space Sci., 24, 45-54. [3] 5 Impact Angle Srnka L. J. (1977) Proc. LSC 8th, 893-895. [4] Schultz P. H. and -:-;0-15-30 Srnka L. J. (1980) Nature, 284, 22-26. [5] Hood L. L. and Huang Z. ~45--60 (1991) JGR, 96, 9837-9846. [6] Crawford D. A. (1992) JGR, 0 -.....-75--90 submitted. [7] SugiuraN. eta!. (1979)Proc.LPSC 10th,2189-2197. 0 I 2 3 4 [8] Cisowski S. M. et al. (1983) Proc. LPSC 13th, in JGR, 88, Time after impact (ms) A691-A704. [9] Collinson D. W. (1984) PEPI, 34, 102-116. [10] Crawford D. A. and Schultz P. H. (1988) Nature, 336,5052. Fig. 3. Spatially averaged magnetic field strength as a function of [11] Crawford D. A. and Schultz P. H. (1991)JGR,96,18807-18817. time after impact. Results from six impact angles are shown. [12] Crawford D. A. and Schultz P. H. (1992) LPSCXXIlI,259-260. Enhanced field production during the first millisecond after impact [13] Schultz P. H. (1988) LPSC IX, 1039-1040. [14] Cisowski S. M. is apparent for oblique impacts at 15° and 30° from horizontal. 20 1992 Sudbury Conference

et al. (1975) Proc. LSC 6th, 3123-3141. [15] Cisowski S. M. et al. (1976) Proc. LSC 7th, 3299-3320. [16] Schultz P. H. and Gault D. E. (1990) GSA Spec. Pap. 247.

IMPACTITE AND PSEUDOTACHYLITE FROM ROTER KAMM CRATER, NAMIBIA. J. J. Degenhardt Jr. I.2, P. C. Buchananl , and A. M. Reidl , IDepartment ofGeosciences, Univer­ sity of Houston, Houston TX 77204, USA, 2Texaco Inc., E&P Technology Division, 3901 Briarpark, Houston TX 77042, USA.

Pseudotachy1ite is known to occur in a variety of geologic settings including thrust belts (e.g., the Alps and the Himalayas) and impact craters such as Roter Karnm, Namibia. Controversy exists, however, as to whether pseudotachylite can be produced by shock brecciation [1] as well as by tectonic frictional melting. Also open to debate is the question of whether pseudotachylites form by Fig. 2. Plane light photomicrograph of opaque clasts entrained in frictional fusion orby cataclasis [2]. Ithas been speculated that the quartz breccia matrix. Large clast (center) is 1.6 mm in length. pseudotachylite at Roter Kamm was formed by extensional settling and adjustmentofbasement blocks during "latemodificationstage" ofimpact [3]. The occurrenceofpseudotachylite in association with A variety of impactites was collected from the crater rim. These rocks resembling quenched glass bombs and melt breccias in a have been described in the literature as "impact-melt breccias," relatively young crater ofknown impact origin offers a rare oppor­ "melt bombs," and "ejecta melts." In this study, three main rock tunity to compare features of these materials. Petrographic, X-ray types have been examined: diffraction, and electron microprobe analyses of the impactites and 1. Meltlike rocks (also called fliidle), common only along the pseudotachylite are being employed to determine the modes of northwest crater rim, are dark gray to black in color and exhibit deformation and to assess the role offrictional melting and commi­ apparent flow patterns characteristic of quenched glass "bombs." nution of adjacent target rocks. The first fmdings are reported here. Because of their smooth, fluidlike shapes, the fliidle have been The Roter Karnm Crater is located in the southern part of the interpreted as impact melt. In spite oftheir meltlike appearance, the Namib Desert about 80 km north of Oranjemund, Namibia. The specimens that were examined exhibited pervasive subparallel craterrim has a crestdiameterofca. 2.5 km with the highestexposed millimeter-wideveinlets containing subhedralmicaand quartz. The point reaching 158 m above the lowest pointofthe crater floor. The fact that these veins are sharply (visibly) truncated by the surfaces impact, which excavated Precambrian granitic-granodioritic of the fliidle brings into question the interpretation that these rocks orthogeneisses of the 1200-900-m.y.-old Namaqualand Metamor­ were melt-derived (see Fig. 1). XRD and electron microprobe phic Complex, has been dated at 3.5-4.0 Ma by Hartung et al. [4]. analyses of the veins confirmed the mica-quartz composition and No volcanic rocks have been discovered in the area [5]. This has showed the mica to be a K, Mg-bearing variety. The finer-grained, been substantiated by others [3]. A detailed account of the crater optically opaque matrix is composed ofmica and quartz similar to geology was published by Reimold and Miller [6]. that of the veins. A Ti-rich mineral was also found to be present in the matrix, mainly in the form ofclustered grains scattered through­ out. Further analysis is being carried out to determine the cause of the opacity. 2. Quartz breccias containing large irregular shocked quartz fragments were also collected along the north crater rim. The fragments are cemented by an aphanitic matrix that resembles pseudotachylite in hand specimen. In contrast to the fliidle, this matrix is not opaque in thin section, but appears fmely crystalline. Of significance is the presence of opaque clasts (som,e containing micaceous veins) located throughout the matrix, which upon pre­ liminary examination appear to be composed of the same material as the fliidle (Fig. 2). Work is currently being done to determine if the chemicalcompositions oftheseentrainedclasts match that ofthe fliidle, and thus whether the fliidle material predates the quartz breccia. 3. Pseudotachylite samples in the form of allochthonous frag­ ments were collected at the south rim. Petrographic examination reveals the presence of angular entrained clasts that have not mdergone recrystallization. However, fmely crystalline quartz resembling the quartz breccia matrix appears in (plastically?) Fig. 1. Photomicrograph offliidle in transmitted light illustrating deformed grains and intermediate zones that often separate the veinlets ofsubhedral mica and quartz set in opaque matrix. Veinlets pseudotachylite from its host rock. The composition of the vein terminate abruptly at edge of sample. material, including nonrecrystallized entrained clasts, was found to LPI Contribution No. 790 21 ..

be essentially that of the host rock. It is apparent from grain SUDBURY IMPACT STRUCTURE deformation that transport distances were on the orderofmillime­ Timpact - 1.85 Ga ters to centimeters along the veins. The textures of flidle, quartz breccia, and pseudotachylite are ftmdamentally different when viewed petrographically. Although eachofthe rock types appears darldy opaque (darkgrayorblack) in hand specimen, the only sample with a matrix that is truly opaque inthinsectionis the flidle. Noclastsrepresentingcompositions and textures of the other impactites have been observed in the pscudotachylite thus far. Present work is directed at determining what textural and compositional changes were involved during formation and whether the pscudotachylite represents material comparable to associated impactitcs. -I' References: [1] SchwamnanB.C.etal.(1983)GSA Bull.,94,

926-935. [2] Wenk H. R. (1978) Geology, 6,509-511. [3] Kocberl _II ...... ~_'-'-_--.J'---_--.J'---_--.J C. et al. (1989) GCA, 53, 2113-2118. [4] Hartung 1. et al. (1991) ... -It 41 .. 141 54th Annu. Met. Soc. MIg., Monterey, CA. [5] Fudali R. F. (1973) eSrua Ga Meteoritics, 8, 245-257. [6] Reimold W. U. and Miller R. MeG. (1989) Proc. LPSC, Vol. 21,711-732. Fig. 1. ENd-ESr diagram for different lithologies of the structure withdatarecalculated to 1.85 Ga, the timeofthe impactevent[14]; data sources [4-6,11,12,15-17,19]. SUDBURY PROJECT (UNIVERSITY OF MUNSTER­ ONTARIO GEOLOGICAL SURVEY): (4) ISOTOPE SYS· TEMATICSSUPPORTTHEIMPACTORIGIN. A. Deutsch1,. Neodymium isotope ratios of the impact melt concur with Nd D. Buhl2, P. Brockmeyer', R. Lakomy', and M. Flucks', 'Institute characteristics of the target lithologies in the Sudbury region, for for Planetology, University ofMilnster, Wilhelm-Klemm-Str. 10, example, the Levackgneiss [12]. The observedspreadinESr reflects 04400, Milnster, Germany, 2Institute for Geology, RU Baehum, the widely varying (!'SrI"Sr)T - 1.85 0. for the Archean basement Postfach 10 21 48, D-463O Bochum I, Germany. [4,15],ProterozoicIntrusives [16,17],andtheHuronianSupergroup [16,18] thatweremixedintothemelt.Distinctfields for thesublayer Introduction: Within the framework of the Sudbury project from different localities [12,19] in Fig. 1 show that the impact melt [1-3] a considerablenumberofSr-Ndisotopeanalyses were carried sheet (SIC) assimilated local bedrocks after its emplacement in the out on petrographically well-defmed samples of different breccia fmal modified crater. Strongly deviating Sr isotope ratios for some units [4-7].Togetherwithisotopedatafrom theliteraturethesedata Onapingrocks inFig. 1with (!'SrI"Sr)T-U5o.as low as 0.700 [6,7] are reviewed in this abstract under the aspect of a self-consistent or 0.67 [20] are due to a reopening ofthe Rb-Sr system during the impact model [5,8-10]. The cmcial point of this model is that the Penokeanorogeny[4,seealso 7]. Thisis demonstratedwithselected Sudbury Igneous Complex (SIC) is interpreted as differentiated growth curves in Fig. 2: Some recrystallized melt particles and impact melt sheet [5,8-11] without any need for an endogenic devitrified glass have enhanced Rb/Srratios but the majority ofthe "magmatic"component such as "impact-triggered" magmatism or material has Isr identical to the granophyre. Together with their ~d' "partial" impact melting of the crust and mixing with a mantle­ this fact supports our view that the melt-breccia on top of the derived magma[e.g., 12]. Forthe terminology usedherewerefer to granophyre and the melt material in the suevitic Onaping breccias the companion abstracts in this volume [1-3]. and in the Green MembCf originated from the same source as the Strontium and Neodymium Isotopes: Impact melt rocks SIC, namely impact-melted crustal material. such as the sublayer [12], the SIC [11,12], and the clast-rich melt Oxygen isotope data [21] support our findings. The norite, the breccia on its top [5,6], as well as melt breccia bodies, matrix, and granophyre, and the matrix ofOnaping breccias all show a consid­ melt particles from the Onaping breccias [5,6], arecharacterized by erable spread in 6110, but typical trends as known from differenti­ eNd between -5 and -12 with Onaping lithologies tending toward atedlayeredintrusions areabsent. The318()valuesoftheselithologies lower ENd values (fig. 1). Their Nd-model ages TOM relative to a are bracketed by oxygen isotopic compositions observed for local depleted mantle [13] cluster around 2.7 Ga, which agrees well with Archean and Proterozoic bedrocks with the Onaping breccias re­ the time of the last major crust-forming event in the Archean flecting a higher input ofHuronian greywackes. To explain the Os Superior Province northwest ofthe Sudbury structure [e.g., 14,15J. isotoperatios for the Sudburyores [22J bymixing between amantle Itis important tonote that ultramafic inclusions in the sublayerplot magma and crust would need up to 90% crustal material. Therefore among other SIC rocks with negative ENd and positive ESr [12J. these data are also in line with a derivation ofthe ores exclusively Figure 1 shows that the SIC has highly radiogenic and variable ISr• from ancient crustal sources by impact melting followed by segre­ All those fmdings fit with the proposed total melting ofthe crust in gation of a sulfide liquid out ofthe melt sheet. the Sudbury region by the impact event that leads to -1.5 x 1()4 km' Summary and Outlook: While in the original contributions of impact melt [10], namely the SIC and the melt breccia, and the SIC isotope systematics were discussed preferentially in terms ofa melt in the suevitic Onaping breccias [3]. In contrast, the data are possiblemixing between a hypothetical mantle component with up incompatible with the input of fresh mantle magma as up to 75% to 75% crustal material, the impact melt model does not have any contamination by upper crustal material would be required [12] to problem explaining the crustal signatures of the SIC, the Onaping explain the Nd-Sr isotope systematics of these units, but an endo­ breccias, and the Sudbury ore&-total melting of basement and genic melt cannot assimilate such a high fraction ofrelatively cold supracrustal lithologies canonly producecrustal signatures. Future material. studies on Sudbury should concentrate on combined analyses of 22 1992 Sudbury Confereru:e

ofrespectability. Also, therecentllpparentsuccessfultying inofthe KIT extinctions totheChicxulubastrobleme in Yucatanencourages the sean:h for an impact event that may have caused the other two majorpost-Paleozoicextinctions (PlTr, Tr/J). This gives us heart to offer two further outrageous hypotheses. NorD'skOres/SiberianBlISlllts: Thecosmogenicconceptfor the Sudbmy ore deposit remains viable because it is giant, nonultramafic. and Wlique (except for Noril'sk). It also has telling geologic relationships; for example, the ore-hosting sOOlayer ap­ pears tobea splash-emplaced target/bolide meltlining the Sudbuly Basincavity like spackle on a bowl that was also injected centrifu­ gallyintotensional cracks (offsets) (sec[3] forfurtherevidence). At Sudbury, endogenic scenarios usually have been assumed, espe­ SIC cially the conceptofthe ring-dike sOOlayer fed from a deep magma reservoir [5]. This view has recently been seriously challenged by 1.0 1.0 1.61 I.aSC. Grieve and Staffler[6], whoexplaintheSudburyIntrusiveComplex as an impact melt sheet. Although the geologic relationships be­ tween the ore and the COWltry rock at Noril'sk remain enigmatic, it Fig. 2. Selected Srgrowth curves for recrystallized Sudbmy melt seems a remarkable Sudbmy look-alike. Their ore mineralogy is particles and devitrified glass from the Gray and Green Member of similar, including platimum group metals, and they are both large the Onaping Formation [3], and the matrix of the melt breccias scale (one Noril'sk sulfide body covers 2 sq km and is 20 m thick.) topping the granophyre; data sources [5,6]. Naldrettetal. [7] believethattheNoril'skores andadjacentSiberian plateaubasalts areintimately related and consanguineous. A similar view was offered by several other authors at the 1991 American Geophysical Union Fall Meeting symposium (Noril'sk Siberia: stable and radiogenic isotopes onpetrographically defmed samples Basalts, Intrusions, and Ores). Using argon/argon laser fusions, in order to Wlderstand the mixing process in impact melts in more Dalrymple et al. [8] assigned a date for the ores and flood basalt of detail. Isotopedatacouldalsohelp todecipher the complexprocess 249 ±1Ma, indistinguishable from the PermianlTriassic boundary. of assimilation by a large hot impact melt sheet for both lithic We therefore suggest that the Noril'sk ores may be ofcosmogenic fragments and crater floor lithologies. parenthood and that this impact also triggered the Siberian plateau References: [I] AvennannM.etal., this volume. [2] Bischoff basalts. An associatedeventthenmightbethegreatextinctionoflife L. etal., this volume. [3] Staffleret al.• this volume. [4] DeutschA. forms at the PlTr boundary, all tied together as an event horizon. et al. (1989) EPSL, 93, 359. [5] DeutschA. etal. (1990)USCXXI, BahamaNexxus: Olsen[9] has attributedtheTriassic/Jurassic 282. [6] Brockmeyer P. and Deutsch A. (1989) USC XX. 113. boundary catastrophic extinctions to the Manicouagan asteroidal [7] Buhl D. et al.• this volume. [8] Staffler D. et al. (1989) impact. butrecentradiometricdating [10] indicates theseevents are Meteoritics, 24,328. [9] Lakomy R. (1990) Meteoritics, 25. 195. diachronous (Manicouagan astrobleJlle 212 ±2 MaandTrlJbound­ [10] GrieveR.A.F. etal.(1991)JGR,96, 22753-22764. [11] Faggart ary 200 Ma). This boundary is also marked by extensive tholeiitic B. E. Jr. et al. (1985) Scieru:e, 230, 436-439. [12] Naldrett A. 1. et basalts (flows, sills,·and dikes) of the rapidly extruded Newark al. (1986) In MetaUogeny ofBasic and Ullrabasic Rocks. 75-91. Supergroup. Radially emplaced dikes on Pangaea (now broken up Inst. ofMiningMetallurgy, London. [13] DePaoloD. 1. (1981 )JGR, into Africa, North America, and South America) focus toward the 86, 10470. [I4] Krogh T. E. et al. (1984) In The Geology and Ore Bahamas [11]. Dietz [12] has previously termed this presumed Deposits ofthe Sudbury Structll1'e (E. G. Pye et al., cds.), 431-446, hidden hot spot (now buried beneath 6 km of shallow water coral Toronto. [15] HurstR. W. andFahartJ.(1977)GCA,41. 1803-1815. reef limestone) as the Bahama Nexxus (a great triple junction [16] Fairbairn H. W. et a1. (1969) Can. J. Earth Sci., 6, 489. connection) that marked the birth of the Atlantic rift ocean. The [17] Gibbins W. A. andMcNuttR.H. (1975) Can. J. EarthSci., 12, Bahama Platfonn might then be a mega coral reef laid down 1970. [18] Fairbairn H. W. et al. (1967) 15th Annu. Rept M./.T.• ensimatically on a subsiding plateau basalt. The floor spreading 1381. [19] Rao B. V. et al. (1984) Misc. Pap. Ontario Geol. SlII'V. process calls for symmetrical repaving of the ocean floor by dike 121,128. [20] FullagarP. D.etal. (1971)Can.J. Earth Sci., 8,435. splitting. This clearly applies to the North Atlantic continental drift [21] Ding T. P. and Schwan:z H. P. (1984) Can. J. Earth Sci., 21, (North America/Africa) from Nova Scotia southward Wltil the 305. [22] Walker R. J. et al. (1991) EPSL, 105,416. Bahama platform is reached. Then the conjugate point between North America and Africa jumps to the eastern tip of the Bahama cresenticplatfonnrather than being at thetipofFlorida. Clearly the NORIL'SK/SIBERIAN PLATEAU BASALTS AND seafloor spreading (dikesplitting) was overprinted by the hot spot, BAHAMA HOT SPOT: IMPACT TRIGGERED? R. S. Deitz causing newly fragmented Gondwana (AfricalSouth America) to and J. F. McHone, Department ofGeology. Arizona State Univer­ remain fixed (relativeto the Earth's spin axis) while North America sity. Tempe AZ 85287, USA. drifted away. (We can observe a modem example by the eastward offsetting ofthe Mid-Atlantic ridge as it transects Iceland hot spot.) Twenty-eight years after oneofus [1] argued that Sudbury was Eventual death of the hot spot allowed the Mid-Atlantic Ridge to an astrobleme. this interpretation has only recently attained wide pave the ocean floor symmetrically. Thus almostthe entire Bahama acceptance; notso for his view that the Sudbury CulNi sulfide ores platfonn was stranded on the North American plate, leaving but a are cosmogenic [2,3]. Papers such as by Alt et al. [4] haveprovided very small conjugate volcanic excrescence attached to Afriea-the the triggering ofplateau basalts by super-large impacts a modicum Bijagos Plateau offPortuguese Guinea. This great magmatic event U'l Contribution No. 790 23

may have been a mantle plwne. but, alternatively. itmay have been most narrow. metal enrichment is lowest and the POEs exhibit triggered by an asteroidal impact. An impact fall-out layer with chondritic ratios. Sections ofthe ejecta horizon with a significantly shocked debris has beenreported at the TrlJ boundary in Italy [13]. wider green alteration envelope are variably enriched in Cu and The synchrooicity ofextrusion and extinction appear established as frequently in Au. Inthese situations. both the ejectahorizon and the an event horizon independent ofradiometric or fossil stratigraphy. green shales that envelope it have strOllg POE enrichments with Ir Ofcourse. enthusiasm for impacts does not score points in the up to 100 times enriched and Ptup to 300 times enriched relative to scientific forwn. but the current evidence of their importance in the hostred shales. CopperandPtare well correlatedwitheachother shaping other terresbial planets (Venus has 900 impact craters as and the POEs exhibit strollg nonchondritic ratios. well as 1500 volcanos) suggests that impacts need not be assigned Thin green shale layers that show no evidence of meteoritic a role oflast resort. contribution and occur at stratigraphic positions above and below References: [1] Dietz (1964)/. Geol.• 72.412~34. [2] Dietz the ejecta horizon in the red shale sequence are similarly enriched (1972) Geol.Assoc. CanadtJ. Spec. Pub. 10.29-40. [3] Dietz (1991) in Ir and Pt as well as Co, V. Zn. and Nt Isolated green reduction Earth, 1.36-41. [4] Altet al. (1988)/. Geol., 96. 647-662. [5] Muir spots in the red shales also have POE enrichments. All thin green (1984) Geology and SlrllClIWe ofSudbury Ore Deposits. 449-490. shale horizons and green reduction spots analyzed have relatively Ont Geol. Surv. Spec. Vol. 1. [6] Grieve and Stiiffier (1991) Geol. high levels of K and other POE regardless of their stratigraphic Assoc. CanadtJ, 16. A48. [7] Naldrett et al. (1991) Geol. Assoc. position. CanadtJ. A88. [8) Dahlrymple (1991) Eos. 570. [9) Olsen (1986) The similarchemisbiesoftheejecta-associatedgreenshales and Lanumt.A1I1IU. Repl. 12. [10] Hodych and Dunning (1992) Geology. green shales at other stratigraphic levels suggest a similarity in the 20.5154. [11] Dietz et al. (1971) GSA, 82.1131-1132. [12] Dietz enrichment process. The very high Pd/Ir. Pt/Ir. and Au/Ir ratios of (1986) 491h A1I1IU. Met. Soc. MIg .• 1-10. [13] Bice and McDauley the green shale and the Cu-enrichedejectasample. togetherwith the (1990) GSA Abstr., A322. Cu-POE correlation. are not totally consistent with an extraterres­ trial origin. The ejecta horizon clearly has a meteoritic component as do the other thin green shale horizons and green reduction spots. which suggests that the elevated values are due to low-temperature MOBIUZATION OFTHEPLATINUM GROUPELEMENTS transport. BYLOW·TEMPERATUREFLUIDS: IMPLICATIONSFOR The element associations and distribution are consistent with a MINERALIZATION AND THE IRIDIUM CONTROVERSY. POE redox entrapment process. It is suggested that the ejecta Kim Dowlingl•Reid R. Keaysl. Malcolm W. Wallacel•and Victor horizon was an aquifer for low Ell fluids derived from deeper inthe A. Gostin2• IDepartment of Geology. University of Melbourne. sedimentary basin. These fluids reduced ferric iron in the redshales Australia. 2Department ofGeology and Geophysics, University of to ferrous iron that was removed in solution. leaving the shales with Adelaide. Australia. their green color. Mixing of the reduced fluids flowing along the aquiferwith oxidizedfluids circulating inthe redshales. from which Geochernicalinvestigationson the widely dispersedLateProtero­ they had leached Au. Co, POE. and other elements. caused metal zoicAcraman impactejectahorizon and its hostmarineshales in the deposition. Adelaide Geosyncline provide slrong evidencefor low-temperature The discovery ofsignificant POE mobility by low-temperature mobilization of the platinwn group elements (POE). including Ir. oxidized fluids has several important implications: Ir, POE. and Au The ejectahorizon was formed when the middle Proterozoic dacitic anomalies may be associated with postdepositional processes, volcanics in the Gawler Ranges. central South Australia. were which is particularly significant given the KIT boundary Ir contro­ impacted by a very large (ca. 4 Ian) meteorite. The resulting versy. Further. it indicates that economically important accumula­ structure. now represented by Lake Acraman. is Australia's largest tions of the metals might be anticipated in environments in which meteorite impact structure. Debris from the impact was blasted for such solutions entered low redox environments. Examples ofsuch many hundreds ofkilometers. some falling into the shallow sea of environments include red-bed Cu and roll-type U deposits. the Adelaide Geosyncline. some 300 Ian to the east of the impact Reference: [1] Wallace M. W. et al. (1990) Geology, 18. site. 132-135. The Bunyeroo Formation (-600 m.y.). which hosts the impact horizon. consists ofmonotonous deep-shelfmaroon and green clay shales. with minor concretionary carbonates. The ejecta horizon is typically 0 to 40emthick and is composed ofa basal clast layer that DOES THE BUSHVELD-VREDEFORT SYSTEM (SOUTH is poorly sorted. angular. anddominated by pebble-sizedfragments. AFRICA) RECORD THE LARGEST KNOWN TERRESTRI· It is overlain by a thin shale layer that contains abundant coarse AL IMPACT CATASTROPHE? W. E. Elston. Department of sand-sized clasts that is in turn overlain by a graded layer that fmes Geology. University of New Mexico. Albuquerque NM 87131­ up from coarse-mediwn sand to afme muddy sand. The largest clast 1116. USA. found to date is 40 cm in diameter. All the clasts and most of the sand-sized grains appear to have been derived from a pink to red The unique 2.05.oa Bushveld and Vredefort complexes cover porphyritic volcanic rock, similar to that currently exposed at the 100.000 Ian2(diameter 400 Ian) on the otherwise stable Kaapvaal Gawler Ranges impact site. The ejecta horizon is almost invariably craton. Since the 1920s. workers have recognized that they are enveloped by green shales that range in thickness from a few bracketed by the same units and were probably formed by related millimeters to several meters. processes. Modern field studies and radiometric dates have pro­ Metal concentration along the horizon is anomalously high vidednocompellingevidenceforditIc:rentages. Hall andMolengraatI though variable,with values up to300times greaterthan averagered [1] and Daly [2] invoked magmatic upthrust. Daly [3] later attrib­ shale background values [1]. Where the green shale envelope is uted Vredefort to impact, but never applied his concept to the 24 1992 Sudbury Conference

Bushveld.Subsequently,VredCfortyieldedshattercones[4],coesite grains in partly digested sandstone clasts were recrystallized to andstishovite[5],andplanarfeatures [6];pseudotachylite(indistin­ tridymite needles at temperatures ~1l75°-1200ee [20,21] and guishable from Sudbury) had long been known [7]. Dietz [8], inverted back to quartz -l100ee [22]. Recrystallization at these Hamilton [9], and Rhodes [10] concluded that at least four simulta­ temperatures would destroy all shock phenomena. Up to 30 m of neous impacts caused the Bushveld-Vredefort system. 11Iree im­ quartzitebeneathRooiberg-Dullstroomflows alsoinvertedtoquartz pacts formed overlapping Bushveld basins; the fourth made the needles and laths, paramorphs after tridymite [23]. Similar ex­ Vredefortdome.1fso,whyhas theBushveldyieldednounequivocal tridymitequartzneedlesoccurinthetransitionfrornmicropegmatite shock phenomena? The nature of intra-Bushveld "fragments" and to basal Onaping quartzite breccia at Sudbury [24]. the properties of Rooiberg Felsite offer clues. The Bushveld lacks "smoking gun" shock phenomena because Cratering of this magnitude would intersect the ambient-pres­ all in sit"exposures are peripheral and all ejectaisrecrystallized 01' sure liquidus isothenns ofboth granite and gabbro. As a result, the melted. In Bushveld-sized impacts, heat effects overwhelm shock Bushveld Complex generated successively the most voluminous effects. A deeplyerodedshockedcoreis exposedonlyinthesmaller siliceous flows, thegreatestaccumulationoflayeredgabbro, andthe and nearly amagmatic Vredefort dome. The Bushveld-Vredefort largest masses of A-type granite on Earth, in a setting of complex eventprobably was thelargestknownmultipleimpactonEarth. The andlong-a>ntinued structural adjustments. Gabbroic sills (Rusten­ alternative would be an as-yet-unknown and unique endogenic burg LayeredSuite, RLS), collectively up to 9 km thick, outline the catastrophe. The event may have had global effects. It coincides Bushveld basins. Up to4.5 kmofearlierRooiberg Felsite foons its with the biogenic transitionfrom reducing to oxidizing atmosphere inttuded roof and locally (as Dullstroom Fonnation) its floor [11], [25] and may correlate with a worldwide 8''<: anomaly [26-28], analogous to the Onaping-sublayer relationship at Sudbury. Late­ greater than those at the Proterozoic-Cambrian, Permian-Triassic, stage Lebowa Granite occupies interiors of Bushveld basins and and K-T boundaries [29-33]. invades the RLS-Rooiberg contact as sills up to 2 km thick. Acknowledgments: David Twist, Joachim Schweitzer (then This simpleschemeis disturbedbydeformationofpre-Bushveld University ofPretoria), and Frikkie Hartter (Geological Survey of rocks around the periphery of the complex [12] and, especially, in South Africa) weremy guides dming field work in 1985, 1987, and 50-km "fragments" within the western and eastern basins [13]. 1991. The facts are theirs, the interpretations are mine. DefonnationoccurredpriortoemplacementofRooibergFelsiteand References: (1] Hall A. L. and Molengraaf G. A. F. (1925) its equivalent "bronzite granophyre" dikes [l4] at Vredefort. Verh. Koninld.A/cQd. Wetensch.Amsterdam,Sect. 2, Vol. 24,No.3. Hamilton [8] and Rhodes [9] interpreted "fragments" as central [2] DalyR. A. (1928) GSA BIIlI.,39,125-152. [3] DalyR. A. (1947) uplifts, whichledtofruitless searchesforshockphenomena[l5,16]. J. Geol., 55, 125-152. [4] Hargraves R. B. (1961) Geol. Soc. S. Based on seismic [17] and field [l2] evidence, I interpret the Africa Trans., 64,147-161. [5] Martini J. E. J. (1978)Natwe, 272, "fragments" as partofa zone between the central initial (transient) 715-717; (1991) EPSL, 103, 285-300. [6] Grieve R. A. F. et al. cavities (reboundedinto Vredefort-typedomes) andthepresentrim (1990) Tectonophysics, 171,185-200.[7] ShandS.J. (1916) Quart. of the Bushveld basins. It is suggested that the Bushveld basins J.Geol.Soc.London, 72,198-217. [8] Dietz R. S. (1963) GSA Spec. becameenlargedbeyondthe initialcavitiesbycollapse, inresponse Pap. 73,35. [9] HamiltonW.(1970) Geol. Soc. S.AfricaSpec. Pllbl. to withdrawal ofsubsurface magma. Today, the central domes are I, 367-379. [10] Rhodes R. C. (1975) Geology, 3, 549-554. totally obscured by Lebowa granite or covered by younger sedi­ (11] Schweitzer J. K., personal communication; (1987) Unpub­ ments. All exposures of RLS and Rooiberg Felsite are on the lished draft, Ph.D. dissertation, Univ. Pretoria. [12] Sharpe M. R. perimetersofthe enlarged collapsebasins, toodistantfrom inferred and Chadwick B. (1982) Geol. Soc. S. Africa Trans. 85, 29-41. central domes for shock phenomena. (13] Hartzer F. 1. (1991) personal COQUIlunication; (1987) Unpub­ In the eastern "fragment," the central (Marble Hall) segment lished M.S. thesis, Rand Afrikaans Univ. (14] French B. M. and exposes pre-Bushveld rocks that are intensely folded, faulted, Nielsen R. L. (1990) Tectonophysics, 171, 119-138. (15] FlCJlch metamorphosed, and boudinaged or brecciated. They are inter­ B. M.andHargravesR.B.(1971)J.Geol.,79,61~20.[l6)FlCJlch pretedaspartofthedeformedcollararound a transientcavity.Inthe B. M. (1990) Tectonophysics, 171,287-301. [17] DuPlessis A., northern (Stavoren) segment, basal Rooiberg-Dullstroom Felsite, personal communication; DuPlessis A. and Levitt J. G. (1987) conformable on unfolded pre-Bushveld quartzite, is interpreted as Indaba,Geol. Soc. S.Africa,Progr. w.Abs.,14-15. (18j'LiprnanP. outflow thatslidinto theexpanding collapsebasin,probablyduring W. (1976) GSA BIIlI., 87, 1397-1410. (19] Twist D. (1985) £Con. emplacement of Lebowa Granite (in the manner of collapse Geol., 80,1153-1165. [20] Eales H. V. (1974) Geol. Soc. S.Africa megabrecciasofignimbritecalderas (18)). Thepre-Bushveldgrani­ Trans., 77,37-51. [21] Schneider H. and Florke O. W. (1982) N. toid core of the southern (Dennilton dome) segment remains enig­ Jahrb.f. MiMral. Abh., 145, 280-290. [22] Hirota K. and 0110 A. matic. Inthewestern"fragment,"thesoutherndeformed (Crocodile (1977) Natruwissensch., 64,39-40. [23] Elston W. E. andSadow J. River) segment and the northern undeformed (Rooiberg) segment (1991) GSA Abs. w. Progr. 23, No.5, A402. [24] Stevenson J. S. play roles similar to the Marble Hall and Stavoren segments (1963) Can. MiMral., 7,413-419. [25] Twist D. and Cheney E. S. respectively. (1986) Precambrian Res., 33, 255-264. [26] Schidlowski et al. No source is known for Rooiberg Felsite, interpreted as several (1976) GCA, 40, 449-455. [27] Master S. et al., Geocongress (S. lithospheric melts (19]. Chemically and physically it differs pro­ Africa) '90, 3pp. [28] BakerA. J. and FallickA. E. (1989) Natwe, foundly from all known volcanic rocks. As a result of quenching 337, 759-762. [29) Kirschvink J. L. et al. (1991) GSA Today, I, from extraordinary temperatures, its feathery textures are more 69-71. [30] MargaritzM. (1989) Geology, 17,337-340. [31] Mar­ appropriate for komatiite than rhyodacite. It incorporates large garitz M. et a1. (1986) Natwe, 320, 258-259; (1988) Natwe, 331, amounts ofsedimentary material, from relict quartz grains to large 337-339; (1985)NewsletteronStratigraphy,15,I00-113.[32] Hol­ (up to50 m)quartzite blocks, commonly brecciated before engulf­ ser W. T. and Margaritz M. (1986) Mod. Geol., 11, 155-180. [33) ment by "felsite" melts. The basal high-temperature Rooiberg­ Tucker M. E. (1986) Natwe, 319, 48-50. Dullstroom flows grade intohigh-energy debris avalanches. Quartz LPI Contribution No. 790 25

RUTHENIUMllRlDIUM RATIOS IN THE CRETACEOUS· DIAPLECTIC TRANSFORMATION OF MINERALS: TERTIARY BOUNDARY CLAY: IMPLICATIONS FOR VOROTILOV DRILL CORE, PUCHEZH·KATUNKI IM· GLOBAL DISPERSAL AND FRACTIONATION WITHIN PACT CRATER, RUSSIA. V. I. Feldman, Geological De­ THE EJECTA CLOUD. Noreen Joyce EvansI, W. D. partment, Moscow University, Moscow, Russia, 119899. Goodfellow2,D.C.Gregoire2,andJ.Veizer',IDivisionofGeological andPlanetaryScience,CalifomiaInstituteofTeclmology,Pasadena Vorotilov core has been drilled in the central uplift of the CA 91125, USA, 2Geological Survey ofCanada, Ottawa, Ontaria, Puchezh-Katunki astrobleme to a depth of 5.1 km. Impactites are CanadaKIAOES, 3Ottawa-CarletonGeoscienceCenter, University revealed in the rocks ofthe core beginning from a depth of366 m: ofOttawa, Ottawa, Ontario, Canada KIN 6N5. suevites (66m), allogenic breccias (112m), and autogenic breccias (deeper than S44 m). These rocks are represented by shocked­ Rutheniwn (Ru) and iridiwn (Ir) are the least mobile platinwn metamorphic gneisses, schists, amphibolites of Archean age, and groupelements(PGEs) withintheCretaceous-Tertiary(K-T)bound­ magmatic rocks (dolerites,olivines, and peridotites) that lie be­ ary clay (BC). The Ru/Ir ratio is, therefore, the most useful POE tween them. interelement ratio for distinguishing terrestrial and extraterrestrial According to thepreliminarydata, theintensity ofthe diaplectic contributions to the BC. The Ru/Ir ratio of marine K-T sections mineral transformation in the crystalline rocks is decreased from (1.77 ±0.53) is statistically different from that of the continental 45-50 GPa at the top to 15-20 GPa at a depth of4 to 4.5 km. The sections (0.92 ±0.28). The marine RuIIrratios are chondritic (C1 = rocks at the upper part of the section are transformed more uni­ 1.48 ±0.09), but the continental ratios are not We discovered an formly. In contrast, one may observe unchanged plagioclase and inverse correlation of shocked quartz size (or distance from the maskelyniteinthesamethinsectionoftherocks from thelowerpart impact site) and Ru/Ir ratio (see Fig. 1). This correlation may arise of the section. Sometimes the thermal metamorphism is superim­ from the difference in Ru and Ir vaporization temperature and/or posed on the shock metamorphism. Due to this superposition, fractionation during condensation from the ejecta cloud. impact glasses and diaplectic minerals are recrystallized to fme­ Postsedimentary alteration, remobilization, or terrestrial POE grained granoblastic aggregates (often monomineralic). The data input may be responsible for the Ru/Ir ratio variations within the obtainedwithSEM (Camscanwithenergy-dispersive analyzerAN­ groups of marine and continental sites studied. The marine ratios 10,000) show the nonisochemical character of the element migra­ could also be attained if ..15% of the boundary metals were tion process, which often accompanies intensive diaplectic contributed by Deccan Trap emissions. However, volcanic emis­ transformation. The result is formation of the aggregated pseudo­ sions couldnothavebeentheprincipalsourceofthePOEs intheBC morphs. They contain components missing from the original min­ because mantle POE ratios and abundances are inconsistent with erals. For example, the shocked, thermally decomposed biotite is those measured in the clay. The Ru/Ir values for pristine Tertiary transformed to a plagioclase-titanomagnetite-pyroxene-glass ag­ mantle xenoliths (2.6 ±0.48), picrites (4.1 ± 1.8), and DeccanTrap gregate; amphibole is decomposed to a andesite-magnetite­ basalt(3.42±1.96)are allstatisticallydistinctfrom thosemeasured clinopyroxene-amphibole aggregate; garnet is decomposed to a in the K-TBC. titanomagnetite-plagioclase-orthopyroxene aggregate. Theprocessofdiaplecticfeldsparanddiaplecticfeldsparglasses formation is also nonisochemical; K-rich, Na-poor diaplectic pla­ gioclases and maskelynites, and Ca-rich, K-poor diaplectic orthoclases (in comparison with the initial feldspars) are revealed. The fmal productofthis process is a melt glass that is composed of 0.7 r------,..------,,....-----,.------, a mixture of both feldspars. Different grains of plagioclases (and different parts ofsome grains) appear to be variously transformed .Roton Bosin becauseofthenonuniformdistributionofthe shockloadintherock. 0.6 Red Deer Volley.L C k Inpolysynthetically twinned plagioclases, transformations always • once ree begin more intensively in one system of the twinning, which is N""'-- t E 0.5 •Morgan Creek revealed under a pressure of20 GPa. Vcry often one system ofthe o E ::J,--, twinning is completely isotropic, i.e., it is transformed into a o Q) 0.4 diaplectic glass, while the other is not yet transformed. In a more '0 .!::! ~(fl transformed system a Nadeficit is observed, i.e., Na is carried out. As a result, anorthite component enrichment and decrease of the g.£ 0.3 .!: 0 swn of cations in the plagioclase formula is observed. Potassiwn (/)(5 Caravaca enrichment ofplagioclase in a more transformed twinning system 0.2 5tevns Klint • is observed for samples containing K-rich minerals. At a load of • Woodside Ck more than 30 GPa, plagioclase is partially or completely trans­ 0.1 Pet~ccio • formed into impact melt glass. Impactmelt plagioclase glasses are Elendgroben often recrystallized into fme-grained secondary plagioclase aggre­

0.0 l...-__---J'-__....-J ~_._____' gate. Sodiwn decrease and more intensive decrease of the swn of 0.5 1.0 1.5 2.0 2.5 cations in the plagioclase formula is observed in the recrystallized Ru/lr impactglasses incomparisonwiththeslightlytransformeddiaplectic plagioclases. This effect may be observed even in a single grain. The investigation ofdiaplectic transformation ofrock minerals Fig. I. in the Vorotilov drill core is continuing. 26 1992 Sudbury Conference

BRECCIA DIKES FROM THE BEAVERHEAD IMPACT dikes consist of angular. slightly deformed clasts in a fme matrix STRUCTURE, SOUTHWESTMONTANA. P. S. Fiskel • S. B. that lacks a fluidal texture. Hougen2• and R. B. Hargraves2• 'DepartmentofGeology. Stanford Type A Brec:cla Dikes: The breccia dikes found in the north­ University. Stanford CA 94305-2115. USA. 2Department of ern three localities are all similar in texture and resemble type A GeologicalandGeophysicalScience.PrincetonUniversity.Princeton breccia dikes described at Sudbury and elsewhere by Lambert [1]. NJ 08544. USA. Thematrixofthedikes is greenish-gray atIslandButte,redtobrown at Erickson Creek. and brown to black at Law Canyon, reflecting While shatter cones are generally accepted as indicators of different compositions of the protolith and different degrees of meteorite impact. other petrologic features are not widely recog­ postimpact alteration. Clasts in all samples are rounded. highly nized in the geologic community. Breccia dikes are one such deformed. and represent local material only. In some clasts the feature: They are found in many large impact sbUctures occurring deformation is most intense at the rims. In the central portion of overan areaatleastasextensively as shattercones [1]. Brecciadikes these clasts. deformation is localized along small shear zones that will survive moderate degrees of metamorphism and tectonism. contain an ultramylonitic material similar in texture to the dike unlike many other microscopic features (shocked quartz grains, matrix. Other clasts are completely pulverized and strung out into high-pressure polymorphs. etc.) and even large-scale features such sinuous schlierenlike bands that fade into the matrix. Some of the as annular or bowl-shaped topographic features. Thus. they are material from Law Canyon contains vesicles. now filled with important diagnostic criteria. especially for large. poorly preserved secondaryquartz andcalcite. Quartz.calcite. and chlorite alsooccur impact structures. as secondary minerals in veins and in rims around some clasts. No The Beaverhead Impact structure is a recently discovered, unambiguous examples of shock lamellae (i.e., multiple Sets of deeply eroded impact structure in southwestern Montana [2]. The planardeformation features) have been found in any samplessofar. remains ofthe structure are delineated by the occurrence ofshatter XRFandmicroprobe studies ofthe brecciasoflslandButteshow cones. found in an area >200 km2•occurring within theCabin thrust thatthematrix isofsimilarcompositionto the wallrock,conftrming plate. part of the Cretaceous Sevier fold and thrust system. The previous studies [2]. butdepleted in Na by a factor of5 andenriched distribution of shattercones is further truncated byTertiary normal in K, Fe, and AI. INAA analyses of the same pairs of samples also faults (Fig. 1). The present remains represent an allochthonous showed the matrix to beenriched in Sc. Co. Rb. and REE. indicating fragment of a larger sbUcture. alterationofthefme-grainedmatrixmaterial.possiblybypostimpact Enigmatic. fluidal-textured breccia dikes have been found in hydrothermalactivity [3]. ICPatomicemissionanalysesofclastand three localities in the areacontaining shattercones. Theserocks are matrix samples from Law Canyon show a similar depletion in Na characterized by rounded, highly deformed clasts of wall rock, and enrichment in K. This trend is seen in impact melts from other suspended in a cryptocrystalline. flow-banded matrix. Twoofthese impact sbUctures [4]. In some samples from Law Canyon clasts outcrops occur in sandstone (Proterozoic GWlSight Formation or have K-rich halos. XRF, INAA. and microprobe analyses are Cambrian Wilbert Formation?) and the third is found in basement currently underway on samples from the other localities. gneiss (Fig. 1). Brecciadikes ofa different nature have been found Despite the presence of vesicles in the samples from Law 15 km south of the region containing shatter cones. Those breccia Canyon and the fluidal texture of the matrix and the schlierenlike deformation ofthe clasts. there is little textural evidence to suggest thatthese rocks weremolten! The intense deformation oftheclasts. the absenceofany melt zones atdikeborders. thecataelastic texture 0; Shanercones (ound ofthe matrix. and thelackofany igneo\lS textures ineven the largest fiI ~bf~ss~~A':!e~~kS of the dikes suggests that these dikes were primarily the product of C2l 5andsone (Cambrian Wilbert Fm.?) extreme cataclasis. Furthermore. cathodoluminescence studies of • Basemenl gneiss the material from Island Butte failed to reveal any evidence for melting [5]. Type B (1) Breccia Dikes: The breccia dikes in the south are very different in texture and composition from those in 'the north. The matrix is cryptocrystalline but does not show any fluidal texture. Clasts within the matrix are angular. slightly deformed. and represent the local lithologies in which these dikes are found. including pieces of brecciated material. Single sets ofplanardefor­ mation features have been found in quartz grains from these breccia dikes. These breccia dikes bear many similarities to type B breccia dikes ofLamben [1]. Their location near the Cabin ~tfault has

Smiles led regional geologists to conclude that these dikes are related to thrusting. Further work is needed to establish the age and origin of these breccia dikes. Confrrmation of the impact nature of these breccias would double the size of the known Beaverhead impact structure. References: [1] Lambert P. (1981) Multi-Ring Basins (p. H. Schultz and R. B. Merrill, cds.). 59-78. [2] Hargraves R. B. et al. Fig. I. Simplified map of the area of the Beaverhead Impact (1990) Geology.1B, 832-834. [3] Kocherl C. and Fiske P. S. (1991) SbUcture, showing thedistribution ofshattercones and the location Meteoritics. 26, in press. [4] DenceM. R. (1971)JGR, 76,5552-5565. of breccia dikes. Geologic map base is modified from [2]. [5] Ramseyer K., personal communication. U'I Contribution No. 790 27

NEW PERSPECTIVES ON THE POPIGAI IMPACT topography ofthe Popigai structure, together with refmed absolute STRUCTURE. J. B. Garvinl and A. L. Dein02, INASA/Goddard, ageestimates, inorder toreconstructthe pre-erosional morphology Geodynamics, Code 921, Greenbelt, MO 20771, USA, ofthebasin,as well as toquantify theerosionorsedimentinfillrates 2Geochronology Center of the Institute of Hwnan Origins, 2453 in the Popigai region. Ridge Road, Berkeley CA 94709, USA. Wehaveassembledan-90-m-resolutiondigitalelevationmodel (OEM) data for the Popigai region (see Fig. 1 for cross sections Therecordoflarge-scalecrateringonEarth isscant, and theonly derived from the 2-D DEM), and are in the process ofattempting to currently "proven" l00-lan-class impact structure known to have reconcile absolute age discrepancies that have resulted from 40Arl formed within the Cenozoic is Popigai, located in the Siberian YfAr radiometric analyses of glass samples provided to U.S. and Arctic at 71.5°N,1l1°E (Masaitis et al., 1975). Popigai is clearly a Canadian investigators over the past five years by Russian impact multiringed impactbasin formed within the crystallineshieldrocks crater specialists such as V. Masaitis (VSEGEJ/St. Petersburg). In (Anabar) and platform sediments of the Siberian taiga, and esti­ 1991 (seeU'SCXXII, pp. 297-298), wereported on 40ArJYfAr laser mates of the volwne of preserved impact melt (i.e., Masaitis and step-heating ages of glass fragments removed from suevite (melt Mashchak, 1986) typically exceed 1700 lan', which is within a breccia)from theintericrcavityofPopigai(providedbyV. Masaitis), factorof2-3ofwhatwouldbepredictedusing scalingrelationships and obtained ages in the -70-60-Marange. Wenow have prelimi­ (Melosh, 1989; Grieve and Pesonen, 1992). In this report, we nary results from the 40ArJYfAr step-heating of six additional glass present the preliminary results of an analysis of the present-day samples from suevite and allogenic breccia from Popigai (again

TABLE 1. Volmnetric analysis of crater..

Model Model Observed Model Interior Volume Melt Interior MaxDepdt Kinetic Dilmeter Age Volume of Excavation Volume Volume to Meltina Energy Crater Name (kin) (Ma) (km') (kin') (km') (kin') (kin') (Meaatone)

Henbury 0.20 0.0040 4.008-04 2.408-04 3.008-08 ? 0.011 1.00B-02 WolfCRlelt 0.94 0.1000 4.208-02 2.508-02 6.008-04 ? 0.065 1.900+00 Darwin 1.00 0.7300 5.108-02 3.008-02 7.008-04 ? 0.070 2.4OE+OO Barrinaer 1.20 0.0490 8.908-02 5.208-02 1.308-03 0.105 0.086 4.500+00 Goat Paddoclt 5.00 55.0000 4.l3E+OO 7.30E+OO l.67B-01 ? 0.431 5.78E+02 BolUllltwi 10.SO 1.3000 2.28E+Ol 4.91E+Ol 2.08E+OO 16.100 1.000 7.200+03 Zltamanaltin 14.40 0.8700 4.71E+Ol 1.08E+02 6.07E+OO 20.100 1.400 2.100+04 Goa_Bluff 22.00 142.0000 1.2SE+02 3.18E+02 2.57E+Ol ? 2.300 8.9OE+04 Popigai· 100.00 34.0000 4.06E+03 1.51E+04 4.42E+03 1300.000 12.700 1.53E+07 •Ale diac:repmtcy: other age is 66 Ma (KfI).

Popigoi : Vo:ume

Popigol••s

d: 0.15km Vi : 728.3gkm'

D: 108.20km Vet,: 2874.60km' '? .;; o o o o

d: O.13km Vi : 480.61km'

D: 99.20km Ve:z,: 2983.78km'

250

Fig. I. 28 1992 Sudbury Conference

provided courtesy ofV. Masaitis). Data analysis is still underway, TABLE 1. ObIerved and model parlllle&en for the ZlF md the RIC u but it is evident that none of the new samples are as well-behaved derived from .wy.i. oftopograpby and _ling relationshilll. in age release patterns as was the original sample, due most likely Parameter 2Jlamanlhin to alteration and the presenceofoldtarget-rock mineral inclusions. Bosumtwi Ref. Predominant ages in these speetta are commonly between -60 and Age (Ma = 10' yr) 0.87 1.3 [2,I4J 40 Ma, butportions ofthe gas release in the -40-30 range are also Apparent diam. Da (Ian) 14.4 10.5 Me... observed. We draw no conclusions as to the age of the Popigai Apparent depth cia (Ian) 0.182 0.300 impact event from these data at this early stage. Planned chemical, ObIerved upcct cWDa 0.013 0.030 hydrogen, and oxygen isotopic analyses may help us sort out the Oba. lit. Rim Ejecta hej (m) 30.3 83.0 Oba. Vol. Cavity Vuv (km') effect alteration has had on the Ar age systematics. It is cwious to 20.1 (max) 16.05 Oba. VuvlSAuv (Ian) 0.018 note thatindependentresults of·Arp'Arlaserstep-heating ofother 0.201 Oba. Vol. Ejecta Vej (km') 16.7 11.6 samples conducted by Bottomley, Grieve, and Yark Grieve, (R. Oba. Tej = VejlSAej (Ian) 0.041 0.049 personal communication, 1992) indicate well-behavedrelease pat­ liV lOll = Vuv-Vej (km') 3.4 4.45 terns that suggest an age in the vicinity of -34 Ma (Eocene­ Tejloll = liV IoIl/SAej (m) 8.3 18.6 Oligocene boundary). At this point, our impression is that a EJER =Tej10lllAge (mmJyr) 0.0095 0.014 combination of analyses of pristine melt glasses and Wlaltered mineralphases isrecommendedinordertoresolvetheagedisparity Model Vol.lnit. Vi (km') 47.1 22.8 Comp. that apparentlyexists withrespecttotheabsoluteageofthe Popigai Model Vol. Excav. Vex (km') 107.9 48.2 .. impact. Model init. depIh di (Ian) 0.436 0.384 Model AIpCCl diJDa 0.030 0.037 Using the high-resolution topography data illustrated in Fig. I, Model Vi/SAi (Ian) 0.289 0.263 we can attempt to reconstruct the initial crater geometry by means Model hej. (m) 360.0 263.0 of standard dimensional scaling relationships, such as those sum­ her = hej·-bej (m) 329.7 180.0 marized in Melosh (1989) and by Grieve and Pesoncn (1992). BRIM = her/Age (mm/yr) 0.38 0.14 Table 1highlightssomeoftheparametervaluesderivedfor Popigai incomparison with a small set ofrepresentative smaller terrestrial IiZ =di-da (Ian) 0.254 0.084 features. Themaximum degree oforiginal reliefatthe crater(floor IiVol. = IVi-Vuvl (km') 27.0 6.75 to rim crest) is between 520 and 960 m (depending on the model Ter =IiVol./SA (Ian) 0.166 0.078 chosen),whilethepresent-daydynamicrangeofreliefis 260-408 m. CER =liV/SA/Ale (mm/yr) 0.19 0.060 This suggests thatbetween260and552mofreliefhasbeen lostdue IiZIAge (mm/yr) 0.29 0.065 to slumping, erosion, and otherprocesses (interiorcavity sediment Erosion Model forTaraet Ie lJ2JU4 leliZUA [3J infill). Ifwe adopt typical erosion models for high-latitude shield Ie in Erosion Model 1.0SE-4 4.2SE-7 [3J terrains (seeGarvinandSchnetzler, this volume),wefmd that upto 0.0052 mm/yr could be eroded at Popigai, which ttanslates into Erosion (mm/yr) @ liZ in m 0.019 0.00016 Comp. -176 m over a 34-Ma lifetime, or 350 m over a 66-Ma lifetime. Erosion (m) for Crater Ale 16.5 0.21 .. Clearly, a refined absolute age for the structure is needed to refme Max. Vol. Eroded (kmS) 2.7 0.018 theseerosionestimates; however, the suggestionisthatPopigaihas experienced up to a factor of5 more erosional infill than the much smaller equatorial shield crater Bosumtwi. (We acknowledge the cooperation of V. Masaitis at the VSEGEI in St. Petersburg for 6km in diameter and atleast4.5 Ma in age [1,10]. However, more providing us with Popigai glass samples on several occasions). recent drilling and geophysical observations at the ZIF have indi­ cated thatits prc-erosional diameter is atleast 13.5 km, and that its age is most probably 0.87 Ma [2,3,7,9,15]. Why the present t0po­ graphic expression of a 13.5-km complex impact crater.less than 1 THE ZHAMANSIUN IMPACT FEATURE: A NEW CLASS m.y. oldmostclosely resembles heavily degradedMesozoic shield OFCOMPLEXCRATER? J. B. Garvinl andC.C.Schnctzlcr, craters such as Lappajarvi is a question of considerable debate INASA/GSFC, Geodynamics Branch, Code 921, Greenbelt MD [6,7,9-11]. Hypotheses for the lack of a clearly defmed craterlike 20771, USA, 2DepartmentofGcography, University ofMuyland, form at the ZIF include a highly oblique impact. a low-strength College Parle MD 20742, USA. "cometuy"projectile,weakorwater-saturatedtargetmaterials,and anomalous erosionpatterns [1,2,6,7,9]. TheproblemremainsWlre­ The record of 10-km-scalc impact events of Quaternary age solved because typical erosion rates within the arid sedimcntuy includes only two ''proven'' impact structures: the Zhamanshin platformenvironment[3] ofcentral Kazakhstan in which theZIFis ImpactFeature(ZIF) and theBosumtwiImpactCrater(BIC). What located are typically low (see Table I); it would require at least a makes these impact landforms interesting from the standpoint of factor of 10greatererosion at the ZIF inorderto degrade the ncar­ recent Earth history is their almost total lack of morphologic rim ejecta typical of a 13.5-km complex crater by hundreds of similarity.inspiteofsimilarabsoluteagesanddimensions.TheBIC meters in only 0.87 Ma, and to partially infill an inner cavity with resembles pristinecomplex craters on the Moon to first order (i.e., 27 km' (an equivalent Wlifonn thickness of infdl of 166 m). Our "U"-shapcd topographic ClOSS section with preserved rim), while analysis ofthe degree oferosion and infill at the ZIFcalls for rates theZIFdisplays virtuallynone ofthetypical morphologic clements in the 0.19 to 0.38 mm/yr range over the lifetime of the landform, of a 13- to 14-km-diamcter complex crater. Indeed, this apparent which are a factor of 10 to 20 in excess of typical rates for the lack of a craterlike surficial topographic expression initially led Kazakhstan semidesert[3]. Ifweapply similarerosional models to Soviet geologists [1] to conclude that the structure was only 5.5 to the BIC, which is located in an equatorial crystalline shield region LPI Contribution No. 790 29

and subject to tropical weathering processes (14), we fmd that the ASTEROIDS AND ARCHAEAN CRUSTAL EVOLUTION: amount of erosion and infill needed to explain its CUITCIlt t0po­ TESTS OF POSSIBLE GENETIC LINKS BETWEEN MAJOR graphic expression is between 0.06 rnm/yr (will) to 0.13 mm/year MANTLE/CRUST MELTING EVENTS AND CLUSTERED (erosion of rim and near-rim ejecta). Of course, the degree of EXTRATERRESTRIAL BOMBARDMENTS. A. Y. Glikson, observed erosion at both the ZIF and the BIC assumes that the pre­ BMR, P.O. Box 378, Canberra, A.C.T., Australia. erosional morphology of these impact· structures can be recon­ structed using established dimensional scaling relationships, such Since the oldest intact terreslrial rocks ofca. 4.0 Ga and oldest as those summarized by Ivanov [4) and Melosh 15). Table 1 ziIcon xenocrysts of ca. 4.3 Ga measured to date overlap with the summarizes the available observational data on the dimensions of lWlar late heavy bombardment, the early Precambrian record re­ the two structures and all our estimates ofparameters that can be quires close reexamination vis a vis the effects of mcgaimpacts. derived on the basis of high-resolution topographic data. Model This includes modeling ofearly megaimpact events (1), examina­ values are listed for comparison on thebasis ofsimple scaling laws tion of the nature and origin of early volcanic activity [2-4], [4,5). A model for teneslrial erosion as a function of geologic examinationofPrecambrian structures [5,6), and closeexamination environment, rock type, and local to regional relief (~) is used to ofthe isotopic age evidence [7). The identification ofmicrotektite­ compute the expected erosionfmfill rates for the regions associated bearing horizons containing spinels ofchondritic chemistIy and Ir with the ZIF and the BIC [3). These model erosion rates are inte­ anomalies in 3.5-3.4-Oa greenstone belts [8,9) provides the fllSt grated throughout geologic time, and as such are upper bounds on direct evidence for large-scale Archaean impacts. The Archaean the rates that would be operational over a time period as short as -1 crustal record contains evidence for several major greenstone­ Ma. Thus, the0.019 rnm/yr thatwould bepredictedfor theZIF does granite-forming episodes where deep upwelling and adiabatic fu­ not take into account that this region of the central Kazakhstan sion of the mantle was accompanied by contemporaneous crustal semidesert has apparently experienced much lower erosion during anatexis. Isotopic age studies suggest evidence for principal age the Quaternary [2). Indeed, the geomorphic record oferosion in the clustersabout3.5,3.0,and2.7(±O.8)Ga.relicsofaCL3.8-Oaevent, ZIP general region has beendominated by eolian redislribution and and several less well defmed episodes. These peak events were deposition ofloess, with probablemaximum accumulation levels in accompanied and followed by protracted thermal fluctuations in the range of 20-70 m within the interior cavity of the ZIF, based intracrustalhigh-grademetamorphic zones. Interpretations ofthese upon Wlpublished drilling results described by Masaitis and Boiko events in terms of internal dynamics of the Earth are difficult to 12). Thus, ourimpression is that itis impossible to reconcile typical reconcile with the thermal behaviour of silicate rheologies in a erosion rates at the ZIP (in the range of0.019 to 0.080rnm/yr) with continuously convecting mantle regime. A lriggering ofthese epi­ what would be predicted (0.19 to 0.38 rnm/yr) given erosion of a sodes by mantle rebound response to intermittent extraterreslrial typical 10- to 15-km-diameter complex impact crater. While the asteroidimpactsissupportedby (1)identificationofmajorArchaean observed erosion attheBIC appears to bewithin afactor of2ofwhat impacts from microtektite and distal ejecta horizons marked by Ir would be predicted using terreslrial erosion models and pre-ero­ anomalies; (2) geochemical and experimental evidence for mantle sional crater dimension scaling laws, that for the ZIP disagrees by upwelling-possibly from levels as deep as the transition zone; and up to a factor of20. We believe that the pre-erosional morphology (3) catastrophic adiabatic melting required to generate peridotitic ofthe initial ZIP cannotbe approximated using traditional complex komatiites. Episodic differentiation/accretion growth ofsial conse- craterscaling relationships, and that the ZIF represents a new class of complex crater form on the Barth that may help to explain the current deficiency ofobserved craters in the 8- to 16-km-diameter range. Furthermore, we believe that it is possible that there are perhaps tens ofZIF-style complex craters preserved, albeit poorly, within the sedimentary platforms ofthe continents [13). Thus, it is important to developmethods forreconstructingZIF-stylecratering events, and for Wlderstanding why sucheventsproduce craterforms with anomalously mundane topographic expressions [11,12). References: (1) F10rensky P. V. and Dabizha A. (1980) ZhanumshinlmpaclCrater, Nauka,Moscow,127pp. [2) BoikoVa. et al. (1991)/mpacICraterZiIomDnshin.Zi:Jpkazgeologiya, 28pp. (guide­ book, V.Masaitis,ed.). [3] MasaitisV. L.etal.(1985) Izveslia Acad. NaukSSSR. Seriya Geol., No. 2,109-114. [4) Ivanov B. A. (1986) CrourinBMechank:s,NASATM-88977,97pp.[S]MeloshH.l.(l989) Impacl Cratering, Oxford, New York, 245 pp. [6) McHone 1. and Greeley R. (1981) In NASA TM-84211, 78-80. [7] Masaitis V. L. (1987)Meleorili/rQ,46,119-123. [8) GrieveR. andPesonenL. (1992) TeclOllOphysics,inpn:ss. [9) Masaitis V. L.etal.(1986)Meteorilika.45, 142-149. (10) Zotkin I. and Dabizha A. (1982) Meleorilika. 40. 82-90. [11) GarvinJ. B.etal. (1992) Teclonophysics. in press. [12] GarvinJ.B.andSchnetzlerC.C. (1988)Eos, 69, 1290.[13] Feldman V. (1991)ThePelrologyoflmpacliles,NASATT-2092S, 399pp. [14] lones W. B. etal. (1981) GSA Bull., 92,342-349. [15) Deino A. and Garvinl. (l990)LPSCXXI, 671-672. Fig. I. Schematic model portraying the conceptofevolution from terreslrial impact basins to greenstone/granite terranes. 30 1992 Sudbury Conference

quent on these events is capahle ofresolving the volume problem The-6OO-MaBWlYeroo Formationconsistsofmaroonandgreen thatarises from comparisons betweenmodemcontinentalcrustand shales, with minor concretionary carbonates, deposited in an outer theestimatedsialproducedbycontinuoustwo-stagemantlemelting marine"shelf setting. The ejecta horizon comprises mainly angular processes. The volume problem is exacerbated by projected high clastsofacidvolcanicsranging from boulder(up to30emdiameter) accretion rates Wlder high Archaean geotherms. In accord with the to fme sand size. All large fragments and most sand-gradematerial model portrayed in Fig. I, it is suggested that impact shock effects were derived from a pink tored porphyritic volcanic rock like that have been largely obscuredby (1) outpouring ofvoluminous basicl at the Acraman impactsite. Theejectasequence varies in thickness ultrabasic lavas, inWldating shock-deformed crust and extending from 0 to 40 em, and is commonly (from base upward) breccia, beyond the perimeters of impact excavated basins; (2) gravity sandy mudstone, and graded sand. Such a sequence probably subsidenceanddownfaultingofterrestrialmaria, accounting for the represents the primary ejectafallout since it(I) is very widespread, burial and anatexis of subgreenstone basement; and (3) extensive (2) displays virtually perfect sorting and normal grading, resulting shearing and recrystallization at elevated temperatures of impact from its settling through a marine watercolumn, and (3) invariably structures, breccias, and mineral deformation features beneath contains a sandy mudstone layer that directly overlies the basal impact-excavated basins, relics ofwhich may be retained in struc­ breccia.Clastsizeanalysisoftheprimaryfalloutscquence indicates tural windows in high-grade metamorphic terranes. Isostatic sub­ that two distinct grain sizepopulations arepresent(gravel andsand sidence and anatexis of thick maria-type piles and Wlderlying sized). These populations may be products of sorting by transport impacted crust resulted in formation of intracrustal comagmatic through the atmosphere or fragmentation processes during impact plutonic and volcanic suites within periods in the order of 15-30 X or subsequent transport. 10' yr,limited by postimpact mantle convection cooling. Repeated Mass flow and stormreworking processes havebeencommonly posttectonic thermal/magmatic fluctuations reflect existence of superimposed on, and in places obliterated this primary sequence. long-term anomalous mantle regions beneath excavated impact To account for various sedimentological features, the following basins, and possibly thermal perturbations related to YOWlger distal sequence of events probably took place: (1) Initial impact occurs, impacts. The broad age zonation of some Archaean terranes sug­ debris is ejected into the atmosphere, and a massive seismic event gestslateral accretionofthemariapiles inaconvection-drivenplate takes place with resulting disruption and slumping of muds in tectonic regime. adjacent marine basins. (2) Ejecta entered the water column, with References: [1] Grieve R. A. F (1980) Precambrian Res., 10, gravel-sized material deposited fmL (3) Deposition of suspended 217-248. [2] Green D. H. (1972) EPSL,15, 263-270. [3] Glikson host muds, together with continued settling of coarse sand, pr0­ A. Y. (1976) Geology, 4, 202-205. [4] Glikson A. Y. (1990) LPI duced the sandy mudstone. (4) Continuedhydrodynamic settling of Contrib. No. 746, 13-15. [5] Goodwin A. M. (1976) In (D. F. sand-sized material produced the graded sand uniL This occurred Windley, ed.), 77-98, Wiley. [6] Weiblen P. W. and Schultz K. J. severalhours afterejectaenteredthe watercolumn,assuming a 200­ (1978) Proc. LPSC9th, 2749-2773. [7] CompstonW. (1990) Third m water depth. Storm waves created during massive atmospheric Inti. ArchaeanSymp., 5-6, Perth. [8] Lowe D. R. andByerly G. R., disruption reached the depositional site during latter stages ofsand Geology, 14, 83-86. [9] Lowe D. R. et al' (1989) Science, 245, deposition and resulted in hummocky and trough cross-stratifica­ 959-962. tion. Evidence supporting an impact origin for the horizon includes the abundance ofshattered mineral grains, the presence ofmultiple THEACRAMANIMPACTANDITSWIDESPREADEJECTA, sets ofshock lamellae inquartz grains, the presenceofsmall shatter SOUTH AUSTRALIA. V. A. Gostinl , R. R. Keays1, and M. W. cones on large clasts, the local abuqdance of altered, tektitelike Wallace1, IDepartment of Geology and Geophysics, University of spherules [6], and anomalous Ir and other POE values [3]. The Adelaide, GPOBox498. Adelaide,500I,Australia,1Departmentof correlation of the Bunyeroo ejecta with the Acraman impact struc­ Geology, University of Melbourne. Parkville, Victoria 3052. ture is further supported by U-Ph ages obtained from severely Australia. shocked, euhedral zircons within theejecta[1]; thedominant ageof 1575 ± 11 Ma for the ejecta is consistent with derivation from the Discovery of a widespread horizon of shock-deformed Gawler Range Volcanics, which has a U-Ph zircon agc'of 1592 ± volcaniclastic ejecta preserved in Late Proterozoic (-600 Ma) 3 Ma. The geographic distribution of the ejecta and the lateral shales in South Australia [2-4,7] and its probable link to the variationofclastsize within the horizon also areconsistentwith the Acraman impactstructure in the Middle Proterozoic Gawler Range Acraman impact site as the source. Volcanics [2,8,9] provide a rare opportunity to study the effects of TheBunyerooejectaisenveloped ingreenshales thatareseveral a major terrestrial impact, including the sedimentology and distri­ centimetersthick [2]. These shales and the sandy layersoftheejecta bution of an ejecta blanket and its precious-metal signature. horizon are enriched in Cu carbonates, barites, and Fe oxides, The ejecta horizon occurs in the Bunyeroo Formation at many minerals that are widespread in sediments ofthe Adelaide Geosyn­ localities within the Adelaide Geosyncline [2,3], including the cline. Geochemical profiles of the ejecta horizon indicate anoma­ Wearing Hills, which are -350kmnortheastofthe Acraman impact lously high Ir, Au, Pt, Pd, Ru, and Cr relative to the host shales of site. Following a search at the same stratigraphic level in other theBunyeroo Formation (Irup to 2.0ppb, Pt upto270ppb).Iridium basins in South Australia, the ejecta has been located within the enrichmentup to 100times the backgroundvaluefor the hostshales Lower Rodda beds ofthe OfficerBasin, extending the limits of the has been recorded. As Ir values for the volcanic rocks that crop out ejecta to -470km northwestofthe Acraman impact structure [4,7]. at the Acraman impactsite are <0.005 ppb,thehigh values forIrand The ejecta is therefore widely dispersed, and provides an important for other POEs and Cr in the ejecta horizon strongly suggest chronostratigraphic marker enabling precise correlation of Late derivation from the impactor itself. The marked enrichment in Ir in Proterozoic sequences in southern Australia. the Bunyerooejectais similartothat in sediments at the Cretaeeous- LPI COtIIributiota No. 790 31

Tertiary boundary, which has been attributed to a major impact Optically, quartz and K-feldspar displayed numerous sets of event. The strong evidence for an impact origin of the B1U1Yeroo planar defonnation features (PDFs) identical to the nondecorated ejecta also points to a cosmic source for its POB signature. PDFs seeninlaboratorysamples andmanynaturalimpactites.Other The shales above and below the B1U1Yeroo ejecta horizon also minerals showed less distinct shock damage, with some fracturing show Ir and Pt enrichments (0.073..0.45 ppb Ir, 3.1-313 ppb Pt), visible in cordierite and hornblende. suggesting postdepositional mobilization ofIrand Pt. Interelement TBM study showed that the POFs in quartz and feldspar corre­ ratios ofthe POBs within the ejecta horizon from different sites are sponded to densely packed wide Il'ansformatioolamellae identical also quite variable, again suggesting postdepositional, low-tem­ to those des<:ribed inlaboratory studies. The transfonnation lamel­ perature mobilization of these elements. Indeed, all green shale laein both minerals were amorphous, withno signofhigh-pressure horizons in the BWlyeroo Formation that were analyzed, regardless phases.InthecaseofK-fcldsparonly,narrowsubllUllellaeextended oftheir stratigraphic position, have relatively high levels ofIr and outward from some wide lamellae. Quartz, which was more abun­ otherPOBs. The diagenetic originofthese anomalies is indicatedby dant and studied more extensively, contained no shock-induced their association with enrichments in Cu-V-Zn-Co-Ni in thin, dislocations. Some planar features were also seen in cordierite, but permeable green-colored reduced beds in a predominantly red bed couldnotbeidentifieddue to rapidbeam damage. No shock defects sequence. A redox precipitation model similar to that invoked for were seen in hornblende in TBM. red bed.Cu-U-V deposits has been proposed to explain the POB The shock-induced defects present at the Sedan site are very anomalies in the green shales [5]. similar to those seen in shock recovery experiments, and also to In summary, the BWlyeroo ejecta is Wlique as the only known those present at certain natural events (e.g., Meteor Crater). This example ofawidely dispersed,coarse-grainedejectablanket thatis, suggests that shock deformation in quartz is notstrongly dependent moreover, strongly linked to a known major impact structure. The on shock pulse duration, and that laboratory recovery experiments marked Ir-POB anomalies in the ejecta horizon provide support for areusefulsimulationsofnaturalimpactevents.Thelackofevidence the hypothesis that meteorite impact events can produce Ir anoma­ for high-pressure phases along transformation lamellae is in agree­ lies in tenestrial sediments. The findings also indicate thatIrcanbe ment with past studies, and supports the idea that direct, solid-state mobilized and concentratedinsedimentsbylow-temperaturediage­ amorphization occurs along transformation lamellae. Finally, no neticprocesses. The identificationofejectahorizonsinsedimentary evidence was seen for decorated PDFs. Presumably decoration is rocks therefore should be based on the coincidence of shock­ due to postshock annealing or weathering. Further work should metamorphic features in the detritus and clear Ir anomalies. focus on processes that lead to decoration oftransfonnationlamel­ References: [1] Compston W. et al. (1987) AIISI. J. Earth lae. Sciencu, 34,435-445. [2] Gostin V. A. etal. (1986) Science, 233, 189-200. [3] Gostin V. A. et al. (1989) NatlU'e, 340, 542-544. [4] Wallace M. W. et al. (1989) Aust. J. Earth Sciences, 36, SIMULATED METEORITE IMPACTS AND VOLCANIC 585-587. [5] Wallace M. W. et al. (1990) Geology, 18, 132-135. EXPLOSIONS: EJECTA ANALYSES AND PLANETARY [6] WallaceM. W.etal. (1990) Meleoritics, 25, 161-165. [7] Wal­ IMPLICATIONS. A. J. Gratz and W. J. Nellis, Lawrence lace M. W. etal. (1990) MwsandEnergy Review, South Australia, Livermore National Laboratory, Livermore CA 94550, USA. 157, 29-35. [8] Williams G. B. (1986) Science, 233, 200-203. [9] Williams G. B. (1987) Search, 18, 143-145. Pastcratering studies have focused primarily oncratermorphol­ ogy. However,importantquestionsremain about thenatureofcrater deposits. Phenomenathat need to bestudied include the distribution OPTICAL AND TEM STUDY OF SHOCK METAMORPH­ of shock effects in crater deposits and crater walls; the origin of ISM FROM THE SEDAN TEST SITE. A. J. Gratz, Lawrence mono- and polymict breccias; differences between local and distal Livermore National Laboratory, Livermore CA 94550, USA. ejecta; deformation induced by explosive volcanism; and the pro­ duction ofunshocked, high-speed ejecta that could form the 11U1ar Thus far, detailed petrologic studies of shock metamorphism and martian meteorites f01U1d on the Barth. To study these phenom­ have been performed onsamples recovered from laboratory experi­ ena, one must characterize ejecta and crater wall materials from ments and on a few natural impactites. The loading history ofthese impacts produced Wlder controlled conditions. samples is quite different: In particular, laboratory experiments New efforts atLLNL simulate impacts and volcanism and study spend only a short time «1 j.1S) at peak pressure, whereas natural resultant deformation. All experiments use the two-stage light-gas impactites may havestress pulses from 0.1-1 ms. Onthe otherhand, gun facility at LLNL to accelerate projectiles to velocities of0.2 to laboratory experiments have known stress histories; natural 4.3 km/s, inducing shockpressures of0.9 to50GPa. We usc granite impactites donoL Natural samples are also subjected to thousands targets and novel experimental geometries to unravel cratering or millions of years ofpostshock annealing and/or weathering. A processes in crystalline rocks. useful intermediate case is that ofnuclear detonation. Stress pulses We have thus far conducted three types of simulations: soft for theseevents canreach 0.1 ms orhigher, and samples are obtained recovery ofejecta, "frozen crater" experiments, and an "artificial in pristine condition. All three types ofloading produce stresses of volcano."Inthe rmtcase,aprojectile impacts agranitedisk,jetting h1U1dreds ofkilobars. ejecta, which is gently recovered in a soft-foam fixture to minimize Samples studied were taken from the Sedannuclear test site, and postejection deformation. In"frozen crater" experiments, a granite consist of a coarse-grained granodiorite containing quartz, K­ block is snugly embedded in a large Al block with a narrow entry feldspar, cordierite, and hornblende. Samples were studied opti­ tunnel for the projectile. The projectile, which deforms on impact, cally in thin section, then were thinned with an ion mill and studied seals most of the ejecta in place, minimizing postimpact material by transmission electron microscopy (TEM). movement and allowing study of the shocked material close to its 32 1992 Sudbury Conference

preshock location. Volcanic simulations use impact projectiles on 8 10 r------.60_ the back surface ofpreheated targets, prodUcing stress waves that Chondrlte Projectiles -.... ". release at the front, unloadingrapidly inmuch the same manner as Gtan/te Target a decompressing magma chamber. ~"~...... -"-Y Ourejectarecoveryexperimentsproduceda useful separationof ..... ~ impactites.Materialoriginallybelowtheprojectileremainedtrapped .~ ... ISlate Islands 30 o>20 Eroded cavityincreaseswiththesizeoftheimpactevent[1-3].Here,weuse W. Clearwater 32 >3~3S Eroded theimpact ofchondriteinto granite at 15,25,and50km S·I tomodel Araguainha 40 >32 Eroded impact-melt volumes at terrestrial craters in crystalline targets and Ctarlevoix S4 >2S Eroded explore the implications for terrestrial craters: details ofthemodel Kara 60 >3S-40 Eroded are given elsewhere [4,5]. Puchezh-KalUn1ti 80 f:::l:::n:i&.U::;@ft:::: Figure 1 illustrates the relationships between melt volume and Manicouagan 100 Eroded Popigai 100 :::::Im~::.i.tm::~~~t:::! fmal crater diameter OR (i.e., after transient-cavity adjustments [5,6]) for observed terrestrial craters in crystalline targets: also Bcucr constrained cstimatcl are shaded. included are model curves for the three different impact velocities. • Simple craten; .II olben arc complex. UI Ct>ntribiUion No. 790 33

1r------,,----"------~_, ~ Chondrite ProjecUles i Chondrite Projectiles Granite Target o c: Granite Target ,.-" 1 ,.­...... I~ i / ~ .. 0.1 ...... / '5 '0 ...... -- ~

~ 0.01 i G) ~ ~ :ii: '0 '5 :6 Terrestrial Gravity ~ 1 Terrestrial Gravity L...... "--~ ~"--~ ~J.-~ ~ 15kms- 2" ...... '--' ~ 0·()()b'-.1~~~'-1 ~~~1'-O~~...... 1'-OO~~~1"t'boo"--' 0 0.1 1 10 100 1000 Final Crater Diameter (km) Final Crater Diameter (km)

Fig. 2. Depth ofmelting along the axis ofpenetration relative to Fig. 3. Volume of melt relative to the volume of the transient the depth of the transient cavity, plotted as a function of[mal crater cavity as a function of [mal crater diameter. The melt volume diamater. Note that. regardless ofthe impactvelocity. melting will approaches that of the cavity at crater diameters above about approach the "base"ofthe cavity even at relatively small diameters. 1000km.

A second implication is that, as the limit ofmelting intersects the large and ancient impact structures. such as those suggested to base of the cavity (Fig. 2). central topographic peaks will be explain meter-thick. areally large spherule beds inthe Archean [14] modified in appearance and ultimately will not occur. That is. the may beunrecognizable in the contextofa classic crater form and its peak will first develop a central depression. due to the flow oflow­ impact deposits. At these sizes. terrestrial impact structures might strengthmelted materials, when the meltvolume beginsto intersect have appeared as low-relief pools of impact melt rocks (107 km3; the transient-cavity base. As the melt volume intersects an increas­ Fig. 1) with little clastic debris and no obvious associated crater ing portion of the transient cavity base, the peak will be replaced structure. Accompanying subsolidus shock effects would be buried upon uplift by a ring. Some of the implications of this mechanism beneath a massive melt sheet and would also tend to anneal out. It for ring formation and observations on other terrestrial planets is would seem. therefore. that such ancient. large structures mightnot given elsewhere in this volume [9]. The morphology of central be recognizable as impact features according to common criteria. structures at complex terrestrial craters was also compiled from the References: (1) Croft S. K. (1983)Proc.LPSC 14th. inJGR, literature [6); again. erosion is a complicating factor as it can both 88.B71. [2) CintalaM. J. and Grieve R. A. F. (1984)LPSCXV, 156. destroy and create topography. Nevertheless. the general trend is [3] Melosh H. J. (1989) ImpQ£t Cratering: A Geologic Process, what would be expected with central depressions at values of DR ~ Oxford. 245 pp. [4] Cintala M. 1. (1992) JGR, 97. 947. [5] Cintala 40 km. and finally rings appearing at DR ~ 100 km. The latter is M. J. and Grieve R. A. F.• this volume. [6] Grieve R. A. F. and equivalent to d,Jdu, values of0.8-0.9 (Fig. 2). and the diameter at Cintala M. J. (1992) Meteoritics. submitted. [7] Robertson P.B. whichrings consistently appear in the terrestrial record is also where (1975) Bull. GSA,86.1630. [8] DresslerB. (1990) Tectonophysu:s, shock pressures in central-upliftstructures record partial melting at 171, 229. [9] Cintala M. J. and Grieve R. A. F.• this volume. (10) ~ ~ 80 km (Table 1). Grieve R. A. F. et al. (1991) JGR, 96. 753. [11] Grieve R. A. F. As crater size increases, the volume of impact melt occupies a (1975) Bull. GSA, 86.1617. (12] PhinneyW. C. et al. (1978)Proc. greaterpercentageofthevolumeofthe transient cavity (Fig. 3).This LPSC9th. 2659. [13] FIoran R. J. et al. (1978) JGR. 83. 2737. implies that less clastic debris is available for incorporation into [14] Lowe D. R. et at (1990) Science. 245. 959. (15] Croft S. K. impact-melt sheets at larger craters. This argument has been used to (1985) Proc. LPSC 15th, in JGR, 88.828. explain. in part. the general lack ofclasts in the bulk ofthe impact­ melt sheet (the Igneous Complex) at Sudbury [10]. There are few detailed studies ofclast-contentvariation in impact-meltrocks. The VENUSIAN IMPACT BASINS AND CRATERED TERRAINS. preserved melt sheets at Mistastin (DR = 28 km) and W. Clearwater Warren B. Hamilton. Mail Stop 964. U.S. Geological Survey. (DR = 32 km) are -100 m thick and have clasts throughout [11,12]. Denver CO 80225, USA. At Manicouagan (DR = 100 km), however, the melt sheet is essentially free of clasts -30 m above its base [13). While this is The consensus regarding interpretation ofMagellan radar imag­ consistent with the implications of the model, it could result from ery assigns Venus ayoung volcanic surface subjected in many areas complete resorption of clasts in the thicker (-200 m preserved to moderate crustal shortening [1-3]. I infer that. on the contrary, thickness) melt sheet at Manicouagan. Ultimately, the volume of ancient densely cratered terrain and large impact basins may be melt could equal or exceed the volume of the transient cavity preserved over more than half the planet and thatcrustal shortening (Fig. 3). In this case (DR - 1000 km) and at larger diameters, the has been much overestimated. I see wind erosion and deposition as resulting final landform would not resemble a classic crater. We far more effective than do others in modifying old structures. venture that terrestrial basins in the lOOO-km size range might have Integration with lunar chronology suggests that most ofthe surface resembled palimpsests, a suggestion made for very large basins on of Venus may be older than 3.0 Ga and much may be older than the Moon and Mercury by Croft [1). Thus, even ifpreserved, very 3.8 Ga. 34 1992 Sudbury Conference

Fig.!. The nearly circular Artemis structure, 2000 km in diameter and centered at 133°E, 35°S, may be the largest impact structure on Venus. The scoured-bedrock ring consists of an inner rim, steepest on the inside, and an outer basin inside a broad, dark outer rim. Other large and small subcircular structures may be the eroded and partly deformed roots ofother impact structures. C2-MIDRP.30S129;I,by JPL.

Broad volcanos, huge volcanic domes, plains preserving lobate Magellan stereo-image pairs. For example, a blowout, longitude flow pattems,and numerous lesservolcanic features, pockedsparsely 073° to 076°E, latitude 2°N to 2°S, stripped deep into the bedrock by impact craters, are indeed obvious on Magellan imagery [4]. of large superimposed craters, is surrounded by a vast swirl of Some of these postvolcanic impact craters have been' slightly connecting erosional canyons, wind streaks, and linear dunes atop extended, but only a small proportion has been flooded by still an eolian plain. younger lavas [5]. Relative ages of the young craters are indicated Numerous possible large magma-flooded impact basins are also by the varying eolian removal oftheir forms and ejecta blankets and preserved. These include many coronae and have nearly circular flow lobes, and the oldest are much subdued [6]. If these young rims, 300-2000 km in diameter, steeper on the inside than the impact craters, maximum diameter 275 km, include all preserved outside. Many are multiring, the inner rims encircled by peripheral impact structures, then their quantity and distribution indicate that basins (some chasmata), outer.broad, subdued rims, and concentric Venus was largely resurfaced by volcanism -0.5 Ga, subsequent and radial fracture systems [8-10]. The interior volcanic plains are eruptions having been at a much reduced rate [5]. commonly higher than plains beyond the rings but lower than the Away from the -0.5-Ga volcanic features, much of Venus is, inner rims. Some large circular basins are superimposed on older however, dominated by circular and subcircular features, 50­ ones. Scaling considerations require that impacts on Venus produce 2000 km in diameter, many of them multiring, that may be mostly larger craters and much more melt than on the Moon [11], and older impact and impact-melt structures substantially modified by venusian basins, like some lunar maria, may be found to have wind action. Eolian erosion scoured to bedrock old ridges and positive gravity anomalies because they are underlain by thick uplands, including those that may be cratered terrains and the rims lopoliths fractionated from impact-melt lakes. The large basins and outer-ring depressions of large impact basins, and removed all have mostly beenregarded as formed by magmawelling upward and surficial deposits to the limits of resolution of the imagery. The outward atop giantplumes but they lack the lobateorirregularforms complementary eolian deposits form not only dunes, wind streaks, to beexpected ofsuchorigins and their abruptcircularorsubcircular and small plains [6,7] but also broad radar-dark plains, commonly rims have yet to be explained in terms of plumes. assumed to be volcanic although lacking flow morphology, whose The inferred heavily cratered terrains consist of arrays of sepa­ materials appear to be thick because they are smoothly compacted rate or overlapping circular to subcircular rims and multiring into buried craters. Plains and erosional features are displayed on complexes 50-1000 km in diameter. Many rims form radar-bright UI Contribution No. 790 3S

stripped-bedrockridgesenclosingradar-darkeolian('l)plains.Other tracts arenowerodedtoalmostcontinuousbedrockdistinguishedby nwnerous much-subdued, large, subcircular rims and basins. The prevailing interpretation of these diverse ring complexes as pro­ duced by crustal shortening and magma upwelling cannot account for their superimposed circular patterns. Misunderstanding of visual illusions in radar imagery detracts from some interpretations. The scale of imagery in the sidelook directionisnothorizontaldistancebutratheris proportional to slant distance. Slopes facing the spacecraft are foreshortened because their tops and bottoms plot close together, whereas slopes facing away are lengthened, an effect opposite to that ofoptical imagery. Symmetrical ridges appear to be hogbacks dipping gently in the directionofradarlook, and suchillusions have been misinterpreted to be thrust-imbricated sheets [2,12]; straight ridges of varying heights can mimic contorted and faulted structures. References: [1]HeadJ.W.etal.(1991)Science,252,276-288. [2] Solomon S. C. et al. (1991) Science, 252, 297-312. [3] BindschadlerD. L. etal.(l992)JGR. 97,inpress.[4] HeadJ.W. etal. (l992)JGR, 97, in press. [5] SchaberG. G. etal. (1992)JGR, 97, in press. [6] Arvidson R. E. et al. (1992) JGR, 97, in press. [7] Greeley R. et al. (1992) JGR, 97, in press. [8] Squyres S. W. et Fig. 1. Southwestern Montana Recess in Sevier Front (from [4]). al. (1992) JGR, 97, in press. [9] Stofan E. R. et al. (1992) JGR, 97, in press. [10] Sandwell D. T. and Schubert G. (1992) JGR, 97, in press. [11] Cintala M. J. and Grieve R. A. F. (1991) USC XXII, Beaverhead shocked rocks come from only the outer part of the 213--216. [12] Suppe J. and Connors C. (1992) USC XXIII, central uplift of what must have been a large (>100 kID diameter) 1389-1390. complex impact structure. Theserocks are allochthonous. They are present in the Cabin thrust plate (one of many in the Cordilleran belt), and are considered to have been tectonically transported50to 150kID east-northeastfrom a sourceineastcentralIdahoduring the WHERE'S THE BEAVERHEAD BEEF? R. B. Hargraves, Laramide orogeny [1,2]. Department of Geological and Geophysical Sciences, Princeton An impacteventofthis magnitude oncontinentalcrust(thought University, Princeton NJ 08544, USA. tohave occurredinlate Precambrianorearly Paleozoic time) could be expected to punctuate local geologic history. Furthermore, Only rare quartz grains with single-set planar (1013) defonna­ although it may now be covered, its scar should remain despite all tionfeatures (PDFs)arepresentinbrecciadikesfoundin association theconsiderablesubsequenterosion/deposition andtectonismsince with unifonnly oriented shatter cones that occur over an area 8 x the impact. The following are three large-scale singularities or 25 kID (see Fiske et al., this volwne). This suggests that the anomalies that may reflect the event and mark its source.

sd-,2O" ,:..:;,·:,..· ...:.;1100 r :"=='~'''COO~' ~11~,..~.·--=....--:'.,',.,.~·----:1.,',.;l!'.-·_-,

4.'

S 10 MUS 100 200 300 KILOMETERS 'i ' II I

Fig. 1. Location ofLemhi Arch (from [3]). Fig. 3. DetailofapexofSouthwesternMontanaRecess(from [4]). 36 1992 Sudbury Conference

REPORT ON THE INTERNATIONAL CAMBODIAN eRATER EXPEDITION-I992. J. Hartungl , C. Kocberl1, P. Lee3, Kuhn Pagnacith4, andTouch Sambath4, 1600 EastFifthStreet, Des MoinesIA50309, USA,1InstituteofGeochemistry, University ofVienna, A-I010 Vienna, Austria, 3404 Space Sciences Building, CornellUniversity,IthacaNY14853,USA,4DepartmentofGeoIogy and Mines, Ministry of Industry, Phnom Penh, Cambodia.

Ithas been proposed thatTonieSap, a lakeinCambodia. 100Ian longand30km wide,marks thelocationofanelongatebasinformed by the oblique impact of a comet or asteroid [1]. The impact is considered to have produced melted ejecta found now as tektites overmuchofsoutheast Asiaand Australia. Thelocation ofthe lake, its approximate age, its size, and the orientation of its long axis (toward Australia) are consistent with this hypothesis. Afterlearningaboutthehypothesis,five individualsvolunteered toparticipateinanexpeditiontoCambodia:JackHartung,Christian Kocberl, Charles Harper, Burkhard Dressler, and Pascal Lee. We Fig. 4. Eastern segment of Aeromagnetic Map of Idaho [5]. agreed that a proper expedition could not be undertaken without a local host or "contact." After a year without progress Hartung 1. The Lemhi Arch (Fig. 1) is a major structural uplift that decided to arrange travel to Phnom Penh, Cambodia, in January occurred in late Proterozoic-early Paleozoic time in East Central 1992. The primary objective of this trip was simply to identify a Idaho and caused the erosion ofatleast4 Ian ofsedimentary cover local contact. A secondary objective was to collect a variety of [3]. This may be directly related to the impact. representative rock samples. In spite of the uncertainty related to 2. Ofthemany thrust sheets comprising the Cordilleranbelt, the getting into the field inCambodia, Lee and Kocberl also decided to Cabin plate that carries the shocked rocks is unique in that it alone make the trip. Early in December 1991, Jolm McAuliff, Directorof intersected the crystalline basement [2]. Italso now marks the apex the U.S.-Indochina Reconciliation Project, infonned us thathe had ofthe SouthwestMontana Recess inthe Sevierfront [4] (Figs. 2 and arranged for us to be received in Phnom Penh by officials of the 3). The basement uplift remaining from the impact may have MinistryofIndustry.Althoughstilluncertainregarding theirstatus, constituted a mechanical obstacle to the advancing thrust sheets in Hartung and Lee arrived in Phnom Penh on January S, 1992, and , Cretaceous time, causing the recess. Perhaps a piece from the were greeted by Kuhn Pagnacith and Touch Sambath, who did all western edge of this uplift was sliced off and transported by the that was possible to make the expedition scientifically profitable Cabin thrust. and personally satisfying. Because we were not from Cambodia. 3. Whatcould be interpreted as a roughly circular aeromagnetic travel into the countryside had to be approved by officials of the anomaly -70Ian indiameter can bediscerned in the state aeromag­ netic map [5] centered about 20 Ian southeast ofChallis, Idaho, in the Lost River range (Fig. 4). It is in approximately the right place, and ignoring the possibility that the anomalies have diverse causes CENTRAL. CAHBOrJ/R and the circular pattern is coincidental, it may mark what remains of the buried central uplift structure! The relevance of these speculations in the search for the source ofthe Beaverhead shocked rocks will beexplored this summerand reported at the meeting. References: [1] Ruppel E. T. and Lopez D. A. (1988) U.S. Geol. Surv.Prof. Paper180,122. [2] SkippB.(1988)Interactionof the Roclcy MOlllllain Foreland and the Cordilleran Thrust Belt (C. 1. Schmidt and W. J. Perry, cds.), 237-266, GSA Mem. 171. [3] Ruppel E. T. (1986) PaleotectonicsandSedimentation inthe Roclcy MOlllllain Region, Uniled States (J. A. Peterson, ed.), 119-130, A.A.P.G. Mem,41. [4] PerryW. 1. etal. (1989) Geol. ResoUl'cesof -,ZON. Montana (D. E. French and R. F. Crabb, cds.), Montana Geol. Soc. 1989 Field Conf. Guidebook. [5] U.S. Geological Survey (1978) o so Aeromagnetic Map ofIdaho.

Af/$tr.li. "-"5ooolr..

Fig. 1. Map of Central Cambodia, modified from [3]. LPI Contribution No. 790 37

TABLE 1. Locations of sites where samples were collected and a brief description ofthe mllCrials colleaed. Site Name Location Roclts Reported

Phnom Krom (Quarry) IOkmSSW Compact cinerites, fermgCl10us scoria, (8 samples) ofSiemReap pink rhyolitic lavas, darlc cinerous sandstones, (AngkorWat) and various volcanic conglomerates

Phnom Basel (Quarry) 22kmWNW Granite, fme-grained, leucocratic, (2 samples) ofPhnom Penh white quartz vein

Phnom OIetares 30bnNW Jointed polychrome jasperites, (I sample) of PImom Penh subvertical schisu

OIealea Village Group 38bnN Devitrified pyromeride, rhyolite, white fluidal (1 samples) of PImom Penh rhyolite, rhyolitic brcc:cias, siliceous darlc dacite

Phnom Batheay 46bnN Upper Indosinias sandstone, vein of (2 samples) of P1mom Penh acidic rhyolitic rock

Tmat Pong Group (Quarry) 26bnW Rhyolitic, weathered whitish, sometimes (2 samples) of PImom Penh fluidal

Phnom OIilO (Quarry) 4SkmS Massive crystalline sandstone, intensely (3 samples) of Phnom Penh eroded black schist

TonIe Bati Massif 33 km SSW Granite, fme-to-medium grained, pseudo- (1 sample) of Phnom Penh vortical biotite flames, black slaty schisu

Cambodian equivalents of the U.S. Departments of Commerce, THE DISTRmUTION AND MODES OF OCCURRENCE OF Interior, State, and Defense. For support in achieving this "diplo­ IMPACT MELT AT LUNAR CRATERS. B. Ray Hawkel and matic" objective we are indebted to Thach Xoval Say, Vice­ 1. W. Head1, IPlanetaryGeosciences,SOEST,UniversityofHawaii, Director, Dept. of Mines and Geology, Sov Chivkun, Director, Honolulu HI 96822, USA, 1Department of Geological Sciences, Dept.ofGeologyandMines,andIthPraing,ViceMinister,Ministry Brown University, Providence RI 02912, USA. of Industry. We were restricted from visiting sites that were "off limits" due either to unknown locations of land mines or known Introduction: Numerousstudiesofthereturnedlunarsamples locations ofmilitary bases. [l-41 as well as geologic and remote-sensing investigations [5,6] . Ourscientific objectives were tofmd impactorshock metamor­ have emphasized the importance of impact melts on the surface of phosed rocks unambiguously related to the Tonle Sap basin, to the Moon. Information concerning the distribution and relative collectsamples ofrocks thatmay representthose melted to produce volumes is importantfor(1) an improvedunderstandingofcratering Australasian tektites, and to learn as much as possible about processes, (2) kinetic energy estimates and energy partitioning Cambodian geology. Using 1:200,OOO-scale geologic maps with studies, (3) the proper interpretation of melt-bearing lunar fairly detaileddescriptions oftherockunits shown [2], we selected samples,and (4) comparative planetology studies. The identifica­ a number ofacceptable "phnoms"(hills thatrise abroptly outofthe tion of major flows of fluidized material associated with impact surroundingplain)thatmaycontainrocks affectedby thepostulated craters on the surface ofVenus has increased interestinimpactmelt TonIe Sap impact A map of central Cambodia is shown in Fig. I, flows ontheotherterrestrial planets. Fora numberofyears, we have and the locations of sites where samples were collected are indi­ been investigating the distribution, modes ofoccurrence, and rela­ cated. A list of those sites, together with a description of the rocks tive and absolute amounts of impact melt associated with lunar reported to be present at each site [2], is given in Table 1. No craters as well as the manner in which melt volumes vary as a obviously shock-metamorphosed or suevite-like rocks were ob­ function ofcrater size, morphology, and targetcharacteristics. The served. Recent alluvium surroundingTonie Sap is judgedtobelake purpose of this paper is to present the results of this effort. sedimentdeposited when the lake surface was ata higher elevation. Method: Impact meltdeposits were identified using the crite­ Acknowledgment: Support provided by the Barringer Crater ria established by Howard and Wilshire [5] and Hawke and Head Company is gratefully acknowledged. [6-8]. Qualitative estimates were made and trends wereestablished References: [1] Hartung1. B.(1990)Meteoritics.25,369-370. utilizing a population of over 100 fresh impact craters that was [2] Dottin O. (1972) Carte GeologiqlU! de Reconnaissanu a 1/ characterized in a previous paper [61, plus additional lunar craters 200,000. Republique Khmere, Phnom Penh, 41 pp., and Siemreap, for which adequate Lunar Orbiter and Apollo photography exists. 16 pp., Editions du Bureau de Recherches Geologiques Minieres, Quantitative determinations ofimpactmelt volumes were made for Paris. [31 Rasmussen W. C. and Bradford G. M. (1977) Ground­ those craters for whichhigh-quality topographic data are available Water Resources o/Cambodia. Geological Survey Water-Supply from Lunar Topographic Orthophotomaps. Paper 1608~P, 122 pp., 3 plates. 38 1992 Sudbury Conference

Results and Discussion: rims oflargercraters: (1) the dominanceoflargeexteriormeltponds Mell occurrence as afunction ofcrater size and morphology. over flows and hard-rock veneer at craters over 50 1an in diameter Impact melt is more common at fresh simple craters (D < -151an) [6,8J; (2) the tendency for melts to occur at greater distances from than has previously been thought. The smallest extensively studied theparentcraters as a function ofcratersize; (3) the observation that craterwith interiormeltis 750m in diameter, butwehavenoted the exteriormeltponds are largerandmorewidespreadatlargercraters; occurrence of even smaller melt-containing craters. At very small and (4) quantitative estimates ofmelt volumes, which indicate that craters (D < 21an), impactmelts typically occuras narrow ponds of relatively moremeltis presenton therims oflargerstructures. Even low-albedo material on crater floors, less common dark streaks on so, this may not imply that a greaterpercentage ofthe total melthas walls, and very thin discontinuous veneers around the rim crests. been ejected since the total amount of melt generated was also Themeltdeposits associatedwithslightlylargersimplecraters (D = relatively greater at larger structures [8J. 2-71an) are similarbutmore abundant than those atsmallercraters. Interior mell volumes as a function ofcrater diameter. There Shallow ponds often occur among the small floor hummocks, and also appears to be a systematic variation in the amounts ofmolten hard-rock veneers cover much of the crater floors. It appears that material in crater interiors. Since the extent and thicknesses of the someofthemeltflowed ontothefloorfrom thelowerportionsofthe ponded material on crater floors tend to increase as a function of crater walls and embayed clastic debris emplaced by mass wasting crater size, more melt may be present in the interiors of larger from the craterwalls. Though some minor wall failure has occurred craters. Support is provided by quantitative estimates of interior at craters in this size range, the positions of these craters on depth­ melt volumes for specific craters where detailed topographic data diameter plots indicate that there has been very little, if any, exist [II,12J. A similar trend has been noted for the impact melt reduction in depth [8-10J. volumes associated with terrestrial impact structures [l3,14J. Interior melt volumes are quite variable in fresh craters from 7 InfllU!nce ofsubstrate onmeltgeneration. Numerouscratering to 12 1an in diameter. These deposits range from unobserved or studies have demonstrated the importance of target characteristics present inonly trace amounts toquite abundant. Extensive deposits indetermining themorphology oflunarcraters [e.g., 15,16J. There­ ofexteriormeltare flfSt observed around cratersnearthe upperlimit fore, we made an attempt todetermine the influence ofsubstrate on of this size range [6,8J. the relative amounts of impact melt associated with craters in Numerous workers havedocumentedthe changes in lunarcrater highland vs. mare terrains. A comparison of the mapped interior morphology and morphomeb'y, which start at a diameter of about melt deposits in similar-sized craters (D < 50 km) suggests that 151an as smaller, simple craters undergo a transition to larger, highlandcraterscontainmeltsinamountseithercqualtoarlessthan complex craters that exhibit central peaks and wall terraces [e.g., the amounts present in mare craters. This observation does not 9,10J. It appears that the crater modification processes operative at necessarily indicate that more melt was generated by impact into craters between 15 and 251an in diameter influence melt deposit mare targets. The observation couldbeexplainedby oneormoreof morphologies and abundances. Whilemostfresh primary craters in the following: (1) for a given impact energy, larger craters may be this diameter range for which adequate photography exists do formed inthe highlands relative tothemare; (2) thestyle and degree contain at least some melt, the amounts are extremely variable. ofwall failure isknown tobedependentonterrain, topography, and Dawes (D=17km) is typical ofcratersin this sizerange. Significant substrate [15J; and (3) a limited amount of evidence suggests that accumulations of impact melt are restricted to a small area imme­ more melt was ejected from highland craters. diately east of the central peak [7J. Additional melt was probably References: [IJ Simonds C. et al. (1976) Am. Mineral., 61, presentinitiallybutwas buriedbyscallopmaterial slumpedonto the 569. [2J Simonds C. et al. (1977) Proc. LSC8th, 1869. [3J TaylorS. crater floor during the modification stage of the impact cratering R. (1982) Planetary Science: A L/Wlr Perspective, 481 pp. [4J event. Freshcraters inthis sizerange that exhibitlittleorno interior SpudisP.etal.(1991)Proc.LPSC, Vol. 21, 151. [5J HowardK.and melt are generally characterized by the presence of extensively Wilshire H. (1975)J. Res. U.s. Geol. Survey,J, 237. [6J Hawke B. scalloped walls and/or swirl-textured floors, features indicative of and HeadJ.(1977)/mpactandExplosionCratering,815. [7JHawke pervasive wall failure [7,8J. The results of our analysis of the B. and Head J. (1977) LSC VJJl. 415. [8J Hawke B. and Head J. interior morphologies of these craters indicate that much of the (1979)LPSC X, 510. [9J WoodC. and Andersson L. (1978)Proc. interior melt was totally buried by scallop material. We conclude LPSC9th. 3669. [10J PikeR.(1977)/mpactandExplosionCratering, that the variable amounts ofinterior melt associated with craters in 489. [lIJ Lange M. and Hawke B. (1979) Conf. LUIIQ1' Highlands this sizerangecanbestbeexplainedbydifferences in thedegreeand Crust, 99. [l2J Lange M. and Hawke B. (1980) Proc. LPSC 10th, style of wall failure. 599. [13J Dence M. et al. (1977) Impact andExplosionCratering, Fresh impactcraters over25-301an in diameter are extensively 247. [14J Grieve R. et aI. (1977) Impact andExplosion Cratering, modified and exhibit terraced walls, central peaks, and flat floors 791. [l5J CintalaM. etal. (1977) Proc. LSC8th, 3409. [l6J HeadJ. with abundant deposits ofimpact melt. Wall failure has been more (1976) Proc. LSC 7th, 2913. extensive anddeep-seated at the larger terraced-walled craters, and little melt appears to have been buried during the modification stage. The results ofdetailed mapping ofinterior and exterior melt LARGEIMPACTSINmEBALTICSIUELDWlTHSPECIAL distributions indicate thatpondedmaterial becomesrelatively more ATTENTION TO mE UPPLAND STRUCTURE. H. Henkel abundant on the floors and rims of these larger craters [8J. and R. Lilljequist, Institute for Fotogrammetty, KTH, 5-10044 Exteriormeltvolumesasaj'UllCtionofcraterdiameter. Previous Stockholm and Department ofGeology, University of Stockholm, work has emphasized the role of oblique impact and preexisting S-I06 91, Stockholm, Sweden. topography in controlling the distribution and amounts ofexterior melts [5,6,8J. While it is clear that these factors do cause variable Within the Baltic Shield several very large structures have been amounts ofmelttobeemplacedoncraterrims, a variety ofevidence identified and are suspected to beofmeteorite impactorigin(Fig. 1 indicates that relatively greater quantities ofmelt are present on the and Table 1). Some ofthese deeply eroded circular features will be LPI Contribution No. 790 39

CIlATEllFORM STRUCTURES IN FENNOSCANDIA Cl....lllc.tlon ". @ It. Impact efllo, JO .1 .. D,C prob.b1e. PO"~ Impact ctl'" :.'5" E !a,oe cilcula, .tructur. "ii';';F ImPtlcl bl'ccct. ,II, 20 .. 'm '---__~__~~__~ ""....' __~_____'ll,X'KN

• OBSERVEO • TONAlITiC 2.81 Mgm" - CALCULATED 0 GRANOOIORITIC. GNEISSIC AND LEPTlTlC 2.71 _ 2.73 '" ,. 1111 . GRANITIC 2.61 - 266

Fig. 3. Gravity profile of the Uppland structure.

H. ~.I' l..J. P.~ ".2 presented with special attention to the Uppland structure, where :u· )0. 55· several indications point toward an impact origin in the mid­ Proterozoic. The structures exceed 100 km in diameter and the topographic expression is inferior or absent. An arcuate arrange­ Fig. 1. Craterfonn structures in Fennoscandia. Nwnbers 45, 46, ment oflithologies occurs aroWld the margin of the structures and and 55 are large suspected impact structures. the central regions show confonn magnetic and positive gravity anomalies. The Uppland structure (Fig. 2) is approximately 320 km in TABLE 1. Location and indication features of large circular structures diameter as expressed by morphological, geological, and geophysi­ in Fennoscandia (compare Fig. 1). cal concentric patterns. The central part is topographically remark­ No Name latIN LatJE Indication Diameter Age ably flat and is characterized by an Wlusual irregular fracture pattern. A subcircular central tonalite with density of 2.81 Mg-3 4S Uppland 60.0 17.0 Top., Grav. 320 Proterozoic gives a positive gravity anomaly of 35 mgal and the gravimetric 46 Nunjes 69.2 20.S Grav, Magn. 200 Proterozoic profile is very similar to that of Manicouagan and Vredefort. The SS Marras 66.9 25.2 Top., Grav., Magn 160 Proterozoic tonalite constitutes a huge antiform, 80 km in diameter, probably representing a 12-km structural uplift of infracrustal rocks (Fig. 3). The flancs of the tonalite are characterized by recrystallized

RADIUS pseudotachylitic breccia dykes and breccia zones. AroWld the 160 km central parts amphibolite-grade metamorphic rocks appear as large

~kfft·. fragments within a fme-grained granite interpreted as a thermally annealed melt rock. Several occurrences of breccia dykes and breccia-bearing melts have been identified about 100 km from the B gravimetriccenterofthe structure. Outsidethe meltring,downslided and eroded crater wall rocks of high metamorphic grade (garnet­ _-6O'N: bearing gneisses and migmatites) occupy most of the terrain. The northwestern quadrant has later been downfaulted and preserves less metamorphic rocks of mainly sedimentary origin and a large quantity of iron deposits and Zn-bearing sulphide mineralizations within huge hydrothermal fields. The northeastern quadrant is affected by a broad shear zone (the Singo Zone) and Rapakivi-type intrusives. The southern sector is overprinted by an east-west striking shear zone (the Sormland Gneiss Zone). lO"' Impact-related ore deposits are located aroWld the margin ofthe

APPROX. lltAMETER ;t~i%it GRANITOID BATHOLITH structure and are interpreted as preexisting downfaulted iron fonna­

... .// GRAVITY PROFilE :~::. GNEISS BELT tions, and deposits fonned from remobilization of these preimpact

f'~ GRAVITY ANOMALY ASPECTS • SHEAR ZONES occurrences. The so-called ball ores are interpreted to have fonned by fluid injection similar to the fonnation of breccia dykes. The -::::: LEPTITES iIi@ TONALITIC DOME extensive hydrothermal alteration along the outer margin of the strueturehavecreatedextremesoda and K-enrichedrocks ("leptites") Fig. 2. Uppland structure. SGB =Sormland Gneiss Belt; SZ = from preexisting gneiss granites and supracrustal sedimentary Singo Shear Zone; PZ =Protogin Zone; VZ =ViisterAs ShearZone; gneisses. As an example, magnetite-skarn deposits are fonned DB =Dala Batholiths; SB =SmAland Batholiths; B =breccia dyke within gneiss granites in hydrothermal cells created during the occurrences. postimpact phase. 40 1992 Sudbury Conference

The other two suspected large impact structures (Fig. 1) have GEOMECHANICAL MODELS OF IMPACT CRATERING: central gravity highs and conformally arranged occurrences of PUCHEZH·KATUNKI STRUCTURE. B. A. Ivanov, Institute metasupracmstalrocks (greenstones) along parts oftheirperiphery, forDynamicsofGcospheres,RussianAcademyofScience,Leninsky here interpreted as parts ofa subsided ring basin. No candidate for Prospect, 38, corp.6, Moscow 117979, Russia. a melt rock has so far been identified in those structures. Impactcratering is a complex natural phenomenonthat involves various physical and mechanical processes [1]. Simulating these processes maybeimprovedusing the data obtainedduring the deep THE PANTHER MOUNTAIN CIRCULAR STRUCTURE, A drilling at the central mound ofthe Puchezh-Katunki impact struc­ POSSmLEBURIEDMETEORITECRATER. Y. W. Isachsenl , ture [2]. S. F. Wright2, F. A. Revetta', and R. J. Dineen4, INew York State A research deep drillhole (named Vorotilovskaya) has been Geological Survey, 2University of Vermont, 'SUNY College at drilled in the Puchezh-Katunki impactstructure (Emopean Russia, Potsdam, 4Roy F. Weston Company. 57°06'N,43°35'E).The ageofthe structureisestimatedatabout 180 to 200 m.y. [1]. The initialrim craterdiameter isestimated at about Panther Mountain, locatednear Phoenicia, New Yark, is partof 40 km. Thecentral upliftiscomposedoflarge blocksofcrystalline theCatskillMountains, which form theeasternendofthe Allegheny basementrocks. Preliminarystudyofthecoreshows thatcrystalline PlateauinNew York. Itis a circularmass dermedphysiographically rocks are shock metamorphosed by shock pressures from 45 GPa by an anomalous circular drainage pattern produced by Esopus nearthe surface to 15-20GPaatadcpthofabout 5 krn [2]. The drill Creek and its tributary Woodland Creek. The mountain is 10km in core allows the possibility ofinvestigating many previouslypoorly diameterandhas a maximumreliefof860m. Itis well displayedon studied cratering processes in the central part of the impact struc­ Landsat images and aerial photographs. Pervasive fluvial cross­ ture. bedding made it impossible to determine whether the structure is As a first step one can use the estimates of energy for the slightly domical, slightly basinal, or unwarped. The circularvalley homogeneous rock target. The diameter of the crater rim may be that rings the mountain is fracture-controlled; where bedrock is estimated as 40krn. The models elaboratedearlier [cf. 3] show that exposed, it shows a joint density 5 to 10 times greater than that on such a cratermay be formed aftercollapseofa transientcavity with either side ofthe valley. Where obscured by alluvial valley fill, the a radius of 10 km. The most probable range of impact velocities bedrock's low seismic velocity suggests that this anomalous frac­ from 11.2to30kmlsmaybeinferredfor theasteroidalimpactor.For turing is continuous in the bedrock underlying the rim valley. the density of a projectile of 2 g/cm' the energy of impact is North-south and east-west gravity and magnetic profiles were estimated as lE28 to 3E28 erg (or about 500,000 Mton TNT). made across the structure. Terrane-corrected, residual gravity pro­ In the case of vertical impact, the diameter of an asteroidal mes show an 18-mgal negative anomaly, and very steep gradients projectile is from 1.5 to 3 krn for the velocity range from 11 to indicate a near-surface source. Several possible explanations ofthe 30 kmls. For the most probable impact angle of45°, the estimated gravity data were modeled. Only one of the computed profiles diameter of an asteroid is slightly larger: from 2 to 4 krn. matched the measured values, namely that of a shallowly buried Forthe homogeneousrock targetonemayexpect40cubickm of meteorite crater with a diameter of 10 km and a breccia lens 3 to impact melt. The depth of such a melt zone is about 3 krn, so two­ 4 km deep, which would pass through the entire Paleozoic section thirds of the probable depth of a melt zone seems to be situated in andperhaps intothecrystallinebasement. Theclosely spacedjoints the limit of the sedimentary layer. Shock heating of the water­ in the rimvalley areinterpreted as the resultofdifferential compac­ saturatedsedimentaryrockstypicallydoesnotproducea continuous tion over the inferred crater rim, leading to bending and dense melt sheet. We need to recalculate the shock attenuation for the fracturing of the bedrock. The magnetic profilell show only small specific geology of the Puchezh-Katunki structure to estimate the variations in intensity over the Panther Mountain area. This is not possible melting in the basement rocks. surprising in view of the significant depth to basement rocks One of the most interesting problems relates to the rock defor­ (-3 km) and the low content of ferromagnetic minerals in the mation history during complex crater formation. In the case of the overlying Paleozoic section. Regional fracture-eontrolled linear Puchezh-Katunki structureonecan usc the level ofshockmetamor­ valleys north and south of Panther Mountain terminate at the rim phism of target rocks as a "label" that marks specific points of the valley. This is consistent with the inferred breccia lens beneath the target. For an estimated projectile energy, the pressure attenuation structure, which would absorb rather than transmit stresses propa­ curve gives the initial length of a vertical column (of 3 krn at the gated upward from the basement. symmetry axis) bounded by the shockpressure 45 GP.and 10GPL We conclude that the Panther Mountain circular structure is When the transient cavity reaches a maximum depth, the column probably a buried meteorite crater that formed contemporaneously seems to be shortened to approximately 1 krn. with marine or fluvial sedimentation during Silurian or Devonian Numerical simulation of the transient crater collapse has been time. An examination ofdrill core and cuttings in the region is now done using severalmodelsofrockrheologyduringcollapse. Results underway to search for ejecta deposits and possible seismic and show thatthe columnat the final position beneath thecentral mound tsunami effects in the sedimentary section. Success would result in is about 5 krn in length. This value is close to the shock-pressure bothdating the impact and furnishing a chronostratigraphic marker decay observed along the drill core. Further improvement of the horizon. model needs totake intoaccountthe blockystructureoftargetrocks revealed by drilling. The model ofcollapse allows theestimation ofthe fmal position ofvariously shockedandheated targetrocks andthe constructionof a thennal modelofthe suberaterspace. Thecompuisonofobserved LPI COIIlributionNo. 790 41

thennalmetamorphic featmes with thethennalmodelmay improve of Ir and other siderophile elements in rocks marking the KIf our knowledge of the geologic process connected with impact boundary and interpreteditas themarkofa giantasteroid(orcomet) cratering. impact, scientists have tried to understand the complexities of the References: [1] Melosh H. J. (1990) Impact Cratering: A KIfboundary event. The impact theory received a critical boostby Geologkal Process, 245 pp. [2] Pevzner L. A. et al. (1992) LPSC the discovery ofshocked minerals that have so far been found only XXIII. [3] Ivanov B. A. (1988) LPSC XIX, 531-532. in association with impact craters [2]. One of the problems of the KIf impact theory was, and still is, the lack of an adequate large crater that is close to the maximum abundance ofshocked grains in THERMOBAROMETRIC STUDIES ON THE LEVACK KIf boundary sections, which was found to occur in sections in GNEISSE8-FOOTWALL ROCKS TO THE SUDBURY Northern America. The recent discovery of impact glasses from a IGNEOUS COMPLEX. R. S. Jamesl , W. PerederyZ, and J. M. KIfsection inHaiti [3,4] has been crucial inestablishing a connec­ Sweeny', IGeology Department, Laurentian University, Sudbury, tion withdocumentedimpactprocesses. Thelocationoftheimpact­ Ontario P3B 2C6, Canada, zINCa B.T.S., Copper Cliff, Ontario glass rrodings and the continental nature of detritus found in all POM INO, Canada, and 'Falconbridge Limited (Exploration), KIf sections supports at least one impact site on or near the North Falconbridge, Ontario POM ISO, Canada. American continent. The Manson Impact Structure is the largest recognized in the Granulite and amphibolite facies gneisses andmigmatites ofthe United States, 35 km in diameter, and has a radiometric age Levack Gneiss Complex occupy a zone up to8 km wide around the indistinguishablefrom thatoftheCretaceous-Tertiary(K/T)bound­ ncxthanpartoftheSudburyIgneousComplex(SIC).Orthopyroxene­ ary [5]. Although the Manson structme may be too small, it may be and garnet-bearing tonalitic and semipelitic assemblages ofgranu­ considered at least one element of the events that led to the lite facies grade occur within 3 km of the SIC together with lenses catastrophic loss oflife and extinction ofmany species at that time. of mafic and pyroxenitic rock compositions normally represented The Manson crater is completely covered by Quaternary glacial by an amphibole ±cpx-richassemblage; amphibolite facies assem­ sedimentary deposits that are underlain by flat-lying carbonate blages dominate elsewhere in this terrain. These 2.711-Gagneisses sediments of Phanerozoic age as well as Proterozoic red clastic, were intruded by (I) the Cartier Granite Batholith during late metamorphic, volcanic, and plutonicrock sequences.In the 35-km­ Archeanto early Proterozoictime and(2) the SIC, at 1.85 Ga, which diameterzone thatmarks theextension ofthe craterthenormal rock produced a contact aureole 1-1.5 km wide in which pyroxene sequence is disturbed due to the impact, and at the center of the hornfelses are common within 200-300 m ofthe contact. structmegraniticbasementrocks arepresentthathavebeenuplifted A suite of12 samples including both theopx-gtand amphibole­ from about4 kmdepth. TheMansonstructure was established as an rich rock compositions have been studied; typical mineral compo­ impact crater on the basis of its geomorphology (circular shape, sitions are OpxXmg =0.55-0.60, GtXpyr =0.12-0.32, PIgAn = central uplift), the presence of shock metamorphic features in 0.25-0.40 in the felsic and pelitic rocks; in the mafic gneisses Cpx minerals (e.g., multiple sets ofplanarlamellae in quartz), Bouguer has Xdi =0.65-0.77 and Al-Tsch =0.036-0.043 and amphibole gravity data, aeromagnetic and ground magnetic data, as well as compositions are Bdenite with (Na+k) =0.52-0.77 and Si(iv) = seismic surveys [6]. 6.4-6.9. Garnets inthesemipeliticgneisses are variably replacedby Detailed studies of the geochemistry of Manson target rocks a plg-bio assemblage. Thennobarometriccalculations using a vari­ (approximatedbythedrillcoresamplesoftheBischeid#1 well,near ,ety of barometers and thennometers reported in the literature the crater) and impact melt rocks and breccia samples [7] have suggestthatthegranulite facies assemblages formed atdepths in the shownthatitis possible to reproduce thechemistryofthemeltrocks 21-28-km range (6-8 kbar). Textures and mineral chemistry in the and breccias by mixing various basement rocks. The elemental garnet-bearing semipeliticrocksindicatethatthis terrainunderwent abundances in the black glasses found at the Haiti KIf boundary a second metamorphic eventduring uplift to depths in the 5-II-km section are not incompatible with the ranges observed for target range (2-3 kbar) and at temperatures as low as 500°-550°C. This rocks andsome impactglasses found atthe Manson crater[7]. Most latterevent is distinct from thennal recrystallization caused by the elemental abundances measured in the black glasses are within the emplacement of the SIC; it probably represents metamorphism range for the Manson rocks, and elemental ratios such as Th/U and attributable to intrusionofthe CartierGraniteBatholith. Thesedata La/Th are also compatible. The Rb-Sr and Sm-Nd isotopic signa­ allow two interpretations for the crustal uplift of the Levack tures of the black glass are compatible with a continental crustal Gneisses: (1) The gneisses were tectonically uplifted prior to the source [3]. Inprinciple, this would apply for Manson rocks, but no Sudbury Bvent (due to intrusion ofthe Cartier Batholith); or(2) the definiteconclusioncanbe madeas theisotopiccharacteristicsofthe gneisses were raised to epizonal.levels as a result of meteorite Manson rocks are not yet known. The yellow glasses, on the other impact at 1.85 Ga. hand, may require a different source material, as norocks with such high Sr or S content have been observed in sufficient quantities in the Manson target rock stratigraphy. However, the target rock mECRETACEOUS-TERTIARY(Kff)IMPACT: ONEOR stratigraphy at Eischeid indicates abundant carbonates. I suggest MORE SOURCE CRATERS? Christian Koeberl, Institute of that a more dermitive answer can be obtained in the near future, Geochemistry, University of Vienna, Dr.-Karl-Lueger-Ring I, A­ when the samples from the newly drilled cores at the Manson 1010 Vienna, Austria. structure are analyzed in more detail. These cores are just now becoming available for studies. The Cretaceous-Tertiary (KIf) boundary is marked by signs of A second candidate for theKIfboundary crater is the Chicxulub a worldwide catastrophe, marking the demise ofmore than 50% of structure, whichwas rlfStsuggested to be animpactcratermorethan all living species. Bversince Alvarez et al. [1] found an enrichment a decade ago. Only recently, geophysical studies and petrological 42 1992 Sudbury Confereru:e

(as well as limitedchemical)analyseshave indicatedthatthis buried At present we can conclude that the Manson crater is the only structure may in fact beofimpact origin [8]. The impact origin was conflDned crater of KIT age, but Chicxulub is becoming a strong recently confinned by the discovery ofunambiguous evidence for contender; however, detailed geochemical, geochronological, and shock metamorphism, e.g., shocked quartz and feldspar [9]. The isotopic data are necessary to provide definitive evidence. stratigraphy ofthe crater and the exact succession and age ofrocks Acknowledgments: I thank J. Hartung, R. R. Anderson, and are not entirely clear at this time, largely because the structure is V. L. Sharpton for Manson and Chicxulub samples and valuable now buried under about 1 km of Tertiary sediments, mainly lime­ discussions. Worksupported by Austrian"Fondszur Forderung der stone, andbecauseoflimitedsample availability due to thedestruc­ wissenschaftlichen Forschung," Project P8794-GEO. tionofcoresamplesina fllC. Thesedimentary sequence (composed References: [1] Alvarez L. W. et al. (1980) Scieru:e, 208, mainly ofcarbonates andevaporites) overlies a basementat3-6km 1095-1108. [2] Bohor B. F. et al, (1984) Scieru:e, 224, 867-869. depth that is inferred to be composed of metamorphic rocks. If [3] SigurdssonH. (1991)Nalwe, 353, 839-842. [4] KoeherlC. and Chicxulubwas formedby impactata time atorbeforetheendofthe Sigurdsson H. (1992) GCA. 56. inpress. [5] KunkM. J. etal. (1989) Cretaceous, thepreimpactsurface consisted largely ofrocks for the Scieru:e, 244, 1565-1568. [6] Hartung J. B. and Anderson R. R. carbonate-evaporitesedimentarysequence,probablyreleasinglarge (1988)LP1Tech. Rept. 88-08, Lunar and Planetary Institute, Hous­

quantities of CO2 and S02 into the atmosphere. ton, 32 pp. [7] Koeherl C. and Hartung J. B. (1992) Proc. LPS, Chicxulubcontains abundant carbonate,limestone, and evapor­ Vol. 22, 111-126. [8] Hildebrand A. et al. (1991) Geology, 19, ite rocks, and the presence of andesitic rocks has been reported 867-871. [9] Marin L. E. (1992) Scieru:e, submitted. (10] Pal D. K. (whichwouldmakeitacandidateforthesourceofthe Haitiglasses), (1982) Scieru:e, 218. 787-789. [11] Glass B. P. and Wu J. (1992) although it is not clear ifthe "andesite" is a real andesitic bedrock, LPSC XXIll. 415-416. (12] Izett G. A. (1991) &S. 72,279. ormakes up theproposedmeltsheet. Thereare someproblems with Chicxulub being the source for the Haiti impact glasses (and therefore for parts ofthe claystones at some KIT boundaries). For this discussion, we need to review the origin oftektites and impact TEKTITE ORIGIN BY HYPERVELOCITY ASTEROIDAL glasses. Rb-SrandSm-Ndisotopic systematics oftektites show that ORCOMETARY~ACT:THEQUESTFORTHESOURCE thesourcematerialwasPrecambriancrustalterrane(fromNdmodel CRATERS. Christian Koeberl, Institute of Geochemistry, ages), and that the sediments that were latermelted to fonn tektites University of Vienna, Dr.-Karl-Lueger-Ring I, A-I0tO Vienna. were weathered and deposited at (for the Australasian tektites, for Austria. example) about 167 Ma ago and probably comprised Jurassic sediments. Further evidence for a sedimentary precursor comes Impact Origin of Tektites: Tektites are natural glasses that from the study of cosmogenic radionuclides. Pal et al. [10] fU'St are chemically homogeneous, often spherically symmetrical ob­ reported that the lOBe content of Australasian tektites cannot have jects several centimeters in size, and occur in four known strewn originated from direct cosmic ray irradiation in space or on Earth, fields onthesurfaceoftheEarth: theNorthAmerican,moldavite(or but can only have been introduced from sediments that have Central European), Ivory Coast, and Australasian strewn fields. absorbed lOBe that was produced in the terrestrial atmosphere. This Tektites found within such strewn fields are related to each other is an extremely important observation. The recent discovery of withrespect totheirpetrological. physical. and chemicalproperties Glass and Wu(11],thatimpactdebrisispresentinthesamedeepsea as well as their age. A theory oftektite origin needs to explain the core layers as microtektites, gives further proofofan impact event similarity oftektites inrespect to age and certain aspects ofisotopic leading to the production of tektites. and chemical composition within one strewn field, as well as the For Chicxulub, a major problem is the production of impact variety of tektite materials present in each strewn field. glasses (or "tektite-like" glasses), which originate, as I have just Inaddition to tektites onland, microtektites (whicharegenerally mentioned, from the surfacelayers ofthe target area. However, any less than 1 mm in diameter) have been found in deep-sea cores. "andesitic" rock or other basement rocks at Chicxulub were obvi­ Tektites are classified into three groups: (1) normal orsplash-form ouslycoveredby carbonates andevaporites ofup to several kilome­ tektites, (2) aerodynamically shaped tektites, and (3) M\J9Ilg NODg­ ters thickness. We therefore cannot conclude, at least not with the type tektites (sometimes also called layered tektites). The aerody­ data presently available, that Chicxulub is the most logical source namic ablation results from partial remelting of glass during for the Haiti glasses. Although the "andesite"present at Chicxulub atmospheric passage after it was ejected outside the terrestrial is similar in composition to the black glasses [8], other rocks that atmosphere and quenched from a hot liquid. Aerodynamically will be mixed in upon impacthave trace-elementsignatures that are shaped tektites areknownmainly from the Australasianstrewnfield notcompatible with any glass composition. Anotherproblem is the where they occur as flanged-button australites. The shapes of obvious lack of quartz-bearing rocks at Chicxulub, which poses splash-form tektites (spheres, droplets, teardrops, dumbbells, etc., problems for the explanation ofthe abundance ofshocked quartz at or fragments thereof) are the result of the solidification ofrotating almost all KIT boundaries. This has ledsomeresearchers [e.g., 12] liquids in the air or vacuum. to propose that two impacts, involving Chicxulub and Manson, Mainly due to chemical studies, it is now commonly accepted might be responsible for the KIT event. In view of the preliminary that tektites are the product ofmelting and quenching ofterrestrial nature ofsome data we refrain from speculating on such an origin. rocks during hypervelocity impact on the Earth. The chemistry of Other proposed impact locations, such as near Kara Crater, which tektites is in many respects identical to the composition of upper was suggested by Russian scientists to beofKIT age, havenotbeen crustalmaterial(1 ,2]. Traceelements areveryusefulforsourcerock conflDned. Precise Ar-Arages ofKarashow thatitis mostprobably comparisons: theratios of,e.g., Ba/Rb, K/U, ThlSm, 8m/Sc,ThlSc, toooldtobeassociated with the KIT boundary,and it is also situated K vs. K/U in tektites are indistinguishable from uppercrustal rocks. onthe wrong side ofthe Earth, as itwas inferred (see above) thatthe The chondrite-normalized REB patterns oftektites are very similar impactcrater(s) are mostlikely nearthe North American continent. to shales or loess, and have the characteristic shape and total U'I COIIlriblllion No. 790 43

abundances of the post-Archean upper crost The determination of Stauffer[8] analyzedthedistributionofAustralasiantektites and the exact source rocks of tektites is complicated because a variety microtektites and found that they do not show a homogeneous ofinhomogeneous targetrocks were sampledbytheimpactMuong distribution. There are radial and concentricpatterns and zones that Nong-type tektites aresimilartoimpactglasses and becauseoftheir do not contain microtektite-bearing deep sea cores. Stauffer sug­ sizeandshapeitis assumedthattheyhavenottraveledfar from their gested• craterthatmaybeconcealedbeneathalluvialdepositsofthe location oforigin, and may thereforeprovide information aboutthe lowerMekong Valley area. A similar analysis was done byKocherl crater location. [9] for the North American strewn field, where I suggested that ThediscoveryofthetektitelocationsatBarbados andDSDPSite tektites show a raylikedistribution, notunlike lunarcraterejecta. A 612 in the North American strewn field is important because possible off-shore impact location (about 175 kID to the eastofthe microtektites and tektites (tektite fragments) as well as shocked Vietnam seashore) was suggested by Schnetzler et al. [10] from minerals are found inthe samelayer. A very importantobservation satellite gravity data. Underwater craters must exist on Earth but, has recently been made by Glass and Wu [3], who showed that withoneexception,havenotyetbeenfound. Hartung [11] proposed several microtektite-bearing layers in cores from the Australasian that the lake Tonie Sap (100 kID long and up to 35 kID wide) in and North American strewn field contain shocked minerals (quartz Cambodiais theresultofthe Australasian tektite sourcecrater. The and feldspar), vesicular impact glass, coesite, and possibly even dimensions areprobably minimum values as the slrUCture is almost .stishovite. This discovery provides an immediate link of tektites completely filled with alluvium. TonIe Sap would be in agreement with an impact event with chemical and isotopic data for tektites, but more detailed Shaw and Wasserburg [4] have shown that the crostal material studies are necessary. The new discovery [3] of impact debris that weathered to form the parent sediments for the Australasian (shocked minerals) in deep-sea cores near the Indochina coast, as tektites have Nd model ages of about 1.15 Ga, and that Rb-Sr data well as the fact that the quantity of both impact debris and micro­ pointtoa finalsedimentationoftheirparentmaterialaround250Ma tektites in the cores increases toward Indochina, is in support of a ago. Recently, Blum et al. [5] have studied the Rb-Sr and Sm-Nd crater in this area. Similarly, Kocherl [9] suggested that the North isotopic systematic ofMuong Nong-type and splash-form tektites, American tektite source crater is in the area ofeastern coast ofthe and found that the source material was Precambrian crostal terrane North American continent, maybe underwater. This was also sup­ (from Ndmodel ages),lJld thatthe sediments thatwere latermelted ported by the fmdings ofGlass and Wu [3]. toformtektites wereweatheredanddepositedabout 167Maagoand The exact mechanism oftektite production during the impact is probably comprised Jurassic sediments, which are not uncommon still not known in detail, but obviously the production of tektites throughoutIndochina. Furtherevidenceforasedimentaryprecursor requires special conditions because otherwise more than just four comes from the study of cosmogenic radionuclides. Pal et al. [6] tektite strewn fields would be associated with the known impact fmt reported that the lOBe content of Australasian tektites cannot craters.Obliqueimpactseemsa possibilitybecauseoftheasymmet­ have originated from direct cosmic ray irradiation in space or on ric distribution oftektites within a strewnfield. Furthermore, in the Earth, butcanonly have been introduced from sediments that have two cases where craters are known to be associated with tektite absorbed lOBe that was produced in the terrestrial atmosphere. fields, the craters are never in the centerofthe strewn field. Jetting WherearetheSource Craters? During thepasthalf-century, mightcontributetoimpactmelts,butitseems thatmaterialoriginat­ numerous suggestions and educated guesses have been made re­ ing from jetting may be composed predominantly of projectile garding the location of the possible source craters for the tektite material, and projectile signatures in tektites are not well pro­ strewn fields. Relatively reliable links between a crater and the nounced, excluding a major projectile component. An initial melt­ respective tektite strewn field have been established between the ing phase during the compression stage, before the formation ofthe Bosumtwi (Ghana) and the Ries (Germany) Craters and the Ivory crater in the excavation stage, is most likely responsible for the Coast and the Central European (moldavite) fields, respectively. tektiteproduction. Tektites havetooriginate from targetrocklayers However, no large crater of the required ages are known for the close to the surface because otherwise it is not possible to explain AustralasianandNorthAmericanstrewnfields. Forthe Australasian their lOBe content. Theexpandingvaporplumeafterthe impactmay field, many proposals for possible craters were made and later be important in distributing the tektite material (which is on the discounted(includingsourcecratersinAntarctica,ortheElgygytgyn order of 10' t for the two larger strewn fields). orZhamanshin Craters). References: [1] TaylorS.R.(1973)EanhSci.Rev.,9, 101-123. Wasson [7] suggests that the tektites in the Australasian field [2] Kocherl C. (1986) Ann". Rev. Earth Planet Sci., 14, 323-350. may have originated in a multitude of small craters scattered over [3] Glass B. P. and Wu J. (1992) IPSC XXIII, 41~16. [4] Shaw allofIndochina.Therearenumerousobjections,including (1) small H. F. and Wasserburg G. J. (1982) Earth Planet. Sci. Leu., 60, craters produce small to negligible quantities ofrelatively inhomo­ 155-177. [5] Blum J. D. et al. (1992) GCA, 56, 483-492. [6] Pal geneous impactglasses, as iswellknownfrom many impactcraters; D. K. et al. (1982) Scien£e, 218, 787-789. [7] Wasson 1. T. (1991) (2) small impact events are unable to provide the energy to launch EPSL, 102, 95-109. [8] StaufferP. H. (1978)3rdRegionaIConj".on the (associated) splash-form and aerodynamically shaped tektites; Geology and Mineral Resources of SOlllheast Asia, Bangkok, (3) the isotopic data do not seem to be in agreement with the 285-289. [9] Kocherl C. (1989) Proc. IPSC 19th, 745-751. multitude ofdifferent source rocks that are required by a multiple [10] Schnetzler C. C. (1988) GLR, 15, 357-360. [11] Hartung J. B. impact theory; (4) the crater problem has been multiplied-instead (1990) Meteoritics, 25, 369-370. of one missing crater, there is a multitude of missing craters. In agreement with most other studies I therefore prefer a single large impact crater. 44 1992 Sudbury ConfereN:e

EARLY ARCHEAN SPHERULE BEDS OF POSSIBLE Acknowledgment: We gratefully acknowledge the support IMPACT ORIGIN FROM BARBERTON, SOUTH AFRICA: from Eastern Transvaal Consolidated (Anglovaal Pty. Ltd.), par­ A DETAILED MINERALOGICAL AND GEOCHEMICAL ticularly Mr. Mauritz van den Berq (ChiefGeologist). STUDY. Christian Koeberll , WolfUwe Reimold2, and RudolfH. References: [I) Lowe D.R. andByerlyG. R. (1986)Geology, Boer2, Ilnstitute ofGeochemislJ'y, University ofVienna, Dr.-Karl­ 14,83-86. (2) Lowe D. R. (1989) ScieN:e, 245, 959-962. [3) Kyte Lueger-Ring 1, A-I0I0 Vienna, Austria, 2Economic Geology F. T. et al. (1992) GCA, 56, in press. Research Unit, University of the Witwatersrand, P.O. Wits, Johannesburg 2050, South Africa.

TheBarbertonGreenstone beltis a 3.5- to 3.2-Ga-old formation U-Ph ISOTOPIC RESULTS FOR SINGLE SHOCKED AND situated in the Swaziland Supergroup near Barberton, northeast POLYCRYSTALLINE ZIRCONS RECORD 550-65.5-Ma Transvaal. South Africa. The belt includes a lower, predominantly AGES FOR A K-T TARGETSITE AND 2700-1850-Ma AGES volcanicsequence,andanuppersedimentary sequence (e.g., theFig FOR THE SUDBURY IMPACT EVENT. T. E. Kroghl , S. L. Tree Group). Within this upper sedimentary sequence, Lowe and Kamol , andB.F.Bohor2,IJackSatterlyGeochronologyDepartment, Byerly [l) identified a series of different beds of spherules with Royal Ontario Museum, 100Queen's Park,Toronto, Ontario, MSS diameters of around 0.5-2 rom. Lowe and Byerly [lJ and Lowe et 2C6,Canada,2U.S. GeologicalSurvey,Box26046,MS972,Denver al. (2) have interpreted these spherules to be condensates of rock CO 80225, USA. vapor produced by large meteorite impacts in the early Archean. This interesting hypothesis is based mainly on the structure of the The refractory mineral zircon develops distinct morphological spherules, which is reported to be similarto quench structures, and features during shock metamorphism and retains these features on the discovery of Ir anomalies of up to several hundred ppb in under conditions that would anneal them in other minerals. In someofthe spherule beds [2). Although Loweetal. [2) reported the addition, weakly shocked zircon grains give primary ages for the abundances oftheplatinumgroupelements (POEs) to be ofroughly impact site, while highly reconstituted (polycrystalline) single chondritic proportions, a more detailed study by Kyte et al. [3) grains give ages that approach the age ofthe impactevenL Datafor showed that the POEs are fractionated relative to the chondritic a series of originally coeval grains will define a mixing line that abundances. They interpreted this to be due to later hydrothermal gives both of these ages providing that no subsequent geological alterations. disturbances have overprinted the isotopic systematics. In this The study ofimpacts early in the history ofthe Earth is ofgreat study, we have shown that the three zircon grain types described by importance and interest; we feel that therefore a more detailed Bohor (this session), from both K-T distal ejecta (Fireball layer, investigation of the Barberton spherule beds is warranted, espe­ Raton Basin, Colorado) and the Onaping Formation, represent a cially becausenodetailedmineralogicalstudy ofthe spherules (and progressive increase in impact-related morphological change that ofall secondary mineralizations such as abundant sulfide mineral­ coincides witha progressiveincreaseinisotopicresetting inzircons ization) and no detailed geochemical stratigraphy (including, e.g., from the ejecta and basement rocks. Unshocked grains are least the rare earthelements) is available so far. The host phase ofthe Ir affectedby isotopic resetting whilepolycrystalline grains are most (and POE) anomaly is also unknown. affected. We have collected a series ofsamples from drill cores from the V-Ph isotopic results for 12 of 14 single zircon grains from the Mt. Morgan and Princeton sections near Barberton, as well as Fireball layer (Fig. 1) plotonorclose to a line recording a primary samples taken from undergroundexposures in the Shebaand Agnes ageof550±10 Maand a secondary ageof65.5 ±3 Ma. Datafor the mines. These samples seem much betterpreserved than the surface least and most shocked grains plot closest to the primary and samples describedbyLoweandByerly [I) andLoweetal. (2). Over secondary ages respectively. The two other grains each give ages a scale ofjust under 30em, several well-defmed spherule beds are between 300 and 350 Ma. This implies that the target ejecta was visible, interspaced with shales and/orlayers ofbanded iron forma­ tion. Some spherules have clearly been deposited on top of a sedimentary unitbecause theshalelayershows indentions from the overlying spherules. Although fresher than the surface samples (e.g., spherule bed S-2), there is abundant evidence for extensive alteration,presumablybyhydrothermalprocesses.Insomesections of the cores sulfide mineralization is common. For our mineralogical and petrographical studies we have pre­ pareddetailed thinsections ofallcore andundergroundsamples (as well as some surface samples from the S-2layer for comparison). For geochemical work, layers with thicknesses in the order of 1-5 rom wereseparatedfrom selectedcoreandundergroundsamples. .a3 Thechemicalanalyses arebeingperformedusingneutronactivation analysis in order to obtain data for about 35 trace elements in each .a2 sample. Major elements are being determined by XRF and plasma Z8? ns spectrometry. To clarify the history of the sulfide mineralization, Pb/ U sulfur isotopic compositions are being determined. We hope to be • 1 .2 .3 .. .S able to identify the host phase of the platinum metal anomaly by separating spherules and matrix. At the time of the conference we will report on the fmt geochemical and mineralogical results and Fig. 1. V-Ph data for single zircons from the K-T boundary their bearing on the impact hypothesis. ftreball Layer, Raton Basin, Colorado. [PI ConlribUlionNo. 790 45

(2.2% discordant). These ages reflect the age spectra ofthe nearby :::; Levack Gneiss :!:7Ma~ Archeanbasement. A volcanicoriginat 1850Maisimpossible since :~ Sudbury Ontario 12711 ?1 ·, ",a the zin:ons wouldhavetobe 100% xenocrysticinorigin and thishas 2sa8 / m ! never been observed in volcanic roclcs.

· .8 • I ::~e" ~...cracked I :H5a m I INFLUENCE OF THE PRESHOCK TEMPERATURE ON m // I · '6 SHOCK EFFECTS IN QUARTZ, F. Langenhorst and A. l'8~ 1'/ i ~ Deutsch, Institut fllr Planctologie, Wilhelm-Klemm-Sir. 10, D­ ." Z3'a / m, • cracked and clear Ii 4400 MOnster, Germany. 2388. //1 • ~ cracked and miley .-U1QIo grain · '2 m-muIII-grain Shock metamorphic features are the prime indicators for recog­ /0 ZI7 US nizing impactphenomenaonEarthandotherplanetarybodies [1,2]. ~ //1850 Ma f'b/ In the past, many shock-loading experiments were performed at · •...."'-----:r---~:8.-----;':;-1----r1;-2----;"·3 1 room temperature [3-5] in order to artificially reproduce shock effects occurring in nature. Results ofsuch experiments have been fig. 2. U-Pb datafor single zircons and multiple zircon fractions extensivelyused as a databaseforshockwavebarometry ofimpact­ from the Lcvack gneiss, Sudbury, Ontario. metamorphosed terrestrial and extraterrestrial roclcs. Although the pressure dependence ofshock features is well known, information about the influence of the preshock temperature is almost lacking dominated by SSQ-Ma roclcs and that the recrystallization features (except [6-8]). Especially in the case of large-scale impacts like ofthezirconwere superimposedduring the impacteventat65.5 Ma. Sudbury,itisexpectedthatdeep-seatedcrustalrocks weresubjected Datafor four ofthe five polycrystalline grains are colinear(33% to shock at elevated temperatures. probability of fit), with data displaced 49%,58%, 62%, and 82% Therefore, we continued to perfonn shock experiments at el­ from 559±5 Matoward the yOWlger event andprovide the bestage evated temperatures on34 GPa. 46 1992 Sudbury ConfereN:e

frequency [%] frequency [%] 30 20 (a) QUARTZ (b) QUARTZ 25 15 20

15 10

10

5

30 60 gO 00 30 60 gO angle between c-axls and poles to PDF angle between c-axis and poles to PDF

[]] 20 OPa/540'C ~ 20 OPa/630'C

Fig. 1. Orientations ofplanardefonnation features in quartz experimentally shocked at a pressure of20 GPa and at preshock temperatures of (a) 540"C and (b) 630"C.

refractive indices density [g/cm3 ] 2. 7 r----.:-...:.:::----.:~------_, 1.56 QUARTZ QUARTZ

,,,, 'i; 2.5 1.52

+. 1.48 23 ''t.,.- ':------+- ---

1.44 L--.J.._-'-_-'-_-'------l_-L_---'-_-'------l_-' 2.1 20 30 40 20 25 30 shock pressure P [GPal shock pressure [GPa]

-+ shocked at 20'C • shocked at 630'C -+. shocked at 20' C • shocked at 630' C -><- shocked at 540' C -><- shocked at 540'C

Fig. 2. Variation of refractive indices lie and fie as function of Fig. 3. Densities ofquartz shocked at 20°C. 540"C, and 630"C vs. shock pressure for preshock temperatures of 20°C, 540"C, and shock pressure. 630°C.

Our investigations indicate that shock metamorphic features are rials (B. M. French and N. M. Short, 008.), 243-253. Mono, stronglydependenton thepreshock temperature. This statementhas Baltimore. [5] Reimold W. U. and Ho12 F. (1986) LPSC XVII, far-reaching implications withrespect toshockwavebarometry that 703-704. [6] Reimold W. U. (1990) LPSCI XIX. 970--971. is based on data from recovery experiments at room temperature. [7] Huffman A. R. etaI. (1989) InShockCompressionofCondensed These datasets might be applicable only to low-temperature target. Matter ( S. C. Schmidt et aI., eds.), Elsevier. [8] Langenhorst F. rocks. Moreover, this study demonstrates that shock recovery et al. (1992) Nature. in press. [9] Schneider H. and Hornemarm U. experiments are defmitely required for understanding the complete (1974)N.Jb.Miner.Mh.• 149-162. [10] RobertsonP.B. andGrieve pressure-temperature regime ofshock metamorphism on planetary R. A. F. (1977) In Impact and Explosion Cratering (Roddy et aI., bodies. 008.),687-702, Pergamon. [II] Engelhardt W. von and Bertsch W. References: [I] Roddy et al.• 008. (1977) Impact Explosion (1969) Contrib. Mineral. Petrol., 20, 203-234. [12] Rehfeldt­ Cratering, Pergamon. New York. [2] StOffier D. et al. (1991) GCA, Oskierski A. and Stoffler D. (1986) LPSCI XVII, 697-698. [13] 55,3845-3867. [3] Stoffler D. (1972) Fortschr. Mineral., 49, 50­ Grothues J. et aI. (1989) LPSCI XX. 365-366. 113. [4] HOn F. (1968) In Shock Metamorphism ofNatural Mate- LPI COIIlriblllionNo. 790 47

SEARCH FOR A METEORITIC COMPONENT AT THE relative to the sandstone, the invoked lIdditional component is BEAVERHEAD IMPACT STRUCTURE, MONTANA. Pas­ unlikely to be chondritic or iron-rich meteoritic material. Instead, cal Leel and Robert W. Kay2, lDepartment of Astronomy, Cornell the low Ni/Cr ratio in the breccia and the melts points either to an University, Ithaca NY 14853, USA, 2Department of Geological achondritic impactor or to some other (terrestrial) source of Cr Sciences, Cornell University, Ithaca NY 14853, USA. enrichment. Meteoriticgeochemicalsignatures atlargeand/orancientterres­ The Beaverhead impact structure, in southwestern Montana trial impactcraters are difficult to fmd [3]. Possiblereasons for this (44036'N;112°58'W), was identified recently by the presence of include the high impact velocities and energies involved in large shattercones and impactites in outcrops ofProterozoic sandstones impacts, the possibility ofchemically inconspicuous (achondritic) oftheBeltSupergroup [1]. Thecones occuroveranarea >100km2. projectiles, weathering and hydrothermal alteration of impactites Because the geologic and tectonic histOly ofthis region is long and and their associated counlly rock, and inadequate sampling. Our complex, the outline of the original impact crater is no longer search for a meteoritic componentin samples from the Beaverhead identifiable. The extent ofthe area over which shatter cones occur impact structure will be discussed in the context of this general suggests, however, that the feature may have beenatleast60km in problem. diameter. TheabsenceofshatterconcsinYOWlger sedimentaryunits References: [1] Hargraves R. B. et al. (1990) Geology, IB, suggests thatthe impacteventoccurredinlate Precambrian orearly 832-834. (2) Kocherl C. and Fiske P. S. (1991) Meteoritics, 26, Paleozoic time. 358-359. [3] Palme H. et al. (1978) GCA, 42, 313-323. We have collected samples of shocked sandstone from the so­ called"Main Site" ofdark-matrix breccias, and ofimpact breccias and melts from the southend ofIsland Butte. The melts, occurring often as veins through brecciated sandstone, exhibit a distinctive ORDOVICIANIMPACTS ATSEAIN BALTOSCANDIA. M. fluidal texture,a greenishcolor,andactyptocrystallinematrix,with Lindstroml, V. Puura2, T. Fl0d6n1, and A. Bruun', IUniversity of small inclusions ofdeformed sandstone. Samplesofthe same type, Stockholm, Sweden, 2Estonian Academy of Sciences. Estonia, along with counlly rock, were analyzed previously for major- and 'Swedish Geological Survey, Sweden. trace-clementabundances[2].ItwasfOWld that,althoughthemajor­ element composition was relatively unifonn, the trace-element Northern Europe has an assemblage of Onlovician probable composition showed variations between the melt material and the impacts that is exceptional because the structures involved are lIdjacent sandstone. These variations were attributed to extensive relatively old yet well preserved because they formed at sea and weathering and hydrothermal alteration. because they formed within a restricted geological time in a In a more specific search for a possible meteoritic signature in relatively small area. The Tvaren, Kardla, and Lockne structures the brecciaand the meltmaterial we haveconducted a newseries of might not be strictly contemporaneous but all formed near the traee-clement analyses onpowders ofourownsamplesby thermal beginning oftheCaradoc Age(about460Ma), whereas theGranby neutron activation analysis. Ourresults indicate thatIrabundances structure is about20Maolder.Therangeofdiametersisfrom about in the breccia, the melts, and the lIdjacent sandstone clasts are no 2 km(Tvaren, Granby) to8 km (Lockne). Thestratigraphic succes­ greater than about 0.1 ppb, suggesting no Ir enrichment of the sion fonned on impact at sea, as unifonnIy documented by these brecciaorthe meltsrelative tothe counllyrock. However, both the structures, begins with a breccia lens consisting ofbasementrocks breccia and the melt material exhibitnotable enrichments inCr (8­ that are intensely crushed. Owing to expulsion ofsea water by the and 10-fold),in U (9- and5-fold), and in the heavy REEs (1.5- and impact, this brecciaformed under essentially dry conditions. Later 3-fold), respectively (normalization carriedoutrelative to La) (see on this breccia was in parthydrothermally altered. It is overlain by Table 1). Such enhancements relative to the counlly sandstone backsurge turbidite that formed from fragments oflocal sedimen­ suggest that weathering and hydrothermal alteration alone might tary bedrock and crystalline basement when the sea water retwned not be sufficient to account for the observed variations in trace­ to the crater site. Either the turbidite is simply a Bowna sequence element abundances, and that the incorporation of a Cr, U, and (although quite thick-as much as over 50 m) from very coarse HREE-rich component in the breccia and the melts is likely. rubble to mud, or it is more complex. Becauseboththebrecciaandthemelts shownoIrorNienrichments Mterdepositionofthebacksurgeturbidite,orturbiditecomplex, the craters still remained as 15O-200-m-deep holes in the seabed. Togetherwith the presence ofrelatively shallow waterovertherim wall, this situation created predictable hydrologic conditions for TABLB I. Trac:e-element concentrllions in ppm extended histories ofsedimentation and biological development at ("±" indiates 2-sigma measurement error). the crater as well as within it. For instance, hypoxia frequently developed in the bottom water before depth had been appreciably BUSS BIIM BHB reduced by sedimentation. The insights acquired at these craters Cr 17.76±0.6 127.88±0.S 71.94±0.4 make it possible to improve paleoecological reconstructions ap­ Ni 14.SO±2.S IS.6S±1.0 7.68±0.9 plied to Ordovician sea beds. U 1.48±0.1 S.S4±O.2 6.99±0.3 The presence ofa concentration ofcraters within a limited area Yb 1.38±O.1 3.23±0.1 1.I3±0.1 of well-preserved and accessible Ordovician deposits raises a Ta 0.39±O.1 1.66±0.1 1.33±0.1 question about the Ordovician, especially its middle portion, as La S2.4%0.5 39.04±0.3 26.49±O.2 potentially an ageofrelativelyintenseimpactactivityeveninwider Ir <0.1 ppb <0.1 ppb <0.1 ppb areas. Inthis connectionitmaybeapposite tomentionthattheonly BHSS = Beaverhead sandstone (counby rock). BHM == Beaverhead impact fossil stony meteorites so far reconled in rocks are from the late mclL bHB = Beaverhead darlt-manix brccciL Early and the Middle Ordovician. 48 1992 Sudbury Conference

DOES THE SEDIMENTOLOGY OF THE CHELMSFORD convincing demonstration of shock metamorphism put the impact FORMATION PROVIDE EVIDENCE FOR A METEORITE mechanism on flJ1J1 ground. Detailed field and laboratorystudies in IMPACT ORIGIN OF THE SUDBURY STRUCTURE? following years byPeredery, Dressler, Guy-Bray, Dence,andmany D. G. F. Long, Department of Geology, Laurentian University, othersgreatlystrengthened theimpacttheory, which is today widely Sudbury, Ontario P3E 2C6, Canada. though not universally accepted by geologists familiar with the Sudbury area. The post ·"event" fJll of the Paleoproterozoic Sudbury Basin Following publication ofthe monumental The Geology andOre consists of at least 600 m of deep-water mudrocks of the Onwatin Deposits o/the Sudbury Structure [5], several lines of study have Formation, overlainby 850m oflithic-arkosicmuddysandstones in clarified the origin and evolution ofthe structure and the intrusive theChelmsfordFormation. WhilemudstonesoftheOnwatinreflect complex. Isotopic analyses by Faggart et al. supported the earlier depositioninadeep-water,anoxic setting,there isnoclearevidence proposalbyDencethat muchoftheSICmightbeimpactmeltrather of local breccias, conglomerates, or sand bodies to support the than internally derived igneous rock. Impact modeling by Grieve conceptthat the basin was protected by the steep walls ofan impact et al. [6] also supports this mechanism, showing that the SIC crater. Carbonatesinthebasal, VermillionMemberareofsedimen­ average composition is close to that of local granite-greenstone tary exhalitive origin and were not derived from a shallow marine terrain with a smallproportionofHuroniancoverrock. Thesizeand shelf. Turbidites in the ChelmsfordFormationshow noevidenceof shape of the original impact crater remain open to debate. An centripetalfill as mightbeexpectedfrom a restricted,circularbasin. imaging radar and field study by Lowman [7] supported the inter­ They appear to have been emplaced by predominantly southwest­ pretation by Rousell that the original crater had been elliptical, erly flowing turbidity currents, which showedlittletono deflection thoughmademoreso,sinceits fonnation, bythePenokeanorogeny. along the depositional axis of an elongate foreland basin that However,primaryelliptical impactcratersdoexistonthe Moonand developed in front of the rising Penokean mountain chain. Mars, and ellipticity should not beconsidered an argument against While the presence ofminor sandstone-filled fractures in parts impact The predeformation shape of the crater was reconstructed of the Chelmsford Fonnation suggests the presence of north- or by Shanks and Schwerdtner using fmite-element modeling meth­ south-directed paleoslopes, no evidence is seen to support the ods, which indicate that the predefonnation structure was 2 to 3 existence ofsubbasins ora central uplift within theSudbury Basin. times wider (northwest-southeast), Le., nearly circular. A LlTHO­ While tilt-eorrected paleocurrent orientations are ambiguous, due PROBE survey carried out in 1991 [8] supported the Shanks­ to postdepositional shortening of strata during cleavage develop­ Schwerdtnerinterpretation totheextent thatthe SouthRange Shear ment,straincorrectionoftheobservations makes littledifferenceto Zone, in which the deformation was concentrated, was traced at the net, south-southwest-directed paleoflow. depth by reflection profiling. The impact theory for theoriginofthe Sudbury structure seems supported by a nearly conclusive body ofevidence. However, even IMPACT ORIGIN OF THE SUDBURY STRUCTURE: assuming an impact origin to be correct, at least three major EVOLUTION OF A mEORY. Paul D. Lowman Jr., Goddard questions requirefurther study: (I)theoriginalsizeandshapeofthe Space Flight Center, Code 921, Greenbelt MD 20771, USA. crater,before tectonicdeformationand erosion, (2)thesourceofthe melt now forming the Sudbury Igneous Complex, and (3) the This paper reviews the origin, development, and present status degree, if any, to which the Ni-eu-platinum group elements are ofthe widely accepted theory, proposed by Robert S. Dietz in 1962 meteoritic. [1], that the Sudbury structure was formed by meteoritic orasteroi­ The history of the impact thOOF)' illustrates several under­ dal impact. The fmtpublicationleading toDietz'sproposal was the appreciated aspects of scientific researeh: (1) the importance of suggestion by Hamilton [2] that the Sudbury Igneous Complex cross-fertilization between space research and terrestrial geology, (SIC) was anextrusivelopolith,coveredby its ownsilicicdifferen­ (2) theroleoftheoutsiderin stimulatingthinking byinsiders,(3) the tiates (the Onaping Formation, then considered a welded tuff). value ofsmall science, at leastin the initial stages of an investiga­ Hamilton's interpretation was applied by Lowman [3] to the prob­ tion, Dietz's first field work having been at his own expense, and lemofhow parentrocks similarto tektites incompositioncould be (4) the value of analogies (here, between the Sudbury Igneous formed ontheMoon,tektites assumedtobeoflunarorigin.Lowman Complex and the maria), which, although incorrect in major as­ proposed thatthelunarmariawereextrusivelopoliths, analogous to pects, may trigger research on totally new lines. Finally, the the Bushveld, Sudbury, and Duluth Complexes, and that tektites Sudbury story illustrates the totally unpredictable and, by implica­ were derived by impact from rhyolites surfacing the maria, analo­ tion, unplannable nature of basic research, in that insight to the gous to the silicic differentiates of lopoliths. The lopolith-mare origin ofthe world's then-greatest Ni deposit came from the study analogy was interpreted by Dietz to imply that the terrestrial of tektites and the Moon. lopoliths, like the circular maria, were large impact craters, and References: [1] Dietz R. S. (1962) Program, AGU, Western should have associated shatter cones. Dietz then searched for, and National Meeting, 445-446. [2J Hamilton W. (1960) Rept. Ins. found, shatterconesinHuronianmetasedimentson the southsideof Geol. Cong., XXI, 59-67. [3] Lowman P. D. Jr. (1962) JGR, 67. the Sudbury structure. It is stressed that this was a predictive 1646, and(1963)Icarus, 2, 35-48.[4} FrenchB. M.(1967)Science, discovery, not a chance one. Dietz then proposed that Sudbury was 156, 1094-1098. [5] Pye E. G. et al. (1984) The Geology and Ore a small terrestrial mare basin. It was shown, however, by B. M. Deposits 0/ the Sudbury Structure, Ontario Geological Survey. French [4] that the Onaping Formation was a metamorphosed [6] Grieve R. A. F. et al. (1992) JGR, 96, 753-764. [7] Lowman fallback breccia rather than a welded tuff. Although contradicting P. D. Jr. (1991)Can.J. RemoteSens.• 17. 152-161. [8J MilkereitB. Dietz's impact-triggered extrusive lopolith hypothesis, French's et al. (1992) Geology, submitted. VI Contribution No. 790 49

IMAGING RADAR INVESTIGATIONS OF THE SUDBURY The Haughton impact- structure, DevonJsland, is a crater of STRUCTURE. P. D. Lowman!, V. H. Singhroy2, and V. R. 24 km in diameter formed in a target ofabout 1700 m of sedimen,­ Slaney2, !Goddard Space Flight Center, Code 921, Greenbelt MD tary rocks (limestones and dolomites) on top of a crystalline 20771, USA, 'Canada Centre for Remote Sensing, 1547 Merivale Precambrian gneiss. Samples were all crystalline fragments from Road, Ottawa, Ontario KIA OY7, Canada. the allochthonouspolymictbreccia[2],whichcoverthecentral area of the crater. Optical microscopy on thin sections does not allow This paperreports preliminary results ofairborne imaging radar identification of minerals by classical criteria: Birefringence of studies of the Sudbury structure carried out in preparation for a most of the phases (actually all but sillimanite) is largely lost, CCRS European Remote Sensing Satellite (ERS-I) investigation. documenting their essentially amorphous character. Nevertheless, Thedatausedweresynthetic apertureradar(SAR)C-band(5.66cm) conservation of primary foliation suggests that total rock melting images acquired from about 6 km altitude in 1987. They cover the did not occur. These crystalline fragments also show large degrees Sudbury area in both wide and narrow swath modes, with east-west of porosity, lying roughly at 40%. Electron microprobe analyses flight paths and north-south illumination directions. Narrow swath reveal five compositional domains: (1) numerous Si02 dominant resolution is 6 m in range and azimuth; wide swath resolution is zones, (2) areas with a biotitelike composition, (3) layers with 20 m in range and 10m in azimuth. feldsparlike composition, (4) areas characterized by Al/Si close to The SAR imagery has proven highly effective for field use, 1, and (5) fracturated sillimanites. providing excellent rendition of topography and topographically Strong chemical heterogeneities were observed within most of expressed structure. Reasons for this include the illumination the compositional domains. To decipher the origin of such hetero­ geometry, notably the look azimuth normal to the long axis of the geneities, an ATEM study was performed, yielding a spatial reso­ Sudbury structure and Penokean fold axes, the good spatial resolu­ lution of 0.5 nm in image mode and of 200 nm for the energy tion, and the short wavelength. Forested areas in the Sudbury area dispersive X-ray microanalyses (EDS). This study reveals the tend to be uniformly rough at C-band wavelength. with backscatter following chemical and structural characteristics: dominated by local incidence angle (i.e., topography). Dielectric 1. Si02 dominant areas are formed by a mixture of pure Si02 properties have relatively little effect on backscatter, except for polycrystalline quartz identified by electron diffraction pattern and targets such as metal or water. chemical analysis (Fig. la) and a silica-rich amorphous phase Field work using the SAR imagery has to datebeenconcentrated containing minor amounts of aluminium, potassium, and iron in the North Range and Superior Province as far north as the Benny (Fig. Ib). greenstone belt. This area was chosen for initial investigation ofthe 2. Areas with biotitelike composition are formed by <200-nm original size and shapeofthe Sudbury structure because the effects grains of iron-rich spinels (Fig. 2a) embedded in a silica-rich of the Penokean Orogeny were minimal there. Field work using amorphous phase (Fig. 2b) that is very similar to the one described SAR indicates that there has been little postimpact deformation of above. the North Range or adjacent Superior Province rock [1]. There 3. Layers with feldsparlike composition are constituted by appears tobeno evidencefor anouterring concentric with the North IOO-2oo-nm-sized alumina-rich grains (the indexation ofthe crys­ Range as indicated by early Landsat imagery [2,3]. The apparent talline structure is under progress) and the silica-rich amorphous ring shown by Landsat is visible on the SAR imagery as the phase. intersection of two regional fracture patterns not related to the 4. Zones characterized by the unusual AVSi ratio close to 1 are Sudbury structure [4]. There is no outerring visiblesouthwestofthe formed by spinel grains (2oo-nm-sized) embedded in the same structure. This can reasonably be explained by Penokean deforma­ silica-rich amorphous phase. tion, but there is also no outer ring to the northeast cutting the 5. The fracturated sillimanites contain domains with a lamellar relativelyundeformedHuroniansedimentsoftheCobaltEmbayment structure, defmed by the intercalation of Ioo-nm-wide lamellae of Further study of these problems is planned with ERS-l imagery. mullite crystals and of a silica-rich amorphous phase (Fig. 3a). References: [1] Lowman P. D. Jr. (1991) Can. J. Remote Figure 3b shows an individual 5OO-nm-sized crystal of mullite. Sens., 17, 152-161. [2] Dressler B. O. (1984) In The Geology and These crystals preserved the crystallographical orientation of the Ore Deposits of the Sudbury Structure (E. G. Pye et al., eds.), preshock sillimanite. OntarioGeological Survey. [3] GrieveR. A. F.etal.(I99I)JGR. 96, All compositional domains, identified at the microprobe scale, 753-764. [4] Lowman P. D. Jr. (1992) Rev. Geophys.• in press. can thus be explained by a mixture in different proportion between the following phases: (1) a silica-rich amorphous phase, with minor Al and K; (2) quartz crystals; (3) spinel crystals and alumina-rich PHASE TRANSFORMATIONS IN 4O-6O-GPa SHOCKED crystals; (4) sillimanite; and (5) mullite. Such mixtures of amor­ GNEISSES FROM THE HAUGHTON CRATER (CANADA): phous phases andcrystals in differentproportions explaindisturbed AN ANALYTICAL TRANSMISSION ELECTRON MICRO­ isotope systems in these rocks and chemical heterogeneities ob­ SCOPE (ATEM)STUDY. I.Martinez, F. Guyot, and U. Scharer, served on the microprobe. UniverSite Paris7 et IPG Paris, 2 place Jussieu, 75251 Paris cedex References: [1] Scharer U. and Deutsch A. (1990) GCA. 54, 05, France. 3435-3447. [2] Metzler A. et al. (1988) Meteoritics, 23• 197-207.

In order to better understand phase transformations, chemical migration, and isotopic disequilibrium in highly shocked rocks [1], we have performed a microprobe and an ATEM study on gneisses shocked up to 60 GPa from the HaughtonCrater. 50 1992 Sudbury Conference

PHASE TRANSFORMATIONS IN 4O-6O-GPa SHOCKED GNEISSES FROM THE HAUGHTON CRATER (CANADA): Martinez I. et al. ,----- a : ; , i CRYSTALS

Y

; ; ,

..iJ ; : , : A

Fig. 1. (a) EDS microanalysis ofpolycrystalline quartz. Oxygen is not detectable with this EDS configuration and some contamination by the copper of the sample support occurs; (b) EDS microanalysis of the silica-rich amorphous phase.

!' .. - ~ --.- .. -' ----- _.----,----.-.------c:, b ' ;I:~a ;~ ~:_ .~~~S~A~ ~'~ ~ ~-_:. -.~.,~ ~.'.~~._I._._ ., GLASS ... : ...... :__...... ' .. · ... " .... ,,,, I;; I ! A: l : ,,- ,,- .,..;,-,..-",,-..,..- -.---1.--.---'( ",1, ,.. ,- ..-.,.,-'"---..,''''' " , ". 1:=fa=Fl-)~~ ~;~~~~_~_ ~_-~:~ ,___ ' "-.:.."."."".-...,..---.".--.,.--".-,,,,-,..-.-..,,-,--i;

I ~'M' It ~. fg~,A II' 1\1 t ... <- KK' r·· 'j: ... -.' ... :-. --.-. ,",'··..·"·,,....'..·"'·-K ± E: E .J_._'~=~o~~"'.".... _._""'_, ... _ _.t~.6.- ._ .. ~) , lJ. .,.,A., ....-, .... ~ ",,__-'....__---'__._"''- .,_._--'-.

Fig. 2. (a) EDS microanalysis ofan iron-rich spinel crystal. The composition is approximately (Mgo,~Feo.~)Alp4' although Fe3+ could not be determined; (b) EDS microanalysis of the silica-rich amorphous phase. Notice me similarity with Fig. lb.

a_

Fig. 3. (a) TEM image of the lamellar structure of intercalated mullite and silica-rich (+Al,K) amorphous phase (see arrows). Scale bar is 200 nm. (b) In some areas, mullite appears as small euhedral crystals surrmmded by the'amorphous phase. TEM, scale bar 200 nm. !PI COIIIributiora No. 790 51

IMPACUTES FROM POPIGAI CRATEL V. L. Masaitis, Karpinsky Geological Research Institute, SL Petersburg, Russia.

Impactites (tagamites and suevites)from Popigai impactcrater, whose diameter is about 100 lan, arc distributed over an area of 5000lan2[1,2].Thecontinuoussheetofsueviteoverlicstheallogenic polymict breccia and partly authogenic breccia, and may also be observed in lenses orirregularbodies. The thickness ofsuevites in the central part of the crater is more than 100m. Suevites may be distinguishedbyc:ontentofvitroclasts,lithoclasts,andcrystalloclasts, by theirdimensions, and by type ofcementation, which reflects the facial settings of ejection of crushed and molten material, its sedimentation and lithification. Tagamites (impactmelt rocks) arc distributedonthesurfacepredominantlyinthe westernsectorofthe crater. The mostcharacteristic arc thick sheetlike bodies overlying the allogenic breccia and occurring insuevites where minor irregu­ lar bodies arc widespread. The maximal thiclmess of separate tagamite sheets is up to 600m. Tagamites, whose matrix is crystal­ KENTLAND ASTROBLEM lized toa differentdegree, includefragments ofminerals andgneiss " ~. blocks, among them shocked and thennally metamorphosed ones. f I Tagamitesheetshavea complex innerstructure; separatehorizontal I zones distinguishin crystallinity andfragmentsaturation. Differen­ , tiation in the impact melt in situ was not observed. The average chemical compositions of tagamites and suevites \ arc similar, and correspond to the composition of biotite-garnet gneisses of the basement [3]. According to the content ofsupplied Ir, Ni, andothersiderophiles, impactmeltwas contaminatedby5% Fig. 1. Map showing Wisconsin epoch glacial transport of two cosmic matter of collided body, probably ordinary chondrite [4]. indicatorrocks (Onaping black suevite and Sudbury breccia) from The total volume of remaining products of chilled impact melt is the Sudbury astrobleme, Ontario, some 820 lan to the Kentland, about 1750lan'. Halfthis amountis representedby tagamitebodies. Indiana,astrobleme. Dashedlines indicatepresumedglacial transport Though impact melt was in general well homogenized, the trend of diamond erratics to the midwest till plain. After [1]. analysis showedthattheconcentriczonationindistributionofSi02, MgO, and N~O and the bandlike distribution of FeO and Alp, content [5] testifies to a certain inheritance and heterogeneity in (Peredery) recognized and sampled in the overlying glacial drift country rock composition laterally and vertically in the melting deposits a distinctive boulder of Sudbury suevite (black member, zone. The radial ray inhomogeneities of content of newly fonned Onaping Fonnation) thatnonnallyoccurs withinthe SudburyBasin high-pressure phases detennined in impactites also reflect the as an impact fall-back or wash-in deposit. The rock was sampled peculiarities ofmelt transportation during excavation. On the other (but later mislaid) from a fanner's cairn next to a cleared field. hand, the irregularity of distribution of supplied siderophile ele­ Infonnal reports of this discovery prompted the other authors to mentsinimpactites shows thatcontaminationbycosmicmatterwas recently reconnoiter the Kentland locality in an attempt to relocate probably associated with the condensation from vapor on cold the original boulder. Several breccia blocks were sampled but fragments engulfed by melt after its homogenization. laboratory examination proved most of these probably to be References (all in Russian): [1] Nedra L. (1980) Geology of diamictites (tillites?) [2] from the Precambrian Gowganda Fonna­ Astroblemes, 231 pp. [2] Masaitis V. L. (1984) In Modem Ideas of • tion, which outcrops extensively in southernOntario [3]. However, TheoreticaIGeology(L. Nedra,ed.), 151-179.[3] RaiIchlinA. I. and one sample was conf1J1Jled as typical Sudbury Breccia, which Mashchak M. S. (1977) Meteoritika, 36, 140-145. [4] Masaitis outcrops in the country rock surrounding the Sudbury Basin. Thus V. L. andRaikhlinA.I.(1989)Meteoritika,48. 161-169.[5] Masaitis two glacial indicators were transported by Pleistocene continental V. L. et al. (1980) In Cosmochemistry ofMeteorites. Moon and glaciers about 820lan overa tightly proscribedpath and, curiously, Planets, 176-191. Naukova dumka, Kiev. from one astrobleme to another. Brecciated boulders in the Dlinoisllndiana till plain arc usually ascribed to the GowgandaorMississagi fonnations in Ontario. But impact-generated rocks need not be confused. The carbonaceous SUDBURY BRECCIA AND SUEVITE AS GLACIAL matrix of the suevite, for example, was sufficiently distinctive to INDICATORS TRANSPORTED 800 KM TO KENTLAND assign itto the upperportion ofthe black Onaping. The unique and ASTROBLEME,INDIANA. John F. McHonel , RobertS.Dietzl , restricted source area of these indicators provide an accurate and and Walter V. Peredery2,IGeoIogy Department, Arizona State reliable control for estimating Pleistocene ice movement. University, Tempe AZ 85287-1404, USA, 2INCO Exploration and References: [1] Flint R. F. (1957) Glacial and Pleistocene Technical Services Inc, Copper Cliff, Ontario POM INO, Canada. Geology, Wiley,553pp.[2] HarkerandGiegengack(1989)Geology, 17,123-126. [3] Young G. M. (1981) In Earth's Pre-Pleistocene A glacial erratic whose place of origin is known by direct Glacial Record International Geological Correlation Progreunme comparison with bedrock is known as an indicator [1]. In 1971, Project 38: Pre-Pleistocene Tillites (M. J. Hambrey and W. B. whilevisiting the known astrobleme at Kentland, Indiana, one ofus Harland, cds.), 807~12. 52 1992 Sudbury Conferent:e

WHAT CAN WE LEARN ABOUT IMPACT MECHANICS (intermediate for Klenova) and outer molUltain rings, possibly FROMLARGE CRATERS ON VENUS?William B. McKinnon representing fault scarps similar to the Outer Rook and, especially, andJ. S. Alexopoulos, DepartmentofBarth and PlanetarySciences theCordilleraringofOrientale,respectively.Furthennore,Klcnova and McDonnell Center for the Space Sciences, Washington exhibits an inner peak ring analogous to Orientale's Inner Rook, University, SL Louis MO 63130, USA. whereas Meitner exhibits a possible additional, exterior, scuplike partialortransitionalring. ArccentMagellanimageofMeitncrdocs More than 50 lUlequivocal peak-ring craters and multiringed not equivocally show this outer segment as a scarp, but the radar impact basins have been identified on Venus from Barth-based geometry was poor for determining this. Mead exhibits no inner Arecibo, Venera 15/16, and Magellan radar images. These ringed peak ring, but may be the structurally clearest example of"mega­ craters are relatively pristine, and so serve as an important new terrace" collapse yet observed on the terrestrial planets. Adjacent dataset that will further lUldcrstanding of the structural and rheo­ ringdiamcterratiosforthethree multiringedbasinsrangefrom -1.6 logical properties ofthe venusiansurface and ofimpactmechanics between the mostprominent rings ofKlenova and Meitner to -1.4 in general. They are also the most direct analogues for craters for Klenova's peak ring, Meitncr's partial ring, and Mead. These fonned on the Barth in Phanerozoic time [1]. ratios arc similar to those observed for Orientale and other lunar The innerrings and craterrims ofvenusianpeak-ringcraters (or multiringed basins. Thus we interpret all three venusian struclUrcs basins) arc morphologically similar to inner rings and crater rims, to represent Orientale-type multiringed implCt basins. Islbclla respectively, ofpeak-ring craters ontheMoon, Mars, andMercury. (-170 km diameter) is also a ringed basin, but it is severely As is observed for mcrcurian ringed basins, ring diameter ratios volcanically flooded; much of the intermediate ring and all of the decreasefrom ~ 3.5 to2(orless)withincreasingcraterdiameter. On inner ring arc dermed by isolated but concentrically arranged. Venus, the lransition from central-peakcratertopeak-ring crater is massifs. well defmed at arolUld 40 km diameter. Decreasing ring diameter Although multiringed basins on Venus occur at much smaller ratios with increasing peak-ring craterdiameter arc consistent with diameters than on the other terrestrial planets (except perhaps the a hydrodynamic origin for peakrings, i.e., collapse and expansion Earth),effectiveviscosities arclowenoughatdepth toallow inwatd of the uplifted central peak. The large but fmite ring ratios at the asthenospheric flow, which in the model of [6] can cause radial lransition and, on other terrestrial planets, the simultaneous pres­ extensional stress and circumferential faulting in the overlying ence of central peaks and rings in many of these transitional lithosphere, block rotation, and ring fonnation. The differingrheo­ structuresimply thatcentralringsrisefrom thecollapsing shoulders logies ofVenus crust and mantle may playarole in the creation of ofcentral peaks, i.e., that they are a wave~brcaking phenomena in Cordilleran-style ring faulting. Thelarge differential stresses asso­ oversteepened and unstable central peaks (cf. photographs in [2)). ciatedwiththedeeptransientcraterprecursors toKlenova,Meitncr, The viscosities offluidized cratermaterial implied by central-peak and Mead arc sufficient to have driven solid-state power-law flow to peak-ring transition diameters on the terrestrial planets are more ofvenusian upper mantle rock at low effective viscosity (not only or less proportional to craterdiameter, D, indicating that the scale low enough to defme an asthenosphere on the collapse timescale, dependence ofviscosity, 11, may dominate gravity in the formation but a lower viscosity than that ofoverlying shock-fluidized lithos­ of peak-ring craters. This approximately linear dependence, 11 .. phere). For the two smaller multiringed basins, sufficiently low 9 (D/km) GPa-s, is compatible with fluidization by strong acoustic viscosities may have only been possible if ICCCSsible mantle tem­ or seismic vibrations. It implies a nearly constant ratio 'A.JD ... 1/80 peratures were close to or at the peridotite solidus. This is not to 1/20, where A..t is the dominant wavelength ofthe acoustic field, geologically implausible, as the ablUldant widespread basaltic vol­ depending on the solUld speed in agitated rubble adopted (0.5­ canismonVenusimpliesthatits mantletemperatureprofilesexceed 2 kmls). In tenns ofscaling, the originally generated ')JD spectrum thesolidus atnumerous locations and the asthenospheremayinfact may beexpected to be roughly invariant with respect to fmal crater be defmed by pressure-release partial melting, but alternative size. For lowersolUld speeds, the dominant wavelength A..t could be possibilities include enhanced flow rates in mantle rocks due to a large fraction of the crater depth, which may imply leakage of plasticity orviscous flow in a weakercrustal asthenosphere. Analo­ lower frequencies and damping ofhigher ones [3]. Impactordiam­ gous argumentscanbemade withrespecttomultiplering fonnation eter and A..t would also be comparable for the lower sound speed on theMoon. Effectiveviscosity estimates imply thatring tectonics above, given presently accepted crater scaling relations. on the terrestrial planets may in fact work most readily within a Important interplanetary differences remain; martian crust in sufficiently thick, soft crustal layer. particularappears tobemoreeasily fluidized by impact. Wealsodo Finite-element simulations ofbasincollapse and ring fonnation notconsiderthe shockmelt hypothesis of[4J here, butnote that the along the lines of [7J were lUldertaken in collaboration with V. J. increasing amolUlt of melt compared to ejected mass in larger Hillgren (University ofArizona). These calculations, summarized structures might be expected to result in decreasing effective in [lJ, used an axisymmetric version of the viscoelastic fmite viscosity with size. The intcrlors of most of the peak-ring craters elementcodeTECTON, modeledstructuresonthe scaleofKlenova appear generally smooth (similar to the surrounding plains), as orMeitner, and demonstrated two majorpoints. First,·viscous flow evidenced by the radar-dark returns in Arecibo and Magellan and ring fonnalion arc possible on the timescale ofcrater collapse images. This may be attributed to the presence of impact melt for the sizes of multiringed basins seen on Venus and heat flows deposits or to postimpact volcanism. The increasing incidence of appropriate to theplanet Second, anclasticlithosphere overlying a bright floors among the youngest craters (i.e., those with dark Newtonian viscous asthenosphere results mainly in uplift beneath parabolic deposits [5)) suggests that volcanism may be the more the crater. Inward asthenospheric flow mainly occurs at deeper correct choice, but postimpact differentiation and eruption from a levels. Lithospheric response is dominantly vertical and flexural. large melt sheet should also be considered. Tensional stress maxima occur and ring formation by normal Of the four largest ringed structures identified on Venus so far, faulting is predicted in some cases, but these predicted rings occur Klenova (144 km diameter), Use Meitner (148 km), and Mead too far out to explain observed ring spacings on Venus (or on the (270 km) exhibit what we interpret to be major asymmetric inner Moon). Similar distant tensional stress maxima were reported by LPI COIIIributioli No. 790 53

[7], who fOWld that in order to generate a ring fault at a distance of helerogenousmatrixconsistingofafme-grainedrock flourdisplays -1.4craterradii, itwas necessary to resmct asthenospheric flow to nonoriented textures as well as extreme flow lines. Chemical achannelatdeptb,oneoverlying astiffermesospbere.Itis tempting analysis substantiates at least some mixing with allochthonous to assign this asthenospheric channel to a ductile lower crust. as material. discussed above. Alternatively, an effectively stiffer mesosphere C. Breccias with a crystalline matrix are a subordinate type of may be a natural consequence of truly non-Newtonian rebound. Sudbury breccia. According to pelrographical and chemical differ­ Much work remains to be done on this problem. ences, three subtypes have been separated. The local origin of the Overall, these estimates and models suggest that multiringed fragments and the close chemical relationship to the COWltry rock basin formation is indeedpossible atthe scales observedon Venus. point to an autochthonous generation probably through in silK Furthermore, due to the slrong inverse dependence of solid-state frictional processes. Fortwo subtypes the geometry ofthe dikes and viscosity onstress, the absence ofCordilleran-style ring faulting in the texture of the matrix indicates that at least some transport of craters smaller than Meitner or Klenova makes sense. The (1) ap­ breccia material has occurred. Breccias with a crystalline matrix parent increase in viscosity of shock-fluidized rock with crater have neverbeenobservedincontactwiththeothertypes ofbreccias. diameter, (2) greater interior temperatures accessed by larger, D. Late breccias with a clastic mattix are believed to represent deeper craters, and (3) decreased non-Newtonian viscosity associ­ the latest phase of brecciation. Two subtypes have been distin­ ated with larger craters may conspire to make the transition with guished due to differences in the fragment contenL Breccias with a diameter from peak-ring crater to Orientale-type multiringed basin low fragment content show. weak lamination and sharp or grada­ rather abrupL tional contacts to country rock. Inclusions of type A breccias are References: [1] Alexopoulos J. S. and McKinnon W. B. observed. Breccias with a high fragment content are characterized (1992)JGR, submitted. [2] GaultD. E. andSonenC. P. (1982) GSA bygradationalcontacts andareonlyknownfrom theoutermostparts Spec. Pap., 190. 69-102. [3] Melosh H. J. andGaffney E. S. (1983) ofthe structure. Fragments of these breccias are oflocal origin. A Proc. LPSC 13th. in JGR. 88. A830-A834. [4] Cintala M. J. and possible correlation of the relative timescale ofbreccia formation Grieve R. A. F. (1991)LPSC XXII. 213-214. [5] Campbell D. B. et with the phases ofcrater formation will be discussed. al. (1991) JGR. in press. [6] Melosh H. 1. and McKinnon W. B. Shock deformation features, which have been recorded within (1978) GRL, 5, 985-988. [7] Melosh H. J. and HiBgren V.J. (1987) breccia fragments up to a radial distance of 9 km from the SIC, LPSC XVl/l. 639-640. represent the shock stage I of the basement rocks. Inclusions exhibiting a higher shock stage, such as melt particles or suevitic fragments, whichareknownfromdikebrecciasof,e.g.,theCarswell impact structure, are lacking. This means that the dike breccias of SUDBURY PROJECT (UNIVERSITY OF MfJNSTER­ Sudbury as presently exposed are from a deeper level of the ONTARIO GEOLOGICAL SURVEY): (5) NEW INVESTI­ subcrater basement than their COWlterparts ofCarswell. GATIONS ON SUDBURY BRECCIA. V. MUller-Mohr, insti­ References: [1] Stamer et al. (1989) Meteoritics. 24. 328. tute of Planetology, University of MOnster, Wilhelm-Klemm-Sir. 10, W-4400 MOnster, Germany.

Sudbury breccias occur as discordant dike breccias within the A HISTORY OF THE LONAR CRATER. INDIA-AN footwall rocks of the Sudbury structure, which is regarded as the OVERVIEW. V. K. Nayak, Department of Applied Geology, possible remnant of a multiring basin [1]. Exposures of Sudbury Indian School ofMines, Dhanbad, India. breccias in the North Range are known up to a radial distance of 60-80 km from the Sudbury Igneous Complex (SIC). The breccias The origin of the circular structure at Lonar, India appearmore frequent within azone of10km adjacentto the SIC and (19°58'N:76°31'E), described variously as cauldron, pit, hollow, a further zone located about 20-33 km north ofthe structure. depression, and crater, has been a conlroversial subject since the From differences in the structure ofthe breccias, as for example early nineteenth century. A history of its origin and other aspects the size ofthe breccia dikes, contactrelationships between breccia from 1823 to 1990 are overviewed. The structure inthe DeccanTrap ,and country rock as well as between different breccia dikes, frag­ Basalt is nearly circular with a breach in the northeast, 1830 m in ment content, and fabric ofthe groWld mass, as seen in thin section, diameter, 150 m deep, with a saline lake in the crater floor. the Sudbury Breccias have been classified into four different types. Since time immemorial, mythological stories prevailed to ex­ A. Early breccias with a clastic/crystalline matrix comprise plain in some way the formation ofthe Lonar structure, which has smalldikes ranging in sizefrom -1em to max. 20em. Characteristic been held in great veneration with several temples within and features of these breccias are sharp contacts to country rock, low outside the depression. Various hypotheses proposed to understand fragment content (20-30%), local origin of fragments, and an its origin are critically examined and grouped into four categories aphanitic, homogenous matrix, which can be related to country as (1) volcanic, (2) subsidence, (3) cryptovolcanic, and (4) meteor­ rock. Locally cOlTOsional contacts to feldspar minerals and small ite impact. In the past, interpretations based ongeological, morpho­ vesicles f1l1ed with secondary minerals are observed. logical, and structural data were rather subjective and dominated by B. Polymict breccias with a clastic matrix represent the most volcanic, subsidence, and, to some extent, cryptovolcanic explana­ common type ofSudbury breccia. The thickness ofthe dikes varies tions [1]. In 1960,experience oftheCanadian craters led Beals etal. from several tens ofcentimeters to a few meters butcan also extend [2] to fIrSt suggest the possibilityofa meteorite impactorigin ofthe to more than 100 m in the case of the largest known breccia dike. Lonar crater, and thus began a new era ofmeteorite impact in the Contacts with COWltry rock are sharp or gradational. Fragment history of the Indian crater. content (60-75%) is usually of local origin but especially in large The last three decades (1960 to 1990) reflect a period of great dikes allochthonous fragments have been observed. Inclusions of excitement and activity of the Lonar crater, perhaps owing to an type Abrecciasreveal thelaterformation ofthis typeofbreccia. The upsurge of interest in exploration of the Moon and other planets. 54 1992 Sudbury Conference

Application of principles of hypervelocity impact cratering has patible- and siderophile-element signatures indicating KREEP and provided overwhelmingevidence for animpictoriginoftheIndian meteoriticcomponents. soa lackofthesecomponents may betaken crater. Among others. shockmetamorphic characteristics ofbasalt. as evidence that a sample preserves its primlUY igneous composi­ impactglasses.mineralogy.chemistry.geochemistry, andcompari­ tion, even though its texture may have been modified by cataclasis son with theMoon's rocks haveclearly demonstrated its formation or annealing. by impact of a meteorite [3-6]. Ifthe SIC represents melt fonned during the impact event that Over the years, the origin of the Lonar structure has risen from created the Sudbury Basin, then ideas of how large melt sheets volcanism. subsidence, and cryptovolcanism to an authentic mete­ behaverequirerevision. TheSICis anoritic-to-gnnophyricmassof orite impact crater. Lonar is unique because it is probably the only crystallizedsilicate liquid withmineral and chemicalcompositions terrestrial craterin basaltand is the closestanalog with the Moon's broadly consistent with closed-system fractional crystallization, craters. Some unresolved questions are suggested. The proposal although greenschist facies alteration has obscured much of the is made that the young Lonar impact crater, which is less than fme-scale record [10]. Despite the Ni and POB (platinum-group­ 50,000 years old,should beconsideredas the bestcraterlaboratory element)sulfideores intheSIC,siderophile-elementabundancesin analogous to those ofthe Moon, be treated as a global monument, the silicates are comparable to those ofthe countryrock [II]. POE andpreservedfor scientists to comprehend more about the myster­ patterns in the ores are not chondritic. as they are in many impact ies of nature and impact cratering, which is now emerging as a melts, but are highly fractionated andsimilarto those ofterrestrial fundamental ubiquitous geological process in the evolution of the basalts [12]. Osmium isotopic compositions in the ores suggest a planets [7]. significant component of continental crust and are difficult to References: [1] La Touche T. H. D. and Christie W. A. K. reconcile with meteoritic contamination [11]. (1912) Rec. Geol. SIU'II.1ndia. 41, 266-285. [2] Beals C. S. et al. A lunar sample with mineralogic and geochemical characteris­ (1960) CurrenlScience, India, 29,205-218. [3] NayakV. K. (1972) tics analogous to those of the SIC probably would be judged as EPSL. 14. 1-6. [4] Fredriksson K. et al. (1973) Science. 180. "pristine,"hence aprimlUY igneous rock.Iflargeimpacteventscan 862-864. [5] Fredriksson K. et al. (1979) Smilhson. Conlrib. Earth create meltrocks with characteristics indistinguishable from those Sci., 22, 1-13. [6] Kieffer S. W. et al. (1976) Proc. LSC 7th. of layered igneous intrusions and with no detectable meteoritic 1391-1412. [7] GrieveR. A. F. and HeadJ. W. (1981) Episodes. 4. contamination, then any or all ofthe pristine lunar highland rocks 3-9. may not necessarily represent endogenous lunar magmatism but fractionated impact melts. Diverse components would still be required in the lunar crust and/or upper mantle to produce the impressivearray oflunarhighlandrocktypes. butthe connection to SUDBURY IGNEOUS COMPLEX: IMPACT MELT OR major mantle reservoirs that could constrain the planet's bulk IGNEOUS ROCK? IMPUCATIONS FOR LUNAR MAG­ composition wouldbeIosL Theremay beeconomicimplications as MATISM. MarcD.Norman.P1anetaryGeosciences, Department well: If lhe SIC is a fractionated impact melt, then large impact ofGeologyandGeophysics,SOEST.UniversityofHawaii,Honolulu structures becomepotential exploration targets. bothterrestrial and HI 96822. USA. extraterrestrial. Despite the somewhat unusual, silica-rich bulk composition of The recent suggestion lhat lhe Sudbury Igneous Complex (SIC) lhe SIC, several characteristics oflhecomplex appearmoreconsis­ is a fractionatedimpactmelt[I]mayhaveprofoundimplicationsfor tentwithendogenousmagmaticprocessesvs. impactmeltingand in understanding thelunarcrustand themagmatichistoryoftheMoon. sit", differentiation. Among these c~acteristics are (l) petrologic A cornerstone of much CUI1'eDt lhought on lhe Moon is lhat the and geophysical evidence suggesting an unexposed mafic orultra­ developmentofthe lunarcrustcan be traced lhrough the lineage of maficmassbeneaththe SIC,(2)COIltaetrelations withintheSIC that "pristine" igneous rocks [2]. However, ifrocks closely resembling suggest multiple intrusive events, and (3) possibly abundant water those from layen:d igneous intrusions canbe produced by differen­ in the SIC magma. In addition. we argue that the silica-rich tiation ofa large impactmeltsheet,thenmuch ofwhat is thought to composition ofthe SIC does notrequire impact melting. butcanbe be known about the Moon may be called into question. This paper accounted for by endogenous magmatic processes although the presents a briefevaluationoftheSIC as a differentiatedimpactmelt impact event may have influenced the course of the magmatic vs. endogenous igneous magma and possible implications for the evolution. These topics are discussed in more detail below. magmatic history of the lunar crust. Petrologic and geophysical evidence favoring an unexposed Petrologic and geochemical studies of terrestrial impact melts mafic or ultramafic mass beneath the SIC includes ultramafic have shownthatmostofthese OCCUI1'eDces cooledquickly. creating xenoliths found in the SIC sublayer [13,14] and gravity and mag­ homogeneous crystalline rocks with compositions approximating netic data [15]. The xenoliths have mineral and trace-element those of the average target stratigraphy [3,4]. Impact melts typi­ compositions suggesting a petrogenetic connection to the SIC. A cally. but not always, have elevated concentrations of siderophile basal ultramafic WIle would suggest lhat the SIC is ~t contained elements relative to the country rock, indicating meteoritic con­ entirely within the impactstructure andwould createa bulkcompo­ tamination [5,6.7]. Applicationofthesestudies tolunarsampleshas sition for the SIC unlike lhat ofthe proposed target stratigraphy. It lead to various criteria thought to be useful for distinguishing would also seem to require a mantle-derived component in the SIC primlUY igneous rocks of lhe lunar highlands crust from lhe mix­ magma,whichmay bemore supportiveofanendogenousmagmatic tures created by impact melting [2,8.9]. Among these criteria are origin rather than incorporation ofmantle material into lhe impact mineralcompositionssuggestingplutonicconditions,non-KREEPy melt. If the excavation cavity of the Sudbury impact event was incompatible trace-element patterns, and low concentrations of 100-150 km diameter [1], the depth of excavation was probably meteoritic siderophile elements. Lunar breccias and impact melts S20 km [16], which would be predominantly orexclusively within identified as polymictonpetrographicgrounds usually haveincom- the continental crust. U'I COIIITiblitiora No. 790 SS

Contactrelatioos between the soblayer and the SIC main mass latedbymoremaficmagmas,whichintmnmayhavebeen produced noriteappeartoreflectmultipleintrusiveevents althoughbothunits by local thermal perturbatioos orpressure-release melting associ­ may have been mobile simultaneously [14]. Multiple intrusions ated with the impact. wouldseemmoreconsistentwithpulses ofendogenousmagmatism Alternatively, crustal material may have been injected into the ratherthana one-shotimpacteventalthoughthemechanicsoflarge­ mantle,producing a mixedsourcethatmeltedtogivethe SICparent scale impact melting remain obscure. Ampht'bole is present in the magmL Nyquist and Shih [33] haveproposed that Mgional hetero­ SIC norite and may be primary [10]. The presence of water in the geneities in the lunar mantle may reflect large impact events that meltinamounts necessary tostabilizeamphibole(2-5wt%)maybe injected crustal material deep into the Moon's interior. more consistent with an endogenous magma rather than a super­ The SIC appears to represent endogenous magmatism although heated impact melt. For example, tektites are among the chiest of probably localized and influenced by a major impact event and terrestrial rocks, but theirsmall volume may not be directly analo­ structure. The role ofpristine lunar highland rocks as products of gous to the SIC. It may also be possible that a dry impact melt endogenous magmatism is correspondingly secure for themoment became hydrated through assimilation of country rock dwing but the effects ofmajorimpactevents inlocalizing and influencing crystallization. that magmatism remains poorly perceived and probably MquiMs The bulk composition ofthe SIC seems to be close to that ofan additional missioos to the Moon to clarify. Regardless, study of average for the upper crustal target stratigraphy [I], which is a impact events remains offundamental importance for understand­ common characteristic ofterrestrial impactmelts. However, endo­ ing the formation and evolution of the planets. genous magmatic processes such as assimilation can incorporate References: [I] Grieveetal. (1991)JGR.96. 22753. [2] War­ significant amounts of continental crust into more mafic magmas renandWasson(1977)Proc.LSC8th. 2215;(1978)Proc.LPSC9th. without superheat [17,18]. Such processes can produce igneous 185. [3] Phinney and Simonds (1977) Impact and Explosion. rocks with compositional characteristics quite similar to that esti­ Cratering. 771. [4] Grieve et al. (1977) Impact and Explosion mated for the bulk composition of the SIC. For example, many Cratering, 791. [5] Morgan et al. (1975) Proc. LSC 6th. 1609. occurrences ofCenozoic volcanic rocks in western North America [6] Palme et al. (1978) GCA. 42, 313. [7] Wolfet al. (1980) GCA. have bulk compositions close to that ofthe SIC [19-22]. 44. 1015. [8] Warner and Bickle (1978) Am. Mineral., 63, 1010. Even ifthe SIC is not a direct impactmelt, there does appear to [9] Ryder et al. (1980) Proc. LPSC 11th, 471. [10] Naldrett and beacloseassociationinspaceandtimebetweentheSICanda major Hewins (1984) Ontario Geol. Surv. Spec. Vol. 1. 235. [11] Walker impact event. Dietz [23] and French [24] described features in the et al. (1991) EPSL. 105,416. [12] Naldrett (1984) Ontario Geol. Sudbury Basin that they attributed to shock. Their arguments that Surv. Spec. Vol. 1. 309. [13] Scribbins et al. (1984) Can. Mineral.• theBasinis an impactstnlctureare persuasive because there areno 22.67. [14] Naldrettetal. (1984) Ontario Geol. Surv. Spec. Vol. 1. knownoccurrences ofsimilarshock features unequivocally associ­ 253. [IS] Guptaetal. (1984) Ontario Geol. Surv. Spec. VoU, 381. ated with volcanic eruptions. H the Sudbury Basin is an impact [16] Pike and Spudis (1987) Earth Moon Planets, 39. 129. stnlcture, it is the largest such structure known on Earth. The [17] Leeman and Hawkesworth (1986)JGR, 91. 5901. [18] Moor­ noncircularityoftheSBhasbeencitedasevidenceagainstanimpact bath and Hildreth (1988) CMP. 98,455. [19] Gerlach and Grove origin, althoughtheoriginalshapeoftheBasinispoorlyconstrained (1982)CMP.80.147.[20]McMillanandDuncan(1988)J.Petrol.• [25]. Although theoriginal shapes ofmostimpactcraters generally 29. 527. [21] Nixon (1988) J. Petrol., 29. 265. [22] Norman and are circular, considerable variation in crater outline and morphol­ Mertunan (1991) JGR .96. 13279. [23] Dietz (1964) J. Geol., 72, ogy can be found. Oblique impacts can produce craters with 412. [24] French (1970) Bull. Volcanol•. 34. 466. [25] Shanks and elongate outlines. as observed on the Moon and Mars [26-29]. An Schwerdlner (1991) Can. J. Earth. Sci.• 28. 411. [26] Wilhelms oblique, skipping impact event that created a series of elongated (1987) U.s. Geol. Surv. Prof. Pap. 1348. [27] Schultz and Lutz­ scars was discovered recently in Peru [30]. Fragmentation of the Garihan (1982) Proc. LPSC 13th. 84. [28] Nyquist (1983) Proc. impactor can produce elongated, noncircular crater patterns or U'SC13th.785; (1984)Proc.LPSC 14th. 631. [29] Mouginis-Mark multiple events as shown by the Henbury cluster. the Cape York et al. (1992) JGR. in press. [30] Schultz and Lianza (1992) Nature. meteorite field. and the East-West Clearwater pair. Erosion and 355.234. [31] Sims et al. (1981) Geol. Surv. Can. Pap. 81-10. 379. deformation can alter the original shape of an impact basin. e.g., [32] Krogh et al. (1984) Ontario Geol. Surv. Spec. Vol. 1. 431. MeteorCraterissomewhatrectangular.Theapparentnoncircularity [33] NyquistL. E. andShihC.-Y. (1992)U'SCXX111.1007-1008. of the Sudbury crater is not a strong argument against an impact origin when stacked against the host of shock features clearly associated with the Basin. MELTINGANDITSRELAT10NSHIPTOIMPACTCRATER Even ifthe SIC is anendogenously produced magma and not an MORPHOLOGY. lohn D. O'Keefe and Thomas 1. Ahrens, impact melt. the association ofimpactevents and magmatism may Lindhorst Laboratory ofExperimental Geophysics. Seismological nonetheless have important implications when considering the Laboratory 252-21. California Institute of Technology, Pasadena locus andstyleofplanetary magmatism. The close correspondence CA 91125. USA. inspaceand time between the impacteventand the magmatism that produced the SIC suggests a broadly genetic connection. especial­ Shock-melting features occur on planets at scales that range ly considering the overall paucity of magmatism of similar age from micrometers tomegameters. Itis the objective ofthis study to (1850 Ma) in the region [31.32]. In order to explain the composi­ determine the extentofthickness, volumegeometryofthe melt,and tional characteristics of the SIC, it appears necessary to invoke relationship with crater morphology. significant mixing of mantle-derived magmas with continental The variation in impact crater morphology on planets is influ­ crust. Spatial variations in mineral compositions away from wall enced by a broad range of parameters: e.g., planetary density (P), rock contacts suggest that the melt was actively assimilating wall thermal state, strength (Y), impact velocity (U), gravitational rock [10]. Intracratermeltrocks orbreccias may have been assimi- acceleration (g),... We modeled the normal impact of spherical S6 1992 Sudbury Conference

-4.,...------4.---.....--+-+-1--+-- -1-"'-1 ---f----+----·---i--1---f- -3 yt.so ~.3.4X'0·· -3 ~=16.5 -2 ..:r,. 6.2X 10" -2 ..::1 pU -1 ~o?** Svmtoll H/Hm - , CJ 0.1Si :l: I. ~O :.. (, 0.310 '-' ~2~ + D.H' ::J::1 3 3: v,E29 1.07 fu2 4~ '" U~ Q3~ .. 5l as 0 2.00 i At UttITern....,. j 01> 2.!50 6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6''---'----' RADIAL DISTANCE (rIa)

FIg.3. Meltlayermorphologyatlatertimesforagravity-dcminated complex crater. 10· ...------, projectilesonasemi-infmiteplanetovera broadrangeofconditions 10" using nwnerical techniques [1J. The scope of the calculations was ... I.g 10" defmed so as to span the range of dimensionless parameters that 10" characterize planetary impacts. These parameters are the inverse Froude nwnber. gaJU2, inverse Cauchy nwnber, Y/(p(J2), melt 10··+-~~....-~~...,...~~ ...... ~~...... j 10" 10· 10' 10' 10' nwnber, H.JU2, and shock weakening nwnber, H..JU2. Here H. is pHm the meltenthalpy (-1010 erg/g), and Haw is the enthalpy required to b -y negate the planet's strength (-I(P-I()' erg/g). Thecratermorphologicalregimeshavebeendescribedassimple bowl-shaped, flat-floored, multiple-ringed, central peak, and cen­ Fig.t. (a)Meltlayermorphologyattimeofmaximwnpenetration tral pit [2]. In the case of simple bowl-shaped craters, the strength for strength-dominated simple crater. (b) Melt layer thickness/ of the planet arrests the growth of the crater and is the dominant crater diameter at time of maximwn penetration as a function of factor in its shape (Fig. I). In all the other regimes, gravity arrests (Cauchy nwnber)-l. the growth and results in a rebounding ofthe depth and a collapse andpropagationofthecraterlip(Fig. 2). Multiplerings, peaks, and -4r--+--+---+--+--<----+--+---+--+--t pits evolve after the rebounding and collapse and are a resultofthe ~.10.2 ..r.. 8.17 X 10" interplaybetweenthe gravitational and strengthfon:es (Fig. 3). The -3 pU -2 planetarystrengthisalteredbytheimpactprocess.Thestrongshock wave melts, thermally weakens, and fractures the planetary mate­ -1 rial. We have modeled the melting !Uld thermal weakening and ~O~--r1W'" '/""",,,,,..... '-' Asphaung et al. [3] are addressing fracturing. We have determined I..... ' .... o '.117 the consequences of these effects on crater morphology scaling f). Ut, using the formalism ofHolsapple and Schmidt [4]. Using theresult + ...." • ...n ofcalculations, we determine the melt layer thickness/crater diam­ * t.07 eter (Taid) for simple and complex bowl-shaped craters. From + '.10 o nwnerical calculations, we fmd for simplecraters that die quantity <00 1.50 Taid is dependent only upon the material properties and scales as 6 -5 -4 -3 -2 -1 1 4 5 6 [Y/(pH".)]-O.2I (Fig. I b). Therelativemeltlayerthickness forcraters a RADIAL DISTANCE (rIa) that are dominatedby a shock weakening scale as (H..JHaw)-O.2I. For

gravity-dominated craters and a given craterdiameter,Tmid DC (HmI 10· y------, synobot , U2)-D2I, sothattherelativethicknessincreaseswithvelocity(Fig. 2b). 10" ____ 10' Inadditionto the melt layerthickness, wedetennined the scaling of the depth and total amount of melting for each of the cratering 10" -....-10... t-I~ regimes. These results will be presented along with a comparison 10" __ I. . with terrestrial [e.g., 5] cratering field results. 10" References: [1] Thompson S. L. (1979) SAND-77-J339, 10··+--~~.--~~...... ,.-~~,...j Sandia National Labs., Albuquerque, New Mexico. [2] Melosh H. 10" 10" Hm 10" 10" J. (I989)/mpactCralering,AGeologicProcess.Oxford,New York, 2 b U 245 pp. [3] Asphaung E. H. et al, (1991) U'SC XXlI. 37-38. [4] Holsapple K. A. and SchmidtR.M.(I987)JGR,92,6350-6376. Fig. 2. (a)Meltlayermorphologyattimeofmaximwnpenetration [5] Grieve R. A. F. et al, (199I)JGR, 96, 22753-22764. for a gravity-dominated complex crater. (b) Melt layer thickness/ crater diameter at time of maximwn penetration as a function of (Froude nwnber)-I. LPI Contribution No. 790 57

IMPACT CRATERING RECORD OF FENNOSCANDIA. 55 km. The impact cratering rate for Fennoscandia in the region L. J. Pesonenl and H. Henke12, ILaboratory for Paleomagnetism, where craters occur is 2.4 X 10-14 km-l a-I and includes 12 proven Geophysics Department, Geological Survey of Finland, Espoo, impact craters with diameters from 3 to 55 km. This amounts to 2 Finland, lDepartment of Photogrammetry, Royal Institute of events per every 100 Ma during the last 700 Ma. Technology, Stockholm, Sweden. There is increasing evidence that some (3, class E) ofthe large circular geological, morphological, or geophysical features [the A compilation (Fig. 1) of circular topographic, morphological, Uppland (No. 45), the Nunjes (No. 46), and the Marras (No. 55) or geophysical structures in Fennoscandia and adjacent areas re­ structures, Fig. 1] represent deeply eroded scars of Early Protero­ veals 62 craterform structures of which 15 (class A or a) appear to zoic impact craters, but impact-generated fQCks or fall-out ejecta beofextraterrestrialorigindue to meteorite impact. Themajorityof layers have not yet been identified with these structures. the structures are probable (class B, 9) and possible (class C, 34) No craterform structures of Archean age have so far been impact craters for which there is not yetsufficientprooffor impact discovered in Fennoscandia although, statistically, remnants of origin. Archean cratering events should be found in the Fennoscandian Four of the proven impact craters (Lappijarvi, No. 31, -77 Ma Shield. New ways of searching for these craters are proposed and old; Dellen, No.5, -90 Ma old; Mien, No. 20, -120 Ma old; and discussed. In addition to changes in the petrophysical properties of Jilnisjilrvi, No. 36, -700 Ma old) contain large volumes of impact rocks, such as density, magnetization, and electrical conductivity, melt and many other features of intense shock metamorphism. The redistribution of large volumes of rocks are associated with large ageoftherecognizedimpactcratersvary fromprehistoric(3500 B.C., impacts. Such changes in structures and rock properties may be No.38,Kaali)tolatePrecambrian(-1210Ma,No.ll,BjOrko). The identified by integrated interpretations of regional high-resolution histogramofthe ages (althoughthenumberofprovenimpactcraters geophysical data. is still very small) shows two possible peaks (Fig. I, inset): one The Siljan impact case shows, however, that the impact over­ group consisting of impact craters less than 150 Ma old and the printing can be very slight in comparison to geophY!lical anomalies second one with ages between 350 and 600 Ma. There is so far a caused by preimpact lithological and structural variations. deficiencyofimpactcraters in Fennoscandiawith ages between200 We review the Fennoscandian impact cratering record giving and 350 Ma. The majority of the proven impact craters have rim examples of geophysical signatures of impact craters. diameters between 5 and 20 km; the largestmeteorite impact crater in Fennoscandia, the Siljan (No.6, age -360Ma), has a diameterof BOHEMIAN CIRCULAR STRUCTURE, CZECHOSLO­ VAKIA: SEARCH FOR THE IMPACT EVIDENCE. Petr CRATERFORM STRUCTURES IN FENNOSCANDIA Rajlich, Geological Institute, Czechoslovak Academy ofSciences, c....llIc.tlon Rozvojova 135,165 00 Prague 6, Czechoslovakia. @ A Impact cra'" • R.C protliIIbIe. pOulble impact crater :45' E large circular atructure Test of the impact hypothesis (1] for the origin of the circular, .... F impact breccia .lte +. "ry Iman crater 260-km-diameter structure of the Bohemian Massif (Fig. 1) led to the discovery of glasses and breccias in the Upper Proterozoic o sequencethatcanbecomparedto autogeneous breccias [2] oflarger craters. The black recrystallized glass contains small exsolution 6$'

Fig. I. Impact craters and other craterform structures in Fennoscandia and adjacent areas. Encircled structures refer to 100 km proven impactcraters (class A); the others refer to classB (probable) '--_-'--_--II and class C (possible) impact craters respectively. The very large 1/' 3 • 4 &a circular patterns refer to class E structures for which impact origin is not yet proven. The class F sites represent locations of breccia Fig. I. CircularstruetureoftheBohemianMassif; l-topographical occurrences without known crater structure. The very small features, 2-important faults with geological contacts of units Quaternary craters (class a) are denoted with a plus sign. (Inset) differing in mobility in Variscan orogenesis, 3-outcrops of Ages and diameters of the proven (classes A and a) impact craters. autogeneous breccias, +-extent of the Upper Proterozoic series. 58 1992 Sudbury Conj'ereru:e

TABLE 1. Chemical composition of fragmenlS and melt veins.

% ~O+ SiOz AlZO] F;O] TiOz Cao MgO Nap KzO MnO PzO, Melt 1.11 59.53 21.52 8.64 1.15 0.62 1.98 0.95 4Jf1 0.053 0.13 CR 1.45 57.76 23.60 8.41 1.19 0.21 1.86 0.61 4.47 0.050 0.12 Melt 1.06 60.61 21.42 8.32 1,(19 0.50 1.91 0.71 3.91 0.058 0.14 Melt 0.99 60.92 21.34 8.37 1.08 0.46 1.89 0.62 3.89 0.053 0.14 ~R 1.36 60.73 18.84 8.34 0.84 1.20 3.04 2.54 2.40 0.310 0.11

ppm Ba Co Cr Nb Ni Rb Sr Y Zn Zr

Melt 958 26 126 24 40 145 173 34 102 209 CR 938 26 153 26 51 132 153 36 99 190 Melt 877 2S 117 22 43 150 150 33 92 194 Melt 874 24 163 21 56 148 146 34 94 193 C·R 392 29 146 13 54 90 95 29 96 126

C·R-country rock from place more distant than other samples.

10

~Co lNb ppm

30 50 '00 200 300 500 900 Fig. 2. Microphotograph of the recrystallized quartz glass with COUNTRY ROCK exsolutions of feldspar and biotite, x 100.

Fig. 4. Correlation graph ofthe chemical composition ofcountry rock and melt.

elements. It is slightly impoverished in water. The proportion of melted rocks to fragments varies from 1:5 to 10:1 (Fig. 2). The mineralogy of melt veins is the function of later, mostly contact metamorphism. On the contact of granitic plutons it abounds on sillimanite, cordierite, and small bullets of ilmenite. Immediately on the contact with syenodiorites itcontains garnets. The metamor­ phism ofthe impactrock meltseems the mostprobable explanation ofthe mineralogy and the dry total fusion ofrocks accompanied by the strong fragmentation. The next rocks in the top of the sequence are largerbodies oftehermakitic metamelagabbros and conglomer­ Fig. 3. Outcrops of the autogeneous breccia with rock fragments ates with the volcanic matrix. Crystalline rock fragments are cemented by the melt. frequently found here. Structurally theyresembleorthogneisses and granites, but with a glassy appearance especially of feldspars. The very-fme-grained textureofrocks is explained tentatively as recrys­ crystals of albite-oligoclase and biotite, regularly dispersed in the tallized shocked rocks. Some parts of the conglomerates can be matrix recrystallized to quartz (Fig. 2). The occurrence of these suevites as well. The breccia formation and conglomerates are rocks is limitedto a l-km2 area. Itisdirectly underlainby thebreccia intercalated between the Moldanubian gneisses that transgrade the (Fig. 3) ofthe pelitic and silty rocks cemented by the meltedmatrix, circular structure and between the Upper Proterozoic sequence, found on several tens of square kilometers. The melt has the same which on the map is crescent shaped, contouring the northern half chemistry as rock fragments (Table 1, Fig. 4) in major and in trace of the circular structure (Fig. 1). The lower part of the Upper LPI CtHltribllliOlaNo. 790 S9

Proterozoic sequence begins with pillow lavas and is tenninated by Sinee 1987 importantnewresults wereprovidedby Hartetal. [3, the sedimentary sequence with shallow water fossils (from <10 m and refs. therein] dealing with thegeochemistry ofthe sraniticcore depth; VavrdovA, oral comm\Ulication), indicating the successive and aspects ofdynamic metamorphism. Reimold [4] evaluated the fdling ofthe hole. Thetotal thicknessofthis formation isnotknown, geochemical database for Vredcfortpseudotachylite, andnew chro­ though geophysicalmodels indicate severalkilometers. Therestric­ nological data were contributedby [5] and [6]. ContinuedSlrUCtural tionofbreccias tothebaseofthis formationprovides age constraints workhadbeendemanded bythe participants ofthe 1987 workshop. that would indicate the age of the impact is 1.8-1.2 m.y. Colliston and Reimold [7] presented the results ofa fust detailed The circular SlrUCture is defmed by the topography, water StlUctural study inthe southernpartofthe Dome and in areas ofthe courses, and geological contacts in the Tertiary through Upper northwestern sector. Minnitt et aI. [8] mapped the Archean green­ Proterozoic sequences. It is visible also on the fault geometries in stone terrane in the southcuternquadrant and completed SlrUCtural the brittle and in the ductile stages from later orogenies as featured analysis ofthe granite-gneiss exposures in the southern part. Both by the half circular Permian and Cretaceous sedimentary basins. studies resulted in similar findings, suggesting that deformation in The rigid conservation ofthe circular form is tentatively explained the basement is mainly ofAIchcan age and related to a stress field, by the later cooling of the upper mantle rocks \Ulder the slrUCture in which the principal stresscs operated in a near-horizontal plane after the impact, enabling them to behave rigidly. Several shearing (cf. Colliston and Reimold, this volume). Latermacroscopic defor­ phenomena CRCO\Ultcred in crystalline rocks of the Moldanubian mationismainlyrestricted tolocalsubvertical shearzoncsscattered can be attributed to the excavation stage. throughout the sraniticcore. Inthecentral partofthe coredeforma­ References: [1] BouSka V. (1990) Vesmir. 69. 9.487-492 (in tion is very limited. New 4OAr_!9Ar stepheating results [9] for Czech). [2] Masaytis V. L. et a1. (1980) Geology ofAstroblemes. mineral separates from host rocks to two pseudotaehylite samples Nyedra, Moscow, 1-231. that were dated by [6] atca. 1.4Gafurthersupported the conclusion that these breccias were formed at post-2-Ga times. Currently several structuralprojects inthe collarare inprogress, mE VREDEFORT OOMF-REVIEW OF GEOLOGY AND with preliminary reports indicating that several deformation events DEFORMATIONPHENOMENA AND STATUS REPORTON sincedepositionoftheWitwatersrandSupersroup(ca.2.7S-3.05 Ga CURRENT KNOWLEDGE AND REMAINING PROBLEM· ago) could be recognized. Consequently, one aspect of utmost ATICS (FIVE YEARS AFTER mE CRYPTOEXPLOSION importance for future research must be to establish a complete WORKSHOp). W. U. Reimold, Econ. Geological Res. Unit at chronological frameworkforthe geologicalevolution in thisregion. the DepartmentofGeology, University ofthe Witwatersrand, P.O. The igneous rocks that intlUded core and collar of the Dome at Wits 2050, Johannesburg, R.S.A. various times since Ventersdorp volcanism (ca. 2.7 Ga ago) are currently being studied as possible candidates for radiometric TheVredefortSIrUCture locatedintl1e centeroftheWitwatersrand dating. 4OAr_39Ar stepheating and laser Ar dating of the various basin in South Africa and the Sudbury slrUCture in Canada are generations ofpseudotaehylite identified in both the stlUcture and widely considered the two oldest and largest impact slrUCtures still the Witwatersrand basin should be continued as well. A detailed evident on Earth. Both stlUcture5 are very similar in a number of metamorphic project, comparing the rocks of different metamor­ geological aspects (e.g., association with major economic ore phic grades in the northwest/west (high) and northeast Oow) sectors deposits, similar ages ofca. 2 Ga, ab\Uldant pseudotaehylite as well respectively with the metamorphic record for the whole Wit­ as shatter cone occurrences, overturned collar), as summarized by watersrand basin, has just been initiated. It is also still \Ulcertain at [1]. However, whereas the geological community generally accepts what times the major metamorphic events took place and whether an impactorigin for the Sudbury slrUCture, anumberofresearchers the enhanced metamorphism in the northwest/west is the result of are still reluctant to accept this for the Vredefort Dome. contactmetamorphisminthevicinity ofalkali granitic intrusions or Five years ago an international workshop focusing on the ofregional metamorphism. The nature ofthe pseudotaehylite-rich Vredefort slrUCture [2] scrutinized the evidence and attempted to and chamockite-bearing transition zone between Outer Granite resolve the differences between impact supporters and protagonists gneiss and Inlandsee Leucogranofels is still controversial: Does it of internal genetic processcs. Clearly. this goal was not achieved, represent apre-Vredefortintraerustallithologicalbo\Uldary, athrust but at least a number of important areas of further research were plane, oradecollement zonepossibly linked tomajorpre-Vredefort defmed. Research in the Vredefortstructuregained new momentum gravity slides in the northern Witwatersrand basin? What is the in 1991,partially in anticipationofthe Sudbury '92Conference. and significance of the chamockite occurrences that to date have not because several mining houses realized how important full \Ulder­ been studied in detail? New quarry exposures in and near this zone standing of the structure and evolution of the Vredefort Dome is are being studied and could reveal the three-dimensional geometry with regard toexploration and mining activities in the surrounding of pseudotaehylite breccia zones. Finally. (sub)planar micro­ Witwatersrand basin. deformations in Vredefort quartz have now become the object of Besides the long-established impact and gas explosion hypoth­ TEM investigations. eses, several other genetic processes have been discussed in recent At this pointin time, the main arguments in favor and against an years: rapid updoming, thrusting, combinations ofseveral tectonic impact origin for the Vredefort slrUCture can be summarized as processes, and an impact event at 2 Ga ago followed by tectonic follows. modification. Reviews ofthe geology andgeophysicsofthe Vredefort ProImpact: (l)Thestructure isregarded as being circularand stlUcture were repeatedly presented in recent years (e.g., [3,4] and (2) surro\Ulded by ring faults. (3) Thedome itselfis considered tobe several papers in [2]). Therefore the aim ofthis review is topresent the central uplift ofa gigantic impact stlUcture with (4) a "crust-on­ new data, tohighlight the most obvious shortcomings in the CWTent edge" configuration of the SIrUclUJ'e, involving both overturned database, and to summarize the major arguments in the genetic collar and basement. (5) The presenceofshattercones. (6) Pseudo­ controversy. tachylite is regarded as an equivalent of impact breccia and (7) the 60 1992 Sudbury Conference

granophyre u melt rock. (8) Shock metamorphism occurs in the structuralorpt-related)evidencethatisnotreadilyexplainedby the form of shock-characteristic planar microdeformation features impact hypothesis. Unfortunately, in recent years these workers (PDFs) in quartz. have been ridiculed in a quite unscientific Wily, e.g., u "academic Contra Impact: (1) The structure is uynunetrical and poly­ dinosaurs" or "reactionary diehards." gonal. (2) The southern equivalent to the collar in other sectors In this paper importantobservationsonVredefortandWitwater­ shows subhorizontal stratigraphy. (3) The"crust-on-edge"modelis srand pseudotachylite are summarized (for more detail, cf. (5». onlyvalidforthenorthernpartofthedome. (4)Thereisonlylimited DIstribution ud Styles of Development: Major pt occur­ structuralevidencefor2-Gadeformation. (5) Deformationintensity rences on the Dome are concentrated along the transition. zone does not increase toward the center, and deformation in the central between Outer Granite gneiss and Jn1andsee Leucogranofels, u area is generally poorly developed. (6) Deformation is magnified well as along a northeut-southwest-trending zone: justsouthofthe along northeut-southwest-trending lineaments. (7) Severalphues Inlandsee. Brecciation in the central core region is extremely of defonnation have been identified. (8) Vredefort deformation limited.Withinthecollatstrata.ptmainlyoccurs intheformof<30­ phenomena are also observed in the northern Witwatersrand buin. em-wide veins along bedding faults, but up to several-meter-wide (9) Temporal relationships between MSJS/shatter cones and zones comprised of intercalated networks of narrow veinlets and pseudotachylite are complex and multiple. (10) Microdeformation more musive melt breccia have recently been observed in mafic isrestrictedtocontroversial"features"inquartzand kinkbandingof intrusives in the collar. Throughout the Dome and thenorthern part mica and occasionally hornblende. No other characteristic shock of the Witwatersrand basin pt is also found u thin melt films on effects have been described from other minerals. (11) Temporal slickenside and shatter cone (MSJS) surfaces. In the granitic core relationships between the various deformation and structural/mag­ onefmds eithermusivedevelopment(networkbreccia)oruptoSO­ matic events are complex and u yet not sufficiently resolved. em-wide veins. Network brecciu are occuionally seen to be References: [1J Dressler B. O. and Reimold W. U. (1988) delimitedatonesidebya thick,straightveinthatpossiblyrepresents Snowbird II, 42-43, LPI Contr. 673. [2J Nicolaysen L. O. and the initial generation vein. Several new quarry exposures indicate ReimoldW.U.(I990)Teclonophysics,l71,422. [3JHartR.J.et al. subhOrlzontal internal structures within major breccia develop­ (1991) Teclonophysics, 192,313-331. [4J Reimold W. U. (199I)N. ments, but individual large-scale breccia zones appear to have Jhrb. Miner. Abh., 161, 151-184. [SJ Allsopp H. L. et al. (1991) S. overall vertical attitude. Thin veins generally resemble tectonic pt Afr. J. Sci., 87, 431-442. [6J Reimold W. U. et al. (l990a) occurrences. Displacement is usually variable inelm to m intervll1s Tectonophysics,l71, 139-152. [7J CollistonW. P. and Reimold W. and ranges normally from <1 mm to >50 em (but

the (sub)planar defonnation structures in quartz represent PDFs or crustal partial melting event that released KREEP lava flows over planar fractures? It is our contention that they generally resemble a wide area. These two events are at present indistinguishable in planar fractures rather than the shorter and closer-spaced PDFs radiometric age. The sample record indicates that such KREEP (compare Figs. 7b,c of [4] with Vredefort microdefonnation). volcanism had not occurred in the region prior to that time, and Other unresolved problematics regarding Vredefort pt are the never occurred again. Such coincidence in time implies a genetic nature and origin of the enigmatic granophyre that, besides being relationship between the two events, and impact-induced partial regionally homogeneous in composition, displays a number of melting appears to be the only candidate process. characteristics similar to those of pt. Major shortcomings in the pt This conclusion rests essentially on the arguments that (1) the database are (1) absolute ages for the several phases ofpt develop­ Imbrium Basin event took place 3.86 ± 0.02 Ga ago; (2) the ment and of structural deformation, (2) understanding of the geo­ Apennine Bench Fonnation postdates Imbrium; (3) the Apollo 15 logical structure in the zones ofmajor ptdevelopment and (3) ofthe KREEP basalts are 3.85 ± 0.03 Ga old; (4) the Apollo 15 KREEP internal structures ofpt-rich (fault?) zones, (4) P-T-x conditions at basalts are derived from the Apennine Bench Fonnation; and (5) the

pt fonnation--also with regard to HP Si02 polymorph generation, Apollo 15 KREEP basalts are volcanic. Thus the Apollo 15 KREEP and (5) the relationships between Vredefort and Witwatersrand pt basalts represent a unique volcanic unit that immediately postdates and Witwatersrand fault structure. the Imbrium event (within 20 Ma, possibly much less). References: [1] SchwarzmanE. C. etal. (1983) Bull. GSA, 94, This abstractsketches the evidence for the links in the argument, 926-935. [2] Dietz R. S. (1992) Earth, 1,22. [3] Reimold W. U. et describes some implications for initial conditions, and briefly al. (1987) JGR, 92, 747-758. [4] Stoffler D. et al. (1992) GCA, 55, explores ramifications of the process for the early history of the 3845-3867. [5] Killick A. M. and Reimold W. U. (1990) S. Afr. J. Earth. Geol., 93, 350-365. [6] Reimold W. U. et al. (1990) Geol. Soc. The Age of Imbrium: Samples collected at the Apennine Austral. Abstr., 27, 82. [7] Allsopp H. L. et al. (1991) S. Afr. J. Sci., Front must be dominantly isotopically reset either by the Imbrium 87,431-442. [8] Reimold W. U. (l991)N. Jhrb. Miner. Abh., 162, event or by older events. Analyses by laser argon relealle methods 151-184. [9]MartiniJ.E.J. (1991) EPSL, 103,285-300. [10] Reim­ [3,4] ofvaried impactmelts from therubble that fonns the Apennines old W. U. (1990) S. Afr. J. Geol., 93, 645-663. constrains the basin to be probably no older than 3.836 Ga, and extremely unlikely to be older than 3.870Ga. Imbrium must also be younger than Serenitatis, reliably dated at 3.87 Ga. A younger limit COINCIDENCE IN TIME OF THE IMBRIUM BASIN IM­ set by the Apennine Bench Fonnation arguments is within uncer­ PACT AND APOLLO 15 KREEP VOLCANIC SERIES: tainty the same age. IMPACT-INDUCED MELTING? Graham Ryder, Lunar and The Stratigraphic Age of the Apennlne Bench Forma­ Planetary Institute, 3600 Bay Area Boulevard, Houston TX 77058, tion: This extensive plains unit inside the Imbrium Basin under­ USA. lies Imbrium-age craters and mare units (Fig. 1). However, it overlaps basin topography, hence is at least slightly younger than On the Earth there may be no finn evidence that impacts can the basin itself [5,6]. inducevolcanic activity [1]. However, the Moondoesprovide avery The Age ofApollo 15 KREEP Basalts: This distinct group of likely example of volcanism induced by an immense impact: The intersertal igneous fragments, widespread at the Apollo 15 landing Imbrium Basin-fonning event was immediately succeeded by a site, gives Ar-Ar and Sm-Nd crystallization ages of 3.85 ± 0.05 Ga

Fig. 1. Orbital photograph ofthe Apollo 15 landing site and relevant environs. Apart from the large area to the left ofthe picture, Apennine Bench Fonnation occurs at the area of the Fresnel Rilles. It probably also underlies the mare near the Apollo 15 lap.ding site and may evell be expOsed at the North Complex (a dimple just to the north of the landing site) and elsewhere nearby. The mountains to the right are !he Apennines, a prominent ring of Imbrium. 62 1992 Sudbury Conference

[summary in 7], indistinguishablefrom the age ofImbrium. TheRb­ LPSC9th, 3379. [6] WilhelmsD. (1987) U.S. Geol.Surv.ProfPap. Sr ages are slightly older (slight Rb loss?) but within uncertainty of 1348,302. [7] Shih C.-Y. et a!. (1992) EPSL,108, 203. [8] RyderG. a 3.85-Gaage. KREEP basalticmaterialofothercharacteroragehas (1987) Proc. LPSC 17th, in JGR, 92, E331. not been identified in the region, thus these basalts appear to represent a unique event in the region. Apollo 15 KREEP =Apennine Bench Formation: Themor­ APOLLO 15 IMPACT MELTS, THE AGE OF IMBRIuM, phology of the Apennine Bench Formation indicates subsidence of ANDTHEEARTH·MOONIMPACTCATACLYSM. Graham a fluidlike material, consistent with volcanic flows [5,6]. The Ryder' and G. Brent Dalrymple2, 'Lunar and Planetary Institute, deconvolvedorbital geochemistry shows the formation to be chemi­ 3600 Bay Area Boulevard, Houston TX 77058-1113, USA, 2U.S. callyverysimilarto the Apollo15 KREEPfragments. Theformation Geological Survey, 345 Middlefield Road, Menlo ParkCA94025, occurs very close to the Apollo 15 landing site (Fig. I), and is USA. inferred to underlie the mare basalts near Hadley Rille; it may even be exposed at the North Complex, intended to be visited on the The early impact history of the lunar surface is of criticlil Apollo 15 mission but missed because of time delays. The Apollo importance in understanding the evolution of both the primitive 15 KREEP basalts are ubiquitous at the Apollo 15 site, and most Moon and the Earth, as well as the corresponding popul3:tions of ·must represent a local, not an exotic, component. The age of the planetesimals in Earth-crossing orbits. Twoendmemberhypotheses KREEP fragments is consistentwithrequirements forthe Apennine call for greatly dissimilar impactdynamics. One is aheavy continu­ Bench. The correlation of the Apollo 15 KREEP basalts with the ous (declining) bombardment from about 4.5 Ga to 3.85 Ga..The Apennine Bench Formation is almost inescapable. other is that an intense but briefbombardment at3:bout 3.85 (±?) Ga Apollo 15 KREEPBasaltsAre Volcanic: The basalts are free was responsible for producing the visible lunar landforms and for ofclasts ormeteoritic siderophile contamination, and have a range the common 3.8-3.9-0a ages of highland rocks. ofcompositions indicating crystal separation (unlike impactmelts) No impact melts among lunar samples have been found with an and lying along the plagioclase-low Capyroxenecotectic [2]. Some age greater than 3.9 Ga [1]. A heavy continuous bombardment demonstrate nonlinear cooling rates inconsistent with cooling of requires such melts to have once been common, and their absence impact melts [8]. They cannot represent an average crustal compo­ requires either that they are present but have not been sampled, or sition such as would be represented by the Imbrium impact melt that they have been reset continuously or terminally at dates later because they are so radiogenic. than 3.9 Ga. The chemical variety of dated impact melts suggests With such a coincidence in age of a giant impact basin and a that at least several impacts have been dated, not just a limited unique flood basalt eruption, the most reasonable conclusion is sample reset by Imbrium and Serenitatis. Most ejecta in an impact cause and effect: The unloading and heat input from the Imbrium is deposited cold and is not radiometrically reset even for Ar Basin impact was directly responsible for thepartial melting ofahot (although it can be disturbed), as shown by studies of both experi­ crust producing the Apollo 15 KREEP basalt floods. The chemical mentally and naturally shocked materials [2-4].Resetting shouldbe and isotopic evidence suggest that a large amount ofpartial melting accomplished only ornearly only by making a new impactmelt, yet of a crustal source is required to produce these basalts. The small lunar samples clearly show that not all ofthe lunar crust has been so gravity field on the Moon shows that the pressurereliefofunloading converted; old melt should remain if it once existed. Furthermore, even 100kmis only0.5 OPa, and brings amass ofsuitable rock only the existence of old basalts and plutonic rocks suggests that old 60 K closerto its melting point [1]. The unloading ofthe lunarlower impact melts should have been preserved, had they existed. These crust would have been less than that, and with latentheatofmelting arguments should be persuasive that no heavy bombardment in the to take into account, not much melting can be expected. Thus, if period from at least 4.3 to 3.9 Ga occurred [1]. Apparently, for impact-induced crustal melting is responsible for the Apollo 15 various reasons they are not persuasive [e.g., 5]. Thus reliable ages KREEP basalis, the sOllrce must have been at or very close to its for impact melt rocks of even more varied composition (hence melting temperature anyway, or melts induced by pressure release potentially distinct origins) are needed to further test the various of the mantle added their heat to the source by upward movement early impact hypotheses, and particularly to establish the relative without actually reaching the surface. abundance ofold impact melts. The oldest Earth roCks of any significant volume have an age The Apennine Front, the main topographic ringpfthe Imbrium similar to that of the lunar cataclysmic bombardment. Older crust Basin, was sampled on the Apollo 15 mission.Th~ rocks in the either did not exist, or was essentially annihilated at that time. A massif represent two main sources: (1) pre-Imbriuin masses that hotter Earth at 3.86 Ga ago was perhaps very susceptible to impact­ have been uplifted by the event itself, and consist of pre-Imbrium induced partial melting, causing very extensive recycling even of rock units, and (2) ejecta from the Imbrium event, consisting of nonsubductable granitic crust. Even larger planetesimals would material melted in the Imbrium event and older material [6]. Either have hit the Earth than the Moon, traveling even faster; the effects way, ifimpactmelts existed in the regionpriorto thehnbrium event, ofpressure release would have been greater because ofthe stronger they should now be part of the Front. Material formed in impacts gravity field, and more material close to its melting temperature. .younger than the ImbriumBasinmust beminor, ofvery localorigin, Such melting could have had drastic effects in remixing and and from smallcraters (which tend to produceglassy meltproducts). assimilating old crust into upper mantle material to add to an The Apollo 15impact melts show a diversity ofchemical composi­ assumed plate-tectonic recycling that could not by itself be very tions, indicating their origin in at least several different impact efficient for granitic material. events [e.g., 7,8,9]. The few attempts at dating them have generally References: [1] Melosh H. J. (1989) Impact Cratering, Ox­ not produced convincing ages, despite their importan<:e. Th1JSc\V~ ford, 245 pp. [2] Ryder G. (1989) LPSC XX, 936. [3] Dalrymple G. chose to investigate the ages of melt rock samples'from' the E. and Ryder G. (1991) GRL, 18, 1163. [4] Dalrymple G. B. and Apennine Front, because of their stratigraphic import:anc~ yet lack RyderG. (1992) Trans.AGU 14Suppl., 362. [5] SpudisP.D.,Ptoc. of previous age definition. .. < ••• LPICotItributioft No. 790 63

Usingacontinuouslasersystan,wehaveobtainedhigh-resolution abundance and size oftektites across the strewn field, variation of • «lArptAr age spectra on single fragments of12 melt rock samples thickness of microtektite layers in ocean cores, nature ofablation from the Apennine Front The melt rocks, all fine-grained, are characteristics across thefield, and, above all, theoccurrence ofthe essentially aluminous basalts, but with a variety ofcompositions, large, blocky, layered Muong Nong-type tektites in Indochina. A e.g.,MgO9to21%,Sm 2to 25 ppm. Webelieve they mustrepresent recent study ofthe location and chemistryofMuong Nong-type and at least several different impact events. A few milligrams ofeach splash-form tektites suggests that the source region can be further sample were crushed to submillimeter sizes and individual frag­ narrowed to alimitedareaineasternThailandand southernLaos [1]. ments, visibly free of clasts and weighing 62 to 620 mg, were There are four lines ofevidence that point toward this area. The irradiated. They were analyzed with a continuous Ar-ion laser fmtisthe observationthat tektitesites inIndochinaarenonrandomly extraction system and mass spectrometer [10,11]. Individual par­ distributed. Many sites seem to be located along linear trends or ticles were incrementally heated, with temperature measured with ''rays'' separated by areas relatively sparse or devoid of samples. an infrared radiometer. We have obtained 26 age spectra on the 12 These rays converge to a small areaalong the Thailand-Laos border melt samples. Someofthese results have beenpreviouslypublished between 15.5ON and 170N latitude. Second, there is a somewhat [11]. larger region, enclosing the area delineated by the rays, where Of the 12 rocks analyzed, 7 have age spectrum plateaus that we Muong Nong tektites predominate and/or there is no mention of interpret as crystallization (impact) ages. Individual plateaus have splash-form-type tektites. Third, a high proportioo of the reported 2-sigmauncertainties of±16 Ma. The «lArptAr plateaus are gener­ sites containing super-sized (>1kg) Muong Nang tektites are inthis ally well dermed in the intennediatetemperature range with40% to area. Lastly, Muong Nong tektites with this area show the largest 70% of the 39Ar released. Six of these ages fall within the narrow chemical inhomogeneity in sites, and· there is a high chemical range of3879 Ma and 3849 Ma, more or less within uncertainty of gradient across the region; these are characteristics one would acommon age. Speclraonfive fragments ofonesamplegave arange expect proximal to the source. The area defmed by the above from 3856 Ma to 3879 Ma. The seventh sample gave a plateau age evidence is centered at 16ON/losoE, with aradius ofapproximately of 3836 Ma. The total span of ages is less than 1%, a very narrow 125km. range. The remaining five samples show spectra that clearly indi­ Satellite multispectral imagery, a digital elevation dataset, and catedisturbance by post-3.8-Ga events, and lackplateaus. None of maps showingdrainage patternshave beenused to searchwithin this the26 age spectrafor the 12 meltrocks show any indicationofolder area for possible anomalous features that may be large degraded meltcomponents. Aconventional«lArptArageof3.85±0.05 Gafor impactcraters. Fourinteresting structureshavebeenidentifiedfrom adifferentimpactmelt from the Apollo 15 landing site was reported these datasets: by [12]. 1. An approximately 3Q-km-diameter, quasicilcular slrUCture in We believe that these data provide ages for a variety ofimpact Laos, resembling a partially infilled impact SlrUCture, centered at melts thatare coeval withorpredatetheImbriumevenL Thus afmt­ 16.35°N/I06.15°E. It has a relatively flat floor sUlTOunded by hills order conclusion is that the Imbrium event is no older than about rising 70 m toseveral hundred meters above the floor, and a central 3870 Ma, and probably no older than 3940 Ma. Independent elevated area rising about 100 m above the floor (Fig. 1). The evidence suggests strongly that Imbrium is not younger than this structure is breached at approximately the cardinal points byrivers. (because ofthe laterKREEP basalts), hence is indeed very close to 2. An approximately 25-km-diametercircularfeature on the east 3840 to 3850 Ma. In that our data show a variety ofmelts at orjust side of the Mekong River, slightly east of Savannakhet, Laos before this time but not older melt, we believe it to be consistent (16.55°N/l04.900E). This feature is not an obvious depression or with avery tightlyconstrainedbombardmentoftheMoon. Serenitatis crater, but is an approximately cilcular area enclosing hummocky (about 3.87 Ga) falls in this same period. We have still no tangible terrain of very low relief. evidence for significant bombardment prior to 3.9 Ga. 3. A9O·km-diameter area, centered at 16.6ON/lOS.soE (directly References: [1] RyderG. (1990)&8,71,313. [2] Jessberger to the east of slrUCture 2). This broad south-sloping feature is Eo andOstertagR. (1982)GCA,46, 1465. [3] BogardD. etal. (1988) rimmed by high hills on the north andeast, rising to 450m above the GCA, 52, 2639. [4] Stephan T. and Jessberger E. (1992) GCA, 56, floor, but only low lying hills to the west and south. The area is 1591. [5] Schmitt H. (1991) Am. MiMral., 76, 773. [6] Apollo 15 drained by two rivers flowing to the south. Preliminary Science Report (1972)NASA SP-289. [7] Ryder G. and 4. An oblong depression on the west side and in a curve of the Spudis P. (1987)Proc.LPSC 17th, inJGR, 92, E432. [8] LaulJ.C. Mekong River, approximately 80kmnortheastofUbonRatchathani, et al. (1988) Proc. LPSC 18th, 203.[9] Lindstrom M. et al. (1990) Thailand. It is approximately 30 km long northwest-southeast, and Proc. LPSC20th, 77. [10] DalrympleG. B. (1989) u.s. Geol. Surv. about 20 km wide southwest-northeast. Hills rise about 75 m in the BII1I., 1890,89. [11] DalrympleG.B. and RyderG. (l991)GRL,18, southwestto over300minthesoutheastabove theflat floored plain. 1163. (12] Bogard D. et al. (1991) LPSC XXll. 117.

SEARCH FOR THE 700,OOO-YEAR·OLD SOURCE CRATER OF THE AUSTRALASIAN TEKTITE STREWN FIELD. C. C. Schnetzlerl and J. B. Garvin1, IGeography Department, UniversityofMaryland,CollegeParkMD20742,USA, 2Qeodynamic Branch, Code 921, Goddard Space Flight Center, Greenbelt MD 20771, USA.

Many tektite investigators have hypothesized that the impact FIg. 1. Profiles across slrUCture I,centered at16.35ON/l06.15°E, crater that was the source ofthe extensive Australasian strewn field in southern Laos. From top to bottom, southeast-northwest, lies somewhere in or near Indochina. This is due to variations in southwest-northeast, west-east, south-north. 64 1992 SIIdb"" COII/erence

All these features lie within a broad region ofMesozoic marine As a result, the distinctive signatures ofcarly energy tranlfcr may sedimentary rocks, the Indosiniu formation; primarily sandstones not be completely consumed by crater growth at planetary scales. interbedded with shales and limestones, which covers much of Twokey diagnostic features ofoblique impacts can be found in central Indochina. The age and composition ofthese sediments arc craters at ~th laboratorY and planetary scales. First, oblique broadly consistent with Australuian tektite composition and age impacts create a distinctive uymmetric crater profile with the [2]. FieldworktoexamincthesestrucnuesandcollectCO\Ultryrocks deepest penetration and steep inner wall uprange and a shallow, for specific comparisons would seem warranted. shelflike wall downrange [3]. Such a profile occurs for impact RefereDc:eS: [1] Schnctzler C. C. (1992) Meteoritics, June. angles from 45° to ISO in strength-eontrollcd craters (e.g., alumi­ [2] Blum J. D. ct al. (1992) GCA, 56, 483-492. num targets) but requires impact angles less than 5° for gravity-

RECOGNIZINGIMPACTORSIGNATURESIN THEPLANE· 3..------...., TARY RECORD. Peter H. Schultzl and Donald E. Oault2, Two-ring basins IDcpartmentofGeologicalScicnccs,BrownUniversity,Providence RI, USA, 2Murphys Center ofPlanetology, Mmphys CA, USA. 2 Cratersizereflects thetargetrcsponseto the combinedeffectsof 1 impactor size, density, and velocity. Isolating the effect of each ~ variable in the cratering record is generally considered muked, if ;r. not lost, during late stages ofcrater modification (e.g., floor uplift ­. andrimcollapse). Importantclues,however,comefrom thedistinc­ tive signatures of the impactor created by oblique impacts. In laboratory experiments, crater diameter exceeds impactor MOlIGIJ diameter by a factor of 40 for vertical impacts into gravity­ o rna• Mal controlledparticulate targets and reduces to 25 for oblique impacts at 15° from the horizontal. Strength-controlled cratering in alumi­ • M-. num reduces these factors to 5 and 2 respectively. As scale in­ -I +-..--...... -,.-----...--...... -,.-----i crcues, crater excavation is limited by gravity and cratering -6 -s -4 -3 -2 efficiency becomes progressively less efficient. A l00-km-diam­ ...... ( .., ..)111 ctercrateronEarth (rim-to-rimdiameterwith25%cnlargcmentdue toslumping)is onlya factorof12greaterthan theimpactordiameter Fig. 2&. Crater diameter (0) scaled to impactor diameter (d) for a vertical impact and reduces to 6 for a 15° impact angle hued inferredfrom peak-ring diameteru a functionofthedimensionless onscaling relations given in [1]. Thecarly compression stage [2] at gravity-scalingparameter~cqualto3.22gr/V2withradius(r)again planetary scales, therefore, comprises a significant fraction of the inferred from peak-ring diameter (g and v refer to gravity and fmal crater size and approaches a value more typical of strength­ impactvelocityrcspcctively;&'/&'rcprcsa1tsprojcc:ailc/tarBetdcnsity controlledlaboratoryexperiments,particularly forobliqueimpacts. ratio). Datafor !heMoon.MereUl)', and Mars <:ollapsc ontoa single relation for impact velocities 14, 16, and 32 km/s respectively. IMPACT ANGLE 3...------, Central Peaks 1.0.------.---...... ----.----.,,-, • turbulent run-out flows 2 .. x laminar run-out flows ­a. 'tJ Q -...... 0.5 Q .9

o 0.0 '-----""-----'""'------1.5 -1.0 -0.5 0.0

Log (sln2tJ) .1 +---.-,r--o--..,.--..--.-,r--r---"'1 -6 -s -2 Fig. 1. Crater diameter (D) scaled to peak-ring diameter (d,.) on ...... ( ..'..,.. Venus u a fw1ction of impact angle estimated from the degree of uymmetry inthe ejectadeposits. Ifpeak-ring diameterdepends on Fig. 2b. Craterdiameter scaled to impactordiameter u inFig. 2a impactor size (given velocity), D/eJ,r should dccrcuc u impact butreferenced to the diameterofthe central peak. Assumed impact angle dcqeascs, u is shown. Different symbols correspond to velocities for the Moon and MereUI)' in this case arc 16and40 km/ different styles oflong run-out flows observed on Venus [6]. s respectively. LPIContribution No. 790 65

controlled targets (sand). Large simple craters on the Moon such as PARADIGM LOST: VENUS CRATER DEPTHS AND mE Messier exhibit the same profile. Larger complex craters, however, ROLEOFGRAVITYlNCRATERMODIFICATION. VirgilL. exhibit an indirect expression of this proftle with more extensive Sharpton, Lunar and Planetary Instiblte, 3600 Bay Area Boulevard, rirnlwall collapse uprange inresponse to the oversteeped innerwall Houston TX 77058, USA. and an uprange offsetofthe centralpeak complex in response to the deepest point ofpenetration (e.g., Tycho and King Craters). Two­ Background: Previous to Magellan, a convincing case had ringed basins with other indications of an oblique trajectOJy also beenassembledthatptedicted thatcompleximpactcratersonVenus exhibit an uprange offset of the inner ring and enhanced uprange were considerably shallower that their COWlterparts on Mars, Mer­ collapse (e.g., Moscoviense on the Moon and Bach on Mercury). cury, the Moon, and perhaps even Earth. 'I1U$ was fueled primarily Second, interior pits created by hypervelocity impacts into by the morphometric observation that, for a given diameter (0), strength-controlled targets become elongate and breached down­ craterdepth (d) seems toscale inversely with surfacegravity for the range as impactangledecreases from 45° to 15°.Thediameterofthe other planets in the inner solar system [e.g., 6]. Thus Venus, which interior pit perpendicular to the trajectOJy is fOWld to depend is similar to Earth in its size and density, shouldyield a very low d­ simply on the impactordiameter (2r), targetfunpactor density ratio D trend, like that reconstructed from the terrestrial impact record (lV3,), vertical component ofimpactor velocity (vsin9), and target (Fig. 1). In addition, deceleration of ejecta through the dense soW1d speed (c) venusian atmosphere [8] and viscous relaxation of crater t0po­ graphy [11] due to Venus' highsurface temperatures wereexpected to contribute to craters perhaps even shallower that those on Earth. Indeed, altimetric data from Pioneer Venus [6] and the Venera Thisrelationcanbedirectlyderivedfrornthepressure-decaylaw orbiters [3] indicated low relief for presumed impact craters on given in [4] where Xo corresponds to a characteristic strength limit Venus, although the large footprint of radar altimeters grossly in the targeL Wldersampled the truetopography for all impact features measured, Planet-scale craters also exhibit elongated and breached central and manyofthe largecircularstructures thought to be impact basins peaks (e.g., King on the Moon) and peak rings (Bach on Mercury) were not. Even the enhanced resolution provided by the Magellan as impact angle decreases [5,6]. If the central structures are radar altimeter probably W1dersamples the relief of all but a few of large-scale analogs for the zones of maximum penetration in the largest impact basins on Venus [4]. laboratory-scale hypervelocity impacts, then the diameter of the Depth Measurements: Useful crater topography can be ex­ gravity-controlledcraterdiameterrelativeto thestrength-controlled tracted fromMagellanradarimages using the distortionsinthe cross central structure should decrease with decreasing impact angle, as track direction imparted by the interaction ofradar incidence angle observedincraters on Venus (Fig. 1). Invery-large-scalebasins that (9) with surface slope. In the general case, two images ofthe same WldergO furtherenlargementbyrim collapseorlithospheric failure, feature, taken at different 9, are required, but crater depths can be thedistinctiveimpactorsignature (e.g.,breacheddownrangecentral estimated by assuming the crater is symmetrical in the cross-track ring) may persist even though its diameter relative to the diameter direction [4]. Using this technique, [9] havecalculatedcraterdepths ofthe outerbasin scarpnolonger follows the same trend as smaller for73craters outofasetofl02large(18 km > D> 175km)complex two-ringed basins (e.g., Orientale on the Moon). craters identifiedinMagellandatathen availableto them, indicating Ifthe central relief in impact structures simply corresponds to a a power law fit d = 0.28 I)OM. The large dispersion in this dataset ZDIlC ofmaximwn compression during initial stagesofpenetration, (R2 = 0.21) probably reflects several effects including (l) errors then it may provide a measure of impactor size for a given impact associated with locating the rim crest and the outer boundary ofthe velocity [6]. This hypothesis can be tested by referencing crater floor; (2) errors associated with asymmetries in the actual crater diameter to the central-ring diameter (in lieu ofimpactor size) and topography; (3) inclusion ofrelatively modified craters, as well u plotting this value againstthe dimensionless gravity-scaling param­ fresh craters; and (4) true variations in crater morphometry. None­ eter gr/V2 where r again is replaced by the central ring dimension. theless, it is clear that venusian craters fall considerably above the Theresulting power-law exponent(Fig. 2a)exhibitsnearly thesame terrestrial doD trend, with the freshest craters (distinguished by value for different planets and can bebrought into line by adopting parabolic deposits of distal ejecta [1]) virtually indistinguishable reasonable average impact velocities for eachplanet. Central peaks from the martian fresh crater trend (Fig. 1). maybesimilarmanifestationsoftheimpactorcompressionzonebut it is uplifted during decompression due to higher peak shock It ,------, pressures and smaller size. Figure 2b tests this hypothesis and reveals a very similar dependence. In swnmary, oblique impacts allow identifying distinctive sig­ natures of the impactor created during early penetration. Such signatures further may allow fust-order testing ofscaling relations for late crater excavation from the planetary surface record. References: .[1] Schmidt R. M. and Housen K. R. (1987) ItIl. 1.I"'I'QCt E1I8118., 5. 543-560. [2] GaultD. E. etal. (1968) In Shock t.5 MetamorphismofNaluralMalerials(B. M. French andN. M. Short, U cds.),87-100,Mono, Baltimore. [3] GaultD.E. andWedekindl.A. t.2 FIene 1: 18 Fresh Venusian Craters (1978) Proc. LPSC 9th. 3843-3875. [4] Holsapple K. A. (1980) with Dart Parabolic Ejecla Proc. LPSC Hth. 2379-2401. [5] Schultz P. H. and Gault D. E. It 1t St I" 2" (1991) Meteoritics, 26. No.4. [6] Schultz P. H. (l992)LPSCXXIII. Diameter (Ian) 1231"-1232. [7] Schultz P. H. (1988) In Mercury (F. Vilas et al.• cds.), 274-335, Univ. of Arizona, Tucson. Fig. 1. Eighteen fresh venusiancraters withdarkparabolicejecta. 66 1992 Sudbury Conferent:e

GuIlber1 Craw Budewkll Crater image shows a narrow ridge, coinciding with topographic profile, u-

·1 $~•• ' •• I ' •• , •• ,I' •• ,•,•, ••• I .'S-.•.. " ... I'" I"""'".,. could contribute todeep craters onVenus include (1) more efficient o N ~ ~ • 100 ,» o ~ ~ ~ • 100 1~ OJ Ckml cts__ (km, excavation on Venus, possibly reflecting rheological effects of the hot venusian environment, (2) more melting and efficient removal ofmeltfrom the cratercavity, and (3)enhanced ejection ofmaterial Corpmoner..... Flaa.der..... ,,- outofthecrater,possibly as aresultofentrainmentin anatmosphere set in motion by the passage of the projectile. l":~ The broaderissue raised by the venusiancraterdepthsis whether .- surface gravity is the predominantinfluenceoncraterdepths onany I -0. ~ -,- planet. There is an apparent d-g-I trend in data from the Moon and .1'-. . .. , ... , . -'6-. . . Mercury, but these planets are all relatively small and, to the first 20 (0 to to ,.. . 20 (0 co eo tOO 110 ..-.... "...... order, ofthe same size. The surfaceproperties and targetclulrKter­ istics ofMars areconsiderably different from those ofthe Moon and Fig. 2. Crater profiles constructed from Magellan image pairs. Mercury and could contribute to its somewhat lower doD relation­ ship. Although shallow depths, in accordance with got scaling, are reported for terrestrial craters [6,2], there are no fresh complex Crater Promes and Morphometry: The four crater profiles craters on Earth from which to directly take these measurements.

(-west-east, tha-ough the crater center) shown in Fig. 2 were pro­ The Venus data in Fig. 2 indicate that Hr is a significant portion of duced using a nonstereo technique on Magellan cycle l-eycle 2 d, and as all terrestrial complex structures are severely modified by imagepairs.GuilbertandBudevskaarecentralpeakcraters;Corpman erosion, d estimates may be up to a factor of 2 too low due to rim and Flagstad are peak ring structures; Budevska is a fresh, bright­ removal alone. In addition, when the crater rim and ejecta blanket floored crater; the rest contain dark-floor deposits that may be have been removed, it becomes more difficult to determine the rim indicative of volcanic or aeolian infilling. Corpman contains the crest diameter. Reconstructions based on structural analysis may most areally extensive dark floor deposits. The depth estimates of overestimate the true diameter iffaulting extends beyond the rim IS these craters support the single-image estimates presented above. shown in Fig. 3. Thus while inverse gravity scaling ofcraterdepths Furthermore,becausethis technique doesnothingeupon symmetri­ has been a useful paradigm inplanetary cratering, the venusian data cal topography, morphometric informationcanbeextended beyond do not support this model and the terrestrial data are equivocal at simple depth constraints. best. Thehypothesis that planetary gravity is the primary influence

The rim height (Hr) of these craters constitutes ...0.3-0.5 d and over crater depths and the paradigm that terrestrial craters are there are slightvariations «0.3 Hr) in the eastern andwestern Hr for shallow should be reevaluated. allcratersmeasured.Crater flapJe~; m!Ultledbybright, blockyejecta, References: [1] Campbell D. B.etal.(1992)JGR. inpress. [2] are typically narrow, ranging from ...0.2-0.5 D. Assuming that Grieve R. etal. (1981) Multi-Ring Basins,Proc.LPS J2A (P.Schultz ejecta constitutes...o.5 Hr [5], the continuous ejectablanket around and R. Merrill, eds.), 37, Pergamon, New York. [3] Ivanov M. N. et BudevskaCratercontains-400km]ofejecta,equivalentto -300 km] al. (1986) JGR, 91. 413. [4] Leberl F. et al. (1991) Photogram. of unfragmented target rock. Engrg. Remote Sens., 57. 1561. [5] Melosh H. 1. (1990) Impact Central peak heights vary considerably from ...0.1 to 1.0 d. The Cralering: A Geologu: Process, Oxford, 245 pp. [6] Pike R.1. lower extent of this range may be due in part to subsequent (1980) Jcarus, 43.1. [7] Schultz P. H. (1972) Moon Morplwlogy, modification of Corpman by extensive dark floor deposits. The 626, Univ. ofTexas. [8] Settle M.(1980) Icarus. 42. 1. [9] Sharpton centralpeakofGuibertprotrudesvirtuallytothelevel oftherim,and V. L. and Edmunds M. S. (1992) GRL, in review. [10] Sharpton V. Budevska's central peak is only slightly shorter. Similar craters L. and Gibson 1. W. (1990)LPSC XIX, 1136. [11] Solomon S. C. et have beennoted onotherplanets, e.g, the lunar farside crater, Icarus al. (1982) JGR, 87. 7763. [7], and the terrestrial Marquez Dome crater [10]. Such craters, however, are relatively rare on other planets; having two such examples within a sampleoffour craters (Fig. 2) suggests that these anomalously tall central peaks might be more common on Venus. KIT BOUNDARY STRATIGRAPHY: EVIDENCE FOR Crater floors are flat in all cases except Budevska, where the radar MULTIPLEIMPACTSAND A POSSmLECOMETSTREAM. E. M. Shoemakert and G. A. Izettl, IU.S. Geological Survey, Flagstaff AZ 86001, USA, lU.S. Geological Survey, Denver CO rl------0.5 q,ceonstructed 80225, USA. Structural Rim Crest Rim 1------0.5 Df~ A critical setofobservationsbearing ontheKII' boundaryevents I has been obtained from several do~ sites in _western North Surface AII... ECosion--- America. Thin strata at and adjacent to the KII'boundary are locally preservedin association withcoalbeds at thesesites. The stratawere laid down in local shallow basins that were either intermittently Fig. 3. Structure Cross section of complex crater half-space. flooded or occupied by very shallow ponds. Detailed examination Crater center is on the right. by [1] of the stratigraphy at numerous sites led to their recognition U'lContribulioft No. 790 67

oftwo distinct strata at the boundary [1-4]. From the time that the reworking. The interlaminated vitrinite shows that the upper stra­ two strata were fIrSt recognized, E. M. Shoemaker has maintained tum had a protracted history of deposition that produced the that they record two impact events. We repQI1 here some of the alternate laminae ofvitrinite and clay. evidence that supports this conclusion. The mostdiagnostic feature ofthe upper stratum is the presence The lower stratum, fIrSt recognized at localities in the Raton ofquartz grains and quartzose lithic fragments, about 30% ofwhich Basin ofNew Mexico and Colorado by [4,5], is referred to as the exhibit shock lamellae [3]. About halfthelithicfragments are chert Kffboundary claystone. Microscopic study ofpolished slabs ofthe and chalcedony and the other halfare quartzite and metaquartzite. boundary claystone, which consists typically of 1-2 em of white, Rare shocked grains of oligoclase and microcline and granitoid gray, or tan kaolinite, reveals that itis speckled with small angular lithic fragments are also present. The shocked grains tend to be to rounded clasts and pellets of white to gray kalolinite and concentrated near the base or in the lowest depositional unit ofthe subparallel flakes ofvitrinite. The claystone commonly appears to upper stratum. No spherules are found in the upper stratum. be a single massive bed,butis found in some places to be acomplex Ofparticular interest for thepresent discussion is the contact of unit when examined in detail. Generally it rests on dark, thinly the upper stratum on the boundary claystone. As noted by [16] and laminated carbonaceous claystone or coal; the lower contact is [3], this contact at some sites is a paleosurface that shows evidence gradational at most places but sharp at some. The lower part ofthe of weathering and reworking or remobilization of the uppermost claystone bed generally is dark colored, evidently owing to rework­ partofthe boundary claystone. The uppermost 1 to few millimeters ing ofdark carbonaceous material from the underlying bed. Up to ofthe boundary claystone generally consists ofirregular claystone four discrete depositional units, most of them bounded by sharp clasts, mostly less than 1 mm across, embedded in a vitrinite­ contacts, an: locally present within the claystone. In an example enrichedmatrix. Following [3], we informallyreferto this reworked illustrated by [3, Fig. 5], angularkaolinite clasts in one depositional zone as the billowy layer. Shocked grainsofquartzderived from the unit an: truncated bya sharpcontactat thebase ofthe nextoverlying upper stratum locally OCCW' in the billowy layer. unit. The occurrence ofthe kaolinite clasts, most ofwhich contain Root casts OCCW' in the boundary claystone at localities in relictvegetalremains in the form ofvitriniteflakes, and the multiple Montana [16] and Colorado, notably at the Clear Creek North site depositional units that rest on sharp truncation surfaces show near Trinidad. Most ofthe recognized root casts are confIDed to the unequivocally that the boundary claystone has not been formed by boundary claystone stratum. Where the root casts can be traced in a single eventofdeposition orfallout ofimpact ejecta. Many clasts polished slabs, they open at the top of the boundary claystone and are ripup clasts of previously deposited claystone [3], and the arefilled by the billowy layer. Ashallow dimplegenerally ispresent presence ofmultiple depositional units indicates multiple episodes at the top of the billowy layer over each root cast. Abundant flakes ofreworking. ofvitrinite OCCW' in the root casts, and shocked quartz canbe found Adistinctive feature ofthe boundaryclaystoneis the presenceof in the billowy mling, commonly along the upper walls of the root very-smooth-surfaced (in somecases shiny) spherules that typically casts. The base ofthe upper stratum is marked in places by a fairly rangefromaboutO.l to 1mm indiameter[3,6-8].Inthe RatonBasin pronounced lamina ofvitrinite that rests on the billowy layer. The the spherules an: composed of gorceixite or kaolinite. At some upper stratum overlies the billowy mling of the root casts and localities in Wyoming the walls of the spherules are composed of therefore postdates them. The stratigraphic evidence reveals quite gon:eixite oranotherphosphate mineral, and the interiors are filled unambiguously thatplants took root in the boundary claystoneprior with kaolinite, gypsum, or, rarely, sulfides. Many spherules are to the deposition of the upper stratum. hollow. Smooth-surfaced forms in the shape ofteardrops, spindles, The strata at the Kff boundary in western North America thus and dumbbells are also present [3,9,10]. These forms are nearly record at least two impact events separated by a time interval long identical to but generally somewhat smaller than spherules, tear­ enough for small plants to grow on the Kff boundary claystone. drops, spindles, anddumbbells found in the Kffboundary claystone Neither the boundary claystone nor the upper stratum, however, in Haiti [9], where remnants of tektite glass are preserved in the were formed simply by airfall of impact ejecta, as each stratum interiors of the larger forms [9,11-14]. It is now clear that the locally consists ofmultiple depositional units and contains clasts of gorceixite spherules in the boundary claystone in western North previously deposited material. The upper stratum, in particular, America arepseudomorphs afterglassy objects. The varietyofthese contains heavy minerals of local provenance [3]; the abundant pseudomorphus fonns is typical ofthose produced by disruption of vitrinite almost certainly represents locally derived plant material. a liquid. Traces of the internal flow bands in the original liquid The clay mineralogy ofthe upper stratum may also be indicative of droplets an: preserved on the surfaces ofsome ofthe pseudomorphs mixing ofmaterials from diverse sources. [9,15]. Sphemlesan: fairly abundantatsomelocalities inWyoming, We interpret the boundary claystone in western North America constituting up to several percentofthe boundary claystone. Where as derived partly and perhaps chiefly from impact ejecta from the they are composed entirelyofkaolinite and embedded ina kaolinite Chicxulub Sb'Ucture, Yucatan [17.9]. Following the suggestion of matrix, they an: often difficult to detect in hand specimen but easily [3], we consider the Manson impactstructure in Iowa[18] the likely detected in thin sections. source of most or all the shocked grains in the upper stratum. The upper stratum of the Kff boundary in western North Shocked quartz grains and quartzose lithic fragments are coarser America, referred to by [3] as the K/I' boundary impact layer, and one to two orders of magnitude more abundant at the sites in generally consistsofafew millimetersofthinly laminatedclaystone western North America than they are atothersites around the world, of mixed clay mineralogy and abundant flakes and laminae of with the possible exception ofsites in Haiti. This global pattern and vitrinite. Nearly everywhere it contains numerous ovoid pellets of the continental afftnity of the grains led to a search for a possible claystone about 0.1 to 1 mm across, commonly referred to as sourcecraterinNorth America and the identification ofthe Manson graupen; in places it contains much largerrounded claystone clasts. stNctureas acandidate source. 4OArl"Ar measurements onshocked Like the underlying boundary claystone, the upper stratum consists microcline from the central uplift ofManson show that the Manson in sOme places of multiple depositional units bounded by sharp structure is synchronous with the Kff boundary within the ±t Ma contacts. Hence the upper stratum also shows clear evidence of precision of the age detennination [19]. Not only is the Manson 68 1992 Sudbury Conference

structure very close to the right age, but the rocks excavated at the centrated along its base; locally, however, they are restricted to crater appear tobe a likely source for the shocked grains u well u centimeter-thick lenses ofpure spherules along a single horizon in most other grains lacking observable shock lamellae in the upper the argillite close beneath the turbidite. The total volume of pre­ stratum. served spherules is estimated at 8 x 1(JI m'. Assuming an original Theoccurrenceoftwo impacts separated in time by atleutpart specificgravityof2.5,typicalofsilicateglus,thelowerlimitofthe ofagrowingseuonappearstobemostreadilyexplainedifthe Earth total original mus ofspherules is about 1.7 x 1014 g. intercepted a compact comet stream at the end of the Cretaceous InthenortheuternpartoftheHamersleyBuin,simiiarspheruies [20]. In repetitive puses through the stream, the Earth may have againoccuratonlyonehorizon, butheretheyarea minorconstituent encounteredmorethantwocrater-formingprojectiles andmayhave ofa dolomitic debris-flow deposit 9.9 to22.7 m thickknown u the swept up substantial amounts of cometary material that did not dolomixtite layer. The dolomixtite layer occurs in the Carawine produce craters. Thepeak Ir abundance, which occurs in the upper Dolomite, which is stratigraphically equivalent to the Wittenoom stratum, may reflect a somewhat protracted accumulation of co­ Formation[3]. Moreover,paleocurrentdata fromcloselyusociated metary material. Iridiwn is relatively low in abundance in the carbonate [4,5] and volcaniclutic [6,7] turbidites indicate the boWldary claystone, possibly u a consequence of blowoff and spherule marker bed wu deposited in deeper-water paleoenvi­ escape of the vaporized projectile that formed the great Chicxulub ronments than the dolomixtite layer. 'Therefore, the dolomixtite impact structure [cf. 21]. layeris believed tobea proximal equivalentofthe spherulemarkel' A comet stream is most likely to have been formed by breakup bed. In addition to spherules, the dolomixtite layer contains par­ of a large SWl-grazing comet. Reexamination ofthe flux of active ticles that also consist of K-feldspar, but are larger than the andextinctEarth-crossingcometssuggeststhatcollisionofperiodic spherules (up to 11 mm across) and have much more internal comets 8CCOWlts for about 25% of the terrestrial impact craters heterogeneity. Some display internal flow banding or schlieren, larger than 20 km in diameter. Periodic cornets initially on orbits while others contain typical spherules u inclusions. These larger withinclinations near90°becomeSWl grazers [22]. More thanone­ particles arein thesizerangeoftruetektites,but ablated forms have ruth of the Earth-crossing periodic comets probably become Sun yet to be observed. grazers that are subject to tidal disruption. Buedon their similarity to microtektites and microlaystites in References: (1] Izett G. A. and Pillmore C. L. (1985) GSA shape, size, andinternaltextures, and theiroccuncnce u a very thin Abstr. withProgr., 17, 617. [2] IzettG. A. (1987) U.s. Geol. SIU'V. layer over a large area. the spherules are interpreted u droplets of Open-FikRept.87-606, 125. [3] IzettG.A. (1990)GSA Spec.Pap. silicatemeltthatweregeneratedanddispersed lICrOSS theHamcrsley 249,100. [4] PillmoreC. L. etal. (1984)Science.233,118~1183. Buinbya majorbolideimpacLThemusofthespherulespreserved [5] Pillmore C. L. and Flores R. M. (1987) GSA Spec. Pap. 209. in the Hamersley Buin is of the same order of magnitude u the 111-130. [6] Pollutro R. M. and Pillmore C. L. (1987) J. Sed. estimated muses of microtektite glus in major Cenozoic strewn Petrol.• 57. 456-466. [7] Bobot B. F. and Triplehorn D. M. (1987) fields, despitethe fact thatthespherulescurrently coveranareathat IPSCXVlll,103-104. [8] BohorB. F. (1990) GSA Spec. Pap. 247, is 2 to 3 orders of magnitude smaller. The layers that host the 335-342. [9] bett G. A. (1991) JGR, 96. 20879-20905. [10] Smit spherules are interpreted to be the deposits of a major sediment 1. (1990)Geol. Mijnbouw. 69. 187-204.[11]Izett G. A.etal.(1990) gravity flow that exhwned and redeposited most of the spherules u.s. Geol.SIU'V. Open File Rept. 90-635. 31. [12] lzettG. A.(l991) after shallow burial. although the flow is not believed to have been IPSCXXI/,625-626. [13] SigurdssonH. etal. (1991) Nature. 349. a direct result of the proposed impact. The internal textures of the 482-487. [14] Sigurdsson H. et al. (1991) Nature. 353. 839-842. spherules suggests the target rocks were mafic in composition, but [15] Izett G. A. (1991) Eos. 72,278. [16] Futovsky D. E. et al. the presence of trace amoWlts ofmictoeline and quartz crystals in (1989) J. Sed. Petrol.• 59. 758-767. [17] Hildebrand A. R. et al. both the spherule marker bed and dolomixtite layer suggests some (1991) Geology. 19. 867-871. [18] HartWlg J. B. etal. (1990) GSA continental basement rocks were also present in the target area. Spec. Pap. 247. 207-221. [19] KWlkM.J.etal.(1989)Science. 244. Given this, plus the fact that the spherules and related particles are 1565-1568. [20] ShoemakerE. M. (1991)Abstroctsforthe /ntema­ largest in the northeutern comerofthe Hamersley Buin, the most tiofIQl Conference ora Asteroids. Comets. Meteors, 199. [21] Vic­ likely site for the proposed impact would have been iR the early kery A. M. andMelosh H. J. (1990) GSA Spec. Pap. 247,289-300. Precambrian ocean close to the northeutern edge of the Pilbara [22] Bailey M. E. et al. (1992) Astrora. Astrophys., in press. Craton. Anotherthin horizon in theoverlying BrockmanIron Formation contains spherules thatagain consist largelyofK-feldsparandhave internal textures strikingly similar to those of the Wittenoom GEOLOGICAL EVIDENCE FOR A 2.6-Ga STREWN FIELD Formation and Carawine Dolomite. They differ in being slightly OF IMPACT SPHERULES IN THE HAMERSLEY BASIN larger on average (up to 1.8 mm) and extensively replaced by iron­ OF WESTERN AUSTRALIA. Bruce M. Simonson, Geology richminerals (particularly stilpnomelane)u they arehostedby iron Department, Oberlin College, Oberlin OH 44074, USA. formation rather than argillite. This horizon is about 250 m higher stratigraphically and persists laterally for atleut 30km. The close Sand-sized spherules up to 1.7 mm across with spherulitic, resemblance of these spherules to those ofthe Wittenoom Forma­ vesicular, and other crystalline textures that consist mainly of K­ tion andCarawine Dolomitesuggests they alsooriginated u impact feldspar help derme a Wlique horizon in the well-preserved 2.6-Ga melt droplets, even though they are admixed with volcaniclutic Wittenoom Formationin theHamersley BuinofWesternAustralia detritus. Using the sedimentation rate of 3-4 m/m.y. typical of the [1,2]. This layer, informally known u the spherule marker bed, is Hamersley Group [8], the stratigraphic separation between the two nowhere thicker than about 1.3 m, yet it persists for more than suggests that a second major impact occurred near the Hamersley 300 km across the buin, blanketing an area ofat leut 13,700 km1. Buin after a time interval of about 75 m.y. elapsed. This suggests Sedimentological evidence indicates the layer is a single turbidite, the record of impacts in early Precambrian strata is richer than is and the spherules are a minor constituent that are usually con- generally appreciated. LPI COItIribution No. 790 69

References: (1) Trendall A. F. (1983) In Iron-Formations: low and indicate that at leastthe friction meltderived from the more FactsandProbkms(A.F. TrendallandR.C.Morris,eds.),69-129, basic protolith would have been highly fluid within the slip zone Elsevier, Amsterdam. (2) Simonson B. M. (1992) GSA Bull., 104, during displacement. The effects of shear thinning at very high in press. (3) Goode A. D. T. (1981) InTM Precambrion Geology of strain rates would reduce these viscosities even further. the Southern Hemisphere (0. R. Hooter, ed.), 105-203, Elsevier, Where friction melts are generated during coseismic slip (to Amsterdam. (4) Simonson B. M. and Goode A. D. T. (1989) form pseudotaehylitcs) this implies that the melt may help to Geology,l7, 269-272. (5) Simonson B. M. et al. (1992) Precam­ lubricate the sliding interface and dissipate stored strain energy. brian Res., in press. (6) Husler S. W. (1991) Ph.D. dissertation, These results are contrary to the views of earlier workers, who University ofCalifornia, Santa Barbara. (7) Hassler S. W. (1992) suggested that any melts generated by frictional heating would Precambrion Res., inpress. (8) Arndt N. T. etal. (1991) Australian possesshighviscosities and soinhibitslip. Partofthis inferencewas J. Earth Sci., 38, 261-281. based on the erroneous uswnption that pseudotachylite generation involved the bulk fusion ofwallrocks. Although a pseudotaehylite matrix plus clasts has a very similar chemislry to the wallrock lithology, the matrix typically possesses amore basicchemislry and SAR IN SUPPORT OF GEOLOGICAL INVESTIGATIONS hence, due to its lower Si02content, a significantly lower viscosity OF THE SUDBURY STRUCTURE, V. Singhroyl, R. than that ofits protolith. On the otherhand, smallerentrained clasts Mussakowski2, B. O. Dressler', N. F. Trowell', and Richard (<1 em diameter) are typically felsic and dominated by quartz. Grievc4, ICanadaCcntrefocRemote Sensing DepartmentofEnergy, These results have implications for the generation of Mines and Resources of Energy, Mines and Resources, Canada, pseudotaehylitic breccias as seen in the basementlithologies ofthe 2Provincial Remote Sensing Office, Ontario Ministry of Natural Sudbury and Vredefort structures and possibly certain dimict looar Resources, Canada,'OntarioGcological SurveyMinistryofNorthem breccias. Many of these breccias show similarities with the more Development and Mines, Canada, 4Geological Survey ofCanada, commonlydeveloped pseudotachylite fault andinjectionveins seen Department of Energy, Mines and Resources, Canada. in endogenic fault zones that typically occurin thicknesses ofa few centimeters or less. The main difference is one of scale: Impact­ Imagingradaris an importantcontributingsourceofinformation induced pseudotaehylitc breccias can attain several meters in thick­ for arangeofgeologicalproblems andenvironments. Airborne SAR ness. This wouldsuggestthatthey weregeneratedooderexceptionally and ERS-l data integrated with other geoscience datasets are being high slip rates and hence high strain rates and that the friction melts used in an attempt to characterize the crustal fracturing associated generated possessed extremely low viscosities. with the Sudbury structure. This presentation highlights examples Reference: (1) Shaw H. R. (1972) Am. J. Sci., 272. 870-893. ofintegrated and composite images aimed at facilitating the inter­ pretation of the Sudbury structure. This work is the result of an ongoing cooperative multidisciplinary SAR study of the basin carried out by the Canada Centre for Remote Sensing, Ontario's THE LARGE IMPACT PROCESS INFERRED FROM THE Provincial Remote Sensing Office, the Ontario Geological Survey, GEOLOGY OF LUNAR MULTIRING BASINS. Paul D. and the Geological Survey ofCanada. Spudis, Looar and Planetary Institute, Houston TX 77058, USA.

The nature ofthe impact process has been inferred through the study of the geology of a wide variety of impact crater types and VISCOSITY DETERMINATIONS OF SOME FRICTIONAL­ sizes. Some of the largest craters known are the multiring basins LY GENERATED SILICATE MELTS: IMPLICATIONS fOood in ancient terrains ofthe terrestrial planets [e.g., 1). Ofthese FORSLIPZONERHEOLOGYDURlNG~ACT-INDUCED features, those fOood on the Moon possess the most extensive and FAULTING. lohnG. Spray, Department ofGeology, University diverse data coverage, including morphological, geochemical, geo­ ofNew Bnmswick, Fredericton, New BrWlSwick, Canada. physical, and sample data. The study ofthe geology oflooar basins over the past 10 years [2-4) has given us arudimentary ooderstand­ Analytical scanning electron microscopy, using combined en­ ing ofhow these large structures have formed and evolved. ergy dispersive and wavelength dispersive spectrometry, has been Basin Morphology: Basins on the Moon begin to form at used to determine the major-clement compositions ofsome natural diameters ofabout300km, the 320-km-diameterSchrodingerbeing and artificial silicate glasses and their crystalline equivalents de­ an example [5,6). At these diameter ranges, only two distinct rings rivedby thefrictional meltingofacid to intermediateprotoliths. The are apparent; the transition diameter atwhich multiple rings appear major-element compositions are used to calculate the viscosities of is uncertain, but appears to be between400and 500km in diameter their melt precursors using the model ofShaw [1) at temperatures (6). Above these diameters, basins possess multiple rings, as few as o 2 of 800 -1400°C, with Fe +/Fe(tot) =0.5 and for 1-3 wt% H20. three and as many as seven [1.5,6). Inevery basin, one ring appears These results are then modified to accooot for suspension effects to be very prominentandis believed tocorrespond structurally to the (i.e., the presence ofmineral and rock clasts) in order to determine topographic rim of complex craters. This ring has various names effective viscosities. (basin rim of (5), Ring IV of (6), MOR of (7», but corresponds to The critical factors in controlling the viscosities of the silicate the Cordillera ring ofthe Orientale Basin. Rings inside and outside melts are Si02and H20 contents and temperature, as has been well this ring are recognized, each having distinct morphology. Basin established for silicate melts ofmagmatic origin. Additionally, for inner rings tend to be clusters or aligned segments of massifs, fault-generated melts, the effects ofshear thinning can reduce the arranged into a crudely concentric pattern; scarplike elements may viscosity to a significant degree. At 1200oC, the viscosities range ormay notbe presenL Basin outerrings tendto bemuch more scarp­ froni 7 p for the more basic melt sample (40 wt% Si02) to 1 x 1()5 P like and massifs are rare to absent. Within a certain subset ofbasins for the more acid melt sample (64 wt% SiOJ.These viscosities are on the Moon (e.g., Crisiwn (8), Humorwn (9», the main topo- .70 1992 Sudbury Conference

graphic rim is not evident. These basins appear to have undergone [21J;thus.themaximumdepthofexcavationisontheorderofabout a different style of postimpact modificatian. possibly related to 50 km, suggesting an effective depth ofexcavation of about 0.1 ± rapidly changing thermal conditions within the Moon 3.9 Ga ago 0.02 times the diameter of the excavation cavity [2-4J. Data from [8.9J. well-studied complex craters on the Earth suggest that the excava­ Basin Ejecta: Basins display texturedejectadeposits.extend­ tioncavity ofcomplex craters is on the orderof0.5 to 0.65 times the ing roughly an apparent craterradius beyond the main topographic diameter of the apparent crater [l9J; the maximum depths of rim. Ejectamaydisplay variousmorphologies.ranging from wormy excavation are ontheorder of0.09-0.12 times theexcavationcavity to hummocky deposits (e.g.• Hevelius Fonnation ofOrientale. Fra diameter [l9J. These numbers compare favorably with the admit­ Mauro FormationofImbrium [5»to knobby texturedmaterial (e.g•• tedly poorly resolvedlWlarvalues [I-3J,a conclusion substantiated Alpes Formation ofImbrium (5». The cause ofthese variations in by certain analytical methods [22J. The relatively shallow effective ejecta morphology are notknown. At Orientale. knobby material is depths ofexcavationpredicted by these various models account for largely confmed within the Cordillera scarp while hummocky the relative paucity ofvery deep ClUStal ormantle materials within materials appear to be mostly restricted beyond this boundary the returned Apollo lunar samples [5,23J. [5.10J. However. at Imbrium, both units are restricted beyond the Basin Ring Formation: A wide variety of mechanisms has topographic rim (Apennine ring) and display a curious "bilateral" been proposed to account for the formation of basin rings (see double symmetry [1 ,3J; this relation remains unexplained. Outside review in [105.7». Inmy opinion,noneofthem areentirelyplausible the limits of the basin textured ejecta are found both fields of and the formation ofrings constitutes the last greatunsolved puzzle satellitic craters (secondaries [5,11» and light plains deposits of multiring basin formation. Ring-forming mechanisms can be (Cayley Formation [5».Thesematerials containbothprimary basin divided into two broad groups (see [1,6]): (I) forcible uplift due to ejecta and local materials, the local materials being predominant fluidizationofthe target [5,24J; (2)concentric,brittle fracturing and [l2J. failure ofthe target on regional (megatemces [25]) to global scales Impact melt sheets are observed on the floors of relatively (lithospheric fracturing [27]). Geological evidence supports por­ Wlfloodedbasins,such as Orientale[l,s,10J. Aclassofimpactmelts tions of all of these models, but none of them completely or in the Apollo sample collections possess basaltic major-element unequivocally. One constraint that has emerged from the examina­ chemislry. have a KREEP trace-element pattern ofvarying concen­ tion of a variety ofbasins on a number ofterrestrial planets is that lration. and all have ages ofabout 3.8-3.9 Ga [I.13J. These rocks. basin rings are spaced at a constant factor, namely the famous ..J2 collectively called"low-K Fra Mauro" basalts. are probably related relation observed between adjacent rings [I,6,27J. Originally pr0­ to basin impact melts [l3-15J. Although the exact number is posed only for the Orientale Basin [27.28J, it has been fmmd to be contentious. at least three major compositional subdivisions of the valid for all of the terreslrial planets and some icy satellites [I,6J. LKFM melt group can be recognized; each may correspond to a Becausegeologicalevidencesupportsdivergentring-forming mod­ different multiring basin, the Imbrium [l3,15J, Serentitatis [l5J. els, itmay be that the ring-locating mechanism is different from the and Nectaris Basins [l6J. Acurious fact about lunar LKFM melts is ring-fanning mechanism [6J. Thus, large-scale crustal fmmdering that they cannot be produced through the fusion of known lunar (megaterraeing) could occur along concentric zones of weakness pristine rock types [l3,14J, suggesting the occurrence ofWlknown createdby some lypeofresonantwavemechanism (fluidization and crustal lithologies on the Moon. The LKFM melts were probably uplift); such immediate cmstal adjustment could then be followed generated at middle to lower crustal levels [l3.15J. by long-term adjustment due to lithospheric fracturing. If the Basin Excavation: The preservation of preexisting topogra­ conundrum ofring genesis can be resolved, we will possess a good phy within the main topographic rim provides some constraints on understanding ofall of the principal phases offonnation ofmulti­ the size of the excavation cavity ofmultiring basins. At Orientale. ring basins. pre-existing craters and basins can be mapped [2.17,18J within the References: [IJ Spudis P. D.• in press, The Geology ofMu/li­ Cordillera scarp (950 km diameter) and some slrUCtures [l7J may ring Impoct Basins. Cambridge Univ. [2J Spudis P. D. et al. (1984) extend inside the outer Rook ring (620 kmdiameter). These obser­ Proc.LPSC 15th. inJGR.89.CI97. {3J Spudis P. D. etal. (1988) vations suggest that the excavation cavity for Orientale must have Proc. IPSC 18th. ISS. [4J Spudis P. D. et al. (1989) Proc. IPSC been less than about 600 km in diameter [2J. The minimum size is 19th. 51. [5J Wilhelms D. E. (1987) u.s. Geol. Surv. Prof. Pop. difficult to constrain; the innermost ring (400 km diameter) may 1348. [6J Pike R.I. andSpudis P. D. (I987)EarthMoonPlanets. 39. provide a lower limit to cavity size [2J. These constraints observed 129. [7J Crofts. K. (1981)InMu/li-RingBasins.Proc.LPS12A (P. at the Orientale Basin are paralleled by similarrelations ofprebasin Schultz and R. B. Merrill, cds.). 207, 227. Pergamon, New York. topographypreservedwithin theImbriumBasin(1160-km mainrim [8J Spudis P. D. et aI. (I989)IPSCXX, 1042. [9J Spudis P. D. et al. diameter). where the prominence ofthe Apennine Bench indicates (1992) IPSC XXIII, 1345. [lOJ McCauley J. F. (1977) PEP1. 15. that the excavation cavity for the Imbrium Basin must be less than 220. [l1J Wilhelms D. E. (1976) Proc. LSC 7th, 2883. [l2J Ober­ about 800 km in diameter [3J. These data indicate that the excava­ beck V. R. (1975) Rev. Geophys. SpacePhys.• 13,337 [l3J Spudis tioncavity ofmultiringbasins is between about0.4and0.6 times the P. D. et al. (199I)Proc. IPS. Vol. 21. lSI. [l4J McCormick K. et diameter ofthe apparent crater diameter [2-4.17J. aI. (1989) Proc. IPSC 19th. 691. [l5J Ryder G. and Wood J. A. Basin depths ofexcavationcanbeinferredfrom the composition (1977) Proc. LSC 81h. 655. (16) Spudis P. D. (1984) Proc. IPSC of basin ejecta. At Orientale, basin ejecta are very feldspathic, 151h. inJGR. 89. C95. [l7J Schultz P. H. and Spudis P. D. (1978) having normative composition of noritic anorthosite, and mafic IPSC IX. 1033. [l8J King J. S. and Scott D. H. (1978)NASATM­ (basaltic) components cannot be presentin quantities greater than a 79729. 153. [19J Grieve R. A. F. et aI. (1981) InMu/li-Ring Basins. few ~nt [2J. ,Because evidence from other basins [l6,20J and Proc. IPSC 12A (P. Schultz and R. Merrill. cds), 37, Pergamon. impact melts from Imbrium and Serentitatis [l5J suggest a more New York. [20J Spudis P. D. and Davis P. A. (1986) Proc. IPSC mafic crustal composition at depth, this basin ejecta composition 17th. inJGR. 92. E188. [21J Bills B. and Ferrari A. J. (l976)Proc. slrongly suggests that basin excavation was limited to uppercrustal LSC 7th, frontispiece. [22J Croft S. K. (1985) Proc. LPSC 151h. in levels [2J. At Orientale, the crust may be as thick as 100 ± 10 km JGR. 90. C828. [23J Taylor S. R. (1982) Planetary Science, LPI. IPI COfIlribUlion No. 790 71

Houston, 481 pp. [24J Baldwin R. B. (1981) In Multi-Ring Basins, ofthe SS ([3J and Fig. 1 of[2]). The relative stratigraphic positions Proc. IPSC J2A (P. Schultz and R. Merrill, eds.), 275, Pergamon, of these units are summarized in Fig. 1. Selected samples were New York. [25J Head 1. W. (1974) Moon,n, 327. [26J Melosh H. analyzedby opticalmicroscopy,SEM,microprobe,XRFandINAA, 1. (1988) Impact Cratering, Oxford. [27J Hartmann W. and Kuiper Rb-Sr and Sm-Nd-isotope geochemistry [4], and carbon isotope G. (1962) Comm. LunarPlanet. Lab., I, 51. [28J FielderG. (1963) analysis. Nature, 198, 1256. This abstract summarizes the results ofpetrographic and chemi­ cal analyses for those stratigraphic units that were considered the main structural elements of a large impact basin (see [1]). SUDBURY PROJECT (UNIVERSITY OF MUNSTER­ Basement and Related Breccias (Fig. 1): The crystalline ONTARIO GEOLOGICAL SURVEY): (3) PETROLOGY, rocks underlying the Sudbury Igneous Complex (SIC), collectively CHEMISTRY, AND ORIGIN OF BRECCIA FORMATIONS. called footwall rocks [5J, display three types of impact-induced D. Swfflerl, A. Deutschl, M. Avennannl,z, P. Brockrneyerl,z, R. effects: (1) An 8-10-krn-wide zone with planar defonnation fea­ Lakomyl.2, and V. Mill1er-Mohrl,z, Ilnstitut fUr Planetologie and tures in quartz immediately below the SIC indicating peak shock 2Qeologisch-Palliontologisches Institut, Universitiit MUnster, pressure up to about 20 GPa [6]. (2) An irregular, mostly lens­ Wilhelm-Klemm-Str. 10 and Correnstr. 24, W-44OQ MUnster, shaped,discontinuous heterolithic breccia zone along the contactof Gennany. the SIC (Footwall Breccia=FB) thatoccasionally occurs indikelike "intrusions" in the footwall rocks. The breccia matrix is crystalline Within the Sudbury Projectofthe University ofMUnster and the with a dioritic composition and intersertal texture in an upper zone Ontario Geological Survey [1 J special emphasis has beenput on the near to the SIC and a tonalitic-to-granitic composition andpoikilitic brecciaformations exposed at the Sudbury structure(SS) because of to granular texture in a lower zone. The matrix texture is caused by their crucial role for the impact hypothesis [2]. They were mapped thennal annealing and partial melting due to the overlying melt and sampled in selected areas ofthe North, East, and South Ranges complex [7--9]. The clast lithologies in this breccia and its chemical

THICKNESS LITHOLOGY INTERPRETATION 800m Hi;~'i~~~.q~~!~~-~~~~~~~·~ f~~~~~~~~~ .. post-impact ~ 600m :-Onwatin Formation ....i sedimentary fill =-:- --=-= ::> 000 0 0 0 0 000 0 0 0 ~ 700-900 m o 0 Upper Blaclc Member 0 0 C1 reworked breccia material 000000000000 P:: . ...Q _IL.ll.. 0 .•0 o•.ll..9. ..D•.o..•o•.···· I:l redeposited suevitic breccias 100·200 m o 0 Lower Blaclc Member 0 0 0 ,.( CDZCOJl/./'O./l///, ..•··• r.77rl 5·70 m s:: final fall-back V VV :tV-t?!~.....l1~~~ ~ 5:l 300-500 m vvvvvvv;l: suevitic breccias Gray Member Il\) v V v v v.;? V inclusions of melt-breccia

clast rich melt-breccia 0-300 m inclusions of suevitic breccias

2300 m

clast free impact melt sheet

clast rich melt rock 0·150 m thermally metamorphosed cla.~tic breccia

dike breccias Fig. 1. Stratigraphic sequence at the brecciated crater floor Sudbury structure with the present genetic interpretation of the lithological units (modified from [12]). 72 1992 Sudbury ConfereN:e

characteristicsindicatethatitisderivedalmostexclusively from the results with regard to the origin and emplacement of these forma­ local Levack gneisses [7,8J. (3) Breccia dikes, collectively named tions are therefore different from most of the previous views as Sudbury Breccia, which are irregularly distributed in the basement expressed mainly in [5]. The present interpretation is as follows: up to a radial distance ofatleast55 km from the SIC [6J. They have The Sudbury Breccias (dike compression ofthe crater basement by a random orientation and range in thickness from millimeters to in situ frictional melting and by shearing processes during the some 300m. Fourgenerations ofdikes havebeen identified [6J that gravity-induced breccias) were formed during shockcollapseofthe differ in matrix characteristics (crystalline to clastic), thickness, transientcavity. The Footwall and Sublayer Breccias including the clast content, and contact relationships. The composition of the offsetdikes were formedinthelatestageofthetransientcavity(TC) matrix typically resembles the compositionofthe country rock and formation as crater floor breccias before the collapse ofthe cavity. theclasts are derived from thelocalrocks exceptforverylargedikes This holds also for the impactmeltsystem and the suevitic breccias where clasts were moved over distances of up to 800 m. Shocked oftheGray Memberthatcoveredthe centralpartandtheupperwalls clasts (stage I) within the dikes occuronly up to a radial distance of ofthe TC. As a consequenceofitscollapse lhe innermost, clast-free 10 km from the SIC. partofthe melt layerwithin theTC pooled to a meltsheetfilling the SIC and Basal Member of OnapIng (Fig, I): The SIC has centraldepression as well as thedepressions outsidethenoweroded been studied only in its contact zones to the Basal Member and to peakring and,immediately afterward,becamecoveredbyclast-rich the Footwall Breccia. The Sublayer is a heterolithic breccia with a meltslumping in from higherregions ofthe TC wall. Similarly, the crystallinemeltmatrixofmafic composition and abundantmafic to suevitic material slumped into the depression from even higher ultramafic lithic clasts. It contains also clasts of FB as the latter portions of the TC wall. It was then covered by airborne fallback carries clasts of the norite and Sublayer. The contacts of the FB to material (Green Member) deposited from the ejecta plume. The thenorite are sharp whereas to the Sublayerthey are eithersharp or Black Member breccias are interpreted as material that was trans­ gradational [8,9J. Similarly, the uppermost section of the SIC ported into the central depression by turbulent slumping from the (granophyre) shows sharp and gradational transitions to the Basal walls of the peak ring and by aquatic sedimentation of ejecta Member, which is a clast-richimpactmeltbreccia. Abundant clasts material covering the ring region. (up to 80 vol%), mainly from the Proterozoic Huronian meta­ References: [1] AvermannM.etal., thisvolume. [2] Dressler sediments,areembeddedin a crystallinematrix withgranophyric to B. O. et al. (1987) In Research in Terrestrial Impact StrllCtures (J. variolitic texture (10J. There is a chemical similarity between the Pohl, ed.), 39, Earth Evolution Series, F. Vieweg, Braunschweig. SIC, especially the granophyre, and the melt breccia of the Basal [3] BischoffL. etal.,this volume. [4J DeutschA. etal.,this volume. Member confmning that both units represent a whole rock melt of [5] PyeE. G. etal.,eds. (1984)The Geology andOre Depositsofthe Precambriantargetrocks.Thereissomeindication foranincreasing Sudbury Structure, 603, Ministry of Natural Resources, Toronto. proportionofProterozoic Huronianrocks toward the top ofthemelt [6] Miiller-MohrV., this volume. [7] DeutschA. etal.(1989)EPSL, system [4J. 93, 359. [8] Lakomy R. (1990) Meteoritics, 25, 195. [9J Dressler Gray,Green,andBlackMembersofOnaplng(FIg,I)[lO,llJ: B. O. (1984) in [5]. [10] Muir T. L. and Peredery W. V. (1984) in TheGray Member is a polymictclasticmatrix brecciacomposedof [5]. [l1J Avermann M., this volume. (12J Brockmeyer P. (1990) lithic clasts ofvariabledegree ofshock and irregularly shapedmelt Doctoral dissertation, University of Munster, 228. inclusions thathavea fme-grainedcrystallinetextureandshow flow structures. Texturally, this breccia layer is similar to the suevitic breccias found inthecraterfill ofmany compleximpactcraters. The clasticbrecciaconstituentsarederivedpredominantlyfrom rocksof "BRONZITE" GRANOPHYRE: NEW INSIGHT ON the HuronianSupergroupand,morerarely, from Archean basement VREDEFORT. A. M. Therriault and A. M Reid, University of rocks. The Gray Member is covered by a thin layer of a rather Houston, Department of Geosciences, Houston TX 77204, USA. uniform, strongly chloritized breccia (Green Member, formerly called chloritized shard horizon). The contact varies from sharp to The Vredefort Dome is located near the center of the Wit­ gradational. This breccia contains relatively fme-grained clastic watersrand Basin, about 120 km southeast ofJohannesburg, South material in a microcrystalline matrix characterized by concavely Africa. Its origin is enigmatic, ranging from a major unpact event shaped vugs filled with chlorite. Green Member is topped by the [1-3] to endogenousprocesses, eitherigneous [4-6] ortectonic [7J. clasticmatrix breccialayeroftheBlackMember, which, as a whole, A unique melt rock, the "Bronzite" Granophyr, occurs in the is characterizedby a carbonaceous matrix. Inits lower unit itshows Vredefort structure as vertical ring dikes along the contact between similarities to the Gray Member. The upper unit displays textural sedimentary collar and core of Archaean granites, and as vertical features ofreworking and sedimentation under aquatic conditions. dikesextendingnorlhwest-southeast andnortheast-southwestin the A major fraction of the carbon is derived from organic matter and granitic core. The granophyre rocks have an unusual composition formedundereuxinicconditions. Thebulkchemicalcompositionof and high contentofrecrystallized sedimentary inclusions compared the whole sequence of clastic matrix breccias (Gray Member to to common intrusive igneous rocks with similarSiOI content (61 to BlackMember)is less siliceous andricherinFe, Mg, K, andNathan 70% by weight). The unique nature of the granophyre has been themeltbrecciaoftheBlackMember[10,11]. Itisnotclearwhether underlined in previous studies and origin hypotheses as an impact this is due to secondary alteration processes or whetheritreflects a melt or as a highly contaminated intrusive mafic magma have also primary change in thecompositionofthe sourcerocks ofthe clastic been discussed [e.g., 8-10J. We present new results obtained from matrixbreccias.TheREEpatternsofallbrecciaunitsoftheOnaping a recentdetailedpetrographic andgeochemical study ofa very large Formation are quite similar and are well within the range of the and texturallydiverse suiteof"Bronzite"Granophyre,tq)resenting potential sourcerocks from the Proterozoic and Archean basement. all dikes occurring at Vredefort. Thedataobtainedfrom thepetrographicandchemicalinvestiga­ Petrography: Themajormin«alphasesobservedinthegrano­ tions of the breccia formations at the SS are compatible with the phyre are hypersthene, plagioclase, orthoclase, quartz, pigeonite, impact basin model summarized in [2]. The interpretation of our augite, biotite, magnetite, and ilmenite. Only rare bronzite grains LPI COIIIriblllUHa No. 790 73

are observed and they occur exclusively as xenoliths. None of the Conduslons: Thematrixtextures,thevariablethermaleffects bronzite grains are in equilibrium with the granophyre melt. in the inclusions, and the chemical variations presented for the Two major types of granophyre are observed: (1) fme-grained, granophyre dikes ofVredefort are compatible with an impact melt clast-richdikes, confinedto thecentral core ofthe structure that are (15,16]. Our observations and results indicate that the granophyre dominatedby a spherulitic texture and textural heterogeneity occur dikes best represent remnants of an impact melt that intruded overdistances ranging from millimeters to tens ofmeters produced fractures ofthe transientcraterfloor ofVredefort. Wethus favor the by four spherulitic subtypes and an ophitic subtype texture; and Vredefort structure as a deeply eroded multiring impact basin. (2) medium- tofme-grainedclast-richgranophyrecore-eollardikes References: [I] Dietz R. S. (1961) J. Geol., 69, 499-516. dominated by hypidiomorphic textures [l1J. Grain size of the [2] Hart R. J. etal. (1991) Tectonophysu:s, 192, 313-331. [3] Mar­ granophyre matrix minerals ranges up to 5 Mm. The mineralogy of tini J. E. J. (1991) EPSL,I03, 285-300. [4] Ully P. A. (1981) JGR, alldikes issimilarwiththeexceptionofthehighermodalabundance 86, 10689-10700. [5J Schreyer W. and Medenbach 0. (1981) of biotite in the core dikes relative to the core-eollar dikes. The Conlrib. MiMral. Petrol., 77, 93-100. [6] Nicolaysen L. 0. and spherulitic texture with skeletal crystal morphologies observed in Ferguson J. (1990) Tectonoplaysu:s, 171, 305-335. [7] Colliston thecoredikesis indicativeofextremeundercoolingconditions (12), W. P. (1990) Tectonoplaysu:s, 171, 115-118. [8] FrenchB.M. etal. whileincreasedtextural homogeneitycharacterizing thecore-eollar (1989) Proc. LPSC 19th, 733-744. [9] French B. M. and Nielsen granophyre dikes indicates moreuniform and slowercooling hislo­ R. L. (1990) Tectonophysu:s, 171, 119-138. (10) Reimold W. U. ries. et al. (1990) Proc. LPSC 20th, 433-450. [11] Therriault A. M. and Numerous monomineralic and lithic fragments, up to 80 em Reimold W. U. (1991) LPSC XXII, 1391-1392. [12] Lofgren G. long, compose up to 20% of the rock volume [11]. All the major (1971)Am.J. Sci., 274, 243-273. [13] GrieveR. A. F. etal. (1990) country rocks arerepresented as inclusions inevery dikeexamined. Tectonophysu:s, 171,185-200. [14] SchreyerW. (1983)J.Petrol., Granite, gneiss, and quartzite are the most abundant, mafic rock 24, 2647. [15] Floran R. J. et al. (1978) JGR, 83. 2737-2759. fragments and metasediments other than quartzite are less abun­ [l6] Phinney W. C. et al. (1978) Proc. LPSC 9th, 2659-2693. dant, and shale inclusions are rare [10,IIJ. These abundant inclu­ sions show intense recrystallization, reactions with the granophyre melt, andmelting. Rareshockfeatures areobservedinquartzgrains and are restricted to remnants ofdecorated planarelements occur­ A COMPARISON OF THE CHEMISTRY OF PSEUDO· ring as one set parallel to the c axis of individual quartz grains. TACHYLYTE BRECCIAS IN THE ARCHEAN LEVACK Chemistry: Chemical homogeneity, on a regional scale, is a GNEISSES OF THE SUDBURY STRUCTURE, ONTARIO, major characteristic of the granophyre dikes of the Vredefort Lucy M. Thompson and John G. Spray, Department of Geology, structure. Homogenization was achievedearly in the melt's history UniversityofNewBrunswick,Fredcricton,NewBrunswick,Canada. and was maintainedas themeltintrudedthe fractured countryrocks where it underwent cooling and crystallization under relatively TheArcheanLevackGneisscsoftheNorthRangehostmillimeter­ undisturbed conditions. thick veins and centimeter-thick lenses ofpseudotachylyte, as well Althoughnogross differences inmajor- andmosttrace-element as substantially larger meter-wide, dykelike bodies of pseudo­ compositions were detected that could be ascribed to regional tachylytic "breccia." The "breccia" occurs up to several tens of positionwithinthestructure,minorchemicalvariations arepresent. kilometers away from the Sudbury Igneous Complex and is com­ The granophyre dikes of the central core have higher SiOz' TiOz' monly sited within ornearjoints andothernatural weaknesses such Alz0" and Kz0 contents thanthecore-collardike, whilecore-collar as bedding, dyke contacts, and lithological boundaries. dikes have higher Fe<> + FezO" MgO, CaO, and NazO contents. The larger "breccia" dykes comprise a generally dark matrix These differences are thought to be due to differences in the containing rounded to subrounded and occasionally angular rock composition and amount of local materials assimilated. Although fragments derived predominantly from Levack Gneiss. Thematrix the granophyre melt is weakly differentiated, this is a minor factor may exhibitflow features andtypically appears aphanitic, although intheevolutionofthegranophyremeltanddifferential assimilation in certain exposures it possesses a fme-grained crystalline texture. is the major cause ofthe chemical variability observed. The "breccia" fragments can be as large as 2-3 m in their long Dlseusslon: Metasedimentsandshaleinclusions,from litholo­ dimension and are typically chaotically arranged within the matrix, gies occurring within units stratigraphically higher than the present showing evidence of both rotation and internal fracturing. More emplacement level of the granophyre dikes, are regionally distrib­ exotic rock fragments, such as amphibolite, also occur and these utedwithinall thesedikes. Thisobservationis hardtoreconcilewith appear to have been transported for somedistance (i.e., atleast tens theprocesses involved in the intrusionofa magma from themantle ofmeters). Theoriginoftheso-calledSudbury Breccias is a subject oruppercrustand indicates thatthegranophyremeltmusthavebeen ofcontroversy, butis generally believedtoberelatedto the 1.85-Ga efficiently and dynamically mixed before being injected intomajor Sudbury event. Field evidence indicates that they are fault-related fractures. Highly heterogeneous clast populations from widely and frictionally induced and are therefore not thedirectproducts of different stratigraphic levels, a complex thennal history, and injec­ shock melting. tion of melt/clast mixtures into dikes are in agreement with pro­ Selectedsamples ofbulkSudbury BrecciaandSudbury Breccia cesses related to impact melt fonnation. The frrst report of rare matrices have been chemically analyzed and compared to existing shock features in xenolithic quartz grains supports the melt origin dataonthe Levack Gneisses and SudburyBreccia. Thematrices are by impact. Two reasons explain why shock planar features, ob­ apparently enriched in Fe and, to a lesser extent, Mg, Ti, and Ca served in the source rocks (13], are rarely seen in inclusions of the compared to the wallrocks and the majority ofclasts. This enrich­ granophyre: (1) they have been annealed (13,14] and (2) shocked ment can be partly explained by the preferential cataclasis and/or fragments are preferentially assimilated in the melt because they frictional melting of hydrous ferrornagnesian wallrock minerals, attained a higher temperatureduring the initial shock event [15,16]. but also appears to require contamination by more basic exotic 74 1992 Sudbury Conference

lithologies.Thissuggests thatc:crtainCCllllponents ofpseudotachylitic SAl-302 Sudbury Breccia have Wldergone significant transport (?kilome­ AGBCJoCA] ters) during their formation.

4tAr-"Ar AGES OF THE LARGE IMPACT STRUCTURES KARA AND MANICOUAGAN AND THEIR RELEVANCE TO THE CRETACEOUS·TERTIARY AND THE TRIASSIC· JURASSIC BOUNDARY. M. Trieloff and E. K. Jessherger, Max-P1anck-lnstitut fOr Kcrnphysik, P.O. Box 103980, W-6900 Heidelberg, Gcnnany.

Since the discovery ofthe Irenrichment in Cretaceous-Tertiary boWldary clays in 1980 by [1] the effects of a 10-1an asteroid _ .. .. ___ _ _ Tr-teJ impacting on the Earth 6S Ma ago arc discussed as the possible reason for the mass extinction- including the extinction of the dinosaurs-atthe endofthe Cretaceous. Butup to now no crater of ' •• 2. this age thatis largeenough(ca.200Ian indiameter)has beenfound. The Manson Crater in north America is 65 Ma old [2], but too this as confumationofearlierproposals that the Kara bolide would sm~n1y35 Ian indiameter. A recently discovered candidate is have been at least one of the K-T impactors. Kocherl et a1. [S] the Chicxulub structure in Yucatan, Mexico, but intensive investi­ determined 4OAr-"Ar ages ranging from 70to 82Maand suggested gations have to be done to identify it as the K-T impact crater. an association to the Campanian-Maastrichtian boundary, another Petrographic signs at the K-T boWldary seem to point to an impact important extinction horizon 73 Ma ago. into the oceans as well as onto the continental crust; multiple We dated four impactmelts, KA2-306, KA2-305, SA1-302,and impacts were considered [3J. AN9-182. All spcclra show well-defined plateaus. They arc shown Another candidate is the Kara Crater in northern Siberia. withastrongextendedagescale(Figs. 1and 2). Ouragesrangefrom Kolesnikovetal. [4] determined aK-Arisochron of65.6 ±0.5 Ma, 69.3 to 71.7 Ma, and itisclearlyvisiblethatourdatasuggestncither indistinguishable from the age ofthe K-T boWldary and interpreted an association with the Cretaceous-Tertiary nor the Campanian­ Maaslrichtian boWldary. Fnors arc given as 10etrOl'S computed of the deviation ofthe plateau fractions. The systematicerror induced by the NL25 hornblende standard is 0.6 Ma. It may be argued that ourages-oldincomparison to the K-Tboundary-c:ou1dbecauscd by relict targetrocks incorporated in the melt. At the first sight this seems tobepossible:Ifonly 1%ofthesample'spotassiumislocated inarelictphaseofpaleozoic age ofSOO Ma, this is enough to liftthe sample's K-Ar age from 65 Ma to 70.S Ma. We consider SAI-302 to test whether the age pattern would then still show a plateau orif therelictphasewould be recognizable. Thedegassing patternshows two distinct reservoirs. We calculated the diffusion parameters, activation energy Qand frequency factor Do, byAnhcnius plots for each reservoir and simulated the gas release by two phases having 1IO differentdiffusion parameters. We assumed an age of65 Ma fOl'the f- 71.7 •.3"'--i two phases and added an ubilrarily chosen relict phase having 70 --

Manlcouallan Anortb1h

SAI-30Z a .,

~ 69.3 ••"'" ~ -L1 r~-.L., I l ~ ~\ -LS

::s oi :r ~ - ~: ! 10 BO 5 I l! - :\-----'=--1 .. - J 10 BO 3C ; ic -I ~\l\oo:::ooo'''-_-_-_''- .-1121111I __1 S ~\t ..Lc o 50 ... F.-.etao..... MAl ...... (Xl Pr.ct.1on.1 ••A.r ...-1._. [X]

FIg. I. FIg. 3. IPI Contriblllion No. 790 1S

diffusion parameters of a typical anorthite, with an age of500 Ma References: [I] Alvarez L. W. et aI. (1980) Science, 208, constituting0.8% ofthesamples' potassium. Thedegassing peak of 1095. [2] Hartung J. B. and Anderson R. R. (l988)!.PI Tecla. Rept. the relict phase coincides by and large with the second reservoir of 88-08. [3] Alvarez L. W. and Asaro F. (1990) Sci. Am., 362. SAI-302 but the degassing feature is broader(Fig. 3). The resulting [4] Kolesnikov B. M. et aI. (1988) In !.PI Contrib. No. 673. calculated age pattern shows a low-temperature plateau of 6S Ma [5] Kocberl C. et aI. (1990) Geology, 18, 50. [6] Nazarov M. A. and an irregular shape in the high-temperature steps. That means et aI. (1991) IPSC XXII, 961. [7] Grieve R. A. F. (1991) Meteorit­ that the measured age pattern can only be interpreted as a result of ics, 26, 175. [8] Wolfe S. H. (1971) JGR, 76, 5424. [9] Harland 6S-Ma age and a relict phase iftherelictphase has exactly the same W. B. et a1. (1982) A Geologic TiIM Scale, Cambridge Univ. gas release pattern-but if this is the case, the relict phase should [10J Palmer A. R. (1983) Geology,lI, 103. have been reset by the cratering event as well as the sample unless it would have been incorporated afterward, e.g., as contamination. A further condition that must be fulfilled to fit this scenario is that 4IAr.BAr DATING OF PSEUDOTACHYLITES FROM THE the K fraction of5()()..Ma old relict phases in the different samples WITWATERSRAND BASIN, SOUTH AFRICA, WITH must be nearly the same (I ±0.2%), oranotherpropercombination IMPLICATIONS FOR THE FORMATION OF THE ofage and K content. We regard this as improbable. VREDEFORTDOME. M. Trieloffl, 1. Kunzl , B. K. Jessbergerl , Althoughthe agespectrashow well-defmedplateaus, theplateau W. U. Reimold2, R. H. Boer2, and M. C. Jackson2, IMax-Planck fractions still exhibit fme-scale structures that deviate from a Institut fUr Kemphysik, P.O. Box 103980, W-6900 Heidelberg, theoretical calculated plateau (Fig. 2, bottom) as expected for a Gennany, 2Bconomic Geology Research Unit, University of the maximum precise measurement of an undisturbed sample. Nor­ Witwatersrand, P.O. Wits 2050, Johannesburg, RSA. mally the mass-dependent diffusion difference of 40Ar and "Ar is considered as negligible,butifweinduceitin ourcalculations in the The formation of the Vredefort dome, a structure in excess of case of SAI-302 we get an age spectra as shown in Fig. 2 with 100 kIn in diameter and located in the approximate center of the slightly increasing ages within each reservoir. This strengthens the Witwatersrand basin, is still the subjectoflively geological contro­ assumption that these deviations are not of statistical, but of versy. It is widely accepted that its formation seems to have taken systematic nature, an artifact induced by the 40Ar-"Ar stepheating place in a single sudden event, herein referred to as the Vredefort teehnique. Inthissense,thespectrumofSAI-302seems torepresent event, accompanied by therelease ofgigantic amounts ofenergy. It the "ideal" spectrum of an undisturbed sample. For the fmc-scale is debated, however, whether this central event was an internal one, deviations ofthe otherthree impact melts"Ar recoil redistribution i.e., a cryptoexplosion triggered by volcanic or tectonic processes, seems toplayamajorrole, but this has to be investigated by further or the impact ofan extraterrestrial body. petrographic studies on grain size and potassium distribution. Ages obtained on rocks from the Vredefort structure cluster We conclude an age of69-71 Mafor the Kara impact structure. largely around 2.0Ga(e.g., review by [1 n. Granophyre, an unusual Hydrogen isotopicmeasurements byNazarovetaI. [6] show thatthe melt rock forming dykes in the Vredefort Dome and thought to be impact occurredon dry lind and the authors concluded amaximum related to the Vredefort event, yielded a Ph-Ph zin:on age of2002 ± age of69-70Ma, thetimeoftheendofthe lastregression within the 52 Ma [2]. Pseudotaehylite, a melt breccia ftrst discovered and crater's region before the end of the Cretaceous. Our data are consistent with an impact a short time after the regression. Figure 4 shows the KlCa and the age spectra of two impact 1,5 metamorphic anorthite samples (IOBD5 and 10BD3C) of the Manicouagan Crater, Canada. As visible in the KlCa spectra (only 0.5 10BDS is shown), the samples consists oftwo differentphases, one IV~ degassing at low temperatures having an age plateau indistinguish­ -0.5 able from the cratering eventof212 Ma(1), the second one showing the signature ofa partially degassed phase, having ages increasing ~ -1.5 up ~.950 Ma (lOBD5), the age of the target rocks [8]. 10BD3C suffered amore completedegassing, having ages ranging only up to 2.4 300 Ma. The low-temperature plateaus are in agreement with the 2:2 I 2.02 Ga I crater age of 212 Ma and do not improve the age of the impact sb'Ucture. Anyway, while thecraterage is quite accurate, the ages of 2.0 - J the adjacent geologic boundaries seem not to be. The last revision 1.8 r [9] of the Triassic-Jurassic bowlClary in 1982 delivered an age of 213 Ma, while a later determination [10] gives a lower age of 1.6 I 208 Ma. We think so far as ages are concerned it is not possible to I conclude orexclude an association ofthe two events until the age of 1.4 the boundary is determined more precisely. 1.2 Our measurements enable us to estimate the intensity of the thennaleventinducedbythecrateringeventforthetwoManicouagan 1.0 samples. Ourresults are consistent with a time-temperature combi­ o 50 100 nation of I Ma at 3370C or 12 hr at llOOOC for IOBD5 and 1Ma at Fractle>nal 39A.r R.elease [%] 361°C or 50 hr at HOO°C for 10BD3C. Future investigations may allow us to infer a cooling model for the Manicouagan impact melt sheet: Fig. 1. Age spectrum for BL-28B. 76 1992 Sudbury Con/ereru:e

extremely abundant in the VrCdefort structure, is, like the grana­ Such recent formation ages would even question the single-stage phyre, widely regarded to be the direct result of the dome-forming formation oftheVredefortdomeand its associatedpseudotaehylites process. [3]. Ages lower than 2 Ga are not restric~ to pseudotachylite We datedeightpseudotachylitesamples from abedding-parallel either: a plagioclase separate of the lamprophyre from the fault zone across the north-central Witwatersrand basin with the Lindequesdrift intrusion dated at 2163 Ma yielded a 1248 ±22 Ma 4OAr_39Ar step heating technique. All age results (Table 1) are age [5]. Several mineral separates from Rietfontein complex rocks indistinguishable from 2.00 Ga within 1 (J errors, except for those gave partial plateaus at about 2004 Ma, but also showedlow ages in obtained for two samples: lOoo5 (2.10 Ga) seems to be severely the low-temperature release steps comprising 20 to 50% ofthe 39Ar disturbed in thelow-temperaturefractions; the other one,BL-30has release, suggestive ofdisturbances at 1170 ± 14 and 725 Ma(6). A an extremely low K contentof20ppm-cornparedto Kcontents of pseudotachylite from the Riefontein complex (USA35) showed an several percent in the other pseudotachylite samples. The relative irregular age pattern suggesting disturbances at 1400 Ma and fractionofexcess argonis significant, whichresults inan unrealistic 800 Ma [3]. In the outer (000) core of the Vredefort structure, K-Ar age of4.7 Ga. pseudotachylite USA30A yielded an age of 1.39 Ga [3], whereas a Figure 1 shows the exemplary spectrum of pseudotachylite Rb-Sr mineral isochron for the host granite resulted in an age of sample EL28B with a plateau over92% ofthe 39Ar release. In light 2002 ± 8 Ma and Ar-dated mineral separates yielded ages ranging ofthe appearanceofthe age spectraand thelargenumberofsamples from 1.39 to 1.76Ga, all significantly lower than 2 Ga(l0). Biotite analyzed toidenticalresults, we donotbelieve thatthe 2-Gaages are separates from a granophyre sample (VVG) from a site less than reset ages causedby thermal overprint. Ourages are consistentwith 3 km from the USA30 locality yielded a possible formation age of the 4OAr_39Ar age of a Vredefort pseudotachylite (USA29, 2.00 Ga 2006 ± 9 Ma, but with evidence in the age spectrum for significant [3]). This indicates that the Vredefort event and the Witwatersrand overprinting at 700 and/or 1200 Ma [6]. Biotite from OGG sample pseudotachylite are related. Taking all our results into accoWlt, the KK55 from the northwestern sector of the core yielded a 1935 ± resulting age for the Vredefort event is calculated to 1997 ± 5 Ma; 8 Maplateau age, still lower than 2 Ga{6]. Finally,pseudotachylite a larger enor of 19 Ma would be realistic if the additional error USA31 provided an age pattern suggestive ofpartial degassing at induced by our NL-2S hornblende standard is considered. 1.4 Ga [3J, whereas its host rock, NW28, gave ages of 2070 Ma A circumstance in favor ofan origin by internal processes is the (biotite) and3030Ma(hornblende). Itshould alsobementionedthat apparent coincidence of the Vredefort dome formation and other four of our Witwatersrand pseudotachylite samples show low­ regional magmatoteetonic processes. These are the emplacements temperature fractions consistent with disturbances at 1400 Ma. ofseveral alkaline granites and ayenites, namely the Schurwedraal Post-2-Ga events could thus have taken place at 1400, 1200, 1000, complex (Rb-Srerrorchronof2016±61 Ma[4]), the Lindequesdrift and 700 Maago. There are twopossiblescenarios thatcouldexplain intrusion (2163 ± 31 Ma 4OAr_39Ar plateau age of hornblende [5]), this record of low ages for Vredefort samples. Firstly, the low the Rietfontein complex (2004 ± 16 Ma 4OAr_39Ar plateau ages of pseudotachylite ages are seen as formation ages that date the result biotite and hornblende [6]), and-in an even wider regional con­ of post-2-Ga tectonic events. This would explain the different text-theBushveldcomplex,especiallybecauseBushveldsynchro­ resetting degrees observed for samples from different localities, as nous intrusives, such as the Losberg complex, occur close to the well as observations ofmultiple geological hints for the existence Vredefort structure. The age of the Losberg complex (Rb-Sr iso­ of several pseudotaehylite generations in the Vredcfort structure chron of 2041 ± 41 Ma [7]), is in agreement with ages thought to represent the main Bushveld activity (2050-2060 Ma [1», but, in general, ages have been determined ranging from 1900 to 2100 Ma, e.g., an 4OAr_39Ar age of Bushveld magnetite gabbro of 2096 ± 12 Ma [8] and an 40Ar_39Ar laserprobe age ofMerensky reefbiotite of 2010 Ma [9]. In agreement with an extended rather than short period of Bushveld activity are our own 4OAr_39Ar step heating results ob­ tained for aBushveldgabbro (wholerock, anorthite, and pyroxene), where two partial plateau can be identified in the age spectra: one corresponding to a mean age of2.00 Ga and the other to a mean age of 2.10 Ga. This could indicate a complex cooling or reheating history of the sample. Investigations are currently in progress to determine whether excess Ar could be the cause of this age di­ chotomy or whether the spectra are the result of a true two-stage geological history. No matler what the result will be, an eventclose to 2.00 Ga will remain indicated. o'-" e.--' e;".., Ifhigh-resolutiongeochronology couldprove that the ages ofthe p ..... O e-. -~ rocks related to the dynamic Vredefort event (granophyre, ~ coo--~ y pseudotachylite)andtheagesoftherocksformedinmagmaroteetonic 111- f-1l"

[10]. Then it should be clear why there are not pseudotachylites Abbas and others [8] conducted additional geophysical studies older than 2 Ga. Secondly, the low ages could be due to post-2-Ga and arrived at similar conclusions: (1) Al Umchaimin was not thennal overprint. Host rocks were then more or less intensely fonned by meteorite impact and (2) probably it represents a solu­ affectedin accordance withthe different Arretentivities ofdifferent tion-collapse feature. Greeley and others [9], in preparation for the minerals, or different closure temperatures in the case of Rb-Sr Magellan mission to Venus, studied shuttle radar images of nine isotope systematics. A major problem is seen in the different maar volcanos, one volcanic caldera dome, one impact structure, intensities of the resetting events at different localities within the and one possible impact structure (AI Umchaimin). Concerning Al dome. This may be explained partly by additional hydrothennal Umchaimin, they wrote: "Although no defmitive impact features activity. have beenreported, the circularity and slightly upliftedrims suggest Future investigations are still needed to completely clarify the an impact origin." nature and the duration ofpost-2-Gaprocesses that took place in the In 1965, the authorand the late RandolphChapman,bothvisiting Vredefort structure. professors at the College ofScience, University ofBaghdad, spent Acknowledgments: Wegratefully acknowledge the generous half a day at Al Umchaimin during which a section was measured support from Anglo American Prospecting Services that pennitted uptheeastwall, samples werecollected forlaterthinsectioning, and the analysis ofWitwatersrandpseudotachylites, and the pennission a search made for meteoritic debris, shatter cones, impact glass, to report the results.. meltbreccia, and so on. Noevidence was found for an impactorigin References: [I] Walraven F. et al. (1990) Tectorwphysics, ofthe depression,nordidstudyofthethinsections from theeastwall 171,23-48.[2] CompstonW. andNicolaysenL.O.in[l]. [3] Reimold of the depression reveal any microscopic evidence of impact. W. U. et al. (1990) Tectorwphysics, 171, 139-152. [4] Nicolaysen It is concluded that, on the basis of the studies that have been L. O. et al. (1963) Gen. Assem. Int. Union. Geol. Geophys. 13th, made ofAl Umchaimin and on the basis ofthe briefsite visit made, Berkeley, Calif., abstract. [5] Hargraves R. B. (1987) S. Afr. J. Al Umchaimin probably is not an impact structure but most likely Geol., 90, 308-313. [6] Allsopp H. L. et al. (1991) Suid afrikaanse resulted from the enlargement and coalescence of sink holes and Tydskrifvir wetenskap, 87, 431-442. [7] Coetzow H. and Kruger eventual collapse of the roof material into the resulting cavity. F. J. (1989) S. Afr. J. Geol., 92, 37-41. [8] Burger A. 1. and References: [I] Suleiman M. W., personal communication. Walraven F. (1977) Annu. Geol. Survey S. Africa, 11, 323-329. [2] Anny Map Service (1952) Wadi Hauran, Iraq,SheetNI37-15, [9] Onstott and Kruger in [1]. [10] Reimold W. U. et al. (1990) scale 1:250,000. [3] AI-Din T. S. et al. (1970) 1. Geol. Soc. Iraq, 3, Geol. Soc. Austr., 27, 82. 73-88. [4] Merriam R. and Holwerda J. G. (1957) Geograph. J., CXXIlI, 231-233. [5] Mitchell R. C. (1958) Geograph. J., CXXW, 578-580. [6] Freeberg J. H. (1966) U. S. Geol. Surv.Bull.1220, 21. [7] AlNaqibK.M. (1967) U.S. Geol.SurveyProf Pap. 560-G,G7. AL UMCHAIMIN DEPRESSION, WESTERN IRAQ: AN [8] AbbasM. J.etal. (1971) J. Geol. Soc. Iraq, 4,25-40. [9] Greeley IMPACT STRUCTURE? James R. Underwood Jr., Department R. et al. (1987) Earth Moon Planets, 37, 89-111. of Geology, Kansas State University, Manhattan KS 66506-3201, USA.

Al Umchaimin, in Arabic "hiding place" or "place of ambush" A LATE DEVONIAN IMPACT EVENT AND ITS ASSO· [1], is located at latitude 320 35.5'N and longitude 390 25'E. It lies CIATION WITH A POSSIBLE EXTINCTION EVENT ON some43km S36° ofthe H-3pumpstation onthe abandoned Kirkuk­ EASTERN GONDWANA. K. WangI and H. H. J. Geldsetzer2, Haifa oil pipeline [2] and 60 km S49°W ofthe western desert town I DepartmentofGeology, University ofAlberta, Edmonton, Alberta of Rutba. The nearly circular depression averages 2.75 km in TOO 2E3, Canada, 2Qeological SurveyofCanada, 3303-33rdStreet, diameter and is 33-42 m deep. It is floored with fme-grained, clay­ N.W., Calgary, Alberta TIL 2A7, Canada. rich deposits, estimated to be 36 m thick [3], the surface of which shows well-developed desiccation fIssures ormudcracks when dry. Evidence from South China and Western Australia for a 365-Ma Because of its nearly circular planimetric shape and its apparent impact event in the Lowercrepida conodont zoneofthe Famennian isolation from other surface and subsurface features, it has been stage of the Late Devonian (about 1.5 Ma after the Frasnian;" considered by some to be a possible meteorite impact structure [4] Famennian extinction event) includes microtektitelike glassy and by others [5] to be a surface collapse feature that originated microspherules [1], geochemical anomalies (including a weak Ir), following removal of magma from the subsurface as the magma a probable impact crater (>70 km) at Taihu in South China [2], and extruded elsewhere. Al Umchaimin was listed in the U.S. Geologi­ an Ir anomaly in Western Australia [3]. A brachiopod faunal cal Survey tabulation of 110 structures worldwide for which a turnover in South China, and the "strangelove ocean"-like l)uC meteorite impact origin had been suggested [6]. It was placed in excursions in both Chinese and Australian sections indicate that at Category VI Structures for which more data are required for least a regional-scale extinction might have occurred at the time of classifIcation. the impact. A paleoreconstruction shows that SouthChina was very K. M. Al Naqib, Iraq Petroleum Company, reported [7] that the close to and facing Western Australia in the Late Devonian [4]. petroleum geology community considered that Al Umchaimin had South China: An Upper Devonian carbonate section exposed originated by fracture-controlled dissolution in the subsurface and at Qidong, Hunan, was studied for biostratigraphy, geochemistry, eventual collapse into the resulting solution cavity. AI-Din and and sedimentology. A brachiopod faunal changeover from the others [3] made geological and geophysical surveys of the depres­ traditional Yunnanellinato Yunnanellafaunas [5] wasrecognizedin sion in 1969 and 1970 and found no evidence for an impact origin. the section. Abundant microspherules were found in a single They concluded, as did AlNaqib, that a solution-collapse originwas stratigraphic horizon immediately below a 3-cm clay with a geo­ likely. chemical anomaly. The microspherule horizon occurs in the Lower 78 1992 Sudbury Conference

crepida conodont zone, based upon the presence of conodont Bai S. (1988) Can. Soc. Petrol. Ceol. Mem.14, 3,71-78. [7] Ii Q. species Pa. quadrantinodosalobata,Po. nodocostatus.J. iowaensis. (1988) Ph.D. thesis, Chin. Acad. Geo!. Sci., Beijing, 140 pp. J. cornutus. J. alternatus, and Pel. inclinatus [6], and Pa. crepida, [8] Nicoll R. S. and Playford P. E. (1988) Ceol.Soc.Austral.Abstr., Pa. minuta minuta, Pa. quadrantinodosalobata, J. iowaensis, 1. 21,296. alternatus helmsi, and J. alternatus alternatus [7]. Petrographic, SEM, XRD, and electron microprobe analyses indicate that the microspherules were like microtektites produced by a bolide im­ ELECTRON PETROGRAPHY OF SILICA POLYMORPHS pact, on the basis of their "splash form" shapes, inner bubble ASSOCIATED WITH PS~UDOTACHYLITE,VREDEFORT vesicles, glassy nature, silica glass inclusions (lechatelierite), and STRUCTURE, SOUTH AFRICA. 1. C. White, Center for chemical compositions that are similar to those of microtektites. Deformation Studies in the EarthSciences, DepartmentofGeology, The geochemical anomaly in the 3-cm clay is characterized by University ofNew Brunswick, Fredericton,NewBrunswick,Canada siderophiles (Jr, Fe, Co, Cr) and chalcophiles (Se, Sb, As) enriched E3B 5A3. by factors of 1 to several orders ofmagnitude over their background values, although Ir abundance in the clay is low (38 ppt). The onc High-pressure silica polymorphs (coesite and stishovite) have maintainsconstantlypositivevalues in thecarbonatesamples below been described from the Vredefort structure [1,2] in association the clay, butshifts suddenly to a minimumof-1.970/00 intheclay and with pseudotachylite veinlets. In addition to the fundamental sig­ above. Although the carbonate rocks we analyzed were altered to nificance of the polymorphs to genetic interpretations of the struc­ some degree by diagenesis (as seen in the thin sections), we believe ture, ithas beenadditionally argued that the type ofpseudotachylite that the trend of the carbon isotopic change is still preserved. A with which they occur forms during the compressional phase ofthe "strangelove ocean"-like onc excursion of 2.70/00 (PDB) in the shockprocess, while the larger, classicpseudotachyliteoccurrences Qidong Section is consistent with the paleontological data suggest­ are barrenofpolymorphs and formed during passage ofthe rarefac­ ing that an extinction might have terminated the Yunnanellina tion wave. This identification of temporal relationships among fauna, which in turn gave rise to the Yunnanella fauna. Taihu Lake, transient shock features at a regional scale is similar to observations a large circular structure, has long been speculated to be a probable from the Manicouagan structure, Quebec [3], where texturally impact crater. Recent work, on the basis of shock metamorphism distinct diaplectic plagioclase glasses formed during both compres­ found in the target sediments, has suggested that it is a probable sional and decompressional phases of the shock process. The impact crater [2]. clarification of such relationships impinges directly on interpreta­ Western Australia: A strong iridium anomaly (20 times the tions of natural shock processes and the identification of high background value) was initially reported in the Famennian Upper probability targets for polymorph searches. triangularis conodont zonein theCanning Basin, Western Australia Detailed analytical scanning (SEM) and transmission electron [3]. An evaluation of the conodont fauna proved that the Iranomaly microscopy (TEM) has been utilized to further establish the nature is actually in the Lowercrepida zone, based upon the occurrence of ofboth the pseudotachylite and the silica polymorph occurrences in Pa. crepida [8]. This strongly indicates that the Canning Basin Ir anomaly occurs at the same stratigraphic level as the Qidong microspherule and geochemical anomaly horizon. Because the Australian Ir anomaly is associated with a Frutexites stromatolite, the interpretation was that the Ir was concentrated biologically by a the cyanobacteria. We acknowledge this scenario but further pro­ pose that there must have been abundantIr available in the environ­ ment for the biological concentration to take place. The most probablesourcefor Iris an impactnearthe region, suchas theimpact in South China. The presenceofFrutexites stromatolites is probably the reason why there is a stronger Ir anomaly in the Canning Basin than in the Qidong area, where no Frutexites stromatolites are present. A negative onc excursionofabout 1.50/00 is coincident with the Ir anomaly in the Canning Basin, and has been suggested to indicate a decrease in biomass [3]. The carbon isotopic excursions, which occur at the same strati­ graphic level in both South China and Western Australia cannot be explained as being coincidental. The onc excursions and the brachiopod faunal turnover in South China indicate that there might b have been at least a regional (possibly global) extinction in the Lower crepida zone. The impact-derived microspherules and geo­ chemical anomalies (especially the Ir) indicate a Lower crepida zone impacteventoneasternGondwana. Thelocation, typeoftarget rocks, and possibly age of the Taihu Lake crater qualify as the probable site of this Late Devonian impact. References: [1] Wang K. (1991) Ceol. Soc. Am. Abstr. Progr., 23. A277. [2] Sharpton V. L. et al., inpreparation. [3] Playford P. E. eta!. (1984)Science. 226,437-439. [4] ScoteseC. R. and McKerrow W. S. (1990) Ceol. Soc. London Mem., 12, 1-21. [5] Tien C. C. (1938) Palaeont. Sinica, New series B, 4,1-192. [6] Wang K. and Fig. 1. U'I Contribution No. 790 79

the Vredefort rocks. This methodology enables the maintenance of structures, however, two alternative lunar analogs (i.e., complex .. sbict control over the spatial, compositional, and crystallographic craters and two-ring basins) can be identified for the Sudbury dike relationships ofthe deformedmaterial. Thebrown, optically isotro­ pattern. The f1J'St compares Sudbury to the central peak crater pic nature of the pseudotaehylite belies an essentially pure silica Haldane in Mare Smythii. Haldane is a multiringed structure with composition. As has been previously noted [2], minor K-feldspar, anouterrim diameterof40km and anuplifted central"floorplate" alwninosilicate(kyanite?) andprimarymicaarethesignificantnon­ separated from this rim by a wide (-5 km) moat structure [5,9]. silicamineral phases.Backscatteredelectronimagingdemonstrates Concenbic fractures occur near the edge of the floor plate, in the complex relationships among the silica phases. Stishovite replace­ moatand in a well-defined annulus 9-16 km beyond the craterrim. ment by quartz (Fig. 1) often takes the form of tensile veinlets Radial fractures aretypicallyresbicted tothecentralplateand moat suggesting reconversion during shock wave relaxation. Coesite regions, butoneseteastofthecraterextends over10kmbeyond the mostcommonlyoccursas aciculargrains and is widely dispersed at crater rim in association with additional concenbic fracturing. a fme scale throughout quartz. Reconverted quartz (Fig. 2) from Crater counts indicate that both the crater and the superposed both preswned melt and polymorphs can be remarkably fme­ fracture systemsformed atnearlythesametime,whereasvolcanism grained, e.g., S200 nm grain diameters. Preservation ofcrystalline in Haldane appears to be coeval with other basalt units in Mare material of such small grain size would appear to preclude signifi­ Smythii. These volcanic units areprimarily located along theouter cant postformation thermal anneals, or otherwise requires ex­ fracture ring and in the crater moat structure [9]. tremely sluggish transformation kinetics. The alternative analogy compares Sudbury with the lunar two­ References: [1] Martini J. E. J. (1978)NatlU'e, 272, 715-717. ring basin Schrodinger. Schrodingeris -300km indiameter with a [2] MartiniJ.E.J. (1991)EPSL, 103,285-300. [3] WhitelC.(1991) broad interior ring surrounding a central floor region -100 km in Geol. Assoc. Can. Prog. Abstr., 16, A13!. diameter. Roughlyconcenbicfractures occuralong the interiorring structure, whereas radial fractures typically extend from the inner ring toward the outercraterwall. Volcanic activity within thebasin FLOOR-FRACTURED CRATER MODELS OF mE SUD­ is limited, but a dark-haloed pit onone ofthe innermost concenbic BURY STRUCTURE, CANADA. R. W. Wichman and P. H. fractures indicates minor pyroclastic activity. Schultz, Department of Geological Sciences, Brown University, Comparison: The dike patterns at Sudbury thus canbe inter­ Providence RI 02912, USA. preted in three ways. First, if the Sudbury Igneous Complex marks the location of the crater rim, the radial offset dikes might corre­ Introduction: The Sudbury structure in Ontario, Canada, is spond to the fractures extending beyond the easternrim ofHaldane. one of the oldest and largest impact slrUCtures recognized in the Theevendisbibutionofthesedikes aroundSudbury, however, then geologicalrecord[1]. Itis alsooneofthemostextensivelydeformed requires a uniformly tensile regional stress field, which is inconsis­ and volcanically modified impact structures on Earth [2-4]. Al­ tentwith theonsetofthe PenokeanOrogenyshortly aftertheimpact though few other terresbial craters are recognized as volcanically [10]. Further, since the Igneous Complex probably represents an modified, nwnerous impact craters on the Moon have been volca­ impact melt unit [11], the edge ofthis unit is more likely to reflect nically and tectonically modified [5] and provide possible analogs the edge ofthe basin floor than the crater rim. for the observed pattern ofmodification at Sudbury. In this study, In the second interpretation,the radialoffsetdikes correspondto we correlate the patternofearly deformation atSudbury to fracture radial fractures observed in the Haldane moat. In this case, the patterns in two alternative lunaranalogs and then use these analogs concenbicdikes and the SudburyIgneous Complex wouldmark the bothtoestimatethe initialsizeoftheSudburyslrUCture and tomodel edge of an uplifted floor plate, while floor uplift could produce a the nature of early crater modification at Sudbury. uniformly tensile stress field within the craterrim [12J. In addition, Structure DescrIptions: Two patterns ofdeformation can be the absenceofan upliftedcentralpeakcomplex atSudburyrequires distinguished at Sudbury: (1) an early sequence centered on the detachmentofthe impact melt from thecentral peaks during uplift. Sudbury Igneous Complex [6] and (2) several later episodes of Although rare, such detachments are observed on the Moon, where regional deformation that cut basin-controlled features, i.e., are afewcraters (e.g., Billy,Camoens)showevidencefor a "foundered insensitive to the impact slrUCture. The Main Igneous Complex central peak complex" [5,12]. presently defmes an elliptical ring about 60 km long and 27 km Third, the Sudbury dike pattern also matches the pattern of across. This norite/micropegmatite layered intrusion has a crystal­ fracturing inthe two-ring basinSchrodinger. Theradialoffsetdikes lization age of-1850 Ma, which is commonly assigned to the time are identified with the pattem of radial fractures occuning in the ofimpact [7]. It also feeds an extensive sequence ofoffset dikes in outer floor region at Schrodinger, whereas the concenbic offset the surrounding basement rocks radial and concenbic to the dikescorrelatewiththe innersequenceofconcenbicfractures along undeformed structure [6]. The radial dikes are the most evenly the interior peak ring- Due to the close proximity of the Igneous disbibuted and, although disrupted by later deformation, they can Complex to the concentric offsetdikes, this interpretation suggests extendupto30kmfrom theedgeoftheMainIgneousComplex.The that the Sudbury basin may preserve much of the original central less extensive concenbic dikes mostly occur south of the structure impact melt sheet consistent with [IIJ. wherethey aretypically about3-10km from the MainComplex [6]. Discussion: The comparison of Sudbury to lunar floor-frac­ Lastly, basin-centered concenbic lineaments can be identified in tured craters thus provides two alternative models for the initial satellite images 20--30 km north and west of the structure [8]. Sudbury slrUCture. Further, the apparent correlation ofcrater floor Unfortunately, deformation along the Grenville front to the south fractures to specific elements ofthe original craterstructure allows and at the Wanapitae Impact to the east masks any similar trends estimation of crater sizes for these two models. For the Haldane elsewhere around the basin. analogy, sinceradial fractures areconfmed to thecraterinterior, the Similar patterns of crater-centered radial and concenbic frac­ extent of the radial offset dikes from the original basin center tures are observed in lunar floor-fractured craters. Since these indicates a minimum Sudbury diameter of -100--120 km. If the patterns appear to be partly controlled by the original impact Sudbury basin contains a down-dropped central peak complex, the 80 1992 Sudbury Conference

size of the Sudbury Igneous Complex then provides a maximwn 22753-22764. [12] Wichman R. W. and Schultz P. H. (1992), in estimate for the crater size. Since strain analyses ofdefonnation in preparation. [13] HaleW. and Head I. W.(1979)Proc.LPSC 10th, the Sudburybasin indicateoriginal basindimensions of60Janby40 2623-2633. [14] LakomyR. (1990)IPSCXXl,678-679. [l5]Gupta km [3], the basal diameter of the Sudbury central peak structures V. K. etal. (1984) In [4], 281-410. [16] Baldwin R. B. (1968)JGR, was probably no more than 35-40 km. Using morphometric rela­ 73,3227-3229. [17] Walker R. I. et al. (1991)EPSL,105,416-429. tions for unmodified IWlar craters [13], these values indicate a maximwn crater diameter of-120-140 km. Alternatively, the Schrodinger analogy indicates a largermulti­ ring structure. Sinceradial fractures areconfmed to the outercrater VARIATION IN MULTIRING BASIN STRUCfURES AS A floor in this model, the extent of the radial offset dikes provides a FUNCTION OFIMPACT ANGLE. R. W. WichmUlandP.H. minimwn basindiameterof-130-140km (corresponding toabasin Schultz, Department of Geological Sciences, Brown University, floor diameterof-100-120Jan). Themaximwn sizeofthe original Providence RI 02912, USA. Sudbury Igneous Complex (-55-60 km), however, also can be related to the basin rim diameter. Ifthis value represents the initial Introduction: Previous studies have demonstrated that the size of the central basin floor, the rim crest diameter becomes impactprocess inthe laboratory varies as a function ofimpact Ulgle approximately 170-180 Jan, which is comparable to the recent [1-3]. This variation is attributed to changes inenergy partitioning estimate of180-200Jan derived from the distribution ofpreserved and projectile failure during the impact [2,3] and, in simplecraters, shock features around Sudbury [11,14]. Although erosional loss of produces a sequenceofprogressively smallerand more asymmetric the Igneous_Complex might accommodate an even larger basin craterfonns as impactangle decreases from -20° [1]. Cratershapes structure, the inferred locationofthe innerbasinring relative to the appear to be constant for higherimpact utgles. Furtherstudies have concentric offset dikes probably precludes any drastic increase in compared the Wlique signatures ofoblique impacts observed in the this estimate. laboratory to much larger impacts on the Moon and Mercury [3-5] The interpretation ofSudbury as a floor-fractured crater ortwo­ as well as Venus [6]. At the largest basin scales, the asymmetry of ring basin also provides two alternative models for early crater the transient cavity profile for highly oblique impacts results in Ul modification at Sudbury. First, most IWlar floor-fractured craters asymmetric lithospheric response [7]. Since transient cavity asym­ apparently reflect deformation over a crater-eentered laccolithic metry should decrease with increasing impact Ulgle, therefore, intrusion [5,12]. Since geophysical studies suggest the presence of comparison of large impact basins produced by slightly different a tabular ultramafic body beneath Sudbury [15], such an intrusion impact angles allows calibration of the effects such asymmetries alsomay bethe causeofdeformationatSudbury. The timing ofdike may have on basin formation. formation atSudbury, however, limits the potential melt sources of Basin Comparison: Although only Crisiwn shows a dis­ such an intrusion and requires direct interaction of the Sudbury tinctly elongated basin outline, both the Crisiwn and Orientale impact with either a contemporaneous orogenic melt or with an basins on the Moon apparently resulted from oblique impacts. Both orogenicthermalanomaly. Second,isostaticupliftofthe basinfloor basins possess an asymmetric basin ejecta pattern (see [8] for a could induce floor fracturing through flexure [16]. In this case, the review), and both basins also exhibit features predicted [7] for dike magmascouldbederivedprimarily from the impactmelt sheet collapse ofan oblique impact cavity: a gravity high offset from the rather than from a mantle melt, but isotope analyses ofthe Sudbury basin center [9-11] and a set of similarly offset basin ring centers ores still suggest a small (10-20%) component ofmantle-derived [7]. Based on the crater outlines and ejecta distributions in labora­ melts [17]. tory impact experiments, Crisium probably represents an impact In either case, interaction of the Sudbury impact with the event at -10-15° off the horizontal [3], whereas the ejecta pattern Penokean Orogeny can be inferred. Since both crater-centered and more circular basin outline at Orientale more closely resemble intrusions and isostatic relaxation should be favored by enhanced laboratory impacts at -15-25°. temperature gradients, this is consistent with the higher heat flows Three primarydifferences canbe identified betweentheCrisium and greater volcanism characteristic of orogenic settings. It also andOrientalebasin structures. First,theCordillerascarpiROrientale may explain why the majorityofterrestrial impactsinmorecratonic has a relief of -4-6 km (12], whereas the outer basin scarp at settings show little evidence of floor fracturing. Since high heat Crisiwn is poorly defmed and has a maximwn relief of only -1­ flows were apparently common during the early Archean, however, 2 km. Second, Orientale shows two distinct massifrings (the inner volcanic crater modification may have been more common at this and outer Rook Montes) separated by a nearly continuous trough time. Such early impact structures, therefore, may notresemble the structure, whereas the massifring at Crisiwn is only disrupted by a more recent impact structures preserved in the terrestrial impact medialsystemofdiscontinuous troughs [13-15].Thedistributionof record. Instead,like Sudbury, they may be preserved primarily as massiftopography is alsodifferent. Thehighestmassifs inOrientale complexes of (possibly anomolous) igneous intrusions. are in the Outer Rooks, butthe highest massifelevations in Crisium References: [1] Grieve R. A. F. and Robertson P. B. (1979) occur in the innermost massifs bounding the central mare. Third, Icarus, 38,212-229. [2] French B. M. (1970) Bull. Volcanol., 34, although mare volcanism in both basins has developed along the 466-517. [3] Rousell D. H.(1984)1n The GeologyandOreDeposils massif troughs and along the base of the outer basin scarps, such o/the Sudbury Structure (Pye et al., eds.), 83-95. [4] The Geology peripheral volcanism appears 10 be less extensive in Orientale andOreDepositso/theSudburyStructure (Pye etal., cds.), Ontario (Lacus Veris, Autumnae) than in Crisiwn (Lacus Bonitatis, Mare Geol. Survey, Spec. Vol. I. [5] Schultz P. H. (1976) Moon, 15, Spumans, Undarum, Anguis). 241-273. [6] Grant R. W. and Bite A. (1984) In [4], 276-300. Despite these differences, both Crisium andOrientale possess a [7] Krough T. E. et al. (1984) In [4],432-446. [8] Dressler B. O. common pattern of basin modification: Ul innermost steplike rise (1984) In [4], 99-136. [9] Wolfe R. W. and EI-Baz F. (1976) Proc. boWlds the central mare; the massif ring(s) are split by a sequence IPSC 7th, 2906-2912. [10] Peredery W. V. and Morison G. G. of concentric troughs, and, in both cases, the outer basin scarp is (1984) In [4],491-511. [11] Grieve R. A. F. et al. (1991) JGR, 96, mostprominentuprange. Further, theoccurrenceofperipheralmare IPI Contribution No, 790 81

volcanism is similar. Although peripheral ponded mare units do Crisium andOrientale, whichisconsistentwithgreaterscarpfailure • occur downrange in Crisium (Mare Spumans. Undarum. Anguis), inthese regions.Inaddition,however,the uprangeoffset inoblique these units canbecorrelated with impacts by hypervelocity decapi­ impacts ofthemantleuplift and deepest impactorpenetration could tated projectile fragments from the Crisium impactor [3J. We produce such an asymmetry by shifting the mantle melt reservoirs interpret these fundamental similarities between Crisium and uprange and byshifting thecenterofbasin uplift and the associated Orientale to be common signatures of basin formation at oblique flexural stress fields. impact angles. whereas the observed differences are attributed to Conclusions: Variations in impact angle can produce differ­ variations in cavity collapse as a function of impact angle. ences in the appearance ofmultiring impact basins. Comparison of Dlscwwlon: As discussed by [3.5.6J. smallercomplex craters Orientale to the more oblique impact structure at Crisium also (20-100 Ian) with asymmetric ejecta patterns typically exhibit suggests thatthesedifferences primarilyreflectthedegreeofcavity offset central peaks and more extensive wall slumps uprange. collapse. The relative changes in massif ring topography, basin reflecting the distinctive asymmetry ofcrater profdes for oblique scarp relief, and the distribution of peripheral mare units are impacts. Similarly, the differences in basin appearance between consistent with a reduction in degree of cavity collapse with Crisium and Orientale provide insight into the effects of impact decreasing impact angle. The prominent uprange basin scarps and angle oncavity collapseduring basin formation. First, the progres­ the restriction of tectonically derived peripheral mare units along sion from discontinuous concentric troughs in the Crisium massif uprange ring structures also may indicate an uprange enhancement ring to a split massifring atOrientale is consistent with increasing offailure during cavity collapse. Finally, although basinring faults failure of the transient cavity rim at higher impact angles. In appear to be preferred pathways for mare volcanism [16,20-22J, addition. the outerCrisium scarp is lowerthan the Cordillerascarp. fault-eontrolled peripheral mare volcanism occurs most readily whereas the inner Crisium massifs are much higher than the inner uprangeofanobliqueimpact; elsewhere suchvolcanism apparently Rook massifs. These topographic differences also support greater requires superposition of an impact structure on the ring fault. slumpingduringcavitycollapsewithincreasedimpactangles. Even References: [1] Gault D. E. and Wedekind I. A. (1978) Proc. the changing expression of the innermost ring structures (from a IPSC9th. 3843-3875. [2J SchultzP. H. andGaultD. E. (1990) GSA marebench-200m high inCrisium [16J to a combinationofscarps Spec. Pap. 247. 239-261. [3] Schultz P. H. and Gault D. E. (1991) and massifs in Orientale [8]) may reflect such variations in basin IPSC XXII, 1195-1196. [4J Schultz P. H. and Gault D. E. (1991) collapse,iftherelationofrim failure tointeriorupliftresembles that Meteoritics, 26.392-393. [5] SchultzP.H.• thisvolume. [6] Schultz observed in smaller terrestrial craters [17J. P. H. (1992) IPSC XXIII, 1231-1232. [7J Wichman R. W. and Second. the greater restriction ofuprange peripheral volcanism Schultz P. H. (l992)LPSC XXIII, 1521-1522. [8J Wilhelms D. E. inOrientale alsocanberelatedtogreatercavity collapse.lfreduced (1987) U.S. Geol. Surv. Prof. Pap. 1348. [9] Millier P. M. and cavity collapse indicates reduced cavity equilibration. isostatic Sjogren W. L. (1969) Appl. Mech. Rev.• 22, 955-959. [10J Millier upliftaftertheimpactshouldincreasewithdecreasing impactangle. P.M. etal. (1974) Moon. 10. 195-205.[11J SjogrenW. L. andSmith Sinceflexural stressesduring suchupliftaretensile atdepthoutside I. C. (1976) Proc. LSC 7th, 2639-2648. [12] Howard K. A. et al. the basin region [7J. conditions for structurally controlled dike (1974) Rev. Geophys. SpacePhys.• 12, 309-327. [13J CasellaC. J. formation will then dependon impact angle. Fora givenbasinsize, and Binder A. B. (1972) U.s. Geol. Surv. Map 1-707. [14] Olson therefore. basins formed by highly oblique impacts should induce A. B. and Wilhelms D. E. (1974) U.S. Geol. SllTV. Map 1-837. greater flexural stresses than higher-angle impacts; hence. magma [15J SchultzP. H. (l979)Proc.Conf.LunarHighlandsCrust.141­ columnsshouldbemorelikely to reach the surface along peripheral 142. [16J Head I. W. et al. (1978) Mare Crisium: The View from basin faults resulting from lower-angle impacts. Luna 24,43-74. [17J Croft S. K. (1981) Mldti-Ring Basins, Proc. Third, thegreaterprominanceofthebasinscarpsuprangeofboth LPSC 12A (P. Schultz and R. Merrill, cds.). 227-258, Pergamon, Crisium and Orientalemayreflectthe asymmetry ofoblique impact New York. [I8J Melosh H. J. and McKinnon W. B. (1978) GRL, 5, cavities.Intheringtectonicmodelofbasinscarp formation [18,19J, 985-988. [19] Melosh H. I. (1982)JGR,87.1880-1890. [20J Head theouterscarpreflects lithospheric failure overmantleflow intothe 1. W. (1976) Rev. Geophys. SpacePhys.• 14. 265-300. [21J Greeley collapsing cavity. For an axisymmetric flow field, therefore, the R. (1976) Proc.LSC7th. 2747-2759. [22] SchultzP.H. andGlicken basin scarp should be equally well developed around the basin H. (1979) JGR, 84,8033-8047. periphery. The uprange offset in deepest projectile penetration at low impact angles [1J, however, should modify the pressure gradi­ ents driving mantle flow during cavity collapse. Since both wall slopes and cavity depths are reduced downrange [1J. the volume of SELF-ORGANIZED ROCK TEXTURES AND MULTIRING mantle flow into an oblique transient cavity may be predominantly STRUCTURE IN THE DUOLUN CRATER. Wu Siben and derived from beneath the basin rim uprange. In addition, since Zhang Iiayun, Institute ofMineral Deposits, Chinese Academy of projectile failure should reduce the energy of later (downrange) Geological Science. Beijing 100037, China. cavity excavation in an oblique impact, peak shock pressures also should be centered closer to the point of initial contact (uprange); The Duolun impact crater is a multiring basin located 200 Ian thus shock disruption and acoustic fluidization could be enhanced northofBeijing [1,2]. From thecenterto the edgeofthe craterthere uprange of the transient cavity. are innermostrim, innerring, outerrim, and outermostring. The 5­ Theobservedsettingsoftheperipheralmareunits inCrisium and km-diameterraised innermostrim, 80to 150m above thesurround­ Orientale provide further support for the inferred asymmetry of ing plain, is located in the cratering center and consists ofvolcanic cavitycollapse. Althoughtectonicallycontrolledmareunits consis­ rock (andesite. etc.). Theprominent70- Ian-diameterring, which is tently develop along uprange scarps in Crisium and Orientale, encircledby the Luanriver. the Shandianriver.and theirtributaries, , peripheral volcanism downrange of Crisium apparently requires is a peripheral trough now occupied by the Lower Cretaceous coal­ intersectionofthering fault with an impact structure. This observa­ bearing formation. etc. The 82-Ian-diameter outer rim, 200 to tion suggests that magma column heights are greatest uprange of 250 m above the plain. consists of Archean metamorphic rocks 82 199" Sudbury Conferent:e

(quartzite,schist),lowerPermianlimestone andimpactmelt,which model of behavior ofimpact melt and vapor~ (2) search ofimpact­ I may indicate the boWldary of excavation. The 170-km-diameter induced geochemical effects in the rock from terrestrial impact outermost ring is marked by monomict megablock breccia. craters; and (3) studies of samples of lWlar regolith. Outside the innermostrim andnearthecratering centerthere are TheflISt approachresulted inthe conclusionthatevenatnotvery some lobes ofimbricate groWld surge consisting ofvolcanic rocks high temperatures a selective vaporization of geologic materials or impact melt. The groWld surges appear to be a lot of festoons occurs andmayleadtonoticeablechanges inelements' contents and consisting of a series of arclike ridges with steep outer slopes and ratios. This effect is controlled by the volatilities of individual gentle innerslopes. Theinclineoftheridgehas acentripetalgradual elements and compoWlds (sequences of volatilities were deter­ decreaseuptoapparentdisappearance. Theouteredgeofthe groWld mined) as well as their interactions, e.g., realized through the so­ surges forms an incomplete ring peak of 18-22 km diameter. The called basicity-acidity effect known in igneous petrology. phenomenon indicates that the groWld surges are products of Thesecondapproachresultedinthe conclusionthatgeochemical oscillatory uplift near the cratering center. effects thatmightbepredictedfrom theexperiments andtheoretical Recently we have fOWld some self-organized textures [3] or work are seen in some impact melts, but these observations are not chaos phenomena in shock-metamorphic rocks from the DuolWl reliable enough and one ofthemajorproblems is the Wlcertainty in impact crater, such as turbulence in matrices of impact glass, estimation of the preimpact target composition. oscillatory zoning, or chemical chaos of spherulites in spherulitic The third approach was successful in fmding such predictable splashedbreccia, fractal wavytextures orself-similarwavy textures impact-induced effects as partial loss of alkalis, Si, and Fe from with varied scaling in impact glass, and crystallite beams shaped regolith and some components ofhighland breccia. like Lorentz strange attractor. Therare phenomena indicate that the The next stage of the studies included experiments on quasi­ shock-metamorphic rocks from DuoIWl crater are fonned far from equilibrium vaporization of geological material in Knudsen cells. equilibirum. If it is considerable that impact cratering generates This study gave reliable data on forms of occurrence of rock­ momentarily Wlder high-pressure and superhigh-temperature, oc­ forming elements in the vaporphase, theirfugacities, andtime-and­ currence ofthose chaos phenomena in shock-metamorphicrocks is temperature sequences of the elements' vaporization, which not surprising. confirmed, in general, the earlierresults from vacuum vaporization This research is supportedby NSFofChina, and by the Chinese experiments. These results provided significant progress in Wlder­ FOWldation for Development of Geological Sciences and Tech­ standing the geochemical effects of impact cratering, but it is niques. evident that impact-induced vaporization is very fast and not References: [1] SibenW. (l987)LPSXVlII. 920-921. [2] Siben equilibrium. W. (1989) Meteoritics. 24,342-343.[3] OrtolovaP. 1. (1990) Earth This is why recently a series ofexperiments using the light gas Sci. Rev., 29. 3-8. gWl were made. The results show that even at a rather low velocity of impact (5 to 6 kmls) the silicate material involved in the shock displays effects ofmelting and partial vaporization. Itis interesting GEOCHEMICAL ASPECT OF IMPACT CRATERING: that with this fast and Wlequilibrium process the associations of STUDIES IN VERNADSKY INSTITUTE. O. I. Yakovlev and vaporized elements may differ drastically from those that resulted A. T. Basilevsky,VernadskyInstituteofGeochemistryandAnalytical from the slow quasiequilibrium process. Forexample, refractory U Chemistry, Russian Academy of Science, Moscow. and Th vaporize quite easily. Effects ofredox reaction between the materials of target and projectile were fOWld. Studies of the geochemical effects of impact cratering at the The mentioned experimental datashow that at small Oaboratory Vemadsky Institute incollaboration with the Institute ofDynamics experiments) scale geochemical effects do occur in high-velocity ofGeospheres, Moscow State University, Leningrad State Univer­ impacts.Buttheirroleinlarge-scalenaturalprocessessuch as heavy sity, and some other institutions were fulfllied by several ap­ bombardment at the early stages of planetary evolution should be proaches. clarified in future studies. At the initial stage, three approaches wereused: (1) experimental studies of high-temperature vaporization of geological materials (basalts, granites, and so on) in vacuum that was considered as a

•